Felix Hüning

The Fundamentals of Electrical Engineering

Felix Hüning

The
Fundamentals
of Electrical
Engineering
for Mechatronics
Author
Prof. Dr. rer. nat. Felix Hüning
FH Aachen, University of Applied Sciences
Faculty for Electrical Engineering and Information Technology
Eupener Strasse 70
52066 Aachen
huening@fh-aachen.de

ISBN 978-3-11-034991-7
e-ISBN 978-3-11-034990-0

Library of Congress Cataloging-in-Publication Data

A CIP catalog record for this book has been applied for at the Library of Congress.

Bibliographic information published by the German National Library

The German National Library lists this publication in the Deutsche Nationalbibliografie;
detailed bibliographic data are available on the Internet at http://dnb.dnb.de.

© 2014 Oldenbourg Wissenschaftsverlag GmbH

Rosenheimer Straße 143, 81671 München, Germany
www.degruyter.com
Part of De Gruyter

Editor: Gerhard Pappert

Production editor: Tina Bonertz
Cover picture: Oksana Kostyushko/iStock/Thinkstock

Printed in Germany
This paper is resistant to aging (DIN/ISO 9706).
Preface

Mechatronics is a multidisciplinary field and synergistic combination of electronic engineer-

ing, mechanical engineering and software engineering. Therefore knowledge of these disci-
plines is a must for mechatronic engineers as well as for mechanical and software engineers.
They should have a basic understanding of electrical engineering to be able to work on
mechatronic systems.
You can find mechatronic systems in nearly every area of modern life, from rather simple
white goods to much more complex systems such as are found in satellites. I have used the
practical applications found in modern vehicles as examples of complex mechatronic sys-
tems. Mechanical, electrical and software components are needed to realise complex func-
tionalities like hybrid and electrical powertrains in hybrid and electric vehicles (HEV/EV),
electric power steering or advanced driver assistant systems. Automotive applications are
therefore ideal for demonstrating the application of the electrical topics introduced in this
textbook.
The aim of this textbook is the introduction of basic concepts and laws of electrical engineer-
ing with an emphasis on mechatronic systems. It is based on a one-semester introduction
course in the “Fundamentals of electrical engineering for mechatronics” at the University of
Applied Sciences in Aachen, Germany.
The target group are students of either mechatronics, or other engineering topics who require
a brief introduction to electrical engineering. In addition, non-electrical engineers working
with mechatronic, or electrical systems can use this textbook as a quick start to understand-
ing electrical engineering. The focus of this book is to help students understand electrical
circuits and to learn different methods of how to analyse them.
The book starts with an introduction to the basic laws and concepts involved in solid state
physics and electric fields. Then basic electric circuit concepts and components like resistors
and sources are introduced. The main part of this book focuses on circuit analysis techniques.
For example, DC circuits are analysed in chapter four using Kirchhoff’s laws. DC analysis is
extended in chapter five using structured methods of analysing circuits using nodal, or mesh
analysis. In chapter six the operational amplifier is introduced as the first more complex
electrical device and circuits with operational amplifiers are analysed.
After DC analysis, time-dependent circuits are analysed. Chapter seven starts with an intro-
duction to capacitors and inductors as energy storage elements. These elements are used to
build circuits with transient behaviour. Afterwards AC circuits are analysed using complex
AC analysis.
In chapter eight simple circuits are combined into more complex building blocks. The analy-
sis of complex circuits can be reduced to the analysis of simpler sub-circuits using the con-
cept of building blocks. Chapter nine deals with AC power and in chapter 10 oscillating
circuits are introduced.
VI Preface

More complex circuit elements are part of chapter 11. It includes semiconductor devices like
diodes, biopolar junction transistors and MOSFETs. The textbook finishes with a short intro-
duction to the important field of circuit simulation.
In addition to this theoretical introduction using a textbook, exercises are very important to
gain a deeper understanding of the subjects. Exercises and solutions to each of the chapters
can be found online under www.degruyter.com.
Last but not least, I would like to thank Prof. Dr. Martin Ossmann for discussing the tech-
nical details and for his very helpful feedback as well as to Gary Evans for editing the text
and to Caroline Huertgen for her support with the images.
Table of contents

1 The fundamentals of solid-state physics 1

1.1 Charge carriers, crystal structure and conductivity .................................................... 1
1.2 The electrical properties of solids .............................................................................. 2

2 Fundamental electrical laws 7

2.1 The basics of electric field theory .............................................................................. 7
2.2 Electric potential and voltage..................................................................................... 8
2.3 Displacement field and electric flux ........................................................................ 10
2.4 Electric current and current density ......................................................................... 12

3 Fundamental circuit elements 15

3.1 Electric circuits ........................................................................................................ 15
3.2 Consumer and generator system .............................................................................. 16
3.3 Voltage sources ........................................................................................................ 16
3.4 Ideal current sources ................................................................................................ 19
3.5 Resistance, resistors and Ohm’s law ........................................................................ 21
3.5.1 Real resistors ............................................................................................................ 24
3.5.2 Short circuit and open load ...................................................................................... 25
3.5.3 Real voltage sources ................................................................................................ 27
3.5.4 Real current sources ................................................................................................. 28
3.5.5 Transformation of sources ....................................................................................... 30

4 Fundamental electrical circuit laws 33

4.1 Kirchhoff’s laws ...................................................................................................... 33
4.1.1 Kirchhoff’s current law ............................................................................................ 33
4.1.2 Kirchhoff’s voltage law ........................................................................................... 36
4.2 Operating point ........................................................................................................ 38
4.3 Wye-Delta transformation ........................................................................................ 40
4.4 Meters and measurements ........................................................................................ 42
4.5 Power and energy..................................................................................................... 44
4.6 Maximum power transfer......................................................................................... 45
4.7 Dependent and independent sources ........................................................................ 47
VIII Table of contents

5 Circuit analysis 51
5.1 Nodal analysis ..........................................................................................................51
5.2 Mesh analysis ...........................................................................................................60
5.3 Linearity and Superposition ......................................................................................65
5.4 Two-terminal circuit and Thévenin’s theorem ..........................................................70
5.5 Norton’s theorem ......................................................................................................75

6 Operational amplifier 79
6.1 Operational amplifier ................................................................................................79
6.2 Operational amplifier circuits ...................................................................................83

7 Time domain circuit analysis 91

7.1 Capacitor...................................................................................................................91
7.2 Inductors ...................................................................................................................99
7.3 Transient effects and switching ..............................................................................102
7.3.1 First order circuit – the natural response ................................................................103
7.3.2 First order circuit – complete response ...................................................................106
7.3.3 Second order circuit – the natural response ............................................................ 112
7.3.4 Second order circuit – the complete response.........................................................120
7.4 AC Analysis ............................................................................................................124
7.4.1 Vector diagram........................................................................................................129
7.4.2 Complex numbers ...................................................................................................134
7.4.3 Application of complex numbers to AC circuits.....................................................136
7.4.4 AC circuits ..............................................................................................................137
7.4.5 Kirchhoff’s laws for AC circuits .............................................................................140

8 Building blocks 149

8.1 High-pass filter .......................................................................................................149
8.2 Bode plot ................................................................................................................152
8.3 Low-pass filter ........................................................................................................153
8.4 Higher order filters .................................................................................................155
8.5 Active filter .............................................................................................................159

9 AC power 161
9.1 AC power of a pure resistive two terminal network ...............................................161
9.2 AC power of a pure inductive two terminal network .............................................162
9.3 AC power of a mixed two terminal network with L, R and C ................................164
Table of contents IX

10 Oscillating circuits 169

10.1 Series configuration ............................................................................................... 169
10.2 Parallel configuration ............................................................................................. 175

11 Semiconductor devices 177

11.1 Diode ..................................................................................................................... 178
11.2 Bipolar transistor.................................................................................................... 186
11.3 MOSFET................................................................................................................ 192

12 Circuit simulation 199

PSPICE .................................................................................................................. 200

References 205

Index 207
1 The fundamentals of solid-state physics

1.1 Charge carriers, crystal structure and conductivity

You should be familiar with some very basic concepts of matter and electrical quantities such
as charge, current and voltage in order to describe the operation of electronic circuits.
Matter consists of atoms made of a nucleus which is orbited by negatively charged electrons.
The nucleus itself is made up of positively charged protons and neutrons without an electrical
charge. Electric charge is measured in coulombs (C). The smallest fundamental unit of elec-
trical charge is a physical constant called the elementary positive charge:

e  1.6021019C
All charges are an integer multiple of this elementary charge. The proton has a charge of +e
while the electron has a charge of –e.
As a consequence of quantum mechanics, the electrons of a single isolated atom (not inter-
acting with other atoms) can occupy only discrete energy levels (atomic orbitals), see Fig.
1.1.

Fig. 1.1: Splitting of discrete energy levels into energy bands for increasing number of atoms.

If several atoms form a molecule by chemical bonding the atomic orbitals split into separate
molecular orbitals with different energy levels. Fig. 1.1 shows this splitting of energy levels:
For an H2 molecule with two atoms the energy levels of the single atoms split into two ener-
gy levels with the lower energy level being occupied by two electrons. In general the outer-
most electrons (valence electrons) can participate in the formation of chemical bonds with
other atoms to form molecules, in solid, liquid or gaseous states.
When a large number of atoms form a solid the number of orbitals (proportional to the num-
ber of valence electrons) becomes very large and the difference in energy between these
orbitals becomes very small. In consequence solids (with about 1023 or more atoms) show
continuous energy bands rather than discrete energy levels. The energy bands can overlap or
2 1 The fundamentals of solid-state physics

are separated by intervals of energy without orbitals (electrons within the solid cannot have
these energies). These forbidden energy intervals are called band gaps and the total of bands
and band gaps is called the band structure.
Fig. 1.2 shows a simplified diagram of the electronic band structure of crystalline solids on
the right. The shape of the band structure depends on the atoms forming the solid and its
crystal structure.

Fig. 1.2: The electronic band structure of solids: metal (left); semiconductor (center); insulator (right).

The electric properties of a solid are mainly determined by the band structure around the so-
called Fermi level (see Fig. 1.2) where the Fermi level in a simplified image divides the band
structure into a region at a low energy level that is occupied by electrons and a region at a
higher energy level that is empty. The highest (almost fully) occupied band is called the va-
lence band and the lowest (almost) empty band is called conduction band. Conductivity oc-
curs if the conduction band is partly filled by electrons or electrons are missing in the va-
lence band.

1.2 The electrical properties of solids

Electrical properties can be distinguished depending on the presence and size of a band gap
at the Fermi level.

Metals
The valence and conduction band overlap (the Fermi level lies within this overlapping band)
or the Fermi level lies within the conduction band (not shown) and therefore the band is
partly filled with electrons regardless of temperature. These electrons form a “sea” of practi-
cally free electrons moving in the background of the positively charged crystal structure (Fig.
1.3). The electron density is of the same magnitude as the density of the atoms. For example
the density of free electrons in copper is about 8·1022 electrons per cm–3. The conductivity is
very high as the electrons can easily absorb energy. Conductivity decreases with increasing
temperature. Classic examples are silver, copper and iron (see Tab. 1.1).
1.2 The electrical properties of solids 3

Fig. 1.3: The crystal structure of a metal: positively charged atomic cores surrounded by delocalized free
electrons.

Insulators
In an insulator the Fermi level is located within a large band gap. The valence band (at abso-
lute zero) is fully occupied and the conduction band is empty, resulting in no conductivity. At
higher temperatures electrons can be excited to the conduction band due to thermal energy
(leaving a hole in the valence band), but at reasonable temperatures the number of excited
electrons is negligible and there is no conductivity.
Examples of insulators are glass or plastic materials, see Tab. 1.1.

Semiconductors
Like for insulators the Fermi level for pure semiconductors, (also known as intrinsic semi-
conductors), is within a band gap of width ΔE, but this time the band gap is smaller. Semi-
conductors are isolators at absolute zero. Thanks to the smaller band gap, electrons can be
excited to the conduction band more easily due to thermal energy. As the excitation of elec-
trons is a thermal effect the number of intrinsic electrons in the conduction band ni is strong-
ly temperature dependent:
3  E
2k BT
ni ~ T 2 e

Here kB=1.38…·10–23 J/K which is the Boltzmann constant.

When an electron is excited to the conduction band this electron is missing in the valence
band. This missing electron is called an electron hole, or defect electron and also contributes
to conductivity. The density of holes is the same as the density of excited electrons. The
conductivity of semiconductors is higher than that of insulators, but still significantly lower
than that of metals.
4 1 The fundamentals of solid-state physics

Fig. 1.4: The crystal structure of silicon.

As every excited electron generates a hole the density of electrons n0 and holes p0 is the same
and corresponds to the intrinsic charge carrier density ni:
n0  p0  ni
The opposite of the generation of electron-hole pairs is called recombination. The recombi-
nation rate depends on the carrier density of electrons and holes. The rate of generation and
recombination of electron-hole pairs is temperature dependent. In thermal equilibrium both
rates are the same and the number of free electron-hole pairs is constant at the given tem-
perature. The equality of the two rates leads to the mass action law:

n0  p0  ni
2

The product of the charge density of the free electrons and holes equals the square of the
intrinsic charge carrier concentration. Mass action law also holds true for doped semiconduc-
tors.
The most important semiconductor used in semiconductor devices is silicon.
The band gap of silicon:
~1.1 eV (where 1 eV = 1.602…·10–19 J is a measure of small scale energy)
Atom density:
~5·1022 atoms per cm–3
The number of intrinsic charge carriers due to thermal activation at room temperature
(293 K):
~1.5·1010 electrons per cm–3
~1.5·1010 holes per cm–3
This example shows that the carrier density of intrinsic semiconductors like silicon is signifi-
cantly lower than of metals. Therefore intrinsic semiconductors are rather poor electrical
conductors.
Other examples of intrinsic semiconductors are germanium (band gap about 0.7 eV), SiC
(band gap about 2.3 eV) or GaN (band gap about 3.2 eV).
1.2 The electrical properties of solids 5

Doped semiconductors
The electron density and conductivity of semiconductors can be varied by so called doping.
By doping small amounts of silicon atoms (with four valence electrons per atom) are re-
placed by other atoms with a different number of valence electrons in a very controlled man-
ner. In p-type semiconductors the number of valence electrons is less than 4 (e.g. boron, three
valence electrons) and holes are easily generated as majority carriers. In n-type semiconduc-
tors the number of valence electrons is more than four (e.g. arsenic, five valence electrons)
and electrons as majority carriers are easily generated. In this way the electron (or hole)
density of the semiconductor can be changed to a desired number.
Fig. 1.5 shows the crystal structure of silicon doped with arsenic (left side) as a donor. The
binding of the additional electron to the arsenic atom is rather weak and, as shown in the
band structure in Fig. 1.5 (right), the energy levels of this electron are within the band gap,
just 0.049 eV below the conduction band. This electron can easily be excited to the conduc-
tion band even at low temperatures and can increase conductivity. Note that no hole is gener-
ated in the valence band.

Fig. 1.5: The crystal structure of arsenic-doped silicon, n-type semiconductor (left) and the band structure of a
n-type semiconductor showing the donator’s energy levels within the band gap (right).

The temperature dependence of a doped semiconductor is depicted in Fig. 1.6. Even at low
temperatures (some 10 K) all impurity atoms are ionized generating electrons in the conduc-
tion band (n-type semiconductors) or holes in the valence band (p-type semiconductors). In a
wide temperature range around room temperature the carrier density is almost constant n0
and equals the density of impurity atoms ND. At higher temperatures intrinsic carriers are
increasingly generated.
6 1 The fundamentals of solid-state physics

Fig. 1.6: The carrier density n of an n-type semiconductor as a function of temperature: solid line = total carrier
density n, dotted line = intrinsic carrier density ni; ND is the density of impurity atoms.

As the mass action law is also valid in doped semiconductors the density of minority charge
carriers can be calculated, e.g. for n-type semiconductors:
2 2
ni n
p0   i
n0 N

Tab. 1.1 lists some examples for metals, semiconductors and insulators and corresponding
conductivity values.

Tab. 1.1: The conductivity values and electrical classification for certain materials.

Classification Material Specific conductivity σ [S/m]

Metal Silver 61·106
Copper 58·106
Iron 10·106
Semiconductor Germanium 1.45
Silicon (pure) 252·10–6
Insulator Plastic material < 10–9
Glas < 10–9
Diamond < 10–9
2 Fundamental electrical laws

2.1 The basics of electric field theory

Any electric charge is responsible for an electric field that surrounds the charge. Forces are
exerted on each other by charges due to this electric field. Charges with same sign repel
whereas charges with an opposite sign attract. The force of charge Q2 on another charge Q1
can be expressed by Coulomb’s law:
 Q Q  Q1  Q2 
F21  1 22 a 21  a 21
4R 4 0 r R 2

As shown in Fig. 2.1 R is the separation of the two point charges Q1 and Q2. a21 is the unit
vector along the line joining the charges. The parameter ε = ε0 · εr is the permittivity of the
surrounding material between Q1 and Q2. Constant ε0 is the permittivity of free-space
(ε0 = 8.854…·10–12 As/Vm) and εr is the relative permittivity of the material. For a vacuum
the relative permittivity is equal to 1.
Of course Q1 exerts a force of the same magnitude, but opposite direction on Q2. The force is
directly proportional to the product of the charges and inversely proportional to the square of
the distance.

Fig. 2.1: An illustration of Coulomb’s law


The electric field strength E describes the force of the electric field onto charged particles.
Consider a very small charge Q1 in the electric field of Q2. The force of Q2 on small charge
Q1 is given by Coulomb’s law and the electric field strength is:

F 21 Q2
E  a 21
Q1 4R 2

The unit of the electric field strength is N/C = V/m.

8 2 Fundamental electrical laws

The electric field strength is parallel to the force for positive charge Q1 and antiparallel for
negative charge Q1. The electric field strength can be depicted by electric field lines as
shown in Fig. 2.2. The field lines normally start at the positive and end at the negative
charge, the direction is indicated by arrows. The strength of the field is given qualitatively by
the separation of the field lines: the closer the lines the higher the strength.

a E
Į
ds b

Fig. 2.2: Electric field lines for two charges (left) and the voltage in an inhomogeneous electric field (right)

2.2 Electric potential and voltage

When electrical forces act on a particle with charge Q1, it will possess potential energy. To
 
move the particle in an electric field of strength E along a distance ds you have to expend
(or you gain) the energy dE:
    
dE  F  ds  Q1  E  ds

Do not mix up electrical field strength E (it is a vector field) with the electrical energy E
which is a scalar. If a positive charge is moved in the direction of the electric field it gains
energy (dE > 0), if it is moved in opposite direction you have to expend energy (dE < 0). If
moved perpendicular to the field on the lines of same potential the energy is zero. These lines
of same potential are called equipotential lines.
The electrical voltage is now defined as the energy needed to move the particle in an electric
field from point A to B divided by the charge Q1:

E AB
B
 
U AB    E  ds
Q1 A

The electrical voltage is a scalar quantity and the unit for the voltage is the volt (V).
In a homogeneous electrostatic field (the electric field strength has the same size and direc-
tion everywhere) the voltage between A and B with distance d is simply:
U AB  E  d
The voltage in case of electro static fields is independent of the path the particle takes within
the field when moving from point A to B (path P1 or P2):
2.2 Electric potential and voltage 9

B
  B
  A
 

A
E  ds 
 E  ds
A

  E  ds
B
P1 P2 P2

Or in other words: The line integral along any closed path (loop) is always zero:
B
  A   

A
 
E  ds  E  ds  E  ds  0
B

This equation is one of Maxwell’s equations for the electrostatic field in integral form. It can
be rewritten in differential using Stoke’s integral theorem.

Excursus: Stoke’s integral theorem


For an arbitrary vector field a the integral along a closed loop around any arbitrary sur-

face A is:
   
 a  ds   rot a  dA   ■
A

Applying Stoke’s integral theorem to the Maxwell equation given above yields:
   
 E  ds   rot E  dA  0  
A

As this equation holds true for any surface A the second integrand has to be zero:

rot E  0
This differential form of Maxwell’s equation states that the electrostatic field has no curls.
The electric potential is the voltage at any point referred to a fixed reference point. It is a
scalar field. As the voltage is always referred to two locations (A and B) the voltage can be
written in terms of the electric potential as:
B
  B


U AB  E  ds   A   B   d
A

A

This equation is true for any path between to arbitrary locations and therefore yields the
correlation of electric field and electric potential:

E   grad 

The electric field is given by the gradient of the electric potential.

10 2 Fundamental electrical laws

2.3 Displacement field and electric flux

As given above Coulomb’s law and the electric field of a charge is strongly dependent on the
surrounding material: by changing the surrounding material the electric field is changed. This
dependence is given by the relative permittivity εr. It is a material parameter without a unit.
In the simplest case εr is just a constant scalar, but in general it can also be non-linear (εr
 
depends on E or D like in ferroelectric materials such as BaTiO3) or anisotripic (εr is a
 
tensor and E and D are not parallel). These three cases are depicted in Fig. 2.3.

Fig. 2.3: The relationship between electric and displacement field : linear with constant εr (left); non-linear with
hysteresis shape (center); anisotropic (right).

If the material is an electrical insulator the material is called a dielectric. An external electric
field causes a shift of electric charges inside the dielectric as shown in a simplified picture in
Fig. 2.4. The origin of the dielectric polarization can be a displacement polarization of the
atoms, or an orientation polarization by the alignment of permanent dipoles with respect to

the external field. In both cases the polarization field P has an opposite direction compared
with the external field and the external field is weakened.

Fig. 2.4: The displacement polarization (left) and orientation polarization (right).

To get rid of the material dependence the electric displacement field can be defined by the
superposition of the external electric field and the polarization field:
    
D  E   r  0 E   0 E  P
2.3 Displacement field and electric flux 11

The displacement field or electric flux density describes the density of field line independent
of the surrounding material. The unit for the displacement field is C/m².
For a point charge q in the origin of the coordinate system the displacement field at location

r yields:
  q  q 
D r   a  r
4r 2 21
4r 3
As the displacement field describes the density of field lines an electric flux is associated
with it. The electric flux ψ is given by the integral of the displacement field perpendicular to

any arbitrary surface A :
 

  D  dA

A

The unit of the electrical flux is C. If the displacement field is homogeneous and the surface

A is perpendicular to the field the electric flux is just:
  D A

  
Fig. 2.5: Electric flux D through a surface A with surface element d A

The displacement field of a point charge is given above and the electric flux for any arbitrary
surface is:
q  
 

A
4r 3
r  dA

Integrating over a closed surface with the point charge inside gives the total electric flux of
the point charge. This closed surface integral is in particularly simple for a sphere (with radi-
us r, see Fig. 2.5) as the displacement field everywhere is perpendicular to the sphere.
  q q
  D  dA  4r   dA  4r
sphere
2
sphere
2
 4r 2  q

The total electric flux through a closed surface like the sphere is equal to the point charge
inside the surface. This result is valid for any surface. In addition any arbitrary charge distri-
bution can be built by point charges and the total charge Q within the closed surface is given

by the charge density ρ( r ):
12 2 Fundamental electrical laws


Q   r dV
volume

Second Maxwell’s equation:

For any arbitrary charge distribution within a volume V the total electric flux through the

closed surface A around the volume is equal to the total charge:
  
   D  dA  Q    r dV
sphere volume

Excursus: Gauss’s integral theorem


For an arbitrary vector field a the integral over a closed surface around an infinitesimal
volume V is:

a r dA

 
div a r   limV 0 A
V
The flux through a closed surface of an infinitesimal volume equals the divergence of

the original vector field ar  . ■

Applying Gauss’s theorem to the second Maxwell’s equation yields its differential form:
  
div D r    r 

Charges are the sources of the displacement field.

2.4 Electric current and current density

The rate of movement of net positive charge dq(t) per unit of time dt through a cross section
of a conductor is known as current:
dq (t )
i (t ) 
dt
For time-dependent values lowercase letters are used, for time-independent values capital
letters. The SI basic unit for current is the Ampere (A) where 1 Ampere corresponds to the
charge flow of 1 C within 1 second:
C
1A  1
s
As we have seen in chapter 1.2 free electrons are the exclusive charge carriers in most metal-
lic conductors such as copper wires. Since the charge of electrons is negative and since the
direction designated for the current, as given above, is that of the net positive charge, the
charges in the wire thus move in opposite direction to the current designation, see Fig. 2.6.
2.4 Electric current and current density 13

Fig. 2.6: The definition of conventional (technical) current flow (flow of positive charge carriers) and the direc-
tion of electron flow.

If a current I is uniformly distributed across a cross-section of a conductor A (like a wire) the

current density J is:
I
J
A
The unit for the current density is A/m² or (more realistic) A/mm².
To calculate the flow speed or drift velocity v of electrons we need the time t an electron
needs to move for a distance Δl. In other words t is the time it takes for the charges in the
grey volume of Fig. 2.7 to pass through cross-section A. The charge Q in the grey volume
inside the blue volume is:
Q  e  n  A  l
Here e is the elementary charge and n is the electron density of the conductor (e.g.
8.5·1019 mm–3 for copper). Corresponding current is:
Q l
I  e n A  en  Av
t t
Finally the drift velocity yields:
I
v
en A
Drift velocity is rather small (~1 mm/s), depending on the current, cross-section and electron
density. Nevertheless the cause of the current propagates with about the speed of light.

Fig. 2.7: The current density J for a current flowing in a wire with cross-section A
3 Fundamental circuit elements

3.1 Electric circuits

A collection of interconnected electronic components such as voltage and current sources,
resistors, capacitors and other active and passive elements that has at least one closed path in
which current may flow is called an electric circuit or network (Fig. 3.1). The behavior of
these networks can be determined using so-called circuit analysis, either time-independent or
time-dependent.
To make circuit analysis easier the so called lumped element model is used. This model sim-
plifies a circuit in a way that the properties of the circuit, like resistance, inductance and
sources, are concentrated into idealized electrical components. These idealized components
(resistors, capacitors, inductors, etc.) are connected by perfectly conducting wires. In addi-
tion, the dimensions of the circuit have to be much smaller than the circuit’s operating wave-
length, or, in other words, the time it takes for signals to propagate around the circuit can be
ignored.

Fig. 3.1: Simple electric circuits or networks.

There are two types of elements in a circuit: sources and loads. A source usually supplies
energy to the circuit. It is a force that drives the current through the circuit, like a battery or a
generator. When the current flows out of the positive terminal of an electric source, it implies
that non-electrical energy has been transformed into electrical energy. In contrast a load ab-
sorbs the energy supplied by a source. The current delivered by the source passes through the
load. When current flows in the direction of a voltage drop, it implies that electrical energy is
transformed into nonelectrical energy.
In a circuit, electrical sources and loads may usually be easily distinguished by a comparison
of their current direction and voltage drop polarities:
– Electrical source: voltage polarity opposite to technical current flow (voltage rise)
– Electrical load: voltage drop parallel to technical current flow (voltage drop)
16 3 Fundamental circuit elements

A source-load combination is depicted in Fig. 3.1. A node in a circuit is a point where two or
more components or devices are connected together. A branch is a part of a circuit containing
only one component between two nodes. A loop or mesh is a closed path through a circuit in
which no electric element or node is encountered more than once. A mesh that contains no
other meshes is called an essential mesh. Both nodes and meshes play a major role in circuit
analysis.

3.2 Consumer and generator system

Both current and voltage have an orientation in electric circuits. As seen above, the current
flows in the direction of the positive charge carriers from the positive terminal of a voltage
source through the circuit to the negative. Voltages usually point from the positive to the
negative terminal of a source as shown in Fig. 3.2.

Fig. 3.2: The direction of current and voltage.

Consider now the circuit on the right of Fig. 3.2. The direction of current depends on voltag-
es of the two sources and the direction indication can be chosen arbitrarily at the beginning.
Consider voltage U1 is higher than U2, then the current flow is as shown. For the right part of
the circuit current and voltage are parallel and the power P = U · I is dissipated. For this reason
it is called a consumer system when current and voltage are parallel. In the left part current
and voltage are antiparallel. This behavior is called generator system as this part provides
power P = –U · I.
During circuit analysis it is common to use the consumer system for resistances to be able to
use Ohm’s law in the way given below without a minus sign.
Before starting circuit analysis, basic elements like current and voltage sources and resistors
are introduced.

3.3 Voltage sources

Any electric circuit needs at least one point where charge carriers are driven by a force. This
point is called the source. The driving force of voltage sources is the electrical voltage. Volt-
age sources can be constant or may be a function of time as depicted in Fig. 3.3.
3.3 Voltage sources 17

Fig. 3.3: Constant (left) and time-dependent (right) voltage source.

Voltage sources can be based on different physical and chemical principles, for example
electro-chemical voltage sources (e.g. batteries) and electro-mechanical voltage sources (e.g.
generators).

Fig. 3.4: A simple electric circuit with an ideal voltage source.

The electromagnetic force within the sources causes current to flow as soon as the loop is
closed. The current flows in a closed loop from one terminal of the source to the other.
Therefore current lines do not have a starting or an ending point. The positive terminal is
called the anode whereas the negative terminal is called the cathode. In case there are both
positive and negative charges inside a conductor these different charge carriers flow in oppo-
site directions. As already mentioned above we need a definition for the direction of the
current flow:
The direction of a current is defined to be same as the direction of the positive charges
and opposite to the direction of negative charges.
This definition implies that flow of electrons in metallic conductors is opposite to the direc-
tion of the current.
18 3 Fundamental circuit elements

Fig. 3.5: A simple electric circuit indicating the direction of current flow I and the direction of electrons.

Ideal voltage sources

For an ideal voltage source the voltage at the terminals is independent of the load connected
to the terminals and the current. It is called a DC source if the voltage is time-independent as
the current is a direct current in this case. If the voltage of the source is a function of time
(like a sinusoidal voltage source) it is called an AC source as it produces an alternating cur-
rent.
Instead of this physical model the following graphical symbols are used for ideal voltage
sources in electronic schematics:

Fig. 3.6: A simple notation of voltage sources: general symbol (left), electrochemical symbol (battery, center),
DC generator (right).

The definition of an ideal voltage source (voltage independent of load connected to the
source) implies that it is not permitted to connect two (or more) ideal voltage sources in
parallel to the same terminal.

Fig. 3.7: Do not connect ideal voltage sources in parallel.

A series connection of ideal voltage sources is possible and for the resulting voltage at the
terminals the voltages of the single sources are simply added.
3.4 Ideal current sources 19

Fig. 3.8: The permitted connection of two ideal voltage sources in series..

Automotive application
The number of electrical systems in modern cars is steadily increasing. The control of these
systems is often achieved using electrical control units (ECUs). Even in conventional cars
with a combustion engine the number of electrical systems and ECUs can be up to 100, or
even higher. ECUs can be found nearly everywhere in the car: for lighting, motor control and
transmission as well as for convenience applications like seat heating, window winders or
multimedia systems. The electrical systems are supplied by a 12 V power circuit and lead
acid batteries are commonly used as electrical energy storage elements. To reach the power
circuit voltage of 12 V 6 lead acid cells with a nominal voltage of about 2 V each are con-
nected in series.
In electric vehicles (EV) with an electric traction motor (and to some extent also for hybrid
electric vehicles (HEV) combining a combustion and electric motor) an additional power
circuit with higher voltages is introduced to provide sufficient power to the electric traction
motor. The voltage level is up to 350–400 V and lithium ion batteries are used for the high
voltage power circuit. Again the battery is built up of a series connection of single cells. For
lithium ion batteries each cell has a voltage of about 3.24 V depending on the technology.
E.g. LiFePO4 cells have an end-of-charge voltage of about 3.6 V and 100 cells are concate-
nated in series to reach a voltage level of 360 V. ■

3.4 Ideal current sources

An ideal current source drives a current I or i(t) regardless of the load connected to the ter-
minals but there has to be a load connected to the terminals. Without the load the current
source cannot produce the current as the circuit has to be closed. Like for ideal voltage
sources the current will be in general a function of time. Fig. 3.9 shows models of ideal cur-
rent sources and waveforms.
20 3 Fundamental circuit elements

Fig. 3.9: Models of ideal current sources (top) and current waveforms (bottom).

For ideal current sources it is not permitted to connect two or more in series. But a parallel
connection of ideal current source is possible and the currents of the two sources are simply
added:

Fig. 3.10: Two ideal current sources connected in parallel.

3.5 Resistance, resistors and Ohm’s law 21

3.5 Resistance, resistors and Ohm’s law

The ratio of the voltage across a material u(t) and the current through it i(t) is called the re-
sistance of the material. If this ratio is constant for the material independent of current or
voltage, is it called a linear resistor (short: resistor) and it’s resistance is:
u(t )
R
i(t )
In case of DC current and voltage sources Ohm’s law yields:
U
R
I

Fig. 3.11: Current I and voltage U for a resistor R.

This law is called Ohm’s law and the SI unit for the resistance is the Ohm or Ω.
The resistor is a simple component, usually considered as linear, concentrated (lumped mod-
el) and is a constant. Symbols for resistors are:

Fig. 3.12: Different models for resistors used in electric circuits: European style (left) and American style (right).

The reciprocal of the resistance R is the conductance G (SI unit: S=1/Ω, siemens):
1
G
R
The value of the resistance of a component is mainly determined by the physical dimensions
of the component and the specific resistivity of the material of which the resistor is com-
posed. For a bar of resistive material of length l and cross-section A the resistance R is given
by
l   (T )
R
A
Here ρ(T) is the specific resistivity of the material in Ω·m. The reciprocal is called the specif-
ic conductivity, given in S/m. Tab. 1.1 lists the specific conductivity values for some materi-
als.
A copper wire of 1 m length and a diameter of 2 mm has a resistance of about 5.5 mΩ at
room temperature.
22 3 Fundamental circuit elements

The (specific) resistivity of conductor metals is temperature dependent and varies approxi-
mately linearly over (normal operating) temperature (see Fig. 3.13). The resistivity at tem-
perature T can be calculated based on the resistivity at a given temperature (e.g. room tem-
perature, R(293 K) and a material dependent constant τ:
T 
R (T )  R ( 293 K )   R ( 293 K )  (1    (T  293 K ))
293 K  
The abbreviation
1

293 K  
is called the temperature coefficient of the resistance. It depends on the material and the
given temperature (here: 293 K). Copper for example has a constant τ = 38 K and a tempera-
ture coefficient at room temperature of α293 K = 3.9·10–3 K–1. In other words, the resistance of
copper increases by ~0.4 % for every degree Kelvin or doubles when heated up to 463 K. For
the 1 m copper wire of 2 mm diameter the resistance at 125 °C (the ambient temperature
within the engine compartment) increases by 40 %.
The temperature dependence of metals has a positive slope and metals are typical examples
for materials with a positive temperature coefficient (PTC): the higher the temperature the
higher the resistance. The opposite of PTC elements are NTC elements (negative temperature
coefficient). For these materials (e.g. semiconductors) the resistance decreases with increas-
ing temperature.

Fig. 3.13: Temperature dependence of the resistance of a conductor.

Automotive application
Everything is resistive in any electrical application and countless resistors are used in all
kinds of electrical systems. Besides resistors in electric circuit resistive devices can also be
used as sensors. In this application they make use of the geometric and temperature depend-
ence of the resistance for example.
Resistors with a well known temperature dependence are used as sensors to measure e.g. air
temperature, water or oil temperature. Very often NTC materials with a dedicated tempera-
3.5 Resistance, resistors and Ohm’s law 23

ture dependence are used as temperature sensors. A small measurement current flowing
through the NTC causes a voltage drop across this element. According to Ohm’s law this
voltage drop together with the measurement current corresponds to a resistance value. This
resistance is in the end a measure of the temperature. In Fig. 3.14 the typical characteristics
of a NTC temperature sensor is depicted. The resistance varies in a wide range and makes a
temperature measurement with high resolution possible.

Fig. 3.14: Typical characteristic of a NTC and circuit symbol.

The geometric dependence of the resistance is used by resistance strain gauges to measure
force, pressure or torque. If a bar of resistive material like silicon is compressed or length-
ened, the geometry of the bar changes as shown in Fig. 3.15 on the left side. Both the length
and the diameter of the bar change slightly if strain is applied to it. Due to the very small
geometric change the change of resistance is rather small. To increase the geometric effect
dedicated structures like a meander are used (Fig. 3.15, right side).

Fig. 3.15: Geometric changes of a bar of material in case of lengthening (left); resistance strain gauge with
a meander like structure.

As the resistance changes are still small the measurement is usually done using four strain
gauges configured in a Wheatstone bridge configuration (Fig. 3.16, left). A Wheatstone
bridge consists of two legs with two resistors in each leg. Both legs build a voltage divider.
Depending on theses four resistors a voltage difference Ua between the middle nodes of the
legs can be measured. In case that all resistors have the same resistance the voltage differ-
ence is zero. If one resistor changes its resistance the voltage will be non-zero and a measure
for the change of the resistance.
For the measurement with strain gauges four elements are used. These elements are mounted
in a way that the change in resistance of the elements amplifies the voltage difference Ua.
Two of the strain gauges are compressed (e.g. the top left and bottom right element) and the
24 3 Fundamental circuit elements

other two are lengthened to increase the voltage difference between the two half bridges of
the Wheatstone bridge.
A pressure sensor as shown in Fig. 3.16 uses four strain gauges to determine the differential
pressure between p1 and p2. The four strain gauges are implemented onto a silicon membrane
fabricated using microsystems technology. One strain gauge element each is located at region
“a” at the edge of the membrane and the other two are located at “b” in the middle of the
membrane. Without a differential pressure each strain gauge has a resistance R. Due to a
differential pressure the membrane is deformed and the mechanical stress is opposite for the
two strain gauges in the middle compared to the elements at the edge. Therefore the re-
sistance of two strain gauges is reduced by ΔR and for the other two it is increased by ΔR.
The voltage difference between the two half bridges is a measure for the resistance change
and hence for the pressure.

Fig. 3.16: A Wheatstone bridge (left) with four strain gauges; pressure sensor for differential pressure measure-
ment (right).
■

3.5.1 Real resistors

Of course resistors can be made in any value and any shape that is needed. But in reality
resistors are manufactured in standard values and a number of different shapes. Examples are
the E series of resistors with standardized resistance values. Types of resistors include com-
position type, wire-wound type and metal-film type. The most common construction tech-
nique for resistors is the composition type, which uses carbon or graphite and is molded into
a cylindrical shape. As the shape of the cylinder is the same the value of the resistance and its
tolerance is color-coded in bands as shown in Fig. 3.17 and Tab. 3.1.
Using these 4 bands the resistance can be calculated using
R  10  b1  b2   10 b3 

Sometimes a fifth band indicates the reliability of the device.

3.5 Resistance, resistors and Ohm’s law 25

Fig. 3.17: Color-coding of resistors.

Tab. 3.1: Color coding of resistors.

Color of the band Value of the band Multiplier Tolerance value [%]
b1, b2 b3
Black 0 100
Brown 1 101 1
Red 2 102 2
Orange 3 103
Yellow 4 104
Green 5 105 0.5
Blue 6 106 0.25
Violet 7 107 0.1
Grey 8 0.05
White 9 –
Gold 10-1 5
Silver 10-2 10
Black/no color

Standard resistance values for the E24 series in the range from 1 to 9.1Ω are listed in
Tab. 3.2. Other available values can be obtained by multiplying these values by factors of 10
ranging from 10 Ω to about 22·106Ω.

Tab. 3.2: Resistance values in Ω for the E24 series of resistors.

1.0 1.8 3.3 5.6

1.1 2.0 3.6 6.2
1.2 2.2 3.9 6.8
1.3 2.4 4.3 7.5
1.5 2.7 4.7 8.2
1.6 3.0 5.1 9.1

Besides the resistance value and the shape of a resistor also its power capability has to be
taken into account when selecting a resistor for an application. Electrical power that is dissi-
pated within a resistor is converted into heat. As excessive heating of the resistor may de-
stroy the device the heat has to be conducted away from the resistor by providing a good
thermal path.

3.5.2 Short circuit and open load

After an introduction to resistance and resistors we will analyze two extremes. Consider an
electric circuit as shown in Fig. 3.18 with a load resistance R. What happens for R = 0 Ω and
R = ∞ Ω?
26 3 Fundamental circuit elements

The first case is called a short circuit. Keeping Ohm’s law in mind we see that voltage drop is
zero for finite currents: a zero-Ohm resistor is equivalent to an ideal voltage source with zero
volts. In other words: if you connect an ideal voltage source to a zero-Ohm resistor, the cur-
rent will rise to infinity. As a conclusion, never place a short circuit, neither intentionally nor
unintentionally, across a voltage source to avoid excessive currents.
The latter with R = ∞ Ω is called an open circuit. Again looking at Ohm’s law it is obvious
that current will tend towards 0 A (as long as the voltage has finite value) in this case. This
behavior is equivalent to a circuit with an opening and no current is flowing. In other words,
as already stated at the beginning of chapter 3.1, it needs a closed loop for currents to flow.

Fig. 3.18: Simple electric circuit: short circuit (left) and open load (right).

Automotive application
Both short circuit and unwanted open load are severe fault conditions in automotive applica-
tions. Consider a lighting application like a headlight as depicted in a simplified circuit in
Fig. 3.19. A voltage source with internal resistance Ri drives the bulb and a current I of some
amperes flows through the circuit. For a given voltage (e.g. 12 V vehicle electrical system)
the current is determined by the internal resistance and the resistance of the bulb. If a short
circuit occurs that shorts the bulb the current will just flow through the short circuit path and
it will only be limited by the internal resistance. Hence the current will be much higher. This
excessive current will rapidly discharge the battery, or even severely damage the circuit and
the battery until total destruction of the system and maybe the vehicle occurs. Therefore a
short circuit has to be detected to prevent damage to the system. In the simplest case a fuse in
the circuit separates the battery from the rest of the circuit if the current gets too high. Alter-
natively the current is measured and a switch is triggered to open the circuit in case of exces-
sive currents without using the fuse.
An open load situation can happen if the filament of the bulb is broken, no current flows
through the bulb and the bulb does not shine anymore. This malfunction of the lighting has to
be detected at least in cases where the lighting is used for a safety critical application like
headlights. A defect headlight is critical for the recognizability and visibility of the vehicle
and hence is a traffic hazard.
3.5 Resistance, resistors and Ohm’s law 27

Fig. 3.19: A simple circuit with voltage source and internal resistance to drive a bulb (left); short circuit (center);
open load by broken filament of the bulb (right).
■

3.5.3 Real voltage sources

An ideal voltage source produces a voltage at the terminals regardless of what is connected
to it (independent of the load and the current). In reality the voltage at the terminals drops if a
load is connected to the source. This voltage drop can be modeled by an internal resistance.
In reality this internal resistance is unavoidable as every voltage source contains internal
resistive elements like wires. Fig. 3.20 shows a simple drawing of a real voltage source with
an internal resistance Ri.

Fig. 3.20: A simple schematic for a real voltage source with internal resistance Ri and voltage vs current charac-
teristic.

The voltage source shown in this simple schematic is divided into an ideal voltage source
(Uq) and an internal resistance Ri, connected in series. Ra is the external load resistance. The
voltage U10 at the terminals of the real voltage source is then:
28 3 Fundamental circuit elements

U10  U q  I  Ri

The two parameters Uq and Ri are in general independent of the load current I. As voltage
U10 is a linear function of the load current I. This type of voltage source is called a linear
source.
In the case of an open circuit (no external load Ra) the voltage at the terminals is called the
open-circuit voltage (OCV). If Ri is zero the real voltage source turns again into an ideal
voltage source with a constant voltage at the terminals (constant voltage source).
In the case of a short-circuit condition (U10 = 0 V) the current is limited by the internal re-
sistance Ri according to:
Uq
I SC 
Ri

In reality an internal resistance Ri as small as possible is often required to come close to an

ideal voltage source (for example for batteries). In this case the short-circuit current ISC might
become very high – be careful of creating short-circuit conditions in your applications.
Even though you deal with real voltage sources in reality we will use ideal voltage sources
during our analysis of electric circuits if not otherwise stated.

Automotive application
A combustion engine needs an electrical starter motor for initial starting. This electrical start-
er motor requires rather high currents of some hundred amperes from the lead acid battery to
generate the torque to start the engine. To provide these high currents during starting the
internal resistance of the battery should be very low. But the internal resistance depends on
many parameters: it increases over lifetime due to corrosion for example and it is higher at
low temperatures, e.g. in winter.
Consider a starter motor with a resistance of 30 mΩ that requires at least 180 A to start the
engine. A new 12 V lead acid battery has an internal resistance of 30 mΩ. During starting the
starter motor is in series with the battery’s internal resistance and a current of 200 A flows
through the starter motor. The terminal voltage of the battery drops down to 6 V (and all
electrical systems supplied by the battery have to keep on operating). For an old battery the
internal resistance at low temperatures might increase up to 60 mΩ. Now the maximum cur-
rent through the starter is just 133 A and the starter motor cannot generate sufficient torque to
start the engine. ■

3.5.4 Real current sources

An ideal current source produces a current regardless of what is connected to it (independent
of the load and the voltage), it is a constant current source. Like for the real voltage source
we model the real current source with an ideal element and an internal resistance, but this
time the ideal element and Ri are connected in parallel as shown in Fig. 3.21.
3.5 Resistance, resistors and Ohm’s law 29

Fig. 3.21: A real current source without (left) and with load (right).

For an open circuit as shown on the left side of Fig. 3.21 the voltage U10 at the terminals is:
Iq
U 10  I q  Ri 
Gi

Connecting a load to the real current source this constant current is divided into two parts
flowing through Ri and Ra:
Iq  Ia  Ii

U 10
Ia  Iq   I q  U 10  Gi
Ri
The voltage at the terminals now depends on the external load resistance and is:
U10 Ra   I a Ra   Ra
According to the current divider rule (based on Kirchhoff’s current law, see below), this
voltage can be calculated as:
Ri  Ra
U 10 Ra   I q 
Ri  Ra
Again the voltage is a linear function of the load current as shown in Fig. 3.22.

Fig. 3.22: Voltage as a function of load current for different external loads.
30 3 Fundamental circuit elements

Linear voltage (see Fig. 3.20) and current (see Fig. 3.21) sources are equivalent and can be
transformed into each other.
At the end of the introduction of real voltage and current sources the following images show
some examples of sources.

Fig. 3.23: Examples of sources:An ideal voltage source (top left); an ideal current source (top right);
a battery (center left); a bipolar transistor (center right); ideal, linear and non-linear current source
of a solar cell (bottom).

3.5.5 Transformation of sources

Consider real voltage and current sources as shown in Fig. 3.20 and Fig. 3.21. If we want to
replace the former (given Uq and Ri) by the latter we have to determine the parameters Iq and
Ri (or Gi) of the real current source.
In case of a short circuit (Ra = 0 Ω) we find for the real voltage source:
Uq
U
Ri
3.5 Resistance, resistors and Ohm’s law 31

And for the corresponding current source:

I  Iq

As these currents have to be the same we conclude for the current source:
Uq
Iq 
Ri

On the other hand for open load (infinite Ra) of the real voltage source we get:
U Uq

And for the corresponding current source:

Iq
U 
Gi

Finally:
1
Gi 
Ri
The internal resistances of the corresponding sources are the same and the relation between
Iq and Uq is as given above.
4 Fundamental electrical circuit laws

4.1 Kirchhoff’s laws

As we have already seen in the previous chapter about real sources, electric circuits in gen-
eral are built up of several different parts. To analyze more complex circuits two basic laws
are fundamental: Kirchhoff’s current law (KCL) and Kirchhoff’s voltage law (KVL). These
laws describe the correlations of currents and voltages in an electric circuit.

4.1.1 Kirchhoff’s current law

Remember that a node is a connection of two, or more elements of a circuit. The first of
Kirchhoff’s laws, the current law describes the currents at any node of the circuit and is
based on the law of conservation of electric charge:
– At any node of a circuit, the currents algebraically sum to zero at any instant of
time.
Here currents flowing into the node are considered to be positive and currents directed out of
the node are negative. In other words, the sum of the currents into the node is equal to the
sum of the currents out of the node. Refer to Fig. 4.1 to see examples for nodes with several
currents flowing into and out of the node.

Fig. 4.1: An electric circuit with one voltage source, two resistors and corresponding current vectors (left); one
node as part of a circuit with six elements connected to the node and corresponding currents (right).

Kirchhoff’s current law can now be written as (see Fig. 4.1):

n

I
k 1
k 0

Example: The circuit on the left side of Fig. 4.1. Here current I1 flows into the node and
therefore is positive whereas I2 and I3 are directed out of the node and are counted negative:
34 4 Fundamental electrical circuit laws

I1  I 2  I 3  0
In a more general way Kirchhoff’s current law is not only applicable for nodes but also for
any closed region of a circuit (see Fig. 4.2). Here KCL of course applies for node 1, but also
for the closed region marked by the dotted line. The algebraic sum of all currents flowing
into and out of the closed region has to be zero. In this case:
n

I
k 1
k  I 10  I 11  I 4  I 5  I 6  I 7  I 8  I 9  0

Fig. 4.2: A more complex part of a circuit with a closed region (dotted line) for which KCL also applies.

Application of KCL: Resistors connected in parallel

Consider two resistors connected in parallel to a voltage source as depicted in Fig. 4.3 on the
left side. Two (or more) parts are connected in parallel if they are connected to the same pair
of nodes. We would like to find the equivalent resistor R (see right side of Fig. 4.3) to replace
the two parallel resistors. How is R related to R1 and R2?

Fig. 4.3: Parallel connection of two resistors (left) and equivalent circuit with one equivalent resistor (right).
4.1 Kirchhoff’s laws 35

According to KCL the current I splits into I1 and I2:

I  I1  I 2
As the voltage across each resistor R1 and R2 is U Ohm’s law for each resistor is:
U
I1 
R1

U
I2 
R2
And for the equivalent circuit:
U
I
R
Using the three equations from Ohm’s law in KCL results in:
U U U
 
R R1 R2
Thus, we see that the parallel connection of two resistors is equivalent to a single resistor
provided that:
1 1 1
 
R R1 R2

R1  R2
R
R1  R2
This rule for parallel resistors can be generalized to any number n of parallel resistors:
n

R
1 1

R k 1 k

Or, using the conductance G and Gk:

n
G G
k 1
k

The resulting conductance of n parallel resistors is the sum of all single conductances.
Coming back to the easy example of two resistors connected in parallel: how is the current I
divided by the resistors? From the equations above we conclude that the voltage U is:
R1  R2
U  R I  I
R1  R2
36 4 Fundamental electrical circuit laws

Replacing the voltage U in Ohm’s laws for the two resistors gives the current for both resis-
tors:
U R2
I1   I
R1 R1  R2

U R1
I2   
R2 R1  R2
These two formulas describe how the current I is divided into two parts through the resistors.
This circuit is often referred to as a current divider. The ratio of the two currents is:
I 1 R2

I 2 R1
From these equations it is obvious that the currents are reciprocally proportional to the re-
sistances. In other words: the smaller the resistance (compared to the other resistance), the
higher the current through this resistance. The current tends to take the path of least re-
sistance.

4.1.2 Kirchhoff’s voltage law

Remember that a loop is a closed path through a circuit in which no electric element, or node
is encountered more than once. Kirchhoff’s second law, the voltage law, describes the volt-
ages within loops and is based on the physical law of the conservation of energy:
– Around any loop in a circuit, the voltages algebraically sum to zero.
In other words: in traversing any loop in any circuit, at every instance of time, the sum of
voltages having one polarity equals the sum of the voltages having the opposite polarity.
KVL is valid for all loops of a circuit, even for open loops (loops with an open circuit) and
loops that do not follow a physical branch in the circuit. But you never encounter any other
node twice except the starting point.
Two loops, I and II, are marked in the circuit shown in Fig. 4.4. In fact these loops are even
meshes. First of all the direction of traversing has to be defined. For both loops in Fig. 4.4
this direction is defined arbitrarily as counterclockwise. All voltages pointing in the same
direction are counted positive. Voltages pointing in opposite direction are counted negative.
Hence we find for the two loops of Fig. 4.4:
– Loop I:
U U 2 U1  0
– Loop II:
U 2 U 4 U 3  0
4.1 Kirchhoff’s laws 37

As KVL is valid for all loops, not only meshes like I and II, we can also write for the outer
loop:
– Outer Loop:
U  U 4  U 3  U1  0
In general we can write for any loop of a circuit:
n

U
k 1
k 0

Here n is the number of voltages within the loop.

Fig. 4.4: A simple circuit with a voltage source and 4 resistors, two loops (meshes) are marked with I and II.

Application of KVL: Resistors connected in series

Consider two resistors connected in series to a voltage source as depicted in Fig. 4.5 on the
left side. Two elements are connected in series if they have a node in common and no other
element is connected to this common node. As a consequence of this definition, the same
current I flows through elements connected in series. We would like to find the equivalent
resistor R to replace the two resistors in series. How is R related to R1 and R2?

Fig. 4.5: Series connection of two resistors.

Applying KVL yields:

n

U
k 1
k  U total  U 1  U 2  0

With Ohm’s law for R1, R2 and R this results in:

R  I  R1  I  R2  I
38 4 Fundamental electrical circuit laws

 R  R1  R2
The resistance R of a single resistor equivalent to a series connection of resistors is just the
sum of the resistors connected in series. In general for n resistors in series:
n
R R
k 1
k

Looking again at Fig. 4.5 shows that the total voltage Utotal is divided by the two resistors
into two parts, U1 and U2. Two voltage divider rules describe how the voltage is divided
between the resistors and two resistors connected in series are therefore often called a volt-
age divider:
R1
U1  U
R1  R2

R2
U2  U
R1  R2
The larger voltage drop will be across the larger resistor.

4.2 Operating point

Consider an easy example for an electric circuit as shown in Fig. 4.6. It consists of two
meshes, the source and the load mesh. Both meshes are closed via the terminal with voltage
U10.

Fig. 4.6: A simple electric circuit with a source and a load mesh.

For mesh I we can write the following equation according to KVL:

U q  Ri  I L  U10  0

For mesh II we can write:

U 10  RL  I L  0
Thus we have two unknown variables (IL, U10) and two equations and this linear equation
system is algebraically solvable:
4.2 Operating point 39

Uq
IL 
Ri  R L

RL
U 10  U q 
Ri  RL
These two values define the operating, or working point of this circuit. For given parameters
(like RL, Ri and Uq) the operating point (or working point, WP) defines the steady state of the
system. In practical problems often small parameter changes around the working point are
considered: small voltage, temperature or resistance variations for example.
Instead of solving these equations algebraically they can also be solved graphically. For this
purpose the equations for the two meshes as given above are transformed to show the de-
pendence of load current IL as a function of terminal voltage U10:
Uq U 10
IL  
Ri Ri

U10
IL 
RL

Fig. 4.7: Characteristic curves of the voltage source and the load. WP indicates the working point of the circuit.

Fig. 4.7 shows the characteristic curves for both the voltage source and the load. The charac-
teristic curve of the voltage source is a falling line that intersects the IL-axis at the short cir-
cuit current ISC (U10 = 0 V) and the U10-axis at the value of the source voltage. The straight
load line rises according to Ohm’s law. Since any point on the source curve satisfies the
source equation and any point on the load curve satisfies the load equation, the intersection
of both plots satisfies both equations simultaneously and the point of intersection is the oper-
ating, or working point (WP).
For linear equations as shown above the graphical solution seems to be inappropriate. But
consider a circuit with a non-linear load where the load current is a non-linear function of the
applied voltage. Semiconductor components like diodes or transistors are examples for such
non-linear components. For these systems with non-linear components the techniques used
for linear, algebraic simultaneous equations cannot be employed and the equation system has
40 4 Fundamental electrical circuit laws

no analytical solution in general. Fig. 4.8 shows an example of a circuit with a non-linear
component, a diode.

Fig. 4.8: A simple circuit with a diode as load (left); Characteristic curves for the source and the diode;
diode current is a non-linear function of the voltage (right).

The current of the diode is a non-linear function of the voltage as shown on the right side of
Fig. 4.8, e.g.

 U10 
I L  I S   e C  1
 
 
Here IS (inverse current) and C are constants. As it is not possible to solve the source and
load equation simultaneously in this case, the WP is determined by the graphical solution as
shown on the right side of Fig. 4.8.
In general for electronic systems the operating conditions have to be set properly to operate
the components and devices in the required functionality. The method for setting proper
operating points, voltages or currents, is also called biasing.

4.3 Wye-Delta transformation

We have seen that certain circuit configurations, serial and parallel connections of resistors,
can be simplified by a single resistor to make circuit analysis easier. Refer to Fig. 4.9 as an
example of simplification:

Fig. 4.9: A circuit with series and parallel connections for the demonstration of simplification.

First of all replace resistors R1 and R2, connected in series, by equivalent resistor
R12 = R1 + R2. This equivalent resistor is parallel to R3 and these two resistors can be re-
placed by R123:
4.3 Wye-Delta transformation 41

R1  R2  R3
R123 
R1  R2  R3
R123 now is connected in series with R4 and the final resistor R, replacing resistors R1-R4 is:
R1  R2  R3
R  R4
R1  R2  R3
But not all configurations can be simplified by these simple laws. In some of such cases a
special transformation, a Wye-Delta transformation can be used to replace three resistors in
Wye configuration by three resistors in Delta configuration or vice versa, so that the circuits
are equivalent as far as the terminals are concerned. Refer to Fig. 4.10 for the two configura-
tions. Both configurations are equivalent if the resistances measured between two of the
terminals 1, 2, 3 are the same.

Fig. 4.10: Delta configuration (left) and Wye configuration (right).

Delta configuration to Wye configuration

Starting with the Delta configuration with given resistors R12, R23 and R31 we are looking for
the equivalent Wye configuration with resistors R1, R2, R3. For the resistance between termi-
nal 1 and 2 to be the same for both configurations it follows that:
R12  R23  R31 
R1  R2 
R12  R23  R31
For the resistance between terminal 1 and 3 and terminal 2 and 3 similar equations can be
obtained:
R23  R31  R12 
R2  R3 
R12  R23  R31

R31  R12  R23 

R3  R1 
R12  R23  R31
Using these three equations we can calculate the three resistors of the Wye-configuration as
follows:
42 4 Fundamental electrical circuit laws

R12  R31
R1 
R12  R23  R31

R23  R12
R2 
R12  R23  R31

R31  R23
R3 
R12  R23  R31
Each resistor of the Wye configuration is given by the product of the two adjacent resistors of
the corresponding Delta configuration divided by the sum of the three resistors of the Delta
configuration.
For a symmetric delta configuration with
R12  R23  R31  R
each resistance of the corresponding Wye configuration is just:
R 2 R
R wye   
3  R 3

Wye configuration to Delta configuration

Going the opposite direction from given resistors in Wye configuration the resistors of the
corresponding Delta configuration can be calculated. The final result for the Delta configura-
tion’s resistors is:
R1  R2
R12  R1  R2 
R3

R2  R3
R23  R2  R3 
R1

R3  R1
R31  R3  R1 
R2

4.4 Meters and measurements

The measurement of parameters like current and voltage in reality is a broad field and cannot
be handled here in detail. However a few basic principles will be introduced to get an idea of
how to measure electrical parameters. Real measurements are made by real instruments. In
general the measurement disturbs the operation of an electric circuit to some extent and
therefore care has to be taken to avoid useless and wrong measurements.
4.4 Meters and measurements 43

Voltmeter
Voltage is measured between two terminals, or nodes of a circuit. For this voltage measure-
ment a voltmeter is connected across these two points as shown in Fig. 4.11. Without looking
at the details of how the measurement is done the voltmeter can be modeled as a parallel
combination of an ideal voltmeter (without current flow) and an internal resistor RV. The
shunt resistor is therefore also parallel to the voltage (resistance) to be measured. As the
current is divided by these parallel combinations of resistances according to KCL the value
of the shunt resistance hast to be very high to avoid disturbance of the measurement as much
as possible. In practice it is of the order of several million Ohms.
If the voltage is measured across a well known resistor, the current flowing through this
resistor can be calculated using Ohm’s law.

Fig. 4.11: The connection of a voltmeter to measure the voltage across the shunt resistor RV.

Ampmeter
In contrast to the voltmeter the ampmeter is connected in series to measure the current
through a line, or wire of a circuit. Therefore the circuit has to be broken to measure its cur-
rent (whereas the circuit doesn’t need to be broken for a voltage measurement, see above).
The ampmeter can be modeled as a series combination of an ideal ampmeter and an internal
resistance RI. According to KVL the internal resistance has to be as small as possible to keep
the disturbance of the circuit as small as possible.
An indirect way of measuring the current without breaking the circuit is to use a current
probe or measuring caliper. These measuring instruments enclose the wire and make use of
the magnetic properties of the current flowing through the wires.

Fig. 4.12: A current probe or measuring caliper.

44 4 Fundamental electrical circuit laws

Oscilloscope
An oscilloscope is used to measure time-varying signals, both voltages and currents. An
oscilloscope samples the time-varying signal at fixed instances of time (e.g. every 10ms) and
displays a graph of the measured parameter as a function of time. This operating mode makes
it possible to observe the general behavior of the voltage as a function of time.

Automotive applications:
Voltage and current are frequently measured by automotive systems and this measurement
has to be done by the system itself. The voltage is usually measured using an analog to digi-
tal converter (ADC). This ADC can be a separate device or it is part of a microcontroller. In
general the ADC has a maximum voltage range (e.g. 0–5 V) that can be measured. If the
voltage to be measured is higher than the measurement range of the ADC can be divided
using a voltage divider to fit to the ADC input requirements.
One indirect way of measuring currents in automotive systems is to use a shunt resistor. This
shunt resistor is designed into the branch of the current flow. Due to the current there will be
a voltage drop across the shunt resistor and this voltage drop can be measured as described
above using an ADC. With the knowledge of the exact value of the shunt resistor the current
can be calculated from the voltage. This simple method has the disadvantage that power is
dissipated in the shunt resistor. Another current measurement method utilizes the Hall effect.
Hall effect sensors measure the magnetic field of the current carrying wire. The output of the
Hall sensor is the Hall voltage that can be measured by an ADC. ■

4.5 Power and energy

When a current i(t) flows through a resistor (voltage drop u(t)) energy is dissipated inside the
resistor and electrical energy is converted into heat as (positive charged) current goes from a
higher potential to lower potential. Indication arrows for current and voltage are parallel. The
electrical energy E during time t1–t2 is given by:
t2

e(t )  u(t )  q(t )   u(t )  i(t )dt

t1

In the case of DC currents and voltages the energy is just (starting at t1 = 0 s):
E  U  I t
The SI unit is the Joule (J) and 1 J = 1 Vas = 1 Ws. For convenience it is usual to calculate
with the unit kWh where 1 kWh = 3.6·106 Ws = 3.6 MJ.
1kWh equals to
– Working for 50 hours on a notebook (20 W power consumption)
– Heating about 10 liters of water from room temperature to 100 °C
– Driving an electric vehicle (EV) about 6–7km (for an EV with 15 kWh / 100 km)
Based on the electrical energy given above the instantaneous electrical power in Watts (W) is
given by:
4.6 Maximum power transfer 45

de(t )
p (t )   u ( t )  i (t )
dt
For DC currents power is time-independent:
U2
P U I  I2 R 
R
So a current of 1 A and a voltage drop of 1 V results in 1 W power dissipated in the resistor.

Fig. 4.13: Power at a resistor with current i(t) and voltage drop u(t).

Efficiency
When talking about the transformation of energy (in a source or load) efficiency is a key
parameter. It is defined as the ratio of transformed power P2 to spent power P1:
P2 E2
 
P1 E1
Efficiency is always < 1 (or < 100 %) as not all the power can be transformed. The difference
P1–P2 is the power loss. Target is to reduce power loss as much as possible and to get close to
1 for the efficiency. For example a generator (which does not actually generate) transforms
mechanical power (P1) into electrical power (P2) and can reach an efficiency of up to 99.5 %
(whereas an automotive combustion engine has something like < 45 % which is not very
good…).

4.6 Maximum power transfer

As already stated above power is provided by a source and consumed by a load such as a
resistor. Of course power is also consumed by the internal resistor of a real source. But what
is the maximum power transfer between a real source (ideal source plus internal resistor) and
a load resistor? What is the maximum power a source can provide? In order to investigate the
power transfer in more detail, consider Fig. 4.14 in which a real voltage source is connected
with a variable load resistor (Ra). Two extremes for the variable resistor were already consid-
ered above, short circuit (Ra = 0 Ω) and open load (Ra = ∞ Ω).
46 4 Fundamental electrical circuit laws

Fig. 4.14: A simple circuit for the investigation of power transfer, Ri is the internal series resistor of the voltage
source, Ra the load resistor.

The current is given by:

Uq
I
Ri  R a

We obtain for the power in the resistor:

U q2
P  I 2  Ra   Ra
Ri  Ra 2
For fixed values of Uq and Ri the value of Ra that maximizes the power absorbed by the load
can be found by setting the first derivative equal to zero:

dP U q  Ri  Ra   2U q  Ra  Ra  Ri 
2 2 2

 0
dRa Ra  Ri 4
 R i  R a   2 R a  R a  R i   0
2

R a  Ri
So maximum power is transferred to the load if the load resistance matches the source re-
sistance or, in other words, both resistances have to be equal to each other. The maximum
power in this case is:

U q2
PL , max 
4 Ri

Fig. 4.15 shows the power transfer to the load (PL) compared to the total power provided by
the source (Pges) and the efficiency of the source η. In power electronics the efficiency is
often maximized, in signal processing it is often the power.
4.7 Dependent and independent sources 47

Fig. 4.15: Power transfer to the load (PL) and total power provided by the source (Pges) in units of the maximum
power transfer (PLmax) (left); efficiency (right).

4.7 Dependent and independent sources

So far we have dealt with sources whose values (current or voltage) are in general time de-
pendent. In more detail these sources were independent sources, i.e. the behavior of the
sources (the current, or voltage they applied to the circuit) was independent of the behavior
of the circuit to which the source belonged. No matter what happens within the circuit, the
independent source supplies a fixed (but time-dependent) value.
On the other hand, for a dependent, or controlled source (current or voltage) the value de-
pends on some variable (usually voltage or current) in the circuit to which the source be-
longs. In electronic circuits dependent sources are represented by the symbols shown in Fig.
4.16.

Fig. 4.16: Symbols of a dependent voltage source (left) and a dependent current source (right).

In general both voltage and current sources can be controlled by a current or a voltage. Fig.
4.17 shows the four types of dependent sources.
48 4 Fundamental electrical circuit laws

Fig. 4.17: Examples for dependent sources.

Example 1: Current controlled current source

The example as depicted in Fig. 4.18 contains a current source that depends on the value of
the current I1 through another branch of the circuit (like for example a bipolar transistor).
Given parameters are the current of the independent source on the left, the two resistors and
the dependence of the current controlled current source. What will the resulting currents I1
and I3 be?

Fig. 4.18: An example of a circuit with a current controlled current source.

The node rule (KCL) provides:

1A  3  I1  I1  I 3  0
4.7 Dependent and independent sources 49

KVL in combination of Ohm’s law for the two resistors provides:

I1  R1  I 3  R2
Substituting I3 into the node rule:
R1
1A  3  I 1  I 1  I 1 
R2

R2
 I 1  1A 
R1  2R2
Using R1 = 3 Ω and R2 = 1 Ω yields I1 = 1 A and I3 = 3 A. ■

Example 2: Bipolar transistor as current controlled current source

A bipolar transistor can be regarded as a current dependent current source as the collector
current is controlled by the base current: the collector current IC (through the transistor) is
given by the base current IB multiplied by DC current gain factor B. Consider a circuit as
shown in Fig. 4.19 on the left side where the bipolar transistor with a DC gain of 100 is used
to make the 21 W-bulb to be operated at a voltage of 12 V. Voltage source is Uq = 14 V. On
the right side of Fig. 4.19 the equivalent circuit with a current controlled current source is
shown. What about the resistors R1 and R2?

Fig. 4.19: A circuit with a biploar transistor acting as a current controlled current source (left); an equivalent
circuit showing the current controlled current source (right).

From the given parameters of the bulb we can calculate the resistance of the bulb and the
current that has to flow through the bulb and that equals the collector current IC:
U L2
PL  U L  I C 
RL

 RL  6.86

 I C  1.75A
50 4 Fundamental electrical circuit laws

As the transistor works as a current controlled current source with a DC gain of B = 100 the
base current has to be:
IC
IB   17.5mA
B
The voltage drop UBE for the bipolar transistor is about 0.7 V and thus the base resistance is:
U BE
RB   40
IB
To avoid excessive currents through the two resistances R1 and R2 we choose R2 to be in the
range of RB, e.g. 100 Ω. Using KCL for node B (I1= I2+ I3) gives:
U q  U BE U BE
  IB
R1 R2

Hence resistor R1 can be calculated as R1 = 543 Ω and the current I1 is 24 mA. ■

5 Circuit analysis

In the previous chapter I presented basic electric circuit concepts like KCL, KVL or Wye-
Delta transformation. Now I want to introduce some more sophisticated circuit analysis tech-
niques for practically and efficiently solving problems associated with circuit operations. We
will start with two basic analysis techniques, nodal and mesh analysis. These two techniques
are based on KCL and KVL and make use of two fundamental facts about electric circuits:
1. In any electric network with n nodes (n–1) independent equations for the nodes can be
found
2. In any electric network with m meshes m independent equations can be found

5.1 Nodal analysis

Nodal analysis can be used for any electric circuit and in particular for circuits with few
nodes (but rather many loops) as the number of equations will be small. It is based on the
definition of voltage: Voltage is the difference between two electrical potentials.

Fig. 5.1: Voltages (including a reference potential U0) in an electric circuit.

It is common to define a reference potential and to refer the other voltages to this reference
potential. The reference potential, or ground potential is marked with a special sign and is
defined to have a voltage of U0 = 0 V. For a circuit with n nodes there are (n–1) nodal voltag-
es referring to ground potential. With nodal analysis the voltages of all nodes referring to the
reference potential can be calculated. The procedure of nodal analysis will be introduced by
an example before the general approach is presented.

An example of a nodal analysis

Refer to Fig. 5.2 for the first example of nodal analysis. The circuit contains 4 nodes, 0–3,
and node 0 is defined as ground potential. The direction of the current vectors can be chosen
arbitrarily (at the end of the calculation the sign of the current will show whether the current
flows in the chosen, or in opposite direction). By using KCL and Ohm’s law the three nodal
52 5 Circuit analysis

voltages U10, U20 and U30 can be calculated. For simplification of the calculation each re-
sistance Ri is substituted by corresponding conductance Gi, i.e. Gi = 1 / Ri.

Fig. 5.2: An electric circuit with corresponding currents and voltages as an example of nodal analysis (left);
node 0 is defined as ground potential; equivalent circuit with the voltage source and resistance R1
transformed into a current source (right).

For node 1-3 in Fig. 5.2 KCL results in:

– node 1:
I1  I 4  I 6  0
– node 2:
I4  I2  I5  0
– node 3:
I5  I6  I3  0
Using Ohm’s law (I = G · U) for the currents give:
I 1  G1  U  U 10   G1 U  G1 U 10

I 2  G2 U 20

I 3  G3 U 30

I 4  G4  U 10  U 20 

I 5  G5  U 20U 30 

I 6  G6  U 10  U 30 
Hence for the KCL of node 1–3:
G1  U U10   G4  U10 U 20   G6  U10 U 30   0
5.1 Nodal analysis 53

G4  U10  U 20   G2 U 20  G5  U 20  U 30   0

G5  U 20 U 30   G6  U10 U 30   G3 U 30  0
So we have 3 linear independent equations for the three unknown variables U10, U20 and U30.
After these voltages have been calculated the currents I1 – I6 can be determined using the
equations for Ohm’s law above.
Sorted by the unknown voltages these equations look like:
G1  G4  G6 U10  G4 U 20  G6 U 30  G1 U
 G4 U10  G2  G4  G5  U 20  G5 U 30  0

 G6 U10  G5 U 20  G3  G5  G6  U 30  0
These equations can also be written in matrix multiplication form:

 G1  G4  G6  G4  G6   U 10   G1 U   I q 
     
  G4 G2  G4  G5  G5   U 20    0    0 
 

  G6  G5 G3  G5  G6  U 30   0   0 

For the last step we make use of the transformation of a voltage source into an equivalent
current source: The voltage source U with resistance R1 in series can be transformed into an
equivalent current source with a parallel conductance G1 with:
1
G1 
R1

U
Iq   G1 U ■
R1

Algorithm of nodal analysis

As we have seen in the example, by applying KCL and Ohm’s law we are able to set up (n–1)
equations for (n–1) unknown nodal voltages of a circuit with n nodes. The following algo-
rithm is one of several slightly different algorithms and it can be used for circuits with n
nodes:
1. The circuit has n nodes. Select one node as reference potential (voltage 0 V).
2. Label the remaining (n–1) nodes with their voltages referring to the reference node (e.g.
U1, U2).
3. Transform all voltage sources (together with the resistors in series) into equivalent cur-
rent source with parallel conductance.
4. Draw the current arrows in the circuit and label the currents. The direction of the arrows
can be chosen arbitrarily.
5. Derive the (n–1) equations by applying KCL to each node.
54 5 Circuit analysis

6. Substitute the unknown currents by the voltage drop across the resistance in the corre-
sponding branch.
7. Solve the n–1 equations for the node voltages.
8. Calculate the branch currents using Ohm’s law.

The application of algorithm for nodal analysis

The following example is used to apply the presented algorithm to an electric circuit as
shown in Fig. 5.3:

Fig. 5.3: Electric circuit with two ideal current sources and three nodes.

1. The circuit has n nodes. Select one node as reference potential (voltage 0 V).
– See node with 0 V in Fig. 5.3
2. Label the remaining (n–1) nodes with their voltages referring to the reference node (e.g.
U1, U2).
– See voltages U1, U2 in Fig. 5.3
3. Transform all voltage sources (together with the resistors in series) into equivalent cur-
rent source with parallel conductance.
– No voltage source
4. Draw the current arrows in the circuit and label the currents. The direction of the arrows
can be chosen arbitrarily.
– See currents I1–I3 in Fig. 5.3
5. Derive the (n-1) equations by applying KCL to each node.
– Two nodes and therefore two equations:
– Node 1:
I q1  I 1  I 2

– Node 2:
I q2  I 2  I 3

6. Substitute the unknown currents by the voltage drop across the resistance in the corre-
sponding branch.
– Node 1:
U1 U1 U 2  R  R2  U2
I q1    U 1   1  
R1 R2  R1  R 2  R2
5.1 Nodal analysis 55

– Node 2:
U1 U 2 U 2 U1  R  R2 
I q2     U 2   3 

R2 R3 R 2  R 2  R3 

7. Solve the (n–1) equations for the node voltages.

– We have two equations for the two unknown node voltages: U1, U2 and we can cal-
culate these node voltages:
– From node 2:
U  R R
U 2   1  I q 2   2 3
 R2  R 2  R3
– Substituting U2 into the equation for node 1 gives:
 R  R2   U1  R3
I q1  U 1   1     I q 2  
 R1  R 2   R2  R 2  R3
– And finally the node voltage U1:
1
 I q 2  R3   R1  R 2 R3 
U 1   I q1   
R 2  R3   R R  R  R  R 
   1 2 2 2 3 
– Node voltage U2 can be calculated afterwards.
8. Calculate the branch currents using Ohm’s law.
U1
I1 
R1

U1  U 2 
I2 
R2

U2
I3 
R3
For given values of Iq1 = 6 A, Iq2 = 12 A, R1 = 1 Ω, R2 = 3 Ω, R3 = 2 Ω the voltages and cur-
rents yield U1 = 1 V, U2 = –14 V and I1 = 1 A, I2 = 5 A, I3 = –7 A. ■

An example of nodal analysis having a voltage source between two non-reference nodes
So far we have dealt with circuits that contain sources between any node and the reference
node. Voltage source with series were transformed into current sources with parallel con-
ductance. In general voltage sources can also be present between non-reference nodes and
without series resistance. To overcome the difficulty of transforming the ideal voltage source
an extension of nodal analysis can be used. This modified nodal analysis (MNA) is for ex-
ample also used in circuit simulation programs like PSPICE.
56 5 Circuit analysis

Consider the circuit given in Fig. 5.4. We want to determine the node voltages for the 4
nodes. It contains two ideal voltage sources between two nodes (Uq1, Uq2) and an ideal cur-
rent source. We select one of the nodes of Uq1 as reference point. The other three nodes and
the currents are labeled as shown in Fig. 5.4.

Fig. 5.4: Circuit for nodal analysis with an ideal voltage source between two reference nodes.

Now we derive the (n–1) = 3 equations by applying KCL and Ohm’s law:
Node 1:
U1  Uq1

Node 2:
I1  I 2  I 4

U q1  U 2 U2
  I4
R1 R2

We cannot immediately derive a requirement from the voltage source Uq2 that determines the
current I4.
Node 3:
U3
I4  I3  Iq   Iq
R3

From node 1 the voltage U1 is immediately given. From the other two nodes we have two
equations but still three unknown variables as the current I4 cannot be determined directly
from voltage source Uq2. However we have the branch voltage for the branch containing Uq2
that provides another equation:
U 2 U3  U q2

From node 2 and 3 we obtain:

U q1  U 2 U2 U
  I4  3  Iq
R1 R2 R3

U q1  1 1  U3
  U 2      Iq
R1  R1 R2  R3
5.1 Nodal analysis 57

Substituting the branch equation gives:

U q1
 Iq 
U3

 U q2  U 3   RR  RR
1 2


R1 R3  1 2 
 1 R1  R2   R  R2 
 U 3      U q 2   1 
 R3 R1  R2   R1  R2 
1
 U q1 R  R2   R  R  R1  R3  R2  R3 
U 3    U q2  1  I1    1 2 
 R1 R1  R2   R1  R2  R3 
Starting from this equation and with given values for Uq1, Uq2, Iq and the resistors we can
calculate the values for the node voltages and the branch currents. ■

Determinants and Cramer’s rule

The algorithm presented above derives the linear equation system from KCL and Ohm’s law.
Using the matrix multiplication form also presented above there is another way of determin-
ing the nodal voltages using determinants and Cramer’s rule.
Determinants are special functions that associate a scalar value to a square matrix A. The
determinant can be used to check whether linear equation systems have a unique solution and
this solution can be calculated using Cramer’s rule.
Determinants for small square matrices (1 × 1, 2 × 2 and 3 × 3) are easy to calculate:
1 × 1 matrix:
det A  deta11   a11

2 × 2 matrix:
a a12 
det A  det 11   a11  a22  a12  a21
 a21 a22 
3 × 3 matrix:
 a11 a12 a13 
 
det A  det  a 21 a 22 a 23   a11  a 22  a33  a12  a 23  a31  a13  a 21  a32
a a33 
 31 a32
 a3  a 22  a31  a12  a 21  a33  a11  a 23  a32

Consider the equation system of the first example for nodal analysis:
G1  G4  G6  U10  G4 U20  G6 U30  G1 U
 G4 U10  G2  G4  G5  U 20  G5 U30  0

 G6 U10  G5 U 20  G3  G5  G6  U30  0

58 5 Circuit analysis

This equation system can also be written in matrix multiplication form:

 G1  G4  G6  G4  G6   U 10   G1  U   I q 
       
  G4 G 2  G 4  G5  G5   U 20    0    0 
  G6  G5 G3  G5  G6   U 30   0   0 

 
G U  I

Here the conductance values are known coefficients listed in conductance matrix G. The
(source) currents on the right side are also known. The unknown parameters we are looking
for are the nodal voltages. Cramer’s rule states that these unknown voltages Ui (here U10,
U20, U30) can be calculated by the determinates as follows (in fact Cramer’s rule does not
care about what is calculated, but is valid for linear equations with as many equations as
unknowns in general):

Ui 
 
det Gi
det G 
G is the conductance matrix and Gi is constructed by replacing column i in the conductance
matrix by the current vector of the right side, e.g. for U10:
 Iq  G4  G6 
 
G1   0 G2  G4  G5  G5 
0  G5 G3  G5  G6 

Remembering how the determinate of a matrix looks like gives for U10:

 
det G1  I q  G2  G4  G5   G3  G5  G6   I q  G52

 
det G  G1  G4  G6   G2  G4  G5   G3  G5  G6   2  G4  G5  G6 
G2  G4  G5   G62  G1  G4  G6   G52  G3  G5  G6  G42

Example of Cramer’s rule

The circuit such as that shown in Fig. 5.5 can be analyzed using Cramer’s rule. The circuit
has four nodes. As node 1 is connected to an ideal voltage source to the reference node, the
node voltage 1 is equal to Uq1.

Fig. 5.5: Circuit with three nodes and the reference node for the application of Cramer’s rule.
5.1 Nodal analysis 59

For nodes 2 and 3 we can apply KCL:

I1  I 2  I 3

I 2  I 4  I5
Converting the resistances into corresponding conductance values, using Ohm’s law and
sorting by U2 and U3 gives:
G1  G2  G3 U2  G2 U3  Uq1  G1
 G2 U 2  G2  G4  G5  U3  0
Or in matrix form:

 G1  G2  G3  G2  U 2  U q1  G1 
     
  G2 G2  G4  G5   U 3   0 

Determinants used for application of Cramer’s rule are:

 G  G2  G3  G2
 
det G  det 1
 G2

 
  G1  G2  G3   G2  G4  G5  G22
G2  G4  G5 


  U  G
det G2  det q1 1
 0
 G2 

  U q1  G1  G2  G4  G5 
G2  G4  G5 


   G  G2  G3 U q1  G1 
det G3  det 1
  G2 0 
  U q1  G1  G2  
Node voltages can be calculated by these determinants as:

U2 
U q1 
 G1  G2  G4  G5 
G1  G2  G3  G2  G4  G5   G22

U3 
U q1 
 G1  G2
G1  G2  G3  G2  G4  G5   G22 ■

Automotive application
Since the introduction of electric lighting in vehicles the number of electric and electronic
components has been steadily increasing. The nominal voltage of a typical automotive elec-
tric system is 12 V. An alternator connected to the internal combustion engine is used to
generate the current needed by the large number of electronic systems. A battery, either a
lead-acid storage battery or a lithium ion battery, is used as a storage element for the electri-
cal energy if the motor and therefore the alternator stop. In this example the battery is placed
in the motor compartment near to the alternator. Therefore the battery can be charged very
well as the voltage drop between alternator and battery can be rather small. On the other
60 5 Circuit analysis

hand the motor compartment is a very harsh environment with respect to temperature, vibra-
tion, dirt. This harsh environment generates lot of stress for the battery.
Several loads are connected to the battery via cables. Here the loads are all electronic sys-
tems of the vehicle. The loads are placed in the motor compartment (e.g. electric power steer-
ing, motor control), in the interior (e.g. dashboard, electric window lifter, seat heating) or the
trunk (e.g. rear lighting).
Fig. 5.6 shows this electrical system and Fig. 5.7 shows the corresponding electric circuit. To
transfer the electrical system into the corresponding electric circuit the elements are mod-
eled: The generator is modeled by a current source and the battery by a voltage source with
internal resistance. According to the lumped element method the cables between the elements
are modeled by resistors RWx. As long as the details of the loads are not relevant the loads are
also summarized as far as possible and modeled by resistors RLx.
By applying nodal analysis the currents and the voltages can be calculated.

Fig. 5.6: Automotive electrical system with the battery in the engine compartment.

Fig. 5.7: The corresponding electric circuit, alternator modeled by a current source.
■

5.2 Mesh analysis

In mesh analysis the meshes of a circuit are the starting point for the calculation. In simple
and small circuits the essential meshes can be used. The circuit in Fig. 5.8 consists of two
essential meshes and these can be used for mesh analysis. After the meshes are found a virtu-
al mesh current and the direction of this virtual mesh current is defined for each mesh. The
labeling of the mesh currents and their direction can be chosen arbitrarily. The real branch
currents are composed of the mesh currents flowing through this branch.
5.2 Mesh analysis 61

The KVL can be applied within the meshes. This means we get an equation for each mesh. In
these equations the voltages are substituted by using Ohm’s law – the corresponding voltage
drop across each component expressed by the mesh current. Then the mesh currents are de-
termined. Once the mesh currents have been found the voltages and branch currents can be
calculated.

Algorithm of mesh analysis

1. Identify the meshes and draw a corresponding mesh current in each mesh (the direction
is arbitrary)
2. Label the voltage drop across each component with an arrow
3. Apply KVL to each of the meshes
4. Apply Ohm’s law to each voltage drop across a component and substitute the voltage by
the corresponding mesh currents
5. Solve the equation system for the mesh currents
6. Calculate the branch currents
7. Determine the voltages across the elements

An example for the application of the algorithm for mesh analysis

Consider an easy example for the presented algorithm as shown in Fig. 5.8. Two meshes
build the circuit and the mesh currents I1, I2 are identified as depicted in a clockwise manner.
Based on the mesh currents the current through resistor R1 is I1, through R2 it is I2 and
through R3 it is I1 – I2.

Fig. 5.8: An example for mesh analysis.

1. Identify the meshes and draw a corresponding mesh current in each mesh (the direction
is arbitrary)
– See mesh currents I1, I2 in Fig. 5.8
2. Label the voltage drop across each component with an arrow
– See voltages U1, U2, U3 in Fig. 5.8
3. Apply KVL to each of the meshes
– Mesh 1:
U1  U 3  U q1

– Mesh 2:
U 2  U 3  U q2
62 5 Circuit analysis

4. Apply Ohm’s law to each voltage drop across a component and substitute the voltages
by the corresponding mesh currents
– Mesh 1:
R1  I1  R3  I1  I 2   U q1

– Mesh 2:
 R2  I 2  R3  I1  I 2   U q2

5. Solve the equation system for the mesh currents

 R1  R3  R3   I 1   U q1 
     
  R3 R2  R3   I 2  U q 2 

det  R1  U  R  R   U  R
I1     q1 2 3 q2 3

det R  R1  R3   R2  R3   R32

det  R 2   U  R  R   U  R
I2    q2 1 3 q1 3

det R  R1  R3   R2  R3   R32

6. Calculate the branch currents
I R1  I1

I R2  I 2

I R3  I 1  I 2
7. Determine the voltages across the elements
U1  R1  I1

U 2  R2  I 2

U 3  R3  I1  I 2 
For given values of R1 = 1 Ω, R2 = 1 Ω, R3 = 2 Ω and Uq1 = 5 V, Uq2 = 10 V the voltages and
currents yield I1 = –1 A, I2 = –4 A, U1 = –1 V, U2 = 4 V, U3 = 6 V. Current I3 = (I1 – I2) = 3 A
through resistor R3 flows parallel to the voltage drop U3.
5.2 Mesh analysis 63

An example for mesh analysis with an ideal current source

The circuit to be analyzed using the algorithm of mesh analysis is given in Fig. 5.9.

Fig. 5.9: Circuit for mesh analysis, mesh currents indicated clockwise by arrows.

Three meshes are identified and the corresponding mesh currents in clockwise direction are
labeled I1, I2, I3. Voltage drops are labeled accordingly. Applying KVL and Ohm’s law to the
three meshes results in:
– Mesh 1:
U1  U 2  U q

– Using Ohm’s law for U1 = R1·(I1 – I3) and U2 = R2·(I1 – I2) we get:
U q  I1  R1  R2   I 2  R2  I 3  R1

– Mesh 2:
U5  U3  U 2  0
– U5 is unknown for the moment as the voltage drop across an ideal current source is
unknown. With U3 = R3·I2 we can express U5 by the mesh currents and the re-
sistances:
U 5  I 2  R3  R2  I1  I 2   0

 U 5  I1  R2  I 2  R2  R3 
– Mesh 3:
U 5  U1  U 4  0
– With U4 = R4·I3 we get for U5:
U 5  I1  R1  I 3  R4  R1 
– Using the two equations from mesh 2 and 3 for U5 gives:
I1  R2  R1   I 2  R2  R3   I 3  R1  R4 
64 5 Circuit analysis

– Up to now we have had two equations for the three mesh currents and we need a
third one to solve for the three currents. This third equation can be derived from the
ideal current source Iq:
I q  I 2  I3

– This leads us to the equation system for the three currents:

I q  I 2  I3

I1  R2  R1   I 2  R2  R3   I 3  R1  R4   0

U q  I1  R1  R2   I 2  R2  I 3  R1

For given values of Uq = 6 V, Iq = 7 A, R1 = 3 Ω, R2 = 2 Ω, R3 = 4 Ω, R4 = 7 Ω the matrix

looks like

 0 1  1   I1   I q 
     
 5  6  10    I 2    0 
 5  2  3   I  U 
   3  q
Using Cramer’s rule this yields for the currents I1 = 2 A, I2 = 5 A, I3 = –2 A. ■

Independent meshes: complete tree of a circuit

So far rather simple circuits have been analyzed using essential meshes. In general the inde-
pendent meshes for mesh analysis of a complex and big circuit can be found using a com-
plete tree of the circuit. A tree of a circuit is a line along the branches of the circuit connect-
ing all nodes without a loop. In general several trees exist for a given circuit and any of these
trees can be used to find the independent meshes. Fig. 5.10 shows a circuit with four nodes.
In addition two complete trees are depicted exemplarily. Both trees connect all four nodes
without forming a loop. After a tree is defined, the branches that are not part of the tree
(links) are used to close the meshes step by step. A mesh must not contain more than one link
and the currents in the links are the mesh currents for the analysis.
In the first example (mid of Fig. 5.10) mesh M2 is closed via the link containing the voltage
source. Meshes M1 and M3 are closed using the links with R1 and R5 respectively. In fact
these are the essential meshes again. In the second example (right side of Fig. 5.10) another
tree is chosen. M1 is again closed using the link with R1. M2 is closed using the link via R3.
For M3 the mesh is closed via the voltage source. This mesh is not an essential mesh. In both
examples all meshes contain just one link and the currents in these links are the correspond-
ing mesh currents.
5.3 Linearity and Superposition 65

Fig. 5.10: Electric circuit with examples of two complete trees for the definition of meshes M1-M3.

Independent meshes of complex circuits can be found by using this method. After the
meshed are defined mesh analysis can be done as described before.

Automotive application
We have again a look at the electrical system of a vehicle. However this time the battery is
placed inside the trunk to reduce the environmental stress. But as can be seen in Fig. 5.11 the
resistance between battery and alternator is now higher and charging the battery is worse as
the voltage drop from alternator to battery is higher. Whether this voltage drop is acceptable
should be calculated, e.g. by applying mesh analysis to determine the currents and the volt-
ages. Of course also by using nodal analysis…

Fig. 5.11: Electric circuit with the battery located inside the trunk.
■

5.3 Linearity and Superposition

Mathematically a function is said to be linear if it satisfies two properties: homogeneity
(scaling) and additivity (superposition). For an arbitrary function f(x) homogeneity is given
by:
f ( Kx)  Kf ( x)
K is a constant scalar value. Additivity is given by:
f ( x1  x2 )  f ( x1 )  f ( x2 )  y1  y2
66 5 Circuit analysis

For a linear circuit (or system) in which excitations x1 and x2 produce responses y1 and y2,
respectively, the application of K1x1 and K2x2 together (K1 and K2 being constants) results in
a response of K1y1 + K2y2.
A circuit consisting of independent sources, linear dependent sources and linear elements
(like resistors) is said to be a linear circuit. For a linear circuit consisting of several inde-
pendent sources the net response in any element, according to the principle of superposition,
is the algebraic sum of the individual responses produced by each of the independent sources
acting only by itself. While each independent source acting on the circuit is considered sepa-
rately, the other independent sources are suppressed. The effect of any dependent source,
however, must be included in evaluating the response due to each of the independent sources.
In brief:
In linear networks we can determine the results of different sources by analyzing the behav-
ior of the circuit independently for each source and the superposition of the results for the
total number of sources. But how are independent sources suppressed, which were not con-
sidered during analysis of another source?
Voltage sources are replaced by short circuits (insignificant internal resistance of the voltage
source, ideal voltage source). Current sources are replaced by open circuits (infinite high
internal resistance).

An example for superposition

An example for the application of principle of superposition is depicted in Fig. 5.12. Both
voltage sources as well as the three resistors are known and we are looking for the current I1
through resistor R1.
We start analyzing the circuit by suppressing all sources but one, e.g. U1. So U2 is suppressed
and replaced by a short circuit as shown in the middle of Fig. 5.12. The branch currents are
now labeled with an extra suffix ‘1’ to indicate that these are the first partial currents. Now
the parallel connected resistors R2 and R3 are in series with resistor R1 and the current I11
through resistor R1 is:
U1 R 2  R3
I 11   U1 
R R R1  R 2  R 2  R3  R3  R1
R1  2 3
R 2  R3

The second step is to analyze the circuit using U2 and suppressing U1, refer to right side of
Fig. 5.12. This time the parallel connected resistors R1 and R3 are in series with resistor R2
and the current I22 through resistor R2 is
U2 R1  R3
I 22  U2 
R R R1  R 2  R 2  R3  R3  R1
R2  1 3
R1  R3
5.3 Linearity and Superposition 67

Fig. 5.12: An example for the application of superposition.

Using the current divider rule for the parallel resistors results in:
R3 R3
I 12  I 22  U2 
R1  R3 R1  R2  R2  R3  R3  R1
In the end the total current through resistor R1 sums up from the two partial currents:
R2  R3 R3
I 1 I 11 I 12  U 1  U 2 
R1  R2  R2  R3  R3  R1 R1  R2  R2  R3  R3  R1 ■

An example for superposition with linear dependent sources

How can we deal with linear dependent sources after introducing superposition with inde-
pendent sources? The effect of any dependent source must be included in evaluating the
response due to each of the independent sources. The following example illustrates the anal-
ysis. The circuit contains an independent voltage source, one independent current source and
one linear voltage controlled current source. The current of the dependent source is given by
the voltage drop across resistor R4 multiplied with a factor of B.
Determine the voltage across the resistor R4 of the circuit depicted in Fig. 5.13 by the appli-
cation of superposition.
68 5 Circuit analysis

Fig. 5.13: An example for superposition for a circuit with a dependent current source; complete circuit (top);
circuit with suppressed current source (open circuit, middle); circuit with suppressed voltage source
(short circuit, bottom).

For superposition we have to suppress the independent sources in turn.

– Step 1: Replace the independent current source by an open circuit as shown in mid-
dle of Fig. 5.13. For node 1 we get by KCL:
 1 1  U 11
 
 R  R   U 21  B
 3 4 

U11 R3  R4
U 21  
B R3  R4
– Applying KVL to the left mesh gives:
R1  R2  I11  U q
Uq
 I 11 
R1  R2

 U11  I11  R2
– Therefore the voltage across resistor R4 from this first part of the solution is:
U11 R3  R4
U 21  
B R3  R4
5.3 Linearity and Superposition 69

– Step 2: Replace the independent voltage source with a short circuit (s. bottom of
Fig. 5.13).
– At node 1, KCL gives
 1 1 
I q      U 12
 R1 R 2 

R1  R2
U 12  I q 
R1  R2
– At node 2, KCL gives
 1 1  U
  U 22  12  I q
R  R  B
 3 4 

U   R R 
U 22   12  I q    3 4 

 B   R3  R 4 
Finally the total net response for the voltage drop across resistor R4 by superposition is:

U 11 R3  R4  U 12   R R 
U 2  U 21  U 22      I q    3 4 

B R3  R 4  B   R3  R 4 
For given values of Uq = 18 V, Iq = 6 A, R1 = 6 Ω, R2 = 12 Ω, R3 = 80 Ω, R4 = 20 Ω and
B=3Ω the voltage across resistor R4 is U22 = 96 V. ■

Automotive application
Superposition can of course also be used to analyze the circuits of the electric system as
given in Fig. 5.7 and Fig. 5.8. ■

An example where the method of superposition fails

The method of superposition is only valid for linear networks. If non-linear elements are part
of the circuit it fails due to the non-linear behavior. Fig. 5.14 shows an example circuit with a
Zener diode (special kind of diode with rather non-linear behavior). First we determine the
current IZ1 for Uq1 and Uq2 = 0 V. Afterwards we determine the current IZ2 for Uq2 and
Uq1 = 0 V. Now we add the voltages UZ1 and UZ2 and compare the resulting current value
with IZ1 + IZ2. As can be seen directly in the diagram the method yields the wrong results.
Don’t use superposition for non-linear components.
70 5 Circuit analysis

Fig. 5.14: An example for a non-linear network with a Zener diode; the method of superpopsition fails in this
case.
■

5.4 Two-terminal circuit and Thévenin’s theorem

Any electric circuit with just two external terminals is called a two-terminal circuit. Without
any current or voltage source inside it is a passive two-terminal circuit. Therefore a passive
two-terminal circuit is an arbitrary configuration of resistors such as depicted in Fig. 5.15.

Fig. 5.15: A passive two-terminal circuit composed of resistors.

This arbitrary configuration of resistors can be replaced by a single, equivalent resistor. In

the example the equivalent resistor Req of the passive two-terminal circuit is:
R2  R3
Req  R1 
R2  R3
In contrast to the passive two-terminal circuit an active two-terminal circuit does contain
sources. Like any resistor, the configuration of a passive two-terminal circuit can be replaced
by an equivalent resistor, any active two-terminal circuit can be replaced by a voltage source
in series with a resistor. This combination of voltage source and resistor is the equivalent
voltage source, or the Thèvenin equivalent circuit of the original circuit. Fig. 5.16 shows a
variable load resistor connected to an active two-terminal circuit with arbitrary internal con-
figuration of linear elements. Consider we don’t care about the details of the internal config-
uration of the two-terminal circuit but we want to know the behavior of the load circuit, e.g.
to know the maximum power transfer to the load resistor. Changing the load resistor results
in a linear dependency of the terminal voltage U of the load current I:
U  U q  I  Ri
5.4 Two-terminal circuit and Thévenin’s theorem 71

Here Uq and Ri are constants and this equation is the same as the equation for a voltage
source Uq with an internal resistor Ri connected in series.

Fig. 5.16: Active two-terminal circuit with arbitrary internal configuration (left); corresponding voltage vs
current diagram (middle); Thevenin equivalent circuit (right).

Thévenin’s theorem says that an arbitrary linear two-terminal circuit (network of linear de-
pendent and independent sources and elements) can be substituted by a real voltage source
(ideal voltage source and internal resistor in series) if just the behavior at the terminals is
regarded. In terms of this theorem, the circuit of voltage source and internal resistor on the
right side of Fig. 5.16 is Thévenin’s equivalent to the active two-terminal circuit on the left
side. If we are just interested in the load circuit (here a single resistor) it can simplify the
analysis if we use this theorem.
Example: We want to determine the maximum power transfer to the load resistor. Using the
original two-terminal circuit connected to the load resistor we have to calculate the power
transfer for changing load resistor values. Depending on complexity of the two-terminal
circuit this might be difficult. Using Thévenin equivalent immediately reveals the solution:
maximum power is transferred to the load resistor if it is equal to the internal resistor of the
Thévenin equivalent.
How can we (easily) determine the two parameters (Uq and Ri) of the Thévenin equivalent?

Algorithm to determine the Thévenin equivalent

1. The load network must not contain dependencies of the supply network
2. Determine the open loop voltage at the terminals of the supply network (i.e. load resistor
RL = ∞ Ω). This yields Uq = Urep = U
3. Determine the inner resistance if the supply circuit Ri = Rrep. Two possibilities:
– First possibility:
• Determine the short-circuit current Isc (i.e. RL = 0 Ω) with
U rep
R rep 
I SC
– Second possibility:
• Short-circuit of all ideal voltage sources
• Remove all ideal current sources (open load)
• Leave depending sources as is
• Look from the outside into the modified arbitrary network and determine the
resistance that you “see from the outside”
• This resulting resistor will be the replacement resistor of the Thévenin equiva-
lent
72 5 Circuit analysis

An example for Thévenin’s theorem

Fig. 5.17: An example for a circuit of a supply network and a simple load network (top); supply network is to be
replaced by the Thévenin equivalent (bottom).

In this example (see Fig. 5.17) we are not interested in the internals of the supply network,
but just in the behavior of the terminals 1 and 0 and we want to know the value of the load
resistor for maximum power transfer from the source. Therefore we simplify the supply
network by applying Thévenin’s theorem. First we regard the supply network without the
load as depicted in Fig. 5.18:

Fig. 5.18: Supply network of the example, currents and voltages are depicted in the figure.
5.4 Two-terminal circuit and Thévenin’s theorem 73

The currents can be expressed by the nodal voltages using Ohm’s law:
U q1  U '
I1 
R1

U 'U10
I2 
R2

U q1  U 10
I3 
R3

– Node 1:
 I 1  I 2  I q 2  I q3  0

U 'U q1 U 'U 10
    I q 2  I q3
R1 R2

– Node 2:
I 2  I 3  I q 3  0

U 'U 10 U q1  U 10
    I q3
R2 R3

Two equations to determine the two node voltages U10 = Urep and U′. Resorting of the two
equations yields the equation system in matrix form:

 R1  R2

 R1
 

  U '   R2  U q1  R1  R2  I q 2  I q 3 
 R3  ( R2  R3 )  U rep    R 2  R3  I q 3  R2  U q1 

The following values are used to calculate Urep and U′:
R1 = 1 Ω, R2 = 2 Ω, R3 = 6 Ω and Uq1 = 20 V, Iq2 = 15 A, Iq3 = 15 A.
 3  1 
R   
 6  8 


det R  18 2

 3  20V 2 / A 
R 2   

 6  220V / A 
2

det R 2   540 2V
 

U rep  30V
74 5 Circuit analysis

After determination of Urep the inner resistor Rrep of the Thévenin equivalent is calculated by
removing the sources from the supply network (replace voltage sources by short-cut and
current sources by open circuit), see Fig. 5.19:

Fig. 5.19: Supply network of the example for determination of Rrep: removal of sources.

The resulting circuit after removal of the sources is just a series combination of two resistors
(R1 and R2) in parallel to a third resistor R3.
R1  R2  R3
Rrep   2
R1  R2  R3
Finally we can draw the Thévenin equivalent circuit in Fig. 5.20.

Fig. 5.20: The Thévenin equivalent circuit with values for Urep and Rrep.

Maximum power PLmax is transferred to the load if the load resistor equals the internal resis-
tor of the voltage source (according to the rule of maximum power transfer):
RL  Rrep  2

RL
U 10  U rep   15V
R L  R rep

U10
IL   7.5 A ■
RL
5.5 Norton’s theorem 75

Automotive application
Consider again the electric system as given in Fig. 5.7. But this time we are just interested in
the analysis of the electrical loads RLx and are not interested in the internal details of the
alternator/battery system. Therefore we would like to replace the alternator/battery system
with a simple real voltage source. By application of Thévenin’s theorem, a replacement can
be achieved and the Thévenin equivalent can be used for the analysis of the load system.

Fig. 5.21: A sub-circuit with alternator, battery and resistances (left) and corresponding Thévenin equivalent
(right).

5.5 Norton’s theorem

Norton’s theorem is the complement to Thévenin’s theorem. While Thévenin’s theorem con-
verts a linear network into a real voltage source, Norton’s theorem converts an arbitrary line-
ar network into a current source:
An arbitrary network of linear elements can be substituted by the Norton equivalent.

Fig. 5.22: Arbitrary (supply) network (left) and Norton’s equivalent (real current source, right).

Algorithm to derive Norton’s equivalent

1. Remove the load resistor in the original network
2. Replace the removed resistance by an electric short-circuit
3. Determine the short-cut current (in the “load” branch): Isc = Irep
4. Determine the replacement conductance Grep
5. Construct Norton’s equivalent with Irep and Grep
6. Determine the load current IL with Norton’s equivalent as the supply circuit
The circuit already analyzed by Thévenin’s theorem is used as an example for Norton’s theo-
rem.
76 5 Circuit analysis

An example for Norton’s theorem

Fig. 5.23: Supply circuit for application of Norton’s theorem, load circuit is short-cut for determination of Irep.

As the load branch is short-cut the terminal voltage U10 = 0 V. The branch currents are:
U q1  U '
I1 
R1

U'
I2 
R2

U q1
I3 
R3

– Node 1:
I 1  I 2  I q 2  I q3  0

U q1  U ' U'
   I q 2  I q3
R1 R2
1
 U q1   1 1 
U '    I q 2  I q 3     
 R1   R1 R 2 
– Using the values from Thévenin’s example and converting the resistor values into
corresponding conductance values the node voltage U′ is:
20
U'  V  6.67V
3
– Node 2:
U q1 U'
I SC  I rep  I 2  I 3  I q 3    I q 3  15 A
R3 R2
5.5 Norton’s theorem 77

– Determination of the replacement conductance Grep:

G1  G2
Grep  G3   0.5S
G1  G2

Fig. 5.24: Replacement circuit of supply network for determination of Grep.

Finally Norton’s equivalent looks like depicted in Fig. 5.25:

Fig. 5.25: Norton’s equivalent with Irep and Grep.

The current of the source is split into a current through the internal resistance and through the
external load. Maximum power is transferred to the load if the internal and external re-
sistances are equal, just like for Thévenin’s equivalent.
For a load resistor of 2 Ω the current is split between the internal and the load resistor ac-
cording to the current divider rule:
GL
I L  I rep   7 .5 A ■
G L  G rep
6 Operational amplifier

6.1 Operational amplifier

The voltage controlled voltage source was introduced in the chapter 4.7. Let’s recall this
circuit and add two resistors to the circuit like shown in Fig. 6.1. The voltage controlled
voltage source amplifies the voltage different U′ = U1 – U2 by amplification factor –A. Be-
tween U1 and U2 there is an open load (R = ∞ Ω). There is one resistor R1 in one of the input
terminal paths and a feedback resistor R2 from one output terminal to the same input termi-
nal. For simplicity we set U2 as the reference point (0 V) and also refer the output voltage Ua
to this reference point.

Fig. 6.1: Voltage controlled voltage source with feedback loop via resistor R2.

What about the output voltage Ua? Is it affected by the resistors? How and why?
According to KCL the currents I1 and I2 sum to zero at input terminal U1. With Ohm’s law
KCL can be written as:
U e U ' U a U '
I1  I 2   0
R1 R2
Using the amplification of the voltage source, Ua = –A · U′ yields:
 U   U 
R 2  U e  a   R1  U a  a   0
 A   A 

 R  R2 
  R2  U e  U a   1  R1 
 A 
80 6 Operational amplifier

1
 R  R2 
 U a  U e  R2   1  R1 
 A 
Due to the feedback loop the output voltage depends not only on the amplification factor A
but also from the resistors R1 and R2. This voltage controlled voltage source with an infinite
input resistance Re and open loop gain A (where Ua = A·(U2 – U1), A positive or negative) is
called ideal voltage amplifier. As Re is infinite no current will enter the input terminals, and
as the output is an ideal voltage source, Ua is driven by the amplifier regardless of load con-
nected to the output. Terminal U2 (labeled with ‘+’) is called the non-inverting input and
terminal U1 (labeled with ‘–’) is called the inverting input.

Fig. 6.2: Circuit and model of an ideal voltage amplifier with gain A.

Considering a very high amplification factor (A → ∞) the output voltage becomes independ-
ent from A and is just determined by the ratio of the resistors:
R2
lim A U a  U e 
R1
The output voltage is inverted compared to the input voltage and amplified by the ratio of the
two resistors. The ideal amplifier can be used to obtain a defined amplification due to the
external resistors R1 and R2. In case of A → ∞ this feedback loop forces the output to stay
finite (just determined by the ratio of the resistors) even though the gain is infinite. This is
due to the fact that the voltage at the internal (infinite) resistor tends to zero:
1
lim A U '  U a  0
A
As the ideal voltage amplifier with infinite gain is very important, it has its own name, the
operational amplifier (or OpAmp). The major property of an OpAmp is the amplification of
an input voltage that can be measured at the output of the device. The symbol that is used in
electric circuits is depicted in Fig. 6.3:
6.1 Operational amplifier 81

Fig. 6.3: Symbol of an OpAmp; UN: voltage at inverting input; UP: voltage at non-inverting input;
UD: differential input voltage; Ua: voltage at output; U+ and U–: power supply terminals.

U+ and U- are the power supply terminals for the operational amplifier, e.g. +15 V and –15 V.
The input voltages must not exceed the supply voltage, otherwise the OpAmp can be de-
stroyed:

U  UP U 

U  UN U 

Under normal conditions the usable output voltage is limited by the power supply voltages
and the range usually is something like:

U   1.5V  U a  U   1.5V

The ideal OpAmp is characterized by an infinite open loop gain A (hence the output has to be
limited somehow by a feedback resistor in most cases), infinite input resistance, zero output
resistance and frequency independent amplification. Real OpAmps differ from this ideal
version as shown in Tab. 6.1.

Tab. 6.1: DC characteristics of ideal and real OpAmps.

Ideal OpAmp Real OpAmp

Open loop gain A=∞ A ≈ 104–107
Input resistance Re = ∞ Ω Re > 1 MΩ
Output resistance Ra = 0 Ω Ra ≈ 1–100 Ω

The input-output characteristic (Ua = Ua(UD) of an OpAmp is depicted exemplarily in the

following diagram (see Fig. 6.4). Notice that due to the very high open loop gain A (or V0)
the output of the OpAmp is already in saturation for rather low input voltage differences in
µV range as shown in this example.
82 6 Operational amplifier

Fig. 6.4: Output voltage of a real OpAmp as a function of the differential input.

Real OpAmps are electronic semiconductor devices composed of several transistors, capaci-
tors and diodes. Fig. 6.5 shows the internal circuit of a µA741 OpAmp from STMicroelec-
tronics.

Fig. 6.5: Circuit of the µA741 OpAmp from STMicorelectronics (µA741 datasheet).

Due to its properties the OpAmp is a very well known device for all kind of applications,
from simple amplification to complex analog calculations.
6.2 Operational amplifier 83

Several standardized packages are available for the packaging of the silicon dies of an
OpAmp. For example the µA741 is housed in a through hole DIP-8 package with 8 pins (see
Fig. 6.6). The package dimensions are 9.5 mm by 7.8 mm and the height of the package is
about 4 mm. The pins have a length of 3.2 mm. A smaller surface mount package type is
SOP-8 with a size of 4.2 mm by 5 mm and a height of 1.5 mm. The pins are just about 1 mm.

Fig. 6.6: Examples of packages for operational amplifier: DIP-8 (package of µA741 OpAmp, left); SOP-8
(right). Package drawings by Infineon Technologies AG.

6.2 Operational amplifier circuits

Comparator
If the OpAmp is used without a feedback loop its function is rather simple: Two voltages are
supplied to the inverting and non-inverting input respectively and the output is, due to infi-
nite (or at least very high) open loop gain just the maximum positive or negative voltage,
depending on which input voltage is higher: Ua = A · (UP – UN) = A · UD. The characteristics
of a comparator is depicted in Fig. 6.7.

Fig. 6.7: Simple comparator circuit (left) and output voltage as a function of UD (right).
84 6 Operational amplifier

Inverting amplifier
One of the easiest circuits with an ideal OpAmp and a feedback loop is given below (also
refer to first example of this chapter with the voltage controlled voltage source):

Fig. 6.8: An inverting amplifier.

As the input resistance is infinite for an ideal OpAmp, KCL at the inverting input P yields
(notice that direction of I2 is opposite to the direction of I2 in the examples of the voltage
controlled voltage source):
I1  I 2  0
Expressing the currents by the voltage drops across the resistors gives:
U e U D U D U a

R1 R2
With the open loop gain Ua = A·UD:
U e  U D U D  A U D

R1 R2
For limit of A → ∞: UD →0 V. Finally:
Ua R
  2
Ue R1
Thus this circuit inverts and amplifies the input voltage and is called inverting amplifier.
Due to the feedback loop of the output to the inverting input the voltage difference UD is zero
and the voltage at the inverting input equals the voltage at the non-inverting input. As the
non-inverting input is connected to ground the voltage at the inverting input is the same. This
voltage is called a virtual ground as it corresponds to the ground voltage without being di-
rectly connected to the ground. Unlike the real ground there is no net current flow to the
virtual ground.

Inverting amplifier with a real OpAmp

The calculations were done so far using an ideal OpAmp. What does this result look like for
a real OpAmp with finite gain and finite input resistance?
6.2 Operational amplifier 85

As we have seen a real OpAmp has a finite input resistance and gain (see Fig. 6.9). Input
voltage u1(t) (lower case if we consider it to be time-dependent) is transferred to output volt-
age u2(t). Resistors R1 and R2 are given as well as the open loop gain V. Five values are un-
known and have to be determined to describe the complete behavior of the real OpAmp:
u2(t), ui(t), ii(t), i1(t), i2(t). Therefore we have to find five equations:
– Functionality of OpApm
u 2 (t )  V  ui (t )
– Voltage drop across input resistance
u i (t )  Ri  ii (t )
– KCL at inverting input
ii (t )  i1 (t )  i2 (t )
– Mesh equation M1
u1 (t )  R1  i1 (t )  ui (t )
– Mesh equation M2
u 2 (t )  R2  i2 (t )  ui (t )

Fig. 6.9: Real OpAmp with input resistance and finite gain V.

We can solve this equation system to get the closed loop gain u2(t)/u1(t). First we replace ii(t)
from the third equation in all other equations:
u 2 (t )  V  ui (t )

ui  Ri  i1 (t )  i2 (t )

u1 (t )  R1  i1 (t )  ui (t )

u 2 (t )  R2  i2 (t )  ui (t )
Afterwards we substitute ui(t) from the second equation:
u 2 (t )  V  R1  i1 (t )  i2 (t )

u1 (t )  R1  i1 (t )  Ri  i1 (t )  i2 (t )
86 6 Operational amplifier


u 2 (t )  R 2  i 2 (t )  Ri  i1 (t )  i 2 (t ) 
From the first equation we get the unknown i1(t) for the other two equations:
u 2 (t ) u (t )
u1 (t )  R1  R1  i 2 (t )  2
V  Ri V

u 2 (t )
u 2 (t )   R 2  i 2 (t )
V

u 2 (t )  V  1 
i 2 (t )   
R2  V 

Using the fifth equation yields for the closed loop gain:

u1 (t ) 1  R1  R  V 1
     1  1   
u 2 (t ) V  Ri  R2  V 
To check the difference to the closed loop gain of the ideal OpAmp let’s consider following
values:
R1 = 10 kΩ, R2 = 30 kΩ, V = 100000, Ri = 1 MΩ
u1 (t ) 1  0.01M  10k  100001  1
    1   
u 2 (t ) 100000  1M  30k  100000  3
So the result for a real OpAmp with realistic values is very close to the result of the ideal
OpAmp ( = 1/3) and we can use the behavior of an ideal OpAmp for most purposes. Similar
calculations can be done to show that a non-zero output resistance Ra changes the behavior of
the real OpAmp just slightly compared to the ideal OpAmp.

Non-inverting amplifier
For the inverting amplifier the input signal is connected to the inverting input. To avoid the
inversion the input signal can be connected to the non-inverting input, keeping the feedback
loop to the inverting input:

Fig. 6.10: A non-inverting amplifier.

6.2 Operational amplifier 87

The differential voltage at the input terminals + (non-inverting) and – (inverting) of the
OpAmp is zero due to the feedback loop to the inverting input. According to voltage divider
rule we get for the inverting input:
R1
UE U U  U A
R1  R2
Hence the closed loop gain of the circuit is
UA R
v  1 2
UE R1

Automotive application
Many sensors are used all over modern vehicles for all kinds of measurements, such as in the
motor compartment (e.g. for rotational speed of the cam shaft, oil pressure, motor tempera-
ture) as well as in the interior (e.g. temperature, light intensity) or on the chassis (e.g. speed,
damping). The sensor output signals are transferred to the corresponding electronic control
unit (ECU, e.g. motor control system). Inside the ECU a microcontroller (µC) uses these data
for the algorithms of the control system.
One way of transferring the measured data to an ECU is to use a simple analog voltage. This
voltage can be read by an analog-to-digital-converter (ADC) of a microcontroller. Unfortu-
nately for some sensors the output voltage is rather small (maybe just a few mV). On the way
to the ECU this small analog signal might be disturbed by the electromagnetic influence of
other electronic systems. A wrong value is then read by the ADC and the control algorithms
do not work correctly any more.
Fig. 6.11 shows the connection from an analog sensor via the non-inverting amplifier to the
ADC input of the microcontroller. Depending on the maximum value of the output voltage of
the sensor, R1 and R2 can be calculated to amplify the voltage to a range that fits to the input
characteristics of the ADC (e.g. 5V maximum).

Fig. 6.11: Amplification of an analog sensor signal by a non-inverting amplifier, measurement by ADC of micro-
controller (µC).
■
88 6 Operational amplifier

Unity gain buffer

A special case of the non-inverting amplifier is the unity gain buffer. Here the output of the
OpAmp is connected directly to the inverting input, i.e. R2 = 0 kΩ (Fig. 6.12). If the resistor
R1 is greater than zero (infinite in limit case) the closed loop gain of the unity gain buffer is
R2
v  1 1
R1

Fig. 6.12: A unity gain buffer.

The output voltage of the unity gain buffer is equal to the input voltage. The purpose of this
OpAmp circuit is to make use of some basic properties of the OpAmp to convert the imped-
ance: the input impedance is very high (infinite for the ideal OpAmp) and the output imped-
ance is very small (zero for the ideal OpAmp). The OpAmp acts like a nearly ideal voltage
source of Ue with very small internal resistance. This eliminates any feedback from the load
connected to the output to the driving circuit (input voltage) as can be seen in a simple ex-
ample.

An example for a circuit with a unity gain buffer

A variable load is connected to a voltage source Uq with internal resistance Ri like depicted
in Fig. 6.13. Depending on the value of the load resistance the terminal voltage U (equal to
the voltage across the load resistor) of the source changes according to the voltage divider
rule. If the load resistance is very high compared to the internal resistance the terminal volt-
age will be about Uq and the current will be very small. If the load resistance is equal to the
internal resistance the terminal voltage will be just half of Uq and the current will be Uq/2Ri.
If finally the load resistor is very small the terminal voltage will be roughly zero and the
current will have its highest value of about Uq/Ri. Thus the load has a major impact on the
behavior of the voltage source and the total circuit. To avoid this feedback from the load to
the source a unity gain buffer can be used. The symbol for the unity gain buffer is a triangle
with a 1 as shown in Fig. 6.13.
After insertion of a unity gain buffer to terminate the voltage source the load voltage will be
independent of the load (see Fig. 6.13, right). The input impedance of the converter is very
high and therefore the input voltage is equal to Uq. Due to its very small output impedance
the OpAmp acts like a nearly ideal voltage source and thus the load voltage is constant and
equal to Uq (at least as long as the load resistor is higher than the output impedance). In gen-
eral a unity gain buffer is used to separate parts of circuits to avoid feedback to other parts.
6.2 Operational amplifier 89

Fig. 6.13: Voltage source with internal resistance and variable load (left); same circuit like on the left side but
with a unity gain buffer to separate the load from the source circuit.
7 Time domain circuit analysis

In previous chapters some concepts for the analysis of electric circuits like mesh or nodal
analysis were introduced. So far only DC circuits have been considered, i.e. circuits with
time-independent sources (DC sources) and after initial disturbances (e.g. switching and
transients) were settled. Even the few examples were sources were time-dependent transient
behavior was not taken into account. If time-dependent parameters like current and voltage
are considered, lower case symbols are used to describe these parameters, e.g. u(t), i(t).
Time domain circuit analysis will be split into two parts:
1. Transient effects (switching events)
2. AC circuits
We will start with the introduction of two new elements in electrical circuits: capacitors and
inductors.

7.1 Capacitor
A capacitor is an electric element that is able to store electrical energy. In a simplified image
an ideal capacitor is built of two plates (electrodes). The electrodes are separated by a non-
conducting space (dielectric) and each electrode is connected to one terminal of the capaci-
tor. A current through a capacitor means that positive charges are accumulated inside the
capacitor on one electrode and negative charges on the other electrode.

Fig. 7.1: A simple image of a capacitor; current i(t) causes positive charges to accumulate on one electrode and
negative on the other; circuit symbol of a capacitor (center) and adjustable capacitor (right).

A separation of charges means there is an electric field generated inside the capacitor storing
electrical energy. The difference of potentials due to the electric field can be measured as
voltage u(t) at the terminals. The ratio of accumulated charges q(t) to created voltage u(t) is
called the capacitance of a capacitor:
q(t )
C
u(t )
92 7 Time domain circuit analysis

The unit of capacity C is Farad:

As
[C ]  1  1F
V
The capacity of 1 Farad of a capacitor means that a stored charge of 1 Coulomb creates a
voltage of 1 Volt at the terminals.
The capacitance C is a constant for a given capacitor depending on the geometric configura-
tion and the dielectric of the capacitor. An ideal capacitor just has a capacitance C, with no
resistance R.
Taking the definition of the voltage and Maxwell’s second equation into account for electro-
static cases the capacitance can be expressed in terms of fields:
 
Q AD  dA
C  B 

 E  ds
U
A

The calculation of the capacitance for arbitrary geometry is in general complex. But for sim-
ple geometries and the neglect of edge effects it can be calculated rather simply, e.g. for a
plate capacitor. As depicted in Fig. 7.2 the capacitor consists of two plates with surface A,
distance d and a dielectric ε. Charges +Q and –Q (same amount, opposite polarity) are accu-
mulated on both plates respectively. The displacement field is homogeneous between the
plates and zero outside the plates (good approximation if the plates are much bigger than the
distance between the two plates).

Fig. 7.2: Plate capacitor with charges +Q and –Q on both plates respectively; left: stray field outside the capaci-
tor; right: simplification: displacement field just inside the capacitor.

The integration to calculate the charge is achieved using the closed surface shown on the
right side of Fig. 7.2. Outside the capacitor the displacement field is zero. Between the plates
it is in x-direction and parallel to the normal of the surface. Thus the charge yields:
7.1 Capacitor 93

 
Q  D  dA  D  A

A

The voltage between the plates is calculated using the electric field and the integration is
done from the left plate (at A = –d/2) to the right plate (at B = d/2). As electric field and inte-
gration path are parallel the voltage is given by:
B   d /2 D Qd
U   E  ds   E  ds  
d /2
 ds 
A d / 2 d / 2  A
Finally the capacitance of a plate capacitor is:
Q A
C 
U d
It is directly proportional to the area of the plates and the dielectric between the plates and
inversely proportional to the distance between the plates.
Recall the definition of electric current:
dq (t ) du (t )
i (t )  C
dt dt
Thus the current entering a capacitor is equal to the rate of buildup of charge on the plate
attached to the terminal and proportional to the buildup of the voltage between the plates.
Integration of the current equation above yields the integral form:
t t
1 1
u (t ) 
C   C 0 
 i( )d   i( )d  u (0)

Here u(0) is the initial capacitor voltage at t = 0 s.

To calculate the energy stored in the electric field of a capacitor we start with the power p(t)
delivered to the capacitor:
du (t )
p (t )  u (t )  i (t )  C  u (t ) 
dt
The energy e(t) stored in the capacitor is obtained by integrating:

du ( )
t t
1 1
e(t )  


p( )d  C  u ( ) 

d
d  C  u 2 (t )  C  u 2 ()
2 2

Assuming the capacitor voltage to be zero at t = –∞ s the energy stored in a capacitor at time
t represents the energy of the electric field between the plates due to the separation of charges
and just depends on the voltage at that time
1
e(t )  C  u 2
2
94 7 Time domain circuit analysis

Some properties of capacitors based on the equations above:

– In the special case that the voltage across the capacitor is constant there is no current
flow through the capacitor any more. In the case of DC (after any switching effects,
s. below) therefore the capacitor behaves like an open load (in fact it is an open load
due to the dielectric between the plates).
– If a capacitor is charged and disconnected afterwards, the current will be zero and
the voltage across the capacitor will stay constant (energy storage element)
– Energy in general cannot be changed instantaneously (this would need infinite high
power). Consequently the voltage u(t) across the capacitor cannot change instanta-
neously. By contrast the current i(t) can change instantaneously.

Series and parallel connection of capacitors

Like resistors capacitors can of course be connected in series and parallel as depicted in Fig.
7.3. For the series connection the voltage drop u(t) across the terminals A-C is split into the
voltages across the capacitors, A-B and B-C:
u(t )  u AB (t )  u BC (t )

du (t ) du AB (t ) du BC (t )
 
dt dt dt
As the same current i(t) is flowing through both capacitors we get:
i (t ) i (t ) i (t )
 
C1 C 2 C eq

Here Ceq is the equivalent capacitance if we replace the two capacitors by a single one. In a
more general manner we can find the equivalent capacitance Ceq for a series connection of n
capacitors with capacitance Ci by:
n

C
1 1

C eq i 1 i

Regarding the parallel connection of capacitors (refer to Fig. 7.3) the voltage drop across
both capacitors is the same and the current i(t) is split into two parts through both capacitors
respectively, i1(t) and i2(t). According to KCL at node B:
du (t ) du (t ) du (t )
i (t )  i1 (t )  i 2 (t )  C1   C2   C eq 
dt dt dt
Again Ceq is the equivalent capacitance if we replace the two capacitors by a single one. For
n capacitors in parallel we can write:
n
C eq  
i 1
Ci
7.1 Capacitor 95

Fig. 7.3: Series (left) and parallel (right) connection of capacitors.

Capacitors and OpAmps

Capacitors can be used also in combination with OpAmps. In this example the feedback loop
of an ideal OpAmp is built up of a resistor R, the inverting input is connected to the input
voltage u1(t) via a capacitor C, see Fig. 7.4.

Fig. 7.4: OpAmp circuit with a capacitor in the input line. The circuit acts as a differentiator.

As the OpAmp is ideal the voltage at the inverting input is 0 V (equal to non-inverting input)
and therefore KCL yields:
i1 (t )  i(t )  i
Using
du 1 (t )
i1 (t )  i (t )  C 
dt

u 2 (t )  R  i2 (t )  R  i(t )
we get:
du1 (t )
u 2 (t )   R  C 
dt
Hence the output voltage is proportional to the negative derivative of the input voltage and
the circuit realizes a differentiator. The term τ = R · C is the time constant of the differentiator.
96 7 Time domain circuit analysis

Real capacitors
Besides the simple capacitor built out of two plane plates (s. above) there are many other
geometric forms for capacitors like cylinder-type shapes (Fig. 7.5). Without derivation the
capacitance for these cylinder-types is given by:
2   0   r  l
C cylinder 
R
ln 
r
Here the parameters are:
– l: length of cylinder
– R: radius of outer electrode
– r: radius of inner electrode

Fig. 7.5: A cylinder-type capacitor.

Capacitors are often made of tightly rolled sheets of metal film with a dielectric material (e.g.
paper or nylon) in between in order to increase the capacitance for a given size. Based on
geometry, dielectric and fabrication process values for the capacitance can range from a few
pico Farads up to the Farad region. Refer to Tab. 7.1 for a list of different types of dielectric
material. The working voltage is the maximum voltage that can safely be applied to the ter-
minals of a capacitor. This value is in general given by the manufacturer. Exceeding this limit
may result in the breakdown of the dielectric (due to the small distance between the elec-
trodes the electric field between the electrodes reaches very high values) and the formation
of an electric path between the capacitor’s plates. Values for the working voltage can range
from a few volts to some thousands volts.

Tab. 7.1: Characteristics of capacitors with different dielectric.

Material Capacitance range Maximum voltage range [V]

Mica 1 pF–0.1 µF 50–600
Ceramic 10 pF–1 µF 50–1600
Paper 10 pF–50 µF 50–400
Electrolytic 0.1 µF–0.2 F 3–600

Due to connections, terminals and internal configuration real capacitors have additional re-
sistive elements in addition to the capacitive behavior. The resistive effect of these parts can
be modeled by a resistor in series (ESR, equivalent series resistance) with an ideal capacitor.
The ESR depends on the capacitor’s type and assembly and is usually in the range of mΩ to
Ω and strongly frequency dependent.
7.1 Capacitor 97

As the dielectric between the electrodes are not perfect isolators there will be a (very small)
amount of current through the capacitor called leakage current. This effect can be modeled
by an ideal capacitor in parallel with a parasitic resistor (Fig. 7.6). These resistors create
losses and the different technical types of capacitors can be distinguished by the amount of
losses.

Fig. 7.6: Model of capacitor with parasitic resistor in parallel to the capacitance due to imperfect dielectric.

There are many different types of capacitors, all with specific pros and cons. Important types
are:
– Ceramic capacitor: the dielectric is a ceramic material (e.g. TiO2, BaTiO3), capaci-
tance values are in the range of 0.5 pF–100 µF and more. Applications are high-
frequency applications as well as storage elements;
– Film capacitor: a dielectric film (e.g. polyester, metalized paper, Teflon) is sand-
wiched between the metal layers, the complete sandwich is wound into a tight roll.
It’s the most common capacitor type with many different forms, capacitance values
ranging from few pF up to 100 µF. Often used in high power applications;
– Electrolytic capacitor: This type uses an electrolyte (ionic conducting liquid) as one
of the electrodes (cathode). The dielectric is formed by a very thin oxide film on the
anode (anode material e.g. Al, Ta or Nb). Due to the very thin dielectric the distance
between the electrodes is very small and due to a coarse surface of the anode the
surface is rather large. These two geometric parameters result in a rather high capac-
itance of 1 µF up to 47 mF. Used in all applications needing high capacitance val-
ues, e.g. DC power supplies. As this type of capacitor is in most cases polarized the
terminals have to be connected with the correct polarity (positive to +terminal, neg-
ative to –terminal, otherwise the capacitor will be destroyed (explode);
– Double layer capacitor (supercap): A special kind of electrolytic capacitor where the
distance between the electrodes is in the nm range. Therefore the capacitance is very
high, up to several hundred Farad or even above. This type of capacitor is used as
storage element e.g. in electric vehicles for short term storage and charging/dis-
charging.
98 7 Time domain circuit analysis

Fig. 7.7: Typical packages for capacitors: ceramic (top left); film (top right); electrolytic, positive terminal
marked with + (center); supercap, positive terminal marked by longer pin (bottom).

Automotive application
Numerous capacitors can be found in almost every electronic system and ECU of modern
vehicles. For example they can be used for filter applications. Or they are used as blocking
capacitors for voltage stabilization as they can provide or absorb high currents in the short
term. This application makes use of the capability of capacitors to store electrical energy.
The combination of energy storage and high current capability makes supercaps very inter-
esting in particular for HEV/EV applications. During braking the electrical motor of
HEV/EV is used as a generator to convert mechanical energy into electrical energy. This
recuperation of braking energy results in high currents. As the battery is not able to cope with
the high currents supercaps can be used as high power storage element as they can absorb
high currents. ■
7.2 Inductors 99

7.2 Inductors
Like the capacitor the inductor is an energy-storage circuit element. However, it is not based
on the electric field, but rather the magnetic field effect: a current flow in a conductor pro-
duces a magnetic field around this conductor. Winding a conductor into a coil (N windings)
increases the magnetic field. This magnetic field is described by the magnetic flux N · Φ(t)
that is directly proportional to the current i(t):
N  (t )  L  i(t )
The constant L is the inductance of the element.

Fig. 7.8: A single inductive coil with N windings (left); American circuit symbol of an inductor (mid) and
European symbol (right).

According to Faraday’s law of induction the voltage across the inductor is proportional to the
change of the total magnetic flux N · Φ(t) and hence to the change of current through the
inductor:
d  (t ) di (t )
u (t )  N   L
dt dt
The unit for inductance is Vs/A = H (= Henry):
An inductor has the inductance of 1 Henry if the induced voltage at the terminals is 1 A as a
reaction to a current change rate of 1 A/s. For the special case of DC current the voltage
across an inductor is zero and an ideal inductor acts like a short circuit.
Integration of the voltage equation above yields the integral form:
t t
1 1
i(t ) 
L   L 0 
 u ( )d   u ( )d  i(0)

Here i(0) is the initial inductor current at t = 0 s.

To calculate the energy stored in the magnetic field of an inductor we start with the power
p(t) delivered to the inductor:
di (t )
p (t )  u (t )  i (t )  L  i (t ) 
dt
The energy e(t) stored in the inductor is obtained by integrating:

di( )
t t
1 1
e(t )   p( )d  L   i( ) 
 
d
d  L  i 2 (t )  L  i 2 ()
2 2
100 7 Time domain circuit analysis

Assuming the inductor current to be zero at t = –∞ s, the stored energy in the inductor at time
t only depends on the current at that time and the inductance of the element and is given by:
1
e(t )  L  i 2 (t )
2
Like for the capacitor (and always in physics) the energy cannot change instantaneously and
according to the relation between energy and current also the current through an inductor
cannot change instantaneously in a step function (but the voltage can). As with the capacitor
the step of a current through an inductor would need an infinitely high voltage at the termi-
nals which cannot be generated.
In other words: A high voltage is induced if a current is switched off very fast. Be careful
with switching off a current through an inductor as this high voltage may damage other com-
ponents of the circuit.

Series and parallel connection of inductors

Like resistors and capacitors inductors can of course be connected in series and parallel as
depicted in Fig. 7.9.
For the series connection the voltage drop u across the terminals A–C is split into the voltag-
es across the inductors, A-B and B-C and same current i(t) flows through the inductors:
u  u AB  u BC

di (t ) di (t ) di (t )
Leq   L1   L2 
dt dt dt

 Leq  L1  L2

Here Leq is the equivalent inductor if we replace the two inductors by a single one. In a more
general manner we can find the equivalent inductance Leq for a series connection of n induc-
tors with inductance Li by:
n
Leq  L
i 1
i

Regarding the parallel connection of inductors (refer to Fig. 7.9) the voltage drop u(t) across
both capacitors is the same and the current i(t) is split into two parts through both inductors
respectively, i1(t) and i2(t). According to KCL at node B:
i(t )  i1 (t )  i2 (t )

di (t ) di1 (t ) di 2 (t )
 
dt dt dt

u (t ) u (t ) u (t )
  
Leq L1 L2
7.2 Inductors 101

1 1 1
  
Leq L1 L2

Again Leq is the equivalent inductance if we replace the two inductors by a single one. For n
inductors in parallel we can write
n

L
1 1

Leq i 1 i

Fig. 7.9: Series (left) and parallel (right) connection of inductors.

Real inductors
Ideal inductors have just the inductance and no resistance or capacitance. Real inductors will
have some associated resistance and capacitance: the wiring of the inductor has some (small
but non-zero) resistance, and sizable capacitances may exist between adjacent turns. A possi-
ble model for a real inductor could be a combination of ideal elements: a combination of
resistance and inductance in series, with a capacitance in parallel. The parasitic resistance
can range from a few Ohms up to several hundred Ohms.
Real inductor (also called coil or choke) values range from about 0.1 µH to several hundred
mH or even several H. Due to the construction of the coils and the storage of energy in the
magnetic field, inductors, in particular for big inductance values, can hardly be miniaturized,
they are rather bulky and expensive. The standardization of inductors is not done to the same
degree as for resistors and capacitors. Some examples of inductor packages are depicted in
Fig. 7.10.

Fig. 7.10: Examples for inductor packages: shielded SMD (surface mount device, left); unshielded SMD (center);
unshielded THD (through hole device, right).
102 7 Time domain circuit analysis

Automotive application

Inductors are often used as chokes to filter out higher frequency AC currents due to the fre-
quency dependence of their impedance. Or they are used as short term energy storage ele-
ment, e.g. in DC/DC converters to convert one DC voltage to another DC voltage. Applica-
tions like an electrical relays make use of the magnetic properties of an inductor (see
Fig. 7.11).
The circuit is split into two parts, a control circuit and a load circuit. Target is to switch the
load without a direct electrical contact to the control circuit. By applying a current to the
inductor of the relays a magnetic field is generated. This field is used to close a magnetic
switch in the load circuit which. Due to the closed magnetic switch the load circuit is electri-
cally closed and the load is switched on. As soon as the control current is switched off, the
magnetic field fades away and the relays opens the load circuit again. The load is switched
off. Notice that there is no direct electrical contact between control and load circuit. Both
circuits are galvanically isolated.

Fig. 7.11: Circuit of an electrical relais to switch a load circuit.

■

7.3 Transient effects and switching

By now we have just regarded DC voltages and currents in our circuit analysis. The circuits
themselves were pure resistive. When starting with time-dependent analysis of circuits with
inductors, capacitors and resistors we have to analyze the behavior of inductors and capaci-
tors as a function of time. However, laws as introduced in previous chapters (like KCL,
KCL) are still valid.
7.3 Transient effects and switching 103

7.3.1 First order circuit – the natural response

We will start with the analysis of switching events and transient effects in first order circuits.
First order circuits contain a single capacitor or inductor and a network of DC sources, resis-
tors and switches. Consider the circuit as shown in Fig. 7.12. The switch is closed for times
t < 0. At t = 0 the switch is opened. What about the voltage, the current and the energy stored
in the capacitor for t < 0 s, t = 0 s and t > 0 s?

Fig. 7.12: A simple circuit with a switch.

t<0s
For t < 0 s the circuit is a DC circuit and the capacitor behaves like an open circuit. The cur-
rent is flowing through the resistor R. According to the voltage divider rule for the two resis-
tors in series the capacitor voltage is
R U
u (t )  u C (t ) 
R  Rg

The current and energy in the capacitor equals:

u (t ) U
i (t )  
R R  Rg
1 1
e(t )  C  u 2 (t )  C  U 2
2 2

t=0s
For t = 0 s the switch is opened instantaneously and the voltage source (as well as the resistor
Rg) are disconnected from the resistor and capacitor connected in parallel. The circuit we are
looking at (the mesh containing the parallel capacitor and resistor) contains no sources any-
more and the result will be called the natural response. As the voltage across a capacitor and
its energy cannot change instantaneously they stay at the value of t < 0 s:
R U
u (0)  u C (0) 
R  Rg
1
e(0)  C  u 2 (0)
2
These values will be the initial conditions for the behavior and solution of times t > 0 s.
104 7 Time domain circuit analysis

t>0s
After the switch opened at t = 0 s it stays open for t > 0 s. As the charged electrodes of the
capacitor are connected via resistor R in a mesh, the capacitor will be discharged via the
resistor. With the currents given as indicated in Fig. 7.12 we can write KCL as:
iC (t )  i R (t )
Using Ohm’s law for the resistor yields:
u(t ) u C (t )
i R (t )  
R R
The current-voltage equation for the capacitor is:
du C (t )
i C (t )  C 
dt
Finally we get:
du C (t ) u C (t )
C  0
dt R
Thus we have a homogenous, first-order, linear differential equation (ordinary differential
equation, ODE) for the voltage across the capacitor (and the resistor) that is to be solved.
Solution:
Rewriting the ODE gives:
du C (t ) 1
  u C (t )  0
dt RC
Separation of variables and substituting t by τ:
duC ( ) 1
 d
u C ( ) RC
Integration from τ = 0 s to τ = t:
t C u (t ) t
1 1 1
 u
 0 C
( )
du C ( ) 
u
u ( 0) C

du C  
RC 
 0
d
C

1
 ln u C (t )  ln u C (0)   t
RC

uC (t ) t
 ln 
u C (0) RC
t

 u C (t )  u C (0)  e RC
7.3 Transient effects and switching 105

So we have the solution for the homogenous ODE with a (so far unknown) constant uC(0).
This constant is determined by the constraints of the initial condition, i.e. the voltage across
the capacitor at time t = 0 s (remember the voltage across a capacitor cannot change instanta-
neously at t = 0 s). This initial value was already determined above (see t = 0 s) and therefore
the final solution for this homogenous ODE is:

R U
t t
 
 u C (t )  u (0)  e RC
  e RC
R  Rg

The function of uC(t) is depicted in Fig. 7.13: The voltage is at u(0)=uC(0) at t = 0 s and de-
creases exponentially with time, so it will never be equal to zero for finite times. The con-
stant R·C in the denominator of the exponential function is called the time constant (unit of
R·C is Ω·F = Ω·s/Ω = s). If depicted in terms of R·C it can be seen that uC(t) decreases to
defined values (e.g. 36.8% of the initial value for t = R·C). So the product R·C gives a direct
measure for the speed of the voltage decrease. The higher R·C is the longer it takes for the
voltage to decrease.

Fig. 7.13: Voltage across the capacitor (and the resistor), time constant τ=R·C determines how quickly the volt-
ages decreases and settles to its final value..

After determination of the voltage the currents through the resistor and the capacitor are:
t t
u C (t ) u (0)  RC U 
i R (t )   e   e RC
R R R  Rg

t t
u (0)  RC U 
iC (t )   e   e RC
R R  Rg

The energy in the capacitor will also decrease:

2
1  
t 2t
1  
e C (t )  C  u C2 (t )  C   u (0)  e RC   1 C  u 2 (0)  e RC
2 2  
 2

After a very long time the capacitor will be completely discharged (never completely but
nearly…) and hence the energy will be zero. As energy cannot vanish and as the resistor is
the only element in the circuit besides the capacitor it is obvious that the energy is dissipated
(converted to heat) in the resistor. The power absorbed by the resistor is
106 7 Time domain circuit analysis

2t
u R2 (t ) u 2 (0)  RC
p R (t )   e
R R
With this power dissipation at time t the total energy absorbed by the resistor from t = 0 s
until t yields:

u 2 (0)  R  C   RC 
t t 2 2
u 2 (0) 
e R (t )   p R ( )d    e RC d    e  1
R 0 R 2  
0  

1   2 
 e R (t )  C  u 2 (0)   e RC  1
2  
 
Thus the total energy from the beginning is conserved and transfers from the capacitor to the
resistor where it is absorbed and dissipated:
1
etotal  eC (0)  e R (t )  eC (t )  C  u 2 (0)
2

Summarizing
At t = 0 s the capacitor was charged to uC(0) (q(0) = C·uC(0)) and it discharges exponentially
after the switch was opened. Until t = 0 s the current through the resistor is driven by the
voltage supply. After disconnection of the voltage supply the current is maintained by the
capacitor at t = 0 s and drops exponentially. The rate at which the voltage decreases is meas-
ured by the time constant τ = R · C. In 5 time constants the voltage is within 1 % of its final
value (steady state value). This behavior of the R · C circuit with no external source of exci-
tation is called the natural response. The capacitor takes the role of a voltage supply with
decreasing voltage.

7.3.2 First order circuit – complete response

After the study of the natural response (no source after the switching event) circuits with a
source as excitation after the switching event such as depicted in Fig. 7.14 are considered.
This time the resistor and the capacitor are in series to the switch and the time-independent
voltage source. The switch is open for t < 0 s, closes at t = 0 s and stays closed for t > 0 s.

Fig. 7.14: A first order circuit with a RC combination and a voltage source as excitation.
7.3 Transient effects and switching 107

Before we analyze the circuit in detail let’s try to figure out qualitatively what will happen,
based on our experience from the natural response. With the switch open no current flows
and the capacitor is uncharged. Closing the switch will make a current flow, hence charging
the capacitor. By charging the capacitor the voltage across it will increase until it reaches the
final value of Uq. At that time the current flow will stop and the circuit behaves like an open
circuit.
The detailed analysis looks like:

t<0s
As the switch is open the voltages u10(t), uR(t) and uC(t) are zero and no current flows
through the resistor and the capacitor.

t=0s
The switch is closed instantaneously. According to KVL for mesh M1 voltage u10(0) equals
to Uq(0). For mesh M2 the voltage across the capacitor cannot change instantaneously and
thus the voltage drop across the resistor equals u10(0). Writing KVL for mesh M2 yields
u10 (0) = U q (0) = u R (0) + u C (0) = u R (0)

And the current i(0) is (remember that the current through a capacitor can change instantane-
ously unlike the voltage):
u R (0) U q (0)
i (0)  
R R

t>0s
The switch stays closed and according to KVL for mesh M1 the voltage will be constant:
u10 (t ) U q(t )

The current i(t) (through resistor and capacitor) will charge the capacitor
du C (t )
i(t )  C 
dt
Writing KVL for mesh M2 yields:
du C (t )
U q (t )  u R (t )  u C (t )  R  i (t )  u C (t )  R  C   u C (t )
dt
Thus in the case of a source in the circuit after the switching event we get a first order inho-
mogeneous ODE:
du C (t ) 1 1
  u C (t )   U q (t )
dt RC RC
1/RC is a constant coefficient and the right side of the equation is a function f, which is in
general time-dependent, f(t).
108 7 Time domain circuit analysis

Excursus: solution of first order inhomogeneous ODE

The general form of a first order inhomogeneous ODE with constant coefficients looks
like:
dx (t )
 a  x (t )  f (t )
dt
To find the solution of this ODE we multiply this equation with eat:

e at  f (t )  e at 
dx (t )
dt
 e at  a  x (t ) 
d at
dt
e  x (t )  
Integration of both sides yields:

 
t t
d a

 0
d
e  x( ) d  e at  f ( )d
 0

t

e
a
 e at  x(t )  e a 0  x(0)   f ( )d
 0

Multiplying with e-at:

t

e
a
 x(t )  e  at   f ( )d  x(0)  e  at
 0

This formula for the solution of a first order ODE with x(0) being determined by the initial
conditions is called the complete response. It consists of two parts that will be discussed in
terms of our problem of transients. ■
Let’s have a closer look at the two terms of the complete response for the switching of the
RC circuit with constant voltage source Uq.
The ODE is:
du C (t ) 1 1
  u C (t )  U q
dt RC RC
With x(t) = uC(t), a = 1/(R·C) and f(t) = Uq/(R·C) the complete response is:

t t  t
 Uq 
 u C (t )  e RC
 
 0
e RC 
RC
d  u Ch (0) e RC

Thus the complete response of this ODE is split into two parts, a solution for the homogene-
ous and a solution for the particular ODE:
uC (t )  uCp (t )  uCh (t )

The solution to the homogeneous ODE (f(t) is set to zero)

du Ch (t ) 1
  u Ch (t )  0
dt RC
7.3 Transient effects and switching 109

was already determined previously and corresponds to the second term of the solution:
t

uCh (t )  uCh (0)  e RC

It is the natural or transient response with an exponential behavior of the capacitor’s voltage.
Constant uCh(0) is determined by the initial conditions of the complete system.
The first term of the solution is determined by the function f(t) which describes the excitation
of the circuit (by the voltage source Uq). It is the solution of the particular ODE and de-
scribes the steady-state behavior of the circuit (t → ∞ s) forced by the excitation (forced
response). For a constant forcing function Uq the steady-state response yields for t → ∞ s:


t t 
Uq Uq 
t
 t 
 lim u Cp (t )  lim e
t  t 
RC
 
 0
e RC 
R C
d  lim
t  R C
 R C e RC
  e RC  1  U q




Looking at the circuit it is obvious that the steady-state response will be just the voltage of
the voltage source (excitation voltage) as in the steady-state the circuit is a DC circuit with-
out any current flowing and the capacitor’s voltage corresponds to the voltage of the source.
The complete response is thus:
t

u C (t )  u Ch (t )  u Cp (t )  U q  u Ch (0)  e RC

uCh(0) is to be determined by the initial conditions. In our example the voltage across the
capacitor uC(0) is zero at t = 0 s. Therefore the equation above yields:
uC (0)  0V  uCh (0)  uCp (0)  Uq  uCh (0)

 u Ch (0)  U q


t
 
t

u C (t )  U q  U q  e RC
 U q  1  e RC 
 
 
The complete response of the inhomogeneous ODE is depicted in with the time scaled by the
time constant τ = R·C is depicted in Fig. 7.15.
110 7 Time domain circuit analysis

Fig. 7.15: The complete response of inhomogeneous first order ODE for an RC circuit.

Using the capacitor’s voltage uC(t) the voltage across the resistor uR(t) and the current i(t) can
be calculated:
t
duC (t ) U q  RC
i(t )  C   e
dt R
t

u R (t )  i(t )  R  U q  e RC
 U q  u C (t )
2
1 1  
t

eC (t )  C  u C2 (t )  C  U q2  1  e RC 
2 2  
 

Example of an RC circuit with a charged capacitor

In the discussion of the RC circuit above the capacitor was uncharged at the beginning and
the initial condition was uC(0) = 0 V. Consider the same circuit like before (see Fig. 7.14),
but this time the capacitor is charged for t < 0 s to a value of uC0. The charging of the capaci-
tor for t < 0 does not change the differential equation of the system:
du C (t ) 1 1
  u C (t )   U q (t )
dt RC RC
The steady-state response of the system and the solution of the homogeneous differential
equation are the same like before:
lim u Cp (t )  U q
t 

t

u Ch (t )  u Ch (0)  e RC

But the pre-charging of the capacitor changes the initial conditions of the system at t = 0 s:
Uq = u R (0) + u C (0) = u R (0)  u C0
7.3 Transient effects and switching 111

Hence uCh(0) can be calculated to be:

uC (0)  uC 0  uCh (0)  uCp (0)  Uq  uCh (0)

 uCh (0)  uC 0  U q

Finally the solution yields:

 
t

uC (t )  U q  uC 0  U q  e RC

The capacitor voltage starts at uC0 and rises up to the steady-state value of Uq.
For given values of C = 75 nF, uC0 = 25 V, Uq = 200 V and R = 10 kΩ the solution is:
t

uC (t )  200V  175V  e 

The time constant is:

  R  C  750µs ■

Example of an RL circuit
Consider an RL circuit like depicted in Fig. 7.16. The switch is open for t < 0 s and it is
closed at t = 0 s. For t < 0 s the circuit is not closed and no current flows. At t = 0 s the cur-
rent cannot change instantaneously and the voltage across the inductor equals the voltage of
the source:
i(0)  0 A

u L (0)  U q

For t > 0 s the current rises to its steady-state value at t → ∞ s. In steady-state the inductor
acts like a short-circuit and the current is given by Ohm’s law
Uq
i ()  i p 
R
The steady-state current corresponds to the particular solution of the first order ODE.

Fig. 7.16: RL circuit with a switch. Switch closes at t = 0 s.

112 7 Time domain circuit analysis

The differential equation of the circuit for t > 0 s can be obtained by KVL and using Ohm’s
law and the inductor relation:
di (t )
U q  u R (t )  u L (t )  R  i (t )  L 
dt

di (t ) R Uq
  i (t ) 
dt L L
The general solution of this ODE for the current is:
R
Uq  t
i (t )  i p  ih (t )   ih ( 0 )  e L
R
The constant ih(0) can be calculated using the initial condition and the final result is:

Uq Uq R
 t Uq  
t

i (t )   e L
  1  e  
R R R  

The time constant τ for the RL circuit in series connection is given by:
L

R
For given values of L = 100 mH, Uq = 200 V and R = 20 Ω the solution is:
 
t

i (t )  20 A  1  e 5 ms 
 
 
The time constant is τ = 5 ms and it takes the circuit about 5·τ ≈ 25 ms to be within 1 % of
the steady-state value. ■

7.3.3 Second order circuit – the natural response

In the previous chapter the circuits contained one energy storing element, a capacitor or an
inductor (besides resistors that do not store but dissipate energy). As a consequence the re-
sulting equations were first order ODEs.
If two energy storing elements are part of a circuit under investigation this circuit is called a
second order circuit as this kind of circuit will be described by linear second order ODEs.
An example for a second order circuit is depicted in Fig. 7.17. No source is part of this series
connection of an inductor, a capacitor and resistor. The voltage across the capacitor is denot-
ed with uC(t) and the current with i(t). The switch is open for t < 0 s and is closed at t = 0 s,
initial conditions are uC(0) and i(0) respectively (For example, the capacitor was charged by
an external voltage source –Uq = uC(0)). Due to the open switch the current is zero for t < 0 s
and due to the inductor it stays zero at t = 0 s.
7.3 Transient effects and switching 113

Fig. 7.17: Series connection of resistor, inductor and capacitor as an easy example for a second order circuit.

Applying KVL to the circuit yields for t > 0 s:

uC (t )  u R (t )  u L (t )  0
In the next step the voltages across the inductor and the resistor are expressed in terms of the
capacitor’s voltage uC(t) using Ohm’s law and the relation between the capacitor’s voltage
and the current:
du (t )
i(t )  C  C

dt

du C (t )
u R (t )  R  i (t )  R  C 
dt

di(t ) d 2 u C (t )
u L (t )  L   LC 
dt dt 2
Finally we get this for the capacitor’s voltage:

duC (t ) d 2 u C (t )
u C (t )  R  C   LC 
dt dt 2

d 2uC (t ) R duC (t ) 1
 2
    u C (t )  0
dt L dt LC
This is a homogeneous second order ODE for the capacitor’s voltage.

Excursus into mechanics: damped harmonic oscillator

A system which exhibits mathematically identical behavior to that of a similar, but phys-
ically different system is analogous to this system. In the case of the RLC-circuit the
homogeneous second order ODE has the same structure (same but the constants) like the
ODE for a damped harmonic oscillator such as given in Fig. 7.18. A mass m is connect-
ed to a spring and the movement x(t) is damped by friction. Balance of forces yields:
d 2 x (t ) dx (t )
m 2
d  c  x (t )  0
dt dt
Here the force for the spring (proportional to the position of the mass) was used:
Fspring  c  x(t )
114 7 Time domain circuit analysis

Also the damping force (proportional to the velocity v(t) = dx(t)/dt) was used:
dx (t )
Fdamp  d   d  v (t )
dt

Fig. 7.18: A mechanical analog to the electrical RLC circuit.

As the mathematical behavior of both mechanical and electrical systems is identical,

these systems are analog. Force causes velocity just as voltage causes current. A damper
dissipates mechanical energy into heat just like a resistor dissipates electrical energy into
heat. Springs and masses store energy in two different ways (potential energy and kinetic
energy respectively) just as capacitors and inductors store energy in two different forms
(electric and magnetic field respectively).
Analog quantities are listed in Tab. 7.2:
Tab. 7.2: Analog quantities of mechanical and electrical systems.

Mechanical Electrical
Force Voltage
Velocity Current
Displacement Charge
Damper (f(t) = d·v(t)) Resistor (u(t) = R·i(t))
Spring (f(t) = c·x(t)=c·∫v(t)dt) Capacitor (u(t) = 1/C·∫i(t)dt
Mass (f(t) = m·dv(t)/dt) Inductor (u(t) = L·di(t)/dt

Due to the analogy to the mechanical system we can expect the behavior of the RLC-
circuit to be the same as known from the damped oscillator: some kind of damped oscil-
lations with different solutions depending on the values of m, c and d. ■
After we found a mechanical analogy to the RLC circuit (and we can expect what the solu-
tion might look like) we have to solve the second order homogeneous ODE for our electrical
oscillating circuit:

d 2 u C (t ) R duC (t ) 1
2
    u C (t )  0
dt L dt LC

Excursus: Solution of linear homogeneous second order ODE

A linear homogeneous second order ODE
d 2 x (t ) dx (t )
 2    n2  x (t )  0
dt 2 dt
7.3 Transient effects and switching 115

can be solved by the following approach:

x (t )  e st

Using this approach for the ODE yields


e st  s 2  2  s   n2  0 
s 2

 2  s   n2  0

This polynomial is called the characteristic polynomial of the corresponding ODE. This
quadratic equation has two solutions:
s1     2   n2

s 2     2   n2

All functions using the given approach and the roots (nulls) s1 and s2 of the characteristic
polynomial are solutions to the given ODE:
x i (t )  e si t

Consequently all linear combinations of these basic solutions are also solutions to the
ODE and the general solution for the ODE is:
x (t )  A1  e s1t  A2  e s2t

A1 and A2 are constant and are determined by the initial conditions. Depending on the
values of α and ωn three different cases have to be distinguished: α > ωn, α < ωn and
α = ωn.
These three cases will be discussed during the analysis of the RLC circuit. ■
Coming back to our original problem of the RLC circuit:

d 2 u C (t ) R duC (t ) 1
2
    u C (t )  0
dt L dt LC
We can determine the solution using:
R

2L

1
 n2 
LC
Depending on the values for R, L and C we have to distinguish three cases.
116 7 Time domain circuit analysis

The overdamped case (aperiodic case): α > ωn

If α > ωn R, L and C have to fulfill:

R 1

2L LC
In this case the roots of the characteristic polynomial s1 and s2 are real and negative:
2
R  R 1
s1      
2L  2L  LC

2
R  R  1
s2      
2L  
2 L LC

Using these values of s1 and s2 the solution for the capacitor’s voltage is:

u C (t )  A1  e s1t  A2  e s2t

Constants A1 and A2 have to be determined by the initial conditions of the system. Initial
conditions are uC(0) and i(0) respectively (e.g. the capacitor was charged by an external volt-
age source Uq= –uC(0)). Due to the open switch the current is zero for t < 0 s and due to the
inductor it stays zero at t = 0 s. So there are two initial conditions to determine the two con-
stants A1 and A2.
For t = 0 s we get:

u C (0)  U q  A1  e s1 0  A2  e s2 0  A1  A2

Due to the current being zero at t = 0 s the voltage change across the capacitor is also zero:
du C (t )
i (0)  C 
dt t 0

du C (t )
 A1  s1  e s1t  A2  s 2  e s2t
dt
At t = 0 s this equation yields:

A1  s1  e s1 0  A2  s 2  e s 2 0  A1  s1  A2  s 2  0

With these two equations A1 and A2 can be determined:

A1  U q  A2

A2  s 2  s1   s1 U q

s1  U q
 A2 
s 2  s1
7.3 Transient effects and switching 117

s1  U q s 2 U q
 A1  U q  
s 2  s1 s 2  s1

 s 2 U q  s1t  s1  U q  s2t
u C (t )    e   
  s  s e
 s 2  s1   2 1
The voltage across the capacitor decreases according to two exponential functions with time
constants 1/s1 and 1/ s2 without any oscillation. This case is called the overdamped or aperi-
odic case and is depicted in Fig. 7.19. With knowledge of the capacitor’s voltage the current
i(t) can easily be calculated by:
du C (t )
i(t )  C 
dt

Fig. 7.19: A capacitor’s voltage as a function of time for the overdamped case.

The underdamped case (periodic case): α < ωn

If α < ωn then the roots of the characteristic polynomial are conjugate complex numbers (j is
the imaginary unit, j² = -1)):
2
R  R  1
s1          j  n2   2    j d
2L  2L  LC

2
R  R  1
s2          j  n2   2    j d
2L  2L  LC

Here the ωd is:

 d   n2   2
118 7 Time domain circuit analysis

Using these two complex number yields for the capacitor’s voltage:

u C (t )  A1  e   j d t  A2  e   j d t u


u C (t )  e t  A1  e  j d t  A2  e j d t 
The first factor again describes the damping with a time constant of 1/α. What about the term
in brackets? As the voltage has to be a real number, A1 and A2 have to be complex conjugates
of each other. We can rewrite A1 and A2 using new constants a and b:
A1  a  jb

A2  a  jb
Using the new constants for A1 and A2 yields for the voltage:
 e jd t  e  jd t 
 
u C (t )  e t   a  e jd t  e  jd t  b  

 j 
With Euler’s formular:

e j  cos( )  j sin( )

The terms in brackets can be rewritten as:

e j d t  e  j d t  2 cos( d  t )

e jd t  e  jd t
 2 sin( d  t )
j

 uC (t )  e t  2a  cos d  t   2b  sin  d  t 

 e t  C1  cos d  t   C 2  sin  d  t 

This expression can be further combined into a single sinusoidal function:


u C (t )  e t  C 0  cos  d  t   
To obtain this last equation the following equations for the new (and last…) constants have
to hold true:

C0  C12  C22

C2
  arctan
C1

Finally the following equation is the general solution to the homogeneous linear second order
ODE given above:

u C (t )  e t  C 0  cos  d  t   
7.3 Transient effects and switching 119

The constants C0 and ψ have to be determined by the initial conditions.

Looking at Fig. 7.20 the parameters can be interpreted as follows:
The capacitor’s voltage and current oscillate with the frequency ωd (damped frequency)
around the steady-state value of zero (both current and voltage). The peak value of the oscil-
lations decreases exponentially with a damping factor α. During the oscillation energy is
transferred from the capacitor to the inductor and vice versa, during the transfer energy is
dissipated in the resistor. α depends linearly on the resistance value R. In the ideal case of
R = 0 Ω there is no damping (α = 0) of the oscillation and the damped frequency ωd equals
the natural frequency ωn that is just determined by the inductor and capacitor. ψ is the phase
angle that describes the shift of the zero crossing of the oscillation with respect to t = 0 s.

Fig. 7.20: A Capacitor’s voltage and current for the underdamped case.

The critically damped case: α = ωn

In case of α = ωn the roots of the characteristic polynomial are equal:
s1  s2  
In this case the second order ODE looks like:
d 2 u C (t ) du (t )
 2  C   2  u C (t )  0
dt 2 dt
The general solution is:

u C (t )   A1  t  A2 e t

A1 and A2 are again constants to be determined by the initial conditions. The capacitor’s
voltage decreases exponentially and the decay is faster compared to all other overdamped
cases.
120 7 Time domain circuit analysis

Fig. 7.21: A capacitor’s voltage for the critically damped case.

7.3.4 Second order circuit – the complete response

After the study of the natural response (no source) circuits with a source as excitation such as
depicted in Fig. 7.22 are considered. A resistor R, an inductor L and a capacitor C are in
series to a switch. At t = 0 s the switch is closed and a voltage source is connected in series.

Fig. 7.22: RLC series connection, voltage source connected at t = 0 s.

Before we analyze the circuit in detail let’s try to figure out what will happen qualitatively,
based on our experience from the natural response. With the switch open no current flows
(i(0) = 0 A) and the capacitor is uncharged (uC(0) = 0 V). No energy is stored in the energy
storing elements L and C. Closing the switch will make a current flow, charging the capacitor
and the inductor. For a constant source U the capacitor will in the end block any current flow.
At that time the current flow will stop, the capacitor’s voltage uC(t) will be equal to the
source voltage U and the energy will be stored in the capacitor. As no current will be flowing
no energy will be stored in the inductor in the end.
The detailed analysis looks like:

t<0s
As the switch is open, all voltages uL(t), uR(t) and uC(t) are zero and no current flows through
the circuit.

t=0s
The switch is closed instantaneously. As the voltage across the capacitor cannot change in-
stantaneously it will be zero, uC(0) = 0 V. As the current through the inductor cannot change
instantaneously, it will also be zero, i(0) = 0 A. These two equations define our initial condi-
tions.
7.3 Transient effects and switching 121

t>0s
The switch stays closed and according to KVL we get:
uC (t )  u R (t )  u L (t )  U
In the next step the voltages across the inductor and the resistor are expressed in terms of the
capacitor’s voltage uC(t) using Ohm’s law and the relation between the capacitor’s voltage
and the current:
duC (t )
i(t )  C 
dt
duC (t )
u R (t )  R  i (t )  R  C 
dt

di (t ) d 2 u C (t )
u L (t )  L   L C 
dt dt 2
Finally we get for the capacitor’s voltage:
d 2 u C (t ) R du C (t ) 1 1
2
    u C (t )  U
dt L dt LC LC
This is an inhomogeneous second order ODE for the capacitor’s voltage. As for the inhomo-
geneous first order ODE we will split the general solution into two parts, a solution to the
homogeneous ODE and a particular solution:
uC (t )  uCp (t )  uCh (t )

The homogeneous solution uCh(t) was already determined in the previous section. The roots
of the characteristic polynomial of the ODE are:
2
R  R  1
s1           2   n2
2L  2 L  LC

2
R  R  1
s2           2   n2
2L  2L  LC

Using the abbreviations:

R

2L
1
 n2 
LC
The homogeneous solution is:
 2  2
    n t     n t
2 2

uCh (t )  A1  e s1t  A2  e s2t  A1  e  

 A2  e  
122 7 Time domain circuit analysis

The particular solution uCp(t) is determined by the excitation of the circuit (by the voltage
source U). It is the solution of the particular ODE and describes the steady-state behavior of
the circuit (t → ∞ s) forced by the excitation (forced response). For a constant forcing func-
tion U the steady-state response yields for t → ∞ s:
uCp (t  )  U

Looking at the circuit it is obvious that the steady-state response will be just the voltage of
the voltage source (excitation voltage) as in the steady-state the circuit is a DC circuit with-
out any current flowing and the capacitor’s voltage corresponds to the voltage of the source.
The complete response is thus:
 2  2 
    n t     n t
2 2

uC (t )  U  A1  e  
 A2  e  

Based on the roots of the characteristic polynomial s1 and s2 three cases can be distinguished
(like of the homogeneous second order ODE) depending on the values of R, L and C (and
therefore α and ωn): the overdamped, the underdamped and the critically damped case. In all
three cases the capacitor’s voltage will tend towards the steady-state value U. Constants A1
and A2 are determined by the initial conditions.
Based on uC(t) all other values can be calculated:

i(t )  C 
duC (t )
dt

 C  A1  s1  e s1t  A2  s2  e s2 t 

u R (t )  R  i (t )  R  C  A1  s1  e s1t  A2  s 2  e s2 t 
u L (t )  L 
di (t )
dt

 L  C  A1  s12  e s1t  A2  s 22  e s2 t 

The overdamped case (aperiodic case): α > ωn

For α > ωn, R, L and C have to fulfill:

R 1

2L LC
In this case the roots of the characteristic polynomial s1 and s2 are real and negative, the
solution is a function that changes by two exponential functions tending towards the steady-
state value U:
 2   2 
     n t      n t
2 2

uC (t )  U  A1  e  
 A2  e  

The behavior is depicted in Fig. 7.23.

7.3 Transient effects and switching 123

Fig. 7.23: A capacitor’s voltage as a function of time for an RLC series connection, voltage source U connected
at t = 0 s : 1: overdamped; 2: underdamped; 3: critically damped.

The underdamped case (periodic case): α < ωn

If α < ωn than the roots of the characteristic polynomial are conjugate complex numbers:
2
R  R  1
s1          j  n2   2    j d
2L  2L  LC

2
R  R  1
s2          j  n2   2    j d
2L  
2 L LC

The general solution is:


u C (t )  U  e t  C 0  cos  d  t   
As shown in Fig. 7.23 the capacitor’s voltage oscillates with an angular frequency ωd around
the steady-state value U. The peak value decreases with an exponential function with time
constant 1/α.

The critically damped case: α = ωn

In case of α = ωn the roots of the characteristic polynomial are same:
s1  s2  
The solution is:

uC (t )  U   A1  t  A2 e t

In this case the capacitor’s voltage tends fastest towards the final steady-state value U as
depicted in Fig. 7.23.

Example of an RLC series circuit

The RLC circuit of Fig. 7.22 should be operated critically damped. The values U, L and C
are given: U = 100 V, C = 80 nF, L = 40 mH. What about the value of the resistor for the
critically damped case?
124 7 Time domain circuit analysis

In critically damped case α = ωn:

R 1

2L LC

L
 R  2  1.41k
C
For a resistor of R = 1.41 kΩ the circuit operates in critically damped mode and reaches the
steady-state value of uC(∞) = U = 100 V in minimum time. If the value of the resistor is
smaller the circuit starts to oscillate (underdamped case), if it is higher it takes more time to
reach the steady-state value. ■

Automotive application
Switching is frequently required in automotive applications. Either single switching events,
e.g. by the driver or frequent and continuous switching within the electronic system. In gen-
eral the switching circuit consists of inductive, capacitive and resistive elements (taking
parasitic effects into account even always). Depending on the size of these elements and the
frequency a detailed analysis of the switching behavior has to be done to avoid an unwanted
behavior, e.g. an oscillation or an overdamped case. Consider a switching event from 0 V to
5 V that has to be detected by a microcontroller (e.g. with an interrupt input pin or even with
an ADC). In case of an oscillation the overshoot of the voltage (see curve 2 in Fig. 7.23) can
disturb or even destroy the microcontroller input pin (and hence the microcontroller) as the
maximum input value of the microcontroller is exceeded. In case of an overdamped case it
may take long time to reach the final value. So the switching from 0 V to 5 V is maybe rec-
ognized to late by the microcontroller. ■

7.4 AC Analysis
During analysis of the oscillating circuits (like RC, RL and RLC circuits) the currents and
voltages turned out not to be constant but time-dependent, either some kind of exponentially
damped dependence or an oscillating behavior. But the sources have been (more or less) time
independent so far. During the following AC analysis just steady state systems will be stud-
ied. All transient effects such as those previously discussed are settled and the system is in a
steady state.
When using AC (alternating current) analysis we will make use of some findings from DC
analysis:
– All events happen at the same time independent of the location within the circuit;
– Kirchhoff’s laws are valid for all instances of time;
– Superposition is still valid for linear elements like resistors, inductors, capacitors
(these elements have linear dependencies (direct linear or derivative) between elec-
trical properties like voltage and current);
7.4 AC Analysis 125

Fig. 7.24: Linear elements and their current-voltage relation.

In this section of AC analysis we will now deal with time-dependent sources (and of course
voltages and currents), in particular with periodically time-dependent elements. Periodically
time-dependent means that the shape u(t) of the time-dependence is repeated periodically
after a time called the period T (k is some arbitrary integer constant):
u(t  kT )  u(t )

Fig. 7.25: Periodical functions u(t): arbitrary shape (left) and sinusoidal shape (right).

The dedicated value at a time t is the instantaneous value. When considering currents or
voltages the arithmetical mean
t0 T  0  2
1 1
u
T  u (t )dt 
t t 0
2  u(t )d (t )
t 0

of an alternating current or voltage is zero. In particular sinusoidal functions like shown in

Fig. 7.25 on the right side have an arithmetic mean of zero:
u(t )  û  sint   

u0
The parameter û of the sinusoidal function is called the peak value and the angular frequency
ω is related to the period T and the frequency f according to
2
  2f
T
126 7 Time domain circuit analysis

Unit for the angular frequency ω is 1/s (whereas for frequency it is Hz). The starting point of
the oscillation is in general not at t = 0 s but shifted for some time indicated by the phase
angle φ. The difference of maximum and minimum value is called the peak-to-peak value.
For sinusoidal shape the peak-to-peak value is 2·û.
Sinusoidal functions play a major role in AC circuit analysis:
The sinusoidal shape stays the same (for same frequency) for addition of sinusoidal functions
and also for differentiation. This is important when using superposition and circuit analysis
techniques like KCL and KVL.
In addition, by using Fourier analysis every periodic function may be represented by a sum
of sinusoidal functions. Therefore the analysis of arbitrary shaped functions can be reduced
to analysis of the sinusoidal functions.
The sinusoidal time-dependence of a current and a voltage looks like:
u(t )  û  sint  u 

i(t )  î  sint  i 
û and î are the peak values of the voltage and the current respectively. The frequency is in
this case the same for both, but the phase angle is different. The phase angle is counted posi-
tive if pointing to the right and negative if otherwise. The phase difference between voltage
and current is
  u  i
As shown in Fig. 7.26 the zero-crossing of the voltage (shifted by φu to the left) is earlier
than the zero-crossing of the current (shifted by φi to the right): the voltage leads the current.
In the opposite case (current earlier than voltage) the phase difference is negative and the
current leads the voltage. In Fig. 7.26 both current and voltage are depicted in one single
diagram even though these two have different values and units. On the y-axis it is denoted
that both current and voltage are used. Even though this labeling of the y-axis will be omitted
in following figures (which is rather common in AC analysis) it should be clear that voltage
and current differ in size and unit.

Fig. 7.26: Sinusoidal voltage and current with different phase angle and same frequency.
7.4 AC Analysis 127

The arithmetic mean of an AC current, or voltage is zero. For certain applications another
value, the rectified value, is used to describe the average effect of current or voltage:
T
1
i
T0 
i(t ) dt

T
1
u
T0 
u (t ) dt

Fig. 7.27: Sinusoidal current, absolute value and rectified value.

For sinusoidal shape (e.g. current î·sin(ωt+φi)) the rectified value is:
2
i  î

Besides the rectified value the root-mean-square value (RMS) is more important, in case of
current and voltage:
T
1 2
I eff  I 
T 0 
i (t )dt

T
1 2
U eff  U 
T 0 
u (t ) dt

By definition:
The RMS value of an AC current is defined as the DC current that leads to the same
power dissipation in a resistor.
For sinusoidal shape (e.g. current î·sin(ωt+φi)) the RMS is:
î
I eff  I 
2

The RMS of a sinusoidal shape is just the peak value divided by 2.

Simple circuits with just one element can easily be calculated using the well known correla-
tions of current and voltage for resistors, capacitors and inductors.
128 7 Time domain circuit analysis

Resistor
Consider a resistor connected to a sinusoidal voltage source:
u(t )  û  sint 
The current through the resistor is given by Ohm’s law:

  sin  t   î  sin t 

u (t ) û
i (t ) 
R R
Current and voltage are in phase (no phase difference between current and voltage). In Fig.
7.28 the line diagram shows both voltage and current. As no scaling for any of the two is
given the size of the curves is unimportant. And both have of course different units. This kind
of line diagram just serves to show the phase difference between voltage and current (no
phase difference for the resistor).

Fig. 7.28: A resistor connected to a sinusoidal voltage source: circuit (left) and line diagram (right) of current
and voltage.

Inductor
Consider an inductor connected to a sinusoidal voltage source:
u(t )  û  sint 
The current through the inductor is given by:
t t
 u ( ) d   û  sin  d    cos t   i (0)
1 1 û
i (t ) 
L 0 L 0  L
û  
  sin  t    i (0)
L  2

Current and voltage are not in phase this time as depicted in Fig. 7.29, the voltage leads the
current by Π/2 or 90 °.
7.4 AC Analysis 129

Fig. 7.29: An inductor connected to a sinusoidal voltage source: circuit (left) and line diagram (right) of current
and voltage.

Capacitor
Consider a capacitor connected to a sinusoidal voltage source:
u(t )  û  sin(t )
The current in the circuit is given by:
 
 C  û  sin t   C  û    cos t   C  û    sin  t  
du (t ) d
i (t )  C 
dt dt  2

Current and voltage are not in phase this time, as the current leads the voltage by Π/2 or 90 °.
This behavior is depicted in Fig. 7.30.

Fig. 7.30: A capacitor connected to a sinusoidal voltage source: circuit (left) and line diagram (right) of current
and voltage.

7.4.1 Vector diagram

Sinusoidal currents and voltages can be shown in line diagrams as depicted in the figures
above and by equations using the sine and cosine functions. But for AC analysis these nota-
tions are complex to use. To simplify the calculation of AC circuits, pointers and vectors and
the complex representation of voltages and currents are used. Just imagine a simple addition
of two currents with same angular frequency but different phase angle as depicted in Fig.
7.31:
130 7 Time domain circuit analysis

Fig. 7.31: Two sinusoidal currents that should be added (left); result of addition (right).

The resulting current will have a new peak value î3 and a new phase angle φ3. Both values
have somehow to be determined.
The addition can be done graphically as shown on the right side of Fig. 7.31. For any change
in frequency of phase this graphical solution has to be repeated.
Another way is to use the representation with sine and cosine functions, e.g. using addition
theorem:
sin     sin   cos   cos   sin 
The addition theorem used for addition of the currents yields:
î3  sint  3   î1  sint  1   î 2  sint  2 

 î 3  sin t   cos 3   cost   sin  3  

î1  cos1   î 2  cos 2   sin t   î1  sin 1   î 2  sin  2   cost 
As the cosine and the sine function are independent of each other the equation has to be true
for both functions and hence the corresponding coefficients have to be equal. This results in
two equations for the values î3 and φ3:

î 3  î12  î 22  2î1  î 2  cos1   2 

î1  sin1   î 2  sin 2 
tan3  
î1  cos1   î 2  cos 2 

Even the simple addition of two currents is a rather complicated matter.

Pointer diagrams can be used to simplify the calculation with sinusoidal currents and voltag-
es. Here currents and voltages are depicted as a vector, or pointer that rotates with an angular
frequency of ω. Fig. 7.32 shows the graphical representation of a sinusoidal function î·sin(ωt)
on the right side. At every instance of time the value of the current is determined by the peak
value î and the angular frequency ω. On the left side of this figure the corresponding vector
diagram is shown. A pointer of length î rotates around the middle of the x-y-diagram with an
angular frequency of ω. At any instance of time the angle φ corresponds to the angle of the
sinusoidal, ωt. The sinusoidal on the right side is nothing other than the projection of the
7.4 AC Analysis 131

pointer in the vector diagram to the y-axis. The left side is the pointer representation of the
sinusoidal.

Fig. 7.32: Line (right) and vector representation (left) of a sinusoidal function.

This representation with rotating vectors makes it much easier to add two sinusoidal func-
tions by just using vector addition. Consider again two currents that should be added as
shown in Fig. 7.31 on the right side. Current i2 leads current i1 by the phase φ2. Transfer of
this information to the vector diagram is shown on the left side of Fig. 7.33. The instance of
time (arbitrarily chosen) is t = 0 s. Current i1 is represented by a vector in the direction of the
x-axis of length î1. The projection to the y-axis is zero in correspondence to the value i1 of at
t = 0 s. At t = 0 s current i2 is non-zero and rotated forward by phase difference φ2 .The
length of the vector is the peak value, î2.

Fig. 7.33: Two currents with the same frequency but phase difference φ: line (right) and vector (left) representa-
tion..

The two vector representations are added by vector addition for any arbitrary time instance,
here t = 0 s, the resulting vector is the total current that rotates around with angular frequency
ω. The addition is done by graphical vector addition as depicted in Fig. 7.34:
132 7 Time domain circuit analysis

Fig. 7.34: Vector addition of two currents (left) and resulting line diagram (right).

The peak value of the total current is î3 and the resulting phase angle is φ3 as depicted in Fig.
7.34. This resulting vector rotates with angular frequency ω. The transfer back to the line
diagram is shown on the right side. So the addition of to time dependent currents with the
same frequency is converted to a simple vector addition. This is still a graphical approach,
not a mathematical calculation that would be preferred (in particular for simulations). How-
ever it is a good starting point for the representation by complex numbers, which is the next
step in the description of AC values.
Of course the vector diagram method can also be used to determine the phase difference
between current and voltage for a circuit element. Application of vector diagrams to the basic
elements R, L and C yields:

Resistor
As we have seen previously current and voltage are in phase (no phase difference between
current and voltage) for the resistor. Therefore the vectors for the current through and the
voltage across the resistor are parallel.

Inductor
As calculated above, current and voltage are not in phase at the inductor, the voltage leads
the current by Π/2 or 90 °. In the vector diagram (Fig. 7.35) the current is rotated clockwise
by 90 ° (π/2) with respect to the voltage vector.

Fig. 7.35: Inductor connected to a sinusoidal voltage source: circuit (left), line diagram (mid) and vector diagram
of current and voltage.
7.4 AC Analysis 133

Capacitor
For the capacitor, the current and voltage are again not in phase, and this time the current
leads the voltage by Π/2 or 90 °. In the vector diagram the current is rotated counterclock-
wise by 90° (π/2) with respect to the voltage vector (Fig. 7.36).

Fig. 7.36: Capacitor connected to a sinusoidal voltage source: circuit (left), line diagram (mid) and vector dia-
gram (right) of current and voltage.

Example for the application of vector diagrams

The circuit depicted inFig. 7.37 is build up of two resistors and one capacitor and is excited
by a sinusoidal voltage source û·sin(ωt). Resistor R2 and the capacitor C are connected in
parallel and together they are connected in series with resistor R1. What about the voltages
and currents in the elements and the total behavior of the circuit?

Fig. 7.37: A circuit with resistors, a capacitor and a sinusoidal voltage source.

The voltage drop u2(t) is the same for the parallel connection of R2 and C. From Ohm’s law
we know that the current i2(t) through the resistor R2 is in phase with the voltage drop u2(t).
Using a vector diagram the magnitude û2 and î2 of the resistor point in the same direction
(here arbitrarily to the right, see Fig. 7.38). For the capacitor the current iC(t) leads the volt-
age u2(t) by 90 ° and therefore points up as indicated in Fig. 7.38.
According to KCL the total current through R2 and C has to be the same like the current i1(t).
Hence the current i1(t) is the vectorial sum of the current vectors as given in Fig. 7.38. Again
using Ohm’s law the voltage across resistor R1 is in phase with current i1(t) through R1 and
both voltage and current of R1 point in the same direction. Current i1(t) is of course the cur-
rent that is provided by the voltage source to the circuit.
134 7 Time domain circuit analysis

Fig. 7.38: Vector diagram of the RC circuit.

Using KVL for the left mesh (sinusoidal voltage source, R1, C) yields the voltage by vectori-
al sum of u1(t) and u2(t) or û1 and û2.
As can be seen in Fig. 7.38 the current (i1(t)) and voltage u(t)of the source are not in phase
but the current leads the voltage. Therefore this circuit has a capacitve behavior. ■

7.4.2 Complex numbers

The vector diagram introduced above resembles the representation of complex numbers.
Recall the imaginary number to be:

j 1

The sum of a real and imaginary number is called a complex number and a complex number
Z can be represented in a Gaussian coordinate system by the rectangular form with the real
part R = Re(Z) on the x-axis and the imaginary part X = Im(Z) on the y-axis:
Z  R  jX

Fig. 7.39: Complex number Z in a Gaussian coordinate system.

The magnitude of the complex number (length of the vector from origin to the point in the
Gaussian coordinate system) and the corresponding angle to the x-axis (real axis) are (see
Fig. 7.39):
7.4 AC Analysis 135

Z  Z  R2  X 2

X
  arctan  
R
Expressing the real and the imaginary part of the complex number by the magnitude and the
angle yields:
R  Z  cos 

X  Z  sin 
Using Euler’s formula (see above) the complex number Z can be written in exponential or
polar form:

Z  Ze j

Both representations of a complex number, rectangular and polar form, are used in AC analy-
sis depending on the purpose. Sometimes the rectangular form is easier to handle, neverthe-
less most of the time the polar form is used as AC analysis deals a lot with differentiation,
multiplication and division.
Some basic calculations with complex numbers:

Summation and subtraction

The easiest way for summation and subtraction of two complex numbers Z1 and Z2 is in
rectangular form, as just the real and imaginary parts are added (subtracted) separately:
Z  Z 1  Z 2  R1  R2  j X1  X 2 

Z  Z 1  Z 2  R1  R2  j X1  X 2 

Multiplication and division

The easiest way for the multiplication and subtraction of two complex numbers Z1 and Z2 is
in polar form, as the magnitude and the phase are treated separately. For multiplication the
magnitudes are multiplied and the phases are added:

Z  Z 1  Z 2  Z 1  Z 2  e j 1  2 

For division the magnitudes are divided and the phases are subtracted:
Z 1 Z1 j 12 
Z  e
Z 2 Z2

A special case of division (and of importance for AC analysis) is the reciprocal value of a
complex number Z:
1 1
  e  j
Z Z
136 7 Time domain circuit analysis

Differentiation of a complex harmonic time function can be easily done in polar form. Con-
sider φ being time dependent, e.g. φ=ωt:

Z  Ze jt

Differentiation yields:
dZ
 j  Ze jt  j Z
dt
So the differentiation in polar form is just a multiplication with jω. In terms of the vector
diagram this multiplication with jω corresponds to counterclockwise rotation of the vector.
For every complex number Z there is a complex conjugate number Z* that differs just by the
sign of the imaginary part:

Z  R  jX  Ze j

Z  R  jX  Ze  j
*

Multiplication of a complex number with its complex conjugate gives the square of the mag-
nitude:

Z  Z  R2  X 2  Z 2
*

The sum of a complex number and the difference of a complex number with its complex
conjugate number yields:

Z  Z  2R
*

Z  Z  2 jX
*

7.4.3 Application of complex numbers to AC circuits

Consider a sinusoidal voltage with peak value û, angular frequency ω and phase angle φu:
u(t )  û  cost  u 
This voltage can be depicted as the real axis projection of a rotating vector of length û in a
vector diagram. Considering this diagram to be a Gaussian coordinate system we can write
the voltage in complex form as:

u (t )  û  cost   u   j sin t   u   ûe j t u 

The momentary value of the complex voltage is given by the real part of the complex volt-
age:
u(t )  Reu(t )  û  cost  u 
7.4 AC Analysis 137

If we are not interested in the actual value of the current or voltage, but just in the relation
between these values (phase difference) the time independent part of the complex quantity
can be considered only:

u  ûe j u 

This is the phase vector or phasor representation of the voltage for any given time (e.g.
t = 0 s). It is useful in particular in all kinds of calculations with a common angular frequency
of all components as it separates the time-dependent term from the time-independent terms.

7.4.4 AC circuits
When using complex numbers to describe AC circuits of course our basic rules hold true:
– All events happen at the same time independent of the location within the circuit;
– Kirchhoff’s and Ohm’s laws are valid for all instances of time;
– Superposition is still valid for linear elements like resistors, inductors, capacitors
(these elements have linear dependencies (direct linear or derivative) between elec-
trical propoerties like voltage and current);
Very basic circuits, circuits with just a sinusoidal source connected to one element, will be
analyzed first. The analysis of these circuits demonstrates the application and benefits of
notation with complex numbers. Afterwards more complex circuits will be studied.

AC circuit with a resistor

Given is a sinusoidal voltage (arbitrarily using φu = 0)

u (t )  ûe j t 

Using Ohm’s law yields for the complex current:

u (t ) û j t 
i (t )   e  îe j t 
R R
As already seen with the trigonometric and the vector approach the current and the voltage at
the resistor are in phase and the peak value of the current is given by the peak value of the
voltage divided by the resistance R.

Fig. 7.40: A simple AC circuit with just a resistor (left), line diagram of current and voltage (center) and vector
diagram (right).
138 7 Time domain circuit analysis

AC circuit with an inductor

Given is a sinusoidal current (arbitrarily using φi = 0)

i (t )  îe j t 

Using the induction law yields for the complex voltage:

d i (t )
u (t )  L   jL  îe j t   jL  i (t )
dt
As the time dependent term is the same for the voltage and the current we can rewrite this
equation using the phasors of the voltage and the current:
u  jL  i
Thus the differentiation yields a multiplication with jω. Multiplication with jω corresponds to
a counterclockwise rotation of 90 ° of the phasor. For the inductor the phasor of the voltage
leads the current by 90 °. The vectors of voltage and current are not in phase but out of phase
by 90 °. Formally this expression equals Ohm’s law and the term ZL = jωL is called the
impedance of the inductor:
u
 j L  Z L
i

In general the ratio u/i is called the impedance of an element. The unit for the impedance is
Ohms (Ω) just as in case of the real resistance.
The reciprocal of the impedance is called the admittance Y and is given by the ratio of cur-
rent phasor to the voltage phasor. The unit for the admittance is Siemens (S):
1 1
YL  
ZL j L

Fig. 7.41: A simple AC circuit with just an inductor (left), line diagram of current and voltage (center) and vector
diagram (right).
7.4 AC Analysis 139

AC circuit with a capacitor

Given is a sinusoidal voltage (arbitrarily using φu = 0)

u (t )  ûe j t 

Using the capacitor’s relation for current and voltage yields for the complex current:
d u (t )
i (t )  C   jC  ûe j t 
dt
In phasor notation it yields:
i  jCu
The differentiation is again a multiplication with jω. Multiplication with jω corresponds to a
counterclockwise rotation of 90 ° of the phasor. For the capacitor the phasor of the current
leads the voltage by 90 °. The vectors of voltage and current are not in phase but out of phase
by 90 °.
Formally this expression again equals Ohm’s law and the factor ZC = 1/jωC=-j/ωC is called
the impedance of the capacitor. The admittance of a capacitor is:
1
YC   j C
ZC

Fig. 7.42: A simple AC circuit with just a capacitor (left), a line diagram of current and voltage (center) and
a vector diagram (right).

Summarizing the results of the simple R, L and C circuits we can write the voltage-current
relations for these linear elements in phasor form as follows:

Fig. 7.43: Impedances for the basic elements resistor, inductor and capacitor in complex form.
140 7 Time domain circuit analysis

Even though the use of complex number phasors seems complicated, this method is used to
simplify the analysis of AC circuits. Fig. 7.44 shows the steps it takes to analyze an AC cir-
cuit, e.g. starting with a sinusoidal voltage. Without the use of complex numbers and phasors,
differential equations have to be solved to calculate the corresponding currents. Depending
on the complexity of the circuit this will be a very difficult task.
Using complex numbers and phasors transforms the differential equations into algebraic
equations which are in general much easier to solve. After the current phasor is calculated it
is transferred back to the time dependent form.

Fig. 7.44: Steps for analysis of an AC circuit using phasors.

7.4.5 Kirchhoff’s laws for AC circuits

In general the behavior of a circuit is calculated using complex numbers and the real currents
and voltages are determined by the resulting real part of the rotating vector associated with
the phasor. The resulting impedance of a circuit can consist of a real and an imaginary part:
Z  R  jX
The real part R is called resistance and the imaginary part X is called reactance. The imped-
ance of a resistor has just a resistance and both an ideal capacitor and an ideal inductor just
have a reactance.
As in the DC case, both Kirchhoff’s current law and Kirchhoff’s voltage law are still valid
for AC circuits. In complex notation these two laws read like:

u
mesh
i 0

i
node
i 0

Consequently (as will be shown in the following section) also the rules for calculation of
elements connected in parallel and in series are still valid.
Series connection of n elements:
n
Z Z i 1
i
7.4 AC Analysis 141

Parallel connection of n elements:

n
Y Y
i 1
i

Series connection of a resistor and an inductor

The analysis of circuits with single elements revealed that there is no phase shift between
current and voltage in case of the resistor circuit and a phase shift of 90 ° with the voltage
leading the current in the case of the inductor. The circuit with a series connection of a resis-
tor and an inductor combines these two elements with different behavior. What about the
current and the voltages?
Consider now Fig. 7.45 with a series connection of a resistor and an inductor. The voltage
drop across the resistor and the inductor given in complex form are:
u R  Ri

u L  jLi
Applying KVL yields for this circuit:
u  u R  u L  Ri  jLi  R  jLi  Z i

Z  R  jL  Z R  Z L

Fig. 7.45: An RL circuit with a sinusoidal voltage source.

The voltage of the voltage source is related to the current via the impedance of the circuit. As
in case of series connection of resistors in DC circuits the total impedance of the circuit is the
sum of the impedances of its elements. In polar form the impedance is:

Z  Z R  Z L  Ze j

The magnitude and the phase angle are given by:

Z  R 2   L 
2

 L 
  arctan 
 R 
Consider a source voltage:
142 7 Time domain circuit analysis

u(t )  û  cost  u 
In complex form this voltage is:

u  ûe j t u 

The resulting current yields:

u û j t u  
i  e
Z Z

 i (t )  Re i    cos t   u   
û
Z
The current lags behind the voltage by a phase shift of φ = arctan(ωL/R).
The voltage drop across the resistor is in phase with the current and yields:
û j t u  
u R  Ri  R e
Z

u R (t )  Re u R   R  cos t   u   
û
Z
The voltage drop across the inductor is given by:
  
û j t u   û j  t  u  
u L  jLi  jL e  L e  2
Z Z

  
u L (t )  Re u L   L
û
 cos  t    u   
Z  2 
As already known, the voltage at the inductor leads the current through the inductor by
π/2 = 90 °.
Graphically this solution is depicted in Fig. 7.46. As there is just one current we use this
current as a starting point for drawing the vector diagram. Current i is drawn horizontal. The
voltage across the resistor uR is in phase with the current and hence also horizontal, whereas
the voltage across the inductor uL leads the current by π/2 and points upwards. The total
voltage u is the graphical sum of the two voltages represented by Z·i. The angle φ is the
phase shift of the voltage leading the current. As long as the voltage vector leads the current
vector the behavior of a circuit is called inductive (if the current vector leads it is called ca-
pacitive).
7.4 AC Analysis 143

Fig. 7.46: Vector diagrams of the series connection of resistor and inductor.

Starting from the vector diagram of the series connection and dividing all terms by the cur-
rent we get a similar triangle formed by the inductances of the elements and the total induct-
ance. As we have seen before the phase angle is given by φ = arctan(ωL/R).

Series connection of a resistor and capacitor

The calculation for a series connection of a resistor and a capacitor is done in the same way
as for the resistor-inductor series connection. The results are:
Impedance:
1
Z R j
C
Magnitude of impedance:
2
 1 
Z  R2   
 C 
Current:
u
i
1
R j
C
Magnitude of the current:
U
I
2
 1 
R2   
 C 

Magnitude of the total voltage:

U  U R2  U C2
144 7 Time domain circuit analysis

Fig. 7.47: Series connection of resistor and capacitor (left), vector diagram for voltages (center) and impedances
(right).

Graphically this solution is depicted in Fig. 7.47. Again the current i is drawn horizontal. The
voltage across the resistor uR is in phase with the current and hence also horizontal, whereas
the voltage across the inductor uL lags the current by π/2 and points downwards. The total
voltage u is the graphical sum of the two voltages represented by Z·i. The angle φ is the
phase shift of the voltage leading the current. Here the total current vector leads the total
voltage and the circuit has a capacitive behavior.

Example of RC series circuit

A bulb should be operated at Ubulb = 230 V and I = 0.5 A using a voltage source of U = 300 V
and f = 50 Hz. What is the value of the capacitance of a capacitor in series with the bulb to
achieve the required operating conditions?
The bulb is a resistive element and the circuit with the series connection of R and C is shown
in Fig. 7.48. The current in this circuit is the same for all elements. According to KVL the
voltage of the source is the vectorial sum of the voltage across the resistor and the capacitor:
u  u R  uC

The magnitudes of the voltage source and the bulb are given and the magnitude of the volt-
age of the capacitor is just:

UC  U 2  U R2  193V

Using this voltage of the capacitor the capacitance is given by the magnitude of the capaci-
tor’s impedance
UC 1
ZC  
I C
I
C   8.25µF
UC
7.4 AC Analysis 145

Fig. 7.48: A bulb operated in series with a capacitor.

■

Parallel connection of a resistor and a capacitor

After the study of series connection of two elements in AC circuits, the parallel connection of
two elements is analyzed, here a resistor and a capacitor. The circuit is depicted in Fig. 7.49.
The total current of the voltage source is split according to KCL:
i  i R  iC
With
u
iR 
R
and
i C  jCu
KCL yields:
u 1 
i  i R  iC   jC u  u    jC 
R R 
The factor (1/R+jωC) is the admittance of the total circuit and is, as expected, the sum of the
admittances of the single elements:
1
Y   jC  Y R  YC
R
The magnitude of the admittance is:
2
1
Y     C 
2

R
And the magnitude of the total current is:
2
1
I  U  Y  U     C 
2

R
146 7 Time domain circuit analysis

Fig. 7.49: Parallel connection of resistor and capacitor (left), vector diagram for currents (middle) and
admittances (right).

Fig. 7.49 shows the vector diagram for the total voltage and the currents. Starting from the
total voltage (in horizontal direction) the current through the resistor is in phase with the total
voltage. The current through the capacitor leads the voltage by 90 ° and points upwards. The
total current of the source is hence the geometrical sum of the two currents. Dividing the
currents by the common total voltage transforms the triangle of the currents to the triangle of
admittances.

Example of RLC circuit

A circuit with series and parallel connections is depicted in Fig. 7.50: A resistor R = 100 Ω is
connected in series with an inductor of L = 10 mH and these two elements are connected in
parallel to a sinusoidal voltage source (U = 50 V, f = 1000 Hz) and a capacitor with
C = 10 µF.

Fig. 7.50: RLC circuit with R and L in series.

To determine the total current i of the circuit we start with the RL series connection. The
impedance of this connection is:

Z RL  Z R  Z L  R  jL   50  j 62.8  80.3  e j 51

As the voltage drop across the series connection of R and L is equal to the voltage of the
source (u) the current i1 through R and L yields:
u 50V
i1    0.62 A  e  j 51  0.39 A  j 0.48 A
Z RL 80.3  e j 51
7.4 AC Analysis 147

The current through the capacitor is given by:

u
i2   u  jC  50V  j  2    1000 Hz  10 µF  j 3.14 A  3.14 A  e90
ZC

According to KCL the total current is:

i  i1  i 2  0.39 A  j 0.48 A  j 3.14 A  0.39 A  j 2.66 A  2.68 Ae82

The total circuit has a capacitive behavior as the current leads the voltage by 82 °.
The total admittance of the circuit is:
1
Y RLC  Y RL  Y C   jC  0.012 S  e  j 51  j 0.0628 S
Z RL
 0.0078 S  j 0.0096 S  j 0.0628 S  0.0078 S  j 0.0532 S  0.0538 S  e j 82 ■

Automotive application
Parallel and series connections of resistors and capacitors are frequently used in automotive
applications, e.g. for high-pass or low-pass filters. Another example is the use of a differen-
tial capacitor in a Wheatstone bridge. This capacitive bridge is used for example in microme-
chanical acceleration or angular rate sensors. The acceleration sensor makes use of the fact
  
that the acceleration a is correlated to a force F  m  a .
A differential capacitor is a series connection of two capacitors with one common electrode,
such as depicted schematically in Fig. 7.51. It acts like a frequency dependent voltage divid-
er. In sensors this common electrode with mass m is free to move if an external force due to
the acceleration is exerted to it. By the movement of the common electrode the capacitances
of the two capacitors change due to the changing distances Δd of the plates. One capacitance
increases and the other decreases:
A
C1 
d  d

A
C2 
d  d
To measure these changes the differential capacitor is one leg of a Wheatstone bridge. The
other leg is built out of two resistors. A sinusoidal voltage source excites the Wheatstone
bridge with frequency ω. In this configuration the voltage difference ua is a direct measure
for the distortion of the differential capacitor and hence for the acceleration:
1
R j C 2 d
u a  u1  u 2  u b  ub   u b 
2R 1 1 2d

jC 2 jC1

The output voltage is constant for constant acceleration and can be measured with an ADC.
148 7 Time domain circuit analysis

Fig. 7.51: The capacitive Wheatstone bridge of an acceleration sensor with a differential capacitor.
■
8 Building blocks

Just as any electric circuit with only two external terminals is called a two-terminal circuit,
two port networks (or four-terminal networks) are circuits with two pairs of terminals such as
shown in Fig. 8.1. In addition two port networks have to fulfill the port condition: current
entering one terminal must be equal to the current flowing out of the other one of the same
port. As for the two-terminal networks two-port networks can be active (containing sources)
or passive (no sources inside). Furthermore, the two port network can be linear (containing
just linear elements like resistors, capacitors, inductors) or non-linear (e.g. with diodes).
The theory of two-port networks is not discussed here. Instead we make use of a special case
of two-port networks: building blocks or system blocks with I1 = 0 A and U2 independent of
I2. This special case can be obtained by adding a unity gain buffer to the input and output
ports respectively. Using this simplification with regard to two-port networks we are able to
analyze building blocks of complex circuits and to determine the transfer characteristics of
these blocks.

Fig. 8.1: An arbitrary two port network fulfilling the port conditions.

The ports connect to other circuits like in the case of the two-terminal circuits. The construct
of building blocks is used to isolate parts of a larger circuit to simplify the analysis of the
complete circuit. In this case often one port is the input port and the other one the output
port, and the main property of the circuit is the transfer function: how does the output voltage
depend on the input voltage and its frequency? If the transfer function is known, the two port
network can be treated as a black box with the internal structure and components being of no
further interest. The building blocks are often used for analysis of filters or transmission
lines.

8.1 High-pass filter

A simple building block of just a resistor R and an inductor L is depicted in Fig. 8.2. It re-
sembles the series connection of R and L treated above. The left port is the input port with an
150 8 Building blocks

input voltage u1(t) whereas the port on the right side is the output port with output voltage
u2(t).

Fig. 8.2: -Two port network consisting of R and L; input voltage is u1(t), voltage across the inductor is the
output voltage u2(t).

What about the transfer function for this building block? The circuit is a frequency dependent
voltage divider and the ratio of the voltages is:
  L  
u2 XL jL L j   arctan   
 R 
   e 2
u1 Z total R  jL R 2  L 
2

The analysis of this ratio can be split into two parts, the ratio of the magnitude and the phase
difference between input and output voltage. The transfer function of the magnitude is also
referred to as voltage gain.

Magnitude
L
u2 L R
 
u1 R  L   L 
2 2 2

1  
 R 

The magnitude of the output voltage depends strongly on the angular frequency ω of the
voltages. For ω → 0 s–1 the magnitude of u2(t) tends to zero. In the DC case (in the limit of
ω = 0 s–1) the output voltage will be totally damped down to 0 V. This is exactly the behavior
of an inductor that is expected in case of a DC circuit: the inductor acts as a short circuit and
there is no voltage drop across the inductor. In this case the full voltage drop of |u1| will be at
the resistor R.
For ω → ∞ s–1 the term ω·L dominates the denominator and the ratio of the voltages tends to
1. In this case the output voltage will be (nearly) undamped and will have the same magni-
tude as the input voltage. The voltage at the resistor will tend to zero.
For frequencies between these two limits the magnitude of the voltage ratio is a steadily
increasing function as depicted in Fig. 8.3 on the left side. How fast the function increases
depends on the ratio of Ω = ω · L / R. Therefore it is very common to rescale the graph loga-
rithmically using this ratio Ω instead of the angular frequency (see Fig. 8.3).
8.1 High-pass filter 151

Fig. 8.3: The transfer function of the voltage; left: scaling using angular frequency; right: logarithmical scaling
using Ω=ωL/R; straight lines show linear approximations for high and low frequencies.

Using the new scaling the functionality of the analyzed circuit becomes obvious: it is a high-
pass filter. High frequencies well above Ω = 1 can pass the circuit with very low damping.
For this frequency range the voltage gain can be approximated by a straight line |u1/u2| ≈ 1.
Low frequencies well below Ω = 1 cannot pass the circuit, they are strongly damped. Here
the transfer function can be approximated by another straight line, | u1/u2| ≈ Ω. These two
lines (approximations of the voltage gain for high and low frequencies) intersect at Ω = 1 and
the voltage gain at this point is:
u2  1
 
u1  1 1  2
2

The corresponding angular frequency is called the cut-off frequency ω0 and is given for the
RL high-pass by:
R
0 
L
Frequencies above the cut-off frequency pass the circuit nearly undamped, and frequencies
below the cut-off frequency are strongly damped and can hardly pass the circuit. In this
blocking region the reduction of the frequency by a factor of 10 also reduces the voltage gain
by a factor of 10.

Phase difference
Besides the voltage gain the phase difference also shows a characteristic behavior given by
the exponential part of the transfer function:
  L  
u2 XL jL L j   arctan   
 R 
   e 2
u 1 Z total R  jL R 2  L 
2
152 8 Building blocks

In the limit of ω → 0 s–1 the arctan term tends to zero and the phase difference between out-
put and input voltage is 90 °. For ω → ∞ s–1 the arctan term tends to Π/2 and the output volt-
age is in phase with the input voltage. In between the phase difference steadily decreases as
shown on the left side of Fig. 8.4.

Fig. 8.4: The phase difference between output and input voltage; left: scaling using angular frequency; right:
logarithmical scaling using Ω = ωL/R.

On the right side of Fig. 8.4 the scaling Ω = ωL/R is used again and the characteristic behav-
ior of the phase difference is clearly visible. At the cut-off frequency ω0 = R/L the phase
difference is 45 °.

8.2 Bode plot

Combination of the two logarithmic diagrams for the magnitude and the phase difference
results in the so called Bode plot. In addition the gain of the magnitude’s diagram is ex-
pressed in decibels:

u2
gain  20 log10 dB
u1
The Bode plot of a high-pass filter is shown in Fig. 8.5. For high frequencies the gain is
approximately constant and equal to 0 dB. At the cut-off frequency ω0 the gain drops down
to –3 dB with regard to the high frequency limit. For low frequencies (ω < ω0) it can be ap-
proximated by:
L 
gain  20 log10 dB  20 log10 dB
R 0
The gain is negative and strongly frequency dependent. As can be seen in the formula above
and Fig. 8.5 the slope of the straight line at low frequencies is –20 dB/decade, where a dec-
ade denotes a change in frequency by a factor of 10. So using Bode plots the properties of the
8.3 Low-pass filter 153

transfer functions for magnitude and phase difference can be easily seen. In addition Bode
plots of complex networks can be constructed by the addition of simpler Bode plots.

Fig. 8.5: A Bode plot of a high-pass filter.

8.3 Low-pass filter

The complementary building block to a high-pass filter is of course the low pass filter. Here
low frequencies can pass the circuit and high frequencies are damped and filtered out. A
simple low-pass filter can be designed by just changing the inductor of the previous high-
pass filter to a capacitor as shown in Fig. 8.6.

Fig. 8.6: An RC low-pass filter.

154 8 Building blocks

The transfer function of this RC low-pass filter is:

1
u2 XC jC 1 1
    e j arctan RC 
u1 Z total 1 jRC  1 1  RC 
2
R
jC

The corresponding Bode diagram is depicted in Fig. 8.7. It clearly shows the filter function-
ality of this building block: low frequencies can pass the circuit, the gain is 0 dB. The cut-off
frequency ω0 is given for the RC low-pass filter by:
1
0 
RC
For the cut-off frequency the attenuation is again –3 dB and the phase shift between output
and input voltage is –45 °. For higher frequencies the gain drops by –20 dB per decade:

gain  20 log 10
1
 2

dB  10 log 10 1  RC  dB  20 log 10 RC 
1  RC 
2

Fig. 8.7: A Bode diagram of a RC low-pass.

8.4 Higher order filters 155

Automotive application
All kind of filters are commonly used in automotive applications. One example is the usage
of a low-pass filter as an anti-aliasing filter. Consider an electronic sensor system as depicted
in Fig. 8.8. A simple sensor like a temperature sensor is connected to an analog input pin of a
microcontroller. The output of the sensor is an analog signal in the range of 0–5 V. An ana-
log-to-digital converter (ADC) inside the microcontroller converts the analog sensor signal to
a digital representation that can be used by the digital logic of the microcontroller. This con-
version takes some time and therefore the sampling of the analog signal is at discrete time
steps (e.g. every 10 µs, so sample rate is 100 kHz). To be able to recover the signal after the
conversion correctly without any aliasing the Shannon-Nyquist criterion has to be fulfilled:
the sample rate has to be at least twice the value of the highest frequency of the signal to be
sampled. Therefore the high frequencies of the analog sensor signal have to be filtered out by
using a low-pass filter like the RC low-pass filter. The cut-off frequency of this filter has to
be adjusted to fit to the sampling rate of the ADC.

Fig. 8.8: A sensor system with microcontroller and low-pass (anti-aliasing) filter for the analog sensor signal.
■

8.4 Higher order filters

So far filters with just one energy storing element, an inductor or capacitor, have been used
to introduce the basic concept of two-port network analysis and filters. These filters are
called first order filters. First order filters have an attenuation far above (or below) the cut-off
frequency of –20 dB per decade. To increase this attenuation, higher order filters can be used.
The order of the filter corresponds to the number of energy storing elements, hence a filter of
nth order contains n energy storing elements. Furthermore the damping is increased by the
order according to n time –20 dB per decade. A low-pass filter of 4th order filters out fre-
quencies well above the cut-off frequency with a damping factor of –80 dB per decade. One
possible way to construct a higher order filter is to concatenate lower order filters (which is
nothing else than connecting building blocks with known transfer function).
However, connecting any element to a building block, in particular to the output, may have a
feedback to its behavior! So care has to be taken to avoid any feedback from one to the other
network if building blocks are concatenated. This kind of feedback has already been dis-
cussed in terms of the unity gain buffer. The example of the two building blocks connected to
each other showed that the load circuit (the second building block) has an influence on the
behavior of the source circuit (the first building block).
156 8 Building blocks

Fig. 8.9 shows an example of concatenated building blocks on the left side. The total network
consists of two pure resistive networks each with resistor R. Separate analysis of these two
network yields a transfer function of 1 for both. In case of the first one there is no current
flow at any terminal, for the second a current enters and leaves via the input port and flows
through the resistor. Any input voltage at the input of each network passes without change to
the output. But after concatenation the behavior changes: the total transfer function is not
again equal to 1 but equal to ½. The concatenated building block is nothing other than a
voltage divider and the second network has an influence on the first one. As soon as the sec-
ond network is connected to the first one a current also flows through the first resistor.

Fig. 8.9: Concatenation of two building blocks: without output termination (left), with an unity gain buffer
at the output (right).

To avoid this kind of feedback the input impedance of the second building block has to be
much higher than the output impedance of the first one. This can be done by terminating the
output of each building block with a unity gain buffer as depicted on the right side of Fig.
8.9. The unity gain buffer makes the output impedance of the first circuit very small and its
output voltage corresponds to the value given by the transfer function of the first network.
The input impedance of the second network is much higher than the output impedance of the
first network and there is no feedback. The transfer function of the total network is equal to 1
(product of the transfer functions of the single networks).
If the output ports are terminated in a proper way, a higher order filter can be obtained by
concatenation. Fig. 8.10 shows how a 2nd order low-pass filter is obtained by concatenation
of two 1st order low-pass filters with the same values of R and C. The output of the first filter
is the input of the second filter. The transfer function of the total 2nd order filter is the product
of the transfer functions of the single filters:
u3 u3 u2 1
   e  j 2 arctanRC 
u1 u 2 u 1 1  RC  2

The cut-off frequency of the 2nd order filter, given by the frequency with a damping of –3 dB
is:

2 1
0 
RC
Well above the cut-off frequency the damping is –40 dB per decade:
8.4 Higher order filters 157

gain  20 log 10
1

dB  20 log 10 1  RC  dB  40 log 10 RC 
2

1  RC 
2

Compared to the used 1st order filter the cut-off frequency is shifted to a lower frequency and
the damping is doubled.
As a general rule Bode diagrams of higher order filters (and in more general of all complex
two port networks), if constructed of lower order filters, can be obtained by the simple addi-
tion of the Bode plots of the lower order filters.

Fig. 8.10: A 2nd order low-pass filter, constructed by concatenation of two 1st order low-pass filters.

Also more complex functionalities can be realized by concatenation of simpler elements.

Consider the circuit given in Fig. 8.11. It consists of a high-pass filter followed by a low-pass
filter, both of 1st order. High frequencies are filtered out by the first element with cut-off
frequency ω1, low frequencies by the second element with a different cut-off frequency ω2. In
total frequencies lower than ω2 and higher than ω1 are filtered out. Just a frequency band
between the two cut-off frequencies can pass the circuit. This kind of circuit is consequently
called a band-pass circuit.
Again, the total transfer function is just the product of the single transfer functions (if output
termination is done properly, e.g. by unity gain buffers). The transfer function is already well
known for the low-pass filter, for the RC high-pass filter it is:
 1 
u2 j arctan  
R1 R1 1 1 
 R1C1 
    e
u1 Z total 1 j 2
R1  1  1 
j C 1 R1 C1 1   
 R1C1 

Hence the total transfer function is given by:

  1  
j  arctan    arctan R 2 C 2  
u3 1 1 
 R1C1



  e  
u1 1  R 2 C 2 
2 2
 1 
1   
 R1 C 1 
158 8 Building blocks

The Bode plot of this band-pass filter is shown on the bottom of Fig. 8.11. At low frequen-
cies the first term of the voltage gain tends to zero, at high frequencies the second term. At a
frequency ωr between ω1 and ω2 it has a maximum. This frequency is called resonant fre-
quency and is given by the mean value of the cut-off frequencies.
Depending on the values of the resistors and capacitors the cut-off frequencies and the reso-
nant frequency can be calculated:
1
1 
R1C1
1
2 
R2 C 2

1
 r  1   2 
R1C1  R2 C 2

The phase difference is 90 ° for ω → 0 s–1 and –90 ° for ω → ∞ s–1. At the resonant frequen-
cy it is 0 ° and output and input signal are in phase:
  1  
j  arctan   arctanR2 C 2  
  
 R1C1 
e  
 e0  1
8.5 Active filter 159

Fig. 8.11: A 2nd order band-pass filter, constructed by concatenation of a 1st order high-pass and low-pass filter
(top); Bode plot of the band-pass filter (bottom).

8.5 Active filter

So far the filters have consisted of passive elements like capacitors and resistors. To avoid
any feedback from a load connected to a passive filter an active element, the unity gain buff-
er, was added to the output terminal. An active filter now uses active elements like OpAmps
to realize the required functionality and the output termination. Fig. 8.12 shows an active
first order low-pass filter using an OpAmp in inverting amplifier configuration. The transfer
160 8 Building blocks

function for this active filter yields (see transfer function of the inverting amplifier, feedback
resistor replaced by parallel connection of R2 and C1):
u2 R2 1
 
u1 R1 1  jR2 C1

The first factor of the transfer function (–R2/R1) corresponds to the inverted amplification of
the inverting amplifier, the second part is similar to the transfer function of the passive low-
pass filter discussed above. In total this active filter combines the low-pass filter functionali-
ty with an amplification of the output voltage.

Fig. 8.12: An active high-pass filter.

9 AC power

Consider an arbitrary linear two terminal network, consisting of resistors, capacitor and in-
ductors as depicted in Fig. 9.1.

Fig. 9.1: Current and voltage of an arbitrary linear two terminal network.

The network has an internal impedance Z and the voltage u(t) and the current i(t) at the ter-
minals. The instantaneous power inside the network is:
p(t )  u(t )  i(t )  û  î  cost   u   cost   i 
Depending on the internal composition of the two terminal network the voltage
u(t) = û · cos(ωt + φu) and the current i(t) = î · cos(ωt + φi) are in general not in phase Hence the
instantaneous power can be positive or negative. In case of p(t) > 0 W power is consumed by
the network, if p(t) < 0 W power is generated by the network.
Before the analysis of the AC power of an arbitrary linear passive network, two limit cases
will be studied: a pure resistive and a pure inductive network.

9.1 AC power of a pure resistive two terminal network

For a pure resistive network voltage and current are always in phase, i.e. φi = φu.Using the
trigonometric relation yields:

cosx    1  cos2 x 
2 1
2
Finally the resulting instantaneous power is:
û î
p(t )  û  î  cost   u    1  cos2t  2 u   U  I  1  cos2t  2 u 
2

2
The instantaneous power oscillates with the double frequency of the voltage and current
(refer to Fig. 9.2) around a finite value with a peak power of p(t)peak = û · î. As current and
162 9 AC power

voltage are in phase the power is always positive and the network consumes power at any
instance of time.

Fig. 9.2: AC power of a resistive network with voltage û·cos(ωt+φu) and current i(t)=î·cos(ωt+ φi).

What about the average power? Due to the sinusoidal shape, both average voltage and aver-
age current are zero. But for the power the average value is:
T
1
pP
T  p(t )dt  U  I
0

The average power is just the product of the effective values of current and voltage:
If an arbitrary time-dependent current (voltage) dissipates the same power within a resis-
tor as a DC current (voltage), then the RMS of the time-dependent current (voltage) is
the same as the DC current (voltage).
This power is called active power (or effective or real power) as it describes the power that is
transferred in one direction, here into the network, and that can be used inside the network.
The unit for the real power is, as usual for power, the Watt (W).

9.2 AC power of a pure inductive two terminal network

If the internal of the network is pure inductive it is the voltage that leads the current by 90 °:
 i   u  90
Therefore the instantaneous power inside the network is:
p (t )  û  î  cost   u   cost   u  90 
û  î  cost   u   sin t   u   U  I  sin 2t  2 u 

Again the instantaneous power oscillates at double the frequency of the voltage and current,
but this time around zero as can be seen in Fig. 9.3. For the first and third quarter of the peri-
od of the voltage, the power is positive and hence power is consumed by the network. The
corresponding energy is stored within the inductor. In the second and fourth quarter, the
9.2 AC power of a pure inductive two terminal network 163

power is negative. Power is generated by the network and the energy stored in the inductor
declines to zero again. 1

Fig. 9.3: AC power of a pure inductive network.

As the oscillation is around zero this time the average value of the power is zero:
T
1
p
T  p(t )dt  0
0

For a pure inductive load no power is transferred to the network on average and the supply
circuit (the circuit connected to the inductive two port network) does not have to provide any
power to the network – on average. The amplitude of this oscillating power that is related to
the temporary storage of energy within the inductor (and that has an average value of zero) is
called reactive power as it is not associated with a permanent power transfer to the network.
Instead the power oscillates to and fro: for half of the time power is transferred to the net-
work and the energy is stored within the inductor. For the other half it is transferred back
from the network and the inductor is discharged again. So the source circuit has to provide
power for half of the time and gets back power the other half. The peak power it has to pro-
vide and readopt is the product of the effective voltage and current, the reactive power Q:
û î
Q  p(t ) peak  U I
2
For a pure capacitive network the situation is similar to the inductive network: current and
voltage are out of phase, this time by –90 °. Energy is temporarily stored in the capacitor, the
power oscillates with double frequency around zero and the reactive power Q is the product
of the effective voltage and current.
So in both cases, pure inductive and pure capacitive, no work at all can be done by the two
port network as no energy is transferred to it on average. But the supply network has to pro-
vide power and hence current for half of the time (and to readopt the same amount of energy
the other time). If the supply network has resistive elements, power will be dissipated and
therewith wasted which is highly unwanted. Fig. 9.4 shows a simple example: a sinusoidal
voltage source with an internal resistance is connected to a pure capacitive network. The
164 9 AC power

average power is zero, but at the internal resistance power is dissipated (converted into heat)
due to the reactive power and the associated the oscillating current flow.

Fig. 9.4: A voltage source with internal resistance connected to a pure capacitive network.

The unit for the reactive power is the var (volt ampere reactive), unlike the Watt for the ac-
tive power.

9.3 AC power of a mixed two terminal network

with L, R and C
Based on the idealized configurations above, mixed networks containing capacitive, induc-
tive and resistive elements can be analyzed. As already seen in the analysis of AC circuits,
mixed networks will have a phase difference between voltage and current that can be any
value between –90 ° and 90 °:
90   u   i  90
Using trigonometric functions the instantaneous power can be transformed:

cos(a)  cos(b) 
1
cos(a  b)  cos(a  b)
2
cos(a  b)  cos(a)  cos(b)  sin(a)  sin(b)

û î
p(t )  û  î  cos(t   u )  cos(t   i )   cos u   i   cos2t   i   u 
2
 U  I  cos u   i   cos2t  2 i   u   i 
 U  I  cos u   i   cos u   i   cos2t  2 i   sin  u   i   sin 2t  2 i 

The final result for the instantaneous power is:

p(t )  U  I  cos u   i   1  cos2t  2 i   U  I  sin u   i   sin2t  2 i 
An example for the instantaneous power of a mixed network is depicted in Fig. 9.5.The two
terms of the instantaneous power of a mixed two terminal network resemble the results of the
pure resistive and inductive network.
9.3 AC power of a mixed two terminal network with L, R and C 165

The first term with a non-zero average corresponds to the active power of the pure resistive
network multiplied by the so called power factor cos(φu – φi). The power factor has a value
between 0 and 1. If the power factor is 1, the voltage and current are in phase and the total
power U · I is transferred from the source to the network. If the power factor is smaller than
one, less power is transferred. The active power P = U · I · cos(φu – φi) is always positive (as
cos(x) = cos(–x)), no matter whether current or voltage is leading (capacitive or inductive
behavior).
The second term with a zero average corresponds to the reactive power of the pure inductive
(or capacitive) network multiplied by sin(φu –φi). The reactive power Q = U · I · sin(φu – φi)
can be positive (inductive network) or negative (capacitive network).

Fig. 9.5: AC power of a mixed resistive-capacitive network.

Reactive and active power have a phase shift of 90 ° and the vector sum of both power com-
ponents results in the so called apparent power S. This yields for the magnitude of the appar-
ent power:

S  P2  Q2  U  I

Even though just it’s just the active power that can be used to do any work within the net-
work all elements of the network and the supply circuit has to be able to cope with the appar-
ent power, e.g. the wires, generators, etc.
The common unit for apparent power is the VA (volt ampere).
A simple visualization of the AC power uses sinusoidal notation. Of course AC power can
also be described in complex notation using complex voltage and current, e.g.:

u  ûe ju

Using complex notation yields an apparent power of:

1
S  u  i*  P  jQ
2
Consequently active and reactive power are just the real and imaginary part of the complex
apparent power respectively:
P  ReS 

Q  ImS 
166 9 AC power

As any two port network can be described by its impedance Z and admittance Y respectively
the apparent power can also be written as:
1 1 1
S  u  i*  i  Z  i*  î 2  Z
2 2 2

Fig. 9.6: An AC power diagram with active (P), reactive (Q) and apparent power (S).

Using apparent power the power factor can be defined as the ratio of active power by reac-
tive power:

cos u   i  
P
S
Using this definition it becomes clear that a high power factor is desirable, as it indicates a
high portion of active power compared to the total apparent power and hence is a measure
for the efficiency of the power transfer. In other words: the higher the power factor, the
smaller the reactive power and therefore the lower the unwanted power losses due to the
reactive power. If a power of 1 kW has to be transferred to the two port network, it takes an
apparent power of 1 VA in case of a power factor of 1 and 2 VA in case of a power factor of
0.5. This additional 1 VA has to be provided by the source and the corresponding currents
generate power losses in resistive elements.
As a high value of the power factor is desired, a lot of effort is spent to increase the power
factor. For linear networks consisting of linear elements only (resistors, capacitors, inductors)
this can be done rather simply by adding the complementary reactive element: In case of a
network with inductive behavior, a capacitor is added and vice versa. This method of power
factor correction is used for example for electric motors such as asynchronous motors: capac-
itors are placed accordingly close to the inductive motor windings. Non-linear loads require
more complex measures for power factor correction.

Example: bulb in series with capacitor

Consider the bulb operated at a sinusoidal voltage source of U = 300 V and f = 50 Hz as
discussed in chapter 7.4.5. The capacitor is 8.25 µF to achieve a voltage drop of UR = 230 V
9.3 AC power of a mixed two terminal network with L, R and C 167

across the bulb for a current of I = 0.5 A. What about the apparent power, the active power
and the power factor?
The total impedance of the circuit is:
1 2
S î  Z  I 2  Z  P  jQ
2
The active power and reactive power are:

P  I 2  R  U R  I  115W

I2
Q  96.5 var
C
These results yield for the apparent power:
S  P  jQ  (115  j96.5)VA
Using the magnitude of the apparent power the power factor can be calculated:

S  P 2  Q 2  150VA  U  I

cos u   i  
115W
 0.77
150VA
The power factor is rather low and a rather high reactive power is oscillating to and fro and
has to be provided by the source.

Fig. 9.7: A bulb in series with a capacitor to be operated by a sinusoidal voltage source.
■
10 Oscillating circuits

10.1 Series configuration

2nd order RLC circuits have been discussed previously and three cases were identified and
analyzed: the overdamped, the critically damped and the underdamped case, depending on
the values of R, L and C. The underdamped case can be obtained with small values of the
resistor and in this case the voltages and the current oscillate with the damped frequency ωd.
In the limit of R = 0 Ω the circuit consists of an inductor and capacitor only and the voltage
and the current oscillates with the natural angular frequency of:
1
n 
LC
Energy is transferred from the capacitor to the inductor back and forth.
If a sinusoidal voltage source (that acts as a driving force for the circuit) is added to the se-
ries RLC circuit (see Fig. 10.1) the voltages across the elements and the current will oscillate
with the frequency of the source and the behavior of the circuit can be analyzed in terms of
the complex impedance of the circuit:

 1  j
Z  R  j  L    Z e
 C 
Magnitude and phase angle are:
2
 1 
Z  R 2   L  
 C 

 1 
 L  
  arctan C 
 R 
 
 
170 10 Oscillating circuits

Fig. 10.1: An RLC oscillating circuit in series configuration (left); vector diagram of voltages and current (right).

The general behavior of the voltages and the current is depicted in the vector diagram on the
right side of Fig. 10.1. As the voltage at the inductor leads the current by 90 ° and the voltage
at the capacitor lags the current by 90 ° these two voltages have opposite directions in the
vector diagram. Accordingly the voltage and the current of the source are in phase if the
magnitudes of uL and uC are equal and therefore if the values of the reactance of the inductor
and capacitor are equal:
1
L 
C
If this condition is true the circuit is in resonance and the corresponding angular frequency is
called the resonance angular frequency. It has the same value such as the natural angular
frequency ωn of the LC circuit:
1
0 
LC
The resonance case of the series RLC circuit has some interesting properties:
– Z = R is purely real
– Source voltage and current are in phase (φ = 0)
– Smallest value of Z for given R, L, C
– ω0 independent of R
– Maximum value of |uR|
– High voltages are possible at the inductor and the capacitor
The last two items can be seen by analysis of the magnitude of the voltages across R, L and
C:
R R
uR  u u u
2 2
 1   L L 
R 2    0 L   R2   
  0 C   C
 C 

0 L 1 L
uL  u u 
2 R C
 1 
R 2    0 L  
  0 C 
10.1 Series configuration 171

1
0C 1 L
uC  u u 
2 R C
 1 
R 2    0 L  

  0C 

At the resonance angular frequency the magnitude of the voltage across the resistor is just the
magnitude of the source voltage. As the circuit acts in resonance, the magnitudes of the in-
ductor and capacitor voltage are the same and equal to the source voltage multiplied by a
factor called the quality factor

1 L
Q 
R C
Depending on the values of R, L and C this factor can be significantly greater than one and
hence the voltages at the capacitor and the inductor will be significantly greater than the
source voltage. For example a RLC circuit with L = 1 mH, C = 1 µF and R = 4 Ω has a reso-
nance angular frequency of about ω0 = 32000 s–1 and a quality factor of about Q = 8.
The magnitude of the voltages at the inductor and capacitor are 8 times higher than the
source voltage and the magnitude of the resistor voltage. This voltage increase has to be
taken into account when designing RLC circuits.
The reciprocal of the quality factor is the damping factor d given by:

1 C
d  R
Q L

As in the mechanical case of a harmonic oscillating system with sinusoidal external force
(e.g. a spring with damping and external excitation), energy for the RLC circuit is provided
by the voltage source to the system. A part of the energy is dissipated by the resistor and the
other part is accumulated in the circuit and resonates between inductor and capacitor. The
energy stored in the circuit corresponds to the maximum energy stored in the inductor when
the current is at its maximum value î (at this moment the voltage across the capacitor and
hence the energy stored in the capacitor is zero due to the –90 ° phase shift):
1
E circuit  E L max  L î 2
2
The energy that is dissipated at the resistor in each cycle is given by the effective value of the
current:
1 2
E Rloss  R I 2   R  î 2   L  C
2 0
172 10 Oscillating circuits

The ratio of the maximum energy stored in the circuit (e.g. in the inductor) to the dissipated
energy in one cycle is related again to the quality factor Q:

1 
2   L  î 2 
2  E max 2  1 L
Q   
E Rloss R  î   L  C
2 R C

The quality factor is therefore a measure for the amount of energy stored in the RLC circuit.
Besides the resonance case, the frequency response of the circuit is also important. Accord-
ing to the formulas for the magnitude of the impedance and the phase angle, the series RLC
circuit shows a very characteristic behavior: capacitive behavior at low frequencies, purely
resistive behavior at the resonance frequency and inductive behavior at high frequencies. Fig.
10.2 shows this characteristic for the magnitude of the impedance and the phase angle.

Fig. 10.2: Magnitude of the impedance of the series RLC circuit with the minimum value of R at resonance
frequency (top); the corresponding phase angle (bottom).

The capacitor dominates both the magnitude of the impedance and the phase angle at low
frequencies: a nearly 1 / (ω · C) behavior for the former one and a phase angle of about –90 °.
In contrast, the almost linear behavior of the magnitude and the phase angle of about 90 °
10.1 Series configuration 173

clearly show the dominating behavior of the inductor at high frequencies. At the resonance
frequency the inductive and capacitive parts cancel each other out and just the resistive part
remains: |Z| = R and the phase angle is 0 °.
The voltages at the resistor, the inductor and the capacitor have their maximum value at the
resonance angular frequency as the impedance is minimal. Close to the resonance angular
frequency the voltages drop more or less sharply as depicted in Fig. 10.3. The cut-off angular
frequencies ω1 and ω2 are the angular frequencies on both sides of the maximum where the
voltage dropped down to 1/√2 or to –3 dB respectively.

Fig. 10.3: The frequency dependence of the voltages across the inductor, capacitor and resistor, ω1 and ω2 are the
cut-off frequencies.

In terms of a two port network the series RLC oscillating circuit acts as a band-pass filter.
Considering the voltage source to be the input voltage of a two port network and the resistor
voltage to be the output voltage just frequencies close to the resonance frequency can pass
the network. The narrower the peak is, the better the filtering of a small frequency band is
around the resonance frequency. A characteristic of the band pass filter is the bandwidth
given by the difference of the cut-off frequencies (cut-off angular frequencies divided by 2π)
2  2
b  f 2  f1 
2
The bandwidth can be calculated using the frequency dependence of |uR|. At the cut-off fre-
quencies the resistor voltage dropped down to 1/√2 and hence the denominator of |uR| has to
be:
2 2
 1   1 
R 2   1 L    R 2    2 L    2R
 1 C    2 C 
174 10 Oscillating circuits

Correspondingly the cut-off angular frequencies yield

2
R 1  R 
1     
2L LC  2 L 

2
R 1  R 
2    
2L LC  2 L 

Finally the bandwidth is related to the quality factor of the circuit:

2  2 1 R 1 0 f 0
b     
2 2 L 2 Q Q

The bandwidth of an oscillating circuit is the resonance frequency divided by the quality
factor Q. By changing the quality factor the bandwidth can be tuned even if the resonance
frequency stays equal. Fig. 10.4 shows the frequency dependence of the resistor voltage for
different values of L and C.

Fig. 10.4: The frequency response of a series RLC circuit with different quality factors but same resonance
frequency.

The resonance angular frequency stays the same for the three parameters sets of R, L and C,
but the quality factor is changed by a factor of 10. The bandwidth his reduced by a factor of
10 accordingly and the filtering functionality of this oscillating circuit is highly enhanced.
The higher the quality factor, the more pronounced the frequency response is, the smaller the
bandwidth is and the higher the voltage amplification is at the energy storing elements (in-
ductor and capacitor).
10.2 Parallel configuration 175

10.2 Parallel configuration

There are many different topologies for oscillating circuits beside the series RLC circuit
given here. They are all described by the parameters derived for the series RLC circuit such
as resonance frequency, quality factor and bandwidth even though the formulas for the calcu-
lation of these parameters differ. Just as an example the parallel RLC circuit such as given in
Fig. 10.5 has the same resonance angular frequency and the same bandwidth as the series
circuit:
1
0 
LC
1 0
b 
2 Q

In contrast to this identity of the formulas the quality factor is the inverse of the quality factor
of the series circuit:

C
Q  R
L

Fig. 10.5: A parallel RLC circuit.

Automotive application
Oscillating circuits are used in numerous automotive applications. Analog radios use these
circuits as tuning circuits. Radio stations transmit at different frequencies and the radio an-
tenna receives a superposition of all these radio signals. The oscillating circuit filters a small
frequency band out of the antenna signal to receive just one radio station. Tuning to different
stations can be done by using a variable capacitor to change the resonance frequency.
Keyless entry systems for vehicles or electronic immobilizer systems are other applications
for oscillating circuits. These systems are closely related to RFID (radio frequency identifica-
tion). Using keyless entry systems vehicles can be unlocked without the use of a (mechani-
cal) key. As soon as the vehicle detects an approximation (e.g. by capacitive or optical prox-
imity sensors) antennas of the keyless entry system start to transmit signals, e.g. with fre-
quencies of some hundred kHz. The key has an oscillating circuit (very often just an LC
oscillating circuit) with a fitting resonance frequency to receive the signal of the antennas.
Afterwards the key sends a response back to the vehicle and in case of a correct response the
vehicle is unlocked. ■
11 Semiconductor devices

Basic elements like resistors, capacitors or inductors are part of almost every electronic cir-
cuit. But besides these elements in particular semiconductor devices are extremely important
to realize any complex electronic circuit. Semiconductors are used for devices like diodes or
transistors as well as for rather complex integrated circuits (IC) like microprocessors and
microcontrollers. These devices in general make use of the properties of doped semiconduc-
tors and combine n- and p-doped semiconductors and metals to realize different functionali-
ties.
The basis for most semiconductor devices is pure silicon and compound semiconductors like
SiC or GaN to some extent. Highly sophisticated processes are used to produce these semi-
conductor devices in the form of small rectangular dies. Depending on the functionality of
the device (MOSFET, microprocessor, etc.) different process technologies are used, but to
some extent these techniques are rather similar. The semiconductor industry is highly innova-
tive in order to continuously improve their technologies. In particular semiconductor struc-
tures have shrunk very rapidly. According to Moore’s law the number of transistor elements
per area doubles every 12–24 months.
The starting point for silicon dies, or chips is an extremely pure (>99,99999999 % purity)
and crystallographically very well defined (very few crystal defects) cylindrical tube of sili-
con called an ingot. The diameter of the ingots ranges from 100–400 mm. From the ingot
thin plates of silicon of some hundred micrometers thickness are cut. The surface of each
wafer is separated into small rectangular areas known as dies. Dedicated process steps like
photolithography, ion implantation for n- and p-doping, chemical etching, oxidation or vapor
deposition are used repetitively to produce the required structures and elements onto the
wafer. Small structures of the elements like the gate length of the transistors are is just about
20 nm in 2014! Up to several billion transistors on one die of some cm² size can be realized
by these techniques.
The connection of the billions of transistors on a die is achieved using many metal lines
deposited on top of the wafer. At the end of the wafer process the wafer is separated into the
single dies of some mm² or cm² size. Finally the silicon dies are mounted into dedicated
packages or modules.
178 11 Semiconductor devices

Fig. 11.1: A wafer with dies, complete dies are marked grey (left); packaged die (light grey) on a leadframe (dark
grey) with bond wires (lines) in a package (right).

11.1 Diode
One of the simplest semiconductor devices is the combination of an n- and p-doped semi-
conductor to form a pn-junction as depicted in Fig. 11.2. The n-doped semiconductor has free
electrons and stationary holes localized at the dopand. For the p-doped semiconductor it is
vice versa. Both the n- and the p-doped semiconductors are electrically neutral. At the pn-
boundary there is a strong concentration gradient of the free charge carriers: free electrons in
the n-doped region and free holes in the p-doped region. Due to the concentration gradient,
free charge carriers will diffuse into the other semiconductor and recombine: electrons will
diffuse into the p-doped region and recombine with the holes of the p-doped region and vice
versa. This diffusion and recombination results in a space-charge region around the junction
as a small region of the p-doped semiconductor is now negatively charged and the n-doped
semiconductor positively.
11.1 Diode 179

Fig. 11.2: From top to bottom: A theoretical pn-junction without electron transfer; a pn-junction with charge
carrier diffusion and space-charge region; an electric field in x-direction and electric potential of the
pn-junction.

These charged regions generate an electric field (see Fig. 11.2). Diffusion takes place as long
as the electric field is not too strong and the potential difference is not too big. For silicon the
diffusion stops at a diffusion voltage between the two regions of about 0.6–0.7 V. The size of
the space-charge region depends on the number of charge carriers that recombine and within
the space-charge region there are no more free charge carriers.
Applying an external voltage to the pn-junction will change the electric potential and the size
of the space-charge region of the pn-junction as the internal and the external potential super-
pose. Depending on the polarity of the external voltage the pn-junction will show a different
behavior.
If the higher potential of the external voltage is applied to the n-type semiconductor, the
internal and the external electric field have the same direction and the electric potentials add
as depicted on the left side of Fig. 11.3. The potential difference at the terminals of the pn-
junction increases and also the space-charge region enlarges. The pn-junction blocks any
current flow.
180 11 Semiconductor devices

Fig. 11.3: An electric potential of a pn-junction with external voltage source.

If the higher potential of the external voltage is applied to the p-type semiconductor the in-
ternal and the external electric fields have the opposite direction. The internal electric poten-
tial is reduced by the external electric potential (right side of Fig. 11.3). The voltage at the
terminals of the pn-junction decreases and also the space-charge region gets smaller. As soon
as the external voltage is greater than the internal voltage, conduction is possible and a cur-
rent can start to flow.
The semiconductor device built out of a pn-junction is called a diode. The two terminals of a
diode are called the anode (p-type semiconductor) and cathode (n-type semiconductor). Fig.
11.4 shows the symbol of a diode with the anode and the cathode. The behavior of a real
semiconductor diode differs slightly from the ideal pn-junction.

Fig. 11.4: Symbol of a diode with anode and cathode.

The characteristic of a diode is depicted in Fig. 11.5. In reverse direction the anode is con-
nected to the lower potential and the diode blocks the current flow almost completely. Due to
small amounts of minority charge carriers that diffuse into the space-charge region a very
small reverse saturation current IS of about some pA or nA can flow in real semiconductor
diodes. This reverse saturation current depends strongly on temperature and on the semicon-
ductor technology. At a high reverse voltage (50–1000 V) the reverse current increases sharp-
ly. This voltage is called the breakdown voltage and depends on the doping concentration,
the semiconductor material and the technology for example. Most diodes should not be oper-
ated in breakdown mode as this operation may destroy the diode. An exception is the Zener
diode (see below).
11.1 Diode 181

Fig. 11.5: A characteristics of a diode.

In the forward direction the anode is connected to the higher potential. For small voltages
(< 0.7 V) only a very small current will flow. For voltages greater than about 0.7 V a signifi-
cant current will start to flow and the current I depends on the voltage across the diode UD in
an exponential manner (ideal Shockley equation):

 UD 
I  I S   e U T  1
 
 
UT is the thermal voltage given by (e is the elementary charge):
k BT
UT 
e
At room temperature the thermal voltage is about 26 mV.
The functionality of the diode corresponds to a valve. In reverse direction any current flow is
(almost completely) blocked. But in the forward direction a current can flow if the applied
voltage is high enough. Based on this functionality diodes are commonly used for any kind
of rectification, or switching. Other applications include light emitting diodes (LED), photo
diodes or voltage protection.
In Fig. 11.6 a schematic cross-section of a vertical diode is shown. A p-doped region is built
up by ion implantation into the n-doped silicon wafer. The boundary of the two regions forms
the pn-junction in the vertical direction. The metallization on top of the p-doped region is the
electric contact for the anode. The other parts of the top surface are coated with SiO2 for
insulation. The bottom surface of the die is also covered with a metallization layer to form
the cathode’s contact.
182 11 Semiconductor devices

Fig. 11.6: Cross-section of a diode.

Several different packages are available for the packaging of the silicon dies of a diode, and
most of these packages are standardized. These packages include cylindrical shape packages
with long wires as well as packages in surface mount and through hole technology (SMD and
THD). Three typical package types for diodes are shown in Fig. 11.7. The small packages
SOD-323 and SC-74 are SMD packages with short pins. The cathode of the SOD-323 pack-
age is marked with a stripe, for the SC-74 package the first pin (out of six pins) is marked
with a dot. The dimensions of these two packages are rather small, just 1.25 mm by 2.5 mm
and a height of 0.9 mm for the SOD-323 and 2.9 mm by 2.5 mm and a height of 1.1 mm for
the SC-74.
The TO-220 is a through-hole device package (THD) for larger die sizes. Packages size is
10.5 mm by 16 mm and a height of 7.7 mm. The pins of this package are 13.6 mm.
Which package is used in an application depends on the power requirements of the applica-
tion, the available space and the assembly technology for example.

Fig. 11.7: Diode packages: SOD-323 SMD package (left), SC-74 SMD package (mid), TO-220 THD package
(right). Package drawings by Infineon Technologies AG.

Application
As the diode blocks the current in one direction it can be used to rectify an AC current as
shown in Fig. 11.8. A sinusoidal input voltage is applied to the circuit of a diode and a resis-
tor. During the negative half of the sinusoidal input voltage the diode blocks the current and
the voltage drop across the resistor is zero. During the positive half the diode conducts if the
11.1 Diode 183

input voltage is greater than about 0.7 V and according to KVL the voltage drop across the
resistor corresponds to:
u R (t )  u i (t )  U D  u I (t )  0.7V

Fig. 11.8: A rectifier circuit with diode and resistor(top); an AC input voltage (bottom left) and a schematic
drawing of rectified voltage at resistor (bottom right)

As depicted in Fig. 11.8 the resistor’s voltage is a periodical function: half of the time it’s a
sinusoidal, the other half zero. The RMS value of the resistor voltage is rather low and hence
the power that is transferred to the resistor. The disadvantage of this kind of rectification is
that half of the period of the input frequency is blocked by the diode. To make use of the total
period a full bridge circuit of four diodes can be used as depicted in Fig. 11.9.
For the positive half of the input voltage’s period, diodes D1 and D4 conduct (if the input
voltage is greater than 2 · 0.7 V) and diodes D2 and D3 block. The current flows via D1
through the resistor and then via D4.The resistor’s voltage has a sinusoidal shape. For the
negative half D1 and D4 block and D2 and D3 conduct. The current flow is D3, resistor, D2 and
it flows again in the same direction through the resistor as in the positive half. In total the
resistor’s voltage is a periodic function again, but the RMS value is higher than in the simple
one way rectifier with just one diode. By adding a capacitor parallel to the resistor the resis-
tor’s voltage can be smoothened after the rectification to get a more DC-like voltage.
184 11 Semiconductor devices

Fig. 11.9: A rectifier circuit with full bridge and resistor(top); an AC input voltage (bottom left) and a schematic
drawing of rectified voltage at resistor.

Another application for diodes is to realize voltage stabilization or overvoltage protection

using a Zener diode. The Zener diode is a special type of diode that is particularly designed
for operation in breakdown mode (special doping and very thin junction). For reverse voltag-
es above the breakdown, or Zener voltage this type of diode is able to conduct high currents
and the voltage drop across the diode stays nearly constant and equal to the Zener voltage. To
achieve a well defined Zener voltage it can be tuned and controlled during the fabrication
process (doping level, size of the very thin pn-junction). Zener voltages may range from of
about 3–100 V. In forward direction the Zener diode behaves like a normal diode.
Fig. 11.10 shows the symbol of a Zener diode and a circuit for overvoltage protection. The
Zener diode is used in parallel to the load resistor. Using this reverse biased Zener diode
limits the voltage across the load resistor to the Zener voltage and hence protects the load
from overvoltage.

Fig. 11.10: The symbol of a Zener diode (left) and circuit for overvoltage protection (right).

Besides purely electrical applications, diodes are also used for optical applications in the
form of LEDs and photo diodes. For LEDs compound semiconductors like AlGaAs or
InGaN are used. The LEDs are forward biased. Electrons from the n-doped region cross the
pn-junction and recombine with the holes in the p-doped region. The energy that is set free
11.1 Diode 185

during the recombination is emitted in form of photons of a dedicated wavelength and hence
color. The luminous flux strongly depends on the current through the LED. Therefore LEDs
are driven by a constant current source. LEDs emit different colors like red, blue or yellow
depending on the semiconductor material. An emission of white light from LEDs is not di-
rectly possible without further optical components. One way to generate white light is to use
a blue LED and to cover it with a photoluminescence material. This material converts the
single color blue into white light. ■

Automotive Application
Diodes are used very often in all kinds of electronic control unit (ECU) in cars, e.g. for recti-
fication, overvoltage and electrostatic discharge (ESD) protection. The use of LEDs ranges
from small signal lights in the interior to high brightness LEDs for headlights. Photo diodes
are used as light sensors. One particular requirement for many automotive ECUs is reverse
polarity protection. Reverse polarity means that the battery is connected in the reverse direc-
tion. This can happen e.g. during maintenance work on the electronic system even though the
connectors are marked with colors or are mechanically different. During reverse polarity
short circuits can occur via elements like internal diodes or transistors. In Fig. 11.11 a simple
ECU is shown with a Zener diode for overvoltage protection. If the voltage is applied in the
correct direction (and is smaller than the Zener voltage) the Zener diode is reverse biased and
the current is limited by the load resistor. If the voltage rises the Zener diode protects the
load by limiting the voltage to the Zener voltage.
In case of reverse applied voltage the Zener diode is forward biased and a short circuit cur-
rent via the Zener diode occurs. This excessive current may damage the ECU.
Reverse polarity protection is required to prevent any damage to an ECU. A simple and
cheap way to realize reverse polarity protection is to insert a diode into the power line of an
ECU as depicted in Fig. 11.11. If the battery is now connected in reverse direction the diode
D prevents any current flow and there is no short circuit via the Zener diode. Hence the addi-
tional diode protects the ECU. If the battery is connected correctly the diode D is forward
biased.
A disadvantage of this solution for reverse polarity protection is the reduction of the voltage
at the ECU by the forward voltage of the additional diode (0.7 V). In addition, the power loss
at this diode reduces the efficiency of the system. The power loss of the diode is:
Pdiode  0.7V  I diode
In case of high currents (e.g. the current in applications like electric power steering, EPS,
might be rather high, at more than 100 A) the power losses can be high and a non-negligible
amount of power is dissipated into heat by the diode. A proper selection of the diode is need-
ed to cope with this heat and to avoid excessive heating of the device, e.g. a package that
provides a good thermal path to conduct the heat from the silicon die to the environment.
186 11 Semiconductor devices

Fig. 11.11: A diode for reverse polarity protection of an ECU.

11.2 Bipolar transistor

The bipolar junction transistor (BJT) is a semiconductor device with two pn-junctions. Two
different types of bipolar junction transistors exist, npn- and pnp-type. This nomenclature
describes the structure of the BJT, e.g. n-doped layer, p-doped layer, n-doped layer like for
the npn-type (see Fig. 11.12). One of the n-doped layers is heavily doped and called the emit-
ter. The other n-doped layer is called the collector and the p-doped layer in between is the
base. For the required functionality (see below) the base has to be very thin. Each of the three
layers is connected to external terminals.

Fig. 11.12: An npn-BJT: layer structure (left); antiparallel diodes (center); circuit symbol (right).

Due to this npn structure there are two antiparallel diodes within the path from collector to
emitter. If a voltage UCE is applied between collector and emitter the base-collector diode
blocks any current flow. If a voltage UBE is additionally applied between base and emitter the
situation changes. As soon as UBE is greater than 0.7 V (and the collector diode is still reverse
biased, UCE > UBE) the pn-diode between base and emitter becomes conductive. A small
current IB starts to flow: holes flow from base to emitter and electrons are emitted from the
emitter towards the base. As the base is very thin, most of the electrons are able to cross the
space-charge region of the base-collector pn-junction (which is still reverse biased). These
electrons form a current IC from the emitter to the collector. Some of the electrons emitted by
the emitter do not cross the base-collector diode, but recombine within the base with the
holes. This recombination would stop any further current flow. To prevent this stopping of
the current, the base current IB removes the electrons. As a consequence the base current IB
can control the collector current IC.
11.2 Bipolar transistor 187

As most of the electrons cross the base into the collector, the collector current is significantly
greater than the base current:
I C  I B
The ratio of the two currents is the current gain:
IC
B
IB

The current gain for a real BJT can be in the range from 4 to 1000. It depends on many tech-
nological and geometrical parameters such as density of donators in the emitter and base, the
size of the base and diffusion parameters.
This is an important functionality of a BJT: an input current (base current IB) controls an
output current (collector current IC) and the output current is the input current amplified by
the current gain. The input current itself is controlled by the base-emitter voltage UBE.

Fig. 11.13: An npn-BJT with external circuit: the base current drives the collector current.

The behavior of a pnp-type BJT is very similar to the npn-type, but the polarities of the ex-
ternal voltages have to be reversed. The structure and symbol of a pnp BJT are given in Fig.
11.14.

Fig. 11.14: A pnp-BJT: layer structure (left) and circuit symbol (right).

The characteristics of the BJT are mainly controlled by the voltages UBE and UCE. In Fig.
11.16 the diode characteristics of the base current is clearly visible. For base-emitter voltages
greater than 0.7 V a small base current IB flows, e.g. in µA range. As the output current IC
depends on the base current its shape is very similar to IB (Fig. 11.16, right). Starting at about
188 11 Semiconductor devices

UBE = 0.7 V (and UCE > UBE) a significant output current starts to flow. Slightly increasing
the base-emitter voltage rises the output current IC sharply. Depending on the current gain of
the BJT the output current is much greater than the control current. In the example given in
Fig. 11.16 the BJT operates in forward mode. The current gain is about 1000 and a control
current in µA range controls the current in the mA range.

Fig. 11.15: Input characteristics of a npn-BJT: base current (control current, left); collector current (output current,
right).

Besides the dependence of UBE the collector current IC also depends on the collector-emitter
voltage UCE. This output characteristics is depicted in Fig. 11.16. For small collector-emitter
voltages up to the saturation voltage UCE, sat the collector current rises sharply. Above the
saturation voltage IC just slightly increases linearly with UCE. Important areas of operating
are the cut-off, forward and saturation regions.
11.2 Bipolar transistor 189

Fig. 11.16: Output characteristics of an npn-BJT: the parameter for the collector current is the base-emitter volt-
age.

In the cut-off region both pn-junctions serve to block and no collector current flows. In this
case UBE is too small (< 0.6 V) to drive a base current. In the output characteristics this oper-
ating mode is a straight horizontal line with IC = 0 A in an ideal case. In reality there will be
small leakage currents. Considering the BJT to be a switch, it is off in this operating mode.
The forward region has already been described in detail above. The emitter diode is forward
biased and the collector diode is reversed biased, UCE > UBE. The collector-emitter voltage is
higher than the saturation voltage UCE,sat. In this operating mode the collector current is given
by the current gain and the base current, IC = B · IB and the BJT acts as an amplifier for a
small current. Small changes in the base current result in large changes in the collector cur-
rent. Fig. 11.17 shows the example already discussed in terms of depending sources. The
base current and the base-emitter voltage are set by resistors R1 and R2 to operate the BJT in
forward mode. By the current amplification of B = 100 the collector current of 1.75 A is
driven by the BJT to light the 21 W-bulb. The collector-emitter voltage is 2 V.
In saturation mode both diodes, emitter and collector diode, are forward biased. In terms of
the circuit in Fig. 11.17 this operating mode can be reached by increasing the base current
(e.g. by changing the resistors R1 and R2): the higher the base current, the higher the collector
current. A higher collector current corresponds to a higher voltage drop across the bulb (re-
sistance of the bulb is about 6.9 Ω). If the base current is increased to 19 mA the voltage drop
across the bulb is 13 V and the collector-emitter voltage of the BJT drops down to 1 V. For a
dedicated base current the saturation voltage UCE,sat of about 0.2 V is reached and both diodes
are forward biased. In this case the collector current does not depend on the base current
anymore and the collector-emitter resistance (= UCE / IC) has its smallest value. Considering
the BJT to be a switch, it is on in this operating mode with smallest resistance value.
190 11 Semiconductor devices

Fig. 11.17: A circuit with a bipolar transistor, the bulb acts as a resistive element with a resistance of 6.9 Ω.

Based on the output characteristics, two major applications for the BJT are amplification and
switching. For amplification the BJT is operated in forward mode as in the example of Fig.
11.17. With a small control current a much higher current is controlled. In the other applica-
tion the BJT is used as a switch. It is operated either in the off, or on mode to switch on and
off a load.
To realize the required functionality the operating point of the BJT has to be set, i.e. the op-
erating voltages UBE and UCE and currents IB and IC. Due to the interdependence of these
values two of these values determine the operating point. In the example above the power
and voltage of the bulb determine the BJT’s operating parameters UCE and IC. With these
values given the other two values UBE and IB were calculated using the BJT’s properties such
as current gain.
In all applications power is dissipated within the BJT due to the two currents, IC and IB. The
total power loss is a sum of the base and the collector losses:
Ptotal  U CE  I C  U BE  I B  U CE  I C
The base loss is much smaller than the collector loss as the base current is much smaller than
the collector current. This electrical power is converted into heat and has to be conducted
from the die to the environment by proper packaging and mounting of the device.
As for the diode the layer structure of a diode is obtained by regions of different doping with-
in a bulk semiconductor. For an npn-BJT a typical layer structure is depicted in Fig. 11.18.
The smaller p- and n-doped regions are implemented within the n-doped bulk semiconductor
by ion implantation. The emitter and base contacts are on the top surface of the die whereas
the collector contact is at the bottom side. Hence the collector current flows in a vertical
direction through the die.
11.2 Bipolar transistor 191

Fig. 11.18: The layer structure of an npn-BJT.

Packages for BJT are manifold and many of these are standardized. Both through-hole devic-
es (THD) and surface mount devices (SMD) are available in different forms. The fitting
device has to be selected depending on application requirements such as build space, mount-
ing technology and electrical and thermal properties. For example the SOT-23 package
(2.9 mm by 2.4 mm) with short pins is significantly smaller than the DPAK package (6.5 mm
by 6.2 mm with a pin length of 3.7 mm). But the maximum collector current for the smaller
package is much smaller than for the bigger package. TO-92 is a THD package with 5.2 mm
by 4.2 mm and a height of 5.2 mm with a pin length of 14.5 mm

Fig. 11.19: Typical packages for BJT: two surface mount devices (SMD), small SOT-23 (left) and DPAK (TO-
252, mid); TO-92 through hole device (THD, right). Package drawings by Infineon Technologies AG.

Automotive application
The use of BJT as current amplifier has already been demonstrated in the example of the
bulb lighting above. The BJT acts as a constant current source to drive the bulb. If the base of
the BJT is driven by a microcontroller the bulb can be switched on and off by the small base
current. Instead of a bulb other loads that require a constant current source, like LEDs can be
connected to this simple constant current source.
If the BJT transistor is used as a switch it is either off (cut-off region), or on (saturation re-
gion). In the on-state the power dissipated within the BJT is rather low as the voltage drop is
just UCE,sat. Seat heating is an application that can be realized with BJT use as a switch. In
this typical convenience application a heating wire is embedded in the seat. As soon as a
current flows through this wire, power is dissipated in the wire. The corresponding heating of
the wire is the required functionality to make the driver feel more comfortable. To switch the
heating wire BJT can be used as shown in Fig. 11.20. A microcontroller controls the switch-
ing of the seat heating. It drives the npn-BJT to operate in the forward region. Thus the small
192 11 Semiconductor devices

output current of the microcontroller is amplified to a much larger current to drive the pnp-
BJT. This BJT operates in saturation mode to drive a rather large current of 5-10 A required
by the heating wire with a low voltage drop UCE,sat and hence minimal power dissipation. The
two resistors are used to set the operating points of the BJT.

Fig. 11.20: BJT in a switching application, e.g. for seat heating.

■

11.3 MOSFET
Like a BJT a MOSFET (metal-oxide-semiconductor field effect transistor) is a semiconduc-
tor device with two pn-junctions. But the structure and the operating principle of a MOSFET
differs significantly from that of a BJT. Like the BJT a MOSFET has three external terminals
called the gate, source and drain as depicted in Fig. 11.21. A fourth connection, the bulk, is
internally connected to the source terminal. Both source and drain are directly contacted to
the semiconductor. But between the gate contact and the semiconductor there is an insulating
layer, in most cases it is silicon oxide. This structure is reflected in the naming of the device,
as it has a metal (gate contact)-oxide (insulator)-semiconductor (MOS) structure to build a
field effect transistor (FET). In modern MOSFETs the metal of the gate contact is replaced
by poly silicon, nevertheless the naming of the device remains. The gate is the switching part
of the device as it controls the current flow from drain to source (or vice versa). Several
types of MOSFET exist but here just the normally-off or enhancement MOSFET will be
introduced. As with BJT (npn- and pnp-type) two different types of enhancement MOSFET
exist: n-type and p-type.
11.3 MOSFET 193

Fig. 11.21: Structure of a lateral n-type MOSFET with the four connections source, drain, gate and bulk (left);
external connections for operation of the MOSFET.

In Fig. 11.21 the basic structure of an n-type MOSFET is depicted. For normal operation a
drain-source voltage UDS > 0 V is applied to the two terminals. As long as this voltage does
not exceed the breakdown voltage of the device (the breakdown voltage depends on the
technology, and is given in the data sheet of the device and should not be exceeded) and the
gate-source voltage UGS is zero, there is no current flowing as the drain-substrate diode is
reverse biased. The gate and the bulk connection form a capacitor that is charged by applying
a charge to it. As the bulk is internally short to the source, the capacitor’s voltage corre-
sponds to the gate-source voltage UGS. If the gate-source voltage rises the electrical field
between gate and bulk (electrical short to source) will attract electrons (minority charge car-
riers in the p-doped substrate) towards the gate. Due to the insulating oxide these electrons
will accumulate beneath the gate. The higher UGS gets the more electrons will be accumulat-
ed. If UGS is sufficiently high, the electrons form an n-type channel beneath the gate from
drain to source. For gate-source voltages above this threshold voltage Uth this n-type channel
enables a current flow from drain to source. The threshold voltage is in the range of 2–3 V
for MOSFET. The size and shape of the n-type channel depends strongly on UGS. The behav-
ior of p-type MOSFETs is similar to the n-type, but the gate source voltage has to be nega-
tive to switch the p-type MOSFET on.

Fig. 11.22: Circuit symbols of an n-type MOSFET (left) and a p-type MOSFET (right).

As can be seen in the structure of a MOSFET there are two antiparallel diodes between the
drain and source contact. As the source is in general short to the bulk (and hence to the sub-
strate), the source-substrate diode has no functionality anymore. In contrast the drain-
substrate diode is functional and forms the intrinsic body diode of a MOSFET. In the sym-
194 11 Semiconductor devices

bols of the MOSFET this body diode is also depicted (see Fig. 11.22). If the device is reverse
biased (UDS < 0 V) it behaves like a diode.
In contrast to the BJT, which is a current controlled device, the MOSFET is a voltage con-
trolled device. The voltages UDS and UGS control the behavior of the MOSFET as depicted in
Fig. 11.23. Four regions of operation can be distinguished in the output characteristics of
MOSFETs.
In the cut-off region the gate-source voltage is smaller than the threshold voltage, UGS < Uth
and the drain-source voltage is forward biased (UDS > 0 V). There is no (or just very small)
drain-source current ID. The MOSFET blocks the current and considering the MOSFET to be
a switch it is off in this operating mode.
In the ohmic region the voltage drop from drain to source is rather small (UDS < UGS – Uth).
The gate-source voltage is above the threshold voltage, UGS > Uth and a conductive n-type
channel is formed. If the gate-source voltage is well above the threshold voltage the drain
current ID is rather independent of UGS but depends in a nearly linear manner from the drain-
source voltage UDS. This behavior corresponds to the behavior of an ohmic resistance. In this
operation mode the MOSFET is switched on and behaves like a resistor with a drain-source
resistance RDS(on). Besides in the cut-off region (no current flow corresponds to no power
loss inside the MOSFET) the power loss of the MOSFET in the ohmic region is lowest.
In saturation mode the drain-source voltage drop is high (UDS > UGS – Uth) and the ID – UDS
characteristics are almost parallel to the UDS axis. Increasing the drain-source voltage has
nearly no effect on the drain current. Instead the drain current can be controlled by the gate-
source voltage. The higher the gate source voltage the higher the current, the MOSFET be-
haves like a voltage controlled current source. In this operation mode high power is dissipat-
ed in the MOSFET due to the high drain-source voltage UDS and high drain current ID.
In the reverse region (UDS < 0 V) the MOSFET behaves like a diode due to the intrinsic body
diode. So in reverse operation the MOSFET does not block a drain current but it starts con-
ducting is the forward voltage of the body diode is exceeded.
11.3 MOSFET 195

Fig. 11.23: Output characteristics of a n-type MOSFET.

Due to the output characteristics, MOSFETs are mainly used in switching applications to act
as a switch. If drain-source voltage is forward biased the MOSFET operates in cut-off and
ohmic mode. In the first mode the resistance of the MOSFET is infinite and the switch is off.
In the ohmic mode it provides a (very low) on-state resistance RDS(on) and the switch is on.
In this mode the power loss is minimal for a conduction state. The on-state resistance for
Power MOSFETs (MOSFETs designed in particular for high power applications) can be less
than 1 mΩ and hence very low. To achieve this low on-state resistance the structure of Power
MOSFETs differ from the structure introduced here. Instead Power MOSFETs have a vertical
trench structure and the drain contact is on the bottom side of the Power MOSFET.
An operation in saturation mode is not desired most of the time. But it cannot be avoided at
least for short times during switching of the device (either on-off or off-on): during switching
the gate capacitance has to be charged (switching on) or uncharged (switching off). During
these switching events the device operates in saturation mode for a short time with signifi-
cant power losses due to the simultaneously occurring drain current and drain source voltage.
The MOSFET is a voltage controlled device and the output is determined by the gate-source
voltage (and the drain-source voltage). If the MOSFET is on or off no current has to be sup-
plied to the gate, just a voltage. To operate a Power MOSFET in on-state a gate-source volt-
age of 5 V (so called logic level MOSFET) or 10 V (standard level MOSFET) has to be ap-
plied. But for switching, the gate capacitor has to be charged or discharged. To keep the
switching time short a suitable gate current has to be provided.
With MOSFETs the power loss is determined by the drain current and the drain-source volt-
age. If used as a switch the total power loss is the sum of the losses during on- and off- state
and during switching:
Ptotal  Poff  Pon  Pswitch
196 11 Semiconductor devices

The power loss in off-state is (nearly) zero and can be neglected. In on-state the MOSFET
acts like a resistor with a resistance RDS(on) and the on-loss is:

Pon  I DS  R DS (on ) 2

The switching losses depend on many device specific parameters and the external operating
conditions and can hardly be estimated in general. A rough estimation shows the dependence
of the power losses of switching time tsw and switching frequency f:
1
Pswitch   I DS  U DS  t sw  f
2
Packages for MOSFETs are manifold and many of these are standardized. Both THD and
SMD packages are available in different forms. Depending on application requirements like
build space, mounting technology and electrical and thermal properties the fitting device has
to be selected. Standard packages for Power MOSFET are DPAK and D2PAK in SMD tech-
nology and TO-220 and TO-262 in THD technology. For small signal MOSFETs also small
packages like SOP-8 or SOT-23 are available.

Automotive application
MOSFETs and in particular Power MOSFET are frequently used in automotive applications,
for example for reverse polarity protection (replacing the diode) or in any kind of switching
application. DC/DC converter is an application where the MOSFET is used as a switch.
The standard automotive supply system on board has a voltage level of 12 V. But many de-
vices, such as microcontroller need another voltage level, e.g. 5 V or 3.3 V. To convert DC
voltages DC/DC converters can be used. A buck converter is a DC/DC converter that gener-
ates a lower output voltage. E.g. it can provide a 5 V output from a 12 V input voltage.
A schematic of a buck converter is depicted in Fig. 11.24. The DC input voltage UE is con-
verted to a lower output voltage UA. It consists of a MOSFET, a diode, an inductor and a
capacitor. The MOSFET switches on and off with a high frequency of some hundred kHz
(e.g. 400 kHz). Using pulse width modulation (PWM) the duty cycle d of the switching can
be adjusted:
t on
d
T
Here T is the period of the switching and ton is the time the MOSFET is switched on.
For the description of the behavior some simplifications can be made: the voltage drop
across the MOSFET in the on-state is neglected (good approximation if a device with low
RDS(on) is used). In addition, the voltage drop across the diode is neglected (this changes the
calculation slightly if the forward voltage of 0.7 V of the diode is taken into account). Also
the current through the inductor is never zero (continuous mode) and a steady state situation
is analyzed.
11.3 MOSFET 197

Fig. 11.24: Schematic of a buck converter.

If the MOSFET is switched on, the diode blocks any current flow and the voltage across the
inductor is according to KVL:
U L  U E U A
During the time the MOSFET is switched on (ton) the current IL through the inductor rises
linearly:
dI L
UL  L
dt
During the off time (toff) of the MOSFET the current keeps on flowing (as it cannot change in
a step function) through the diode and the voltage drop across the inductor is:
U L  U A
Accordingly the current decreases linearly. In steady state operation the rise of inductor cur-
rent during on time equals the decrease during off time:
U E  U A   t on U A  t off
I L  
L L
Using this steady state condition the output voltage can be calculated:
t on
UA UE   d U E
t on  t off

The output voltage just depends on the duty cycle (and the input voltage of course). By mod-
ulation of the duty cycle (that’s why PWM is used) the output voltage can be changed over a
wide range.
Both the output voltage and the inductor current are not constant, but do change with the
PWM frequency as depicted in Fig. 11.25. The capacitor is used to filter the output voltage to
get a more DC-like behavior.
198 11 Semiconductor devices

Fig. 11.25: Signals of a buck converter in continuous mode.

■
12 Circuit simulation

Circuit analysis can be achieved using the techniques introduced so far. Depending on the
circuit under investigation, equation systems can be derived. Whether these equation systems
can be solved at all depends on the complexity of the circuit, the size of the circuit, the ele-
ments used (e.g. linear, non linear) and the problem. Besides analytical calculations by hand,
another way of finding the solution to a given problem of any circuit is circuit simulation.
Simulation in general transforms a complex system into an adequate model representation
and analyzes the model. The result of this analysis is then transferred back to the original
system. Key topics for simulation are the development of a proper model and the usage of
the correct analysis and calculation methods.
In circuit simulation a real system is modeled by a circuit of lumped elements. These models
can be as simple as a linear resistor with just a resistance, or very complex like semiconduc-
tors with parasitic inductances, capacitances, etc. Even for the simple elements the level of
idealization has to be considered, depending on the purpose of the simulation: is a capacitor
just an ideal capacitor, or do parasitic elements like an ESR or a parallel resistor have to be
taken into account? So setting up a suitable representation of the circuit under investigation
is a major task. Once the model is developed the calculations can be made by computer pro-
grams like PSPICE which is introduced here.
Circuit simulation can be used for different purposes. One purpose is visualization: to ob-
serve a general behavior of a circuit, e.g. the frequency response of a two port network. It can
also be very useful for teaching and learning. Another purpose is for supporting circuit de-
sign for determining the behavior of a new circuit, checking for alternative solutions, deter-
mining working points and fitting parameters for the elements used. Or it can be used for
design validation, to prove that a given design behaves as required and specified.
No matter what the purpose of simulation is, two points are always valid: A simulation is not
reality and cannot replace reality, but it can help to improve reality. And a simulation without
knowledge is worthless, or even dangerous.
Most circuit simulation programs are based on the SPICE (Simulation Program with Inte-
grated Circuit Emphasis) software developed by the Electrical Engineering and Computer
Sciences department at the University of California in Berkley in the early 1970s. This soft-
ware can be used for all kinds of DC or AC circuit analysis, time or frequency domain analy-
sis or power analysis. In SPICE the circuit is described in a netlist, and an ASCII text file that
describes the circuit elements and their interconnection. The circuit elements are described
by models, either simple ones such as for a resistor with just a resistance value or more com-
plex ones such as for a MOSFET. The topology of the circuit and its elements determine the
differential equations. Finally the algorithms of the SPICE software are used to solve these
differential equations.
200 12 Circuit simulation

As the SPICE software is an open source software, several companies offer simulation soft-
ware with additional features based on SPICE. Additional features are, for example, a GUI
for the schematic entry of circuits, model editors to create own models or a graphical output
of the simulation data. Examples of simulation software are PSPICE by Cadence Design
Systems, LTspice by Linear Technology or Multisim by National Instruments. The examples
in this chapter use PSPICE as a simulation program. For students a free student version of
PSPICE is available for download and simulation.

PSPICE
The workflow for a simulation with PSPICE is shown in Fig. 12.1. It is split into several
parts: Circuit editors like Capture or Schematics are used to design the circuit in a graphical
way. The schematic of a circuit is designed by drag and drop of models of circuit elements
and the wiring of these elements. The models of circuit elements are stored in libraries in-
cluding a graphical representation and its electrical behavior. Examples of models are any
kind of sources, resistors, capacitors, or semiconductor devices. In addition a model editor
can be used to describe own models if necessary. There is only one mandatory element that
has to be used in all schematics: the ground or reference potential.

Fig. 12.1: Workflow of PSPICE simulation.

Afterwards the graphical schematic can be automatically translated into a netlist that can be
used by the SPICE algorithms for circuit simulation. Different types of analysis are available:
– Bias point: determination of DC operating point;
– DC sweep: variation of a DC parameter in a given range (e.g. voltage source from
0–10 V in steps of 0.1 V);
– AC sweep: variation of operating frequency (e.g. for transfer functions);
– Time domain: simulation of time dependent effects (e.g. transient effects).
The graphical schematic is translated into a netlist that can be used by the SPICE algorithms
for circuit simulation. The results of the simulation are graphically visualized in another
PSPICE module called PSPICE AD. Besides graphs of electrical parameters like currents and
PSPICE 201

voltages, derived parameters like power or any mathematical value can be calculated and
displayed.
A simple AC circuit with a sinusoidal voltage source, a resistor and a diode is used as an
example (see Fig. 12.2). The mandatory ground element is denoted with 0 V. Simple ele-
ments like the ideal AC voltage source, or a resistor are described by a circuit element name
(here V1, R1) and the corresponding parameter like the peak voltage of the source (10 V)
and the frequency (50 Hz) or 2 Ω for the resistor. These parameters can easily be changed
after the model is placed to the schematic. More complex element like the diode D1 in the
circuit use more complex models with given parameters. These models are in general provid-
ed by the manufacturer of the device. Here the diode D1N914 is to be used in the design and
the corresponding model is placed into the schematic. After the models are place the wiring
can be done by just connecting the elements with lines.
After the circuit is completed the simulation setup has to be done. This includes the selection
of simulation type and simulation parameters like simulation time.

Fig. 12.2: A schematic of a simple AC circuit with probe marks in PSPICS capture.

The next step is the generation of the netlist for the simulation. The netlist of the AC circuit
is depicted in Fig. 12.3. The three non-ground elements are listed: in the first column the type
and name of the element, in the next columns (here 2nd and 3rd) the nets, or wires that are
connected to the element are listed. The wiring information is followed by the information
about the parameters of the element, e.g. 2 for the 2 Ω. The parameter information may ex-
tend to the next line like for the voltage source.

1: * source TEST START

2: V_V1 N00603 0
3: +SIN 0 10 50 0 0 0
4: D_D1 N00759 0 D1N914
5: R_R1 N00603 N00759 2

Fig. 12.3: The netlist of the circuit depicted in Fig. 12.2.

202 12 Circuit simulation

Before finally starting the simulation probe marks can be set in the circuit to probe voltages
or currents. The values of these marks are in the end graphically displayed in the simulation
output. The time domain simulation of the example circuit yields the expected behavior (see
Fig. 12.4): The diode blocks the current during the negative half period of the source voltage.
If the source voltage is above about 1 V the diode starts to conduct and the current causes a
voltage drop across the resistor.

Fig. 12.4: Time domain simulation with PSPICE.

The models that are used can be rather simple (e.g. resistor, capacitor) or rather complex (e.g.
diode, transistor). Complex models of dedicated elements that should be used are most of the
time available from the manufacturer of this device. Examples are all kind of semiconductor
devices like bipolar transistors or MOSFETs. Fig. 12.5 shows the electrical PSPICE model of
the NP50N04YUK Power MOSFET by Renesas Electronics. The model is generated to re-
flect the real behavior of the device as well as possible. Besides the basic property of the on-
resistance RDS(on) it takes capacitances like the gate-source capacitance (CGS), and parasitic
resistances like the gate resistance RG or the body diode into account.
PSPICE 203

.SUBCKT NP50N04YUK 1 2 3
**************************************
* Model Generated by Renesas *
* All Rights Reserved *
*Commercial Use or Resale Restricted *
**************************************
* Model generated on December 1, 2012
* MODEL FORMAT: SPICE2G.6
* POWER MOSFET Model (Version 3.1)
* External Node Designations
* Node 1 -> Drain
* Node 2 -> Gate
* Node 3 -> Source
***************************************
M1 4 5 3 3 NMOS W=5198515.2u L=0.4u
DDS 3 1 DDS
CGS 5 3 5.880E-10
RG 2 5 3.57
RD 1 4 RTEMP 0.805264E-3
FGD 1 5 VFGD 1
EVGD 7 0 1 5 1
DDG1 8 7 DD1
DDG2 8 0 DD1
EGD1 9 0 7 8 1
EGD2 10 0 8 0 1
COX 10 11 9.07886E-10
DCRR 11 9 DDG
VFGD 11 0 0
**************************************************************************
.MODEL NMOS NMOS (LEVEL = 3 TOX = 500E-10
+ XJ = 0.14E-06 LD = 0 WD = 0
+ TPG = 1 RS = 0.9E-3 RD = 0.8235604E-3
+ RG = 0 NSUB = 2.811E17 IS = 0
+ UO = 600 KAPPA = 0.006
+ NFS = 0.146E12 THETA = 0.241
+ KP = 2.4061E-5 PHI = 0.87296 VMAX = 1.51E5
+ CGSO = 0 CGDO = 0 CGBO = 0
+ XQC = 1.0 AF = 1 CBD = 0
+ CBS = 0 CJ = 0 CJSW = 0
+ FC = 0.5 JS = 0 KF = 0
+ MJ = 0.5 MJSW = 0.33 PB = 0.8
+ RSH = 0)
*************************************************************************
.MODEL DDS D (CJO=3.06687E-9 VJ=1.542717618 M=1.027746599
+RS=0.001593006 IS=2.543E-12 TT=0.9876E-8 N=1.012594482 BV=40)
*************************************************************************
204 12 Circuit simulation

.MODEL DDG D (CJO=7.93101E-10 VJ=0.483405441 M=0.45294397 IS=1E-32 N=50

FC=1E-08)
*************************************************************************
.MODEL DD1 D (CJO=0 N=1)
*************************************************************************
.MODEL RTEMP RES (TC1=15.825349E-03 TC2=4.84641E-05)
*************************************************************************
.ENDS NP50N04YUK

Fig. 12.5: PSPICE model of a Power MOSFET NP50N04YUK by Renesas Electronics.

Automotive Application
Like for any design of electronic systems circuit simulation is carried out to a large extent for
automotive applications. As with all circuit simulations it serves as a design support tool, a
virtual circuit prototyping and circuit validation tool. This speeds up the design phase and
increases the quality of the design as many elements can be tested in advance. Besides the
simulation of electrical properties circuit simulation programs can perform thermal simula-
tions to some extent. In particular for power electronics this additional feature can be very
helpful in finding suitable designs and solutions. ■
References

Albach, Manfred, Elektrotechnik, Pearson Studium, 2011

Bobrow, Leonard S., Fundamentals of Electrical Engineering, 2nd edition, Oxford University Press,
1996
Dorf, Richard C., Circuits, Signals, and Speech and Image Processing, 3rd edition, CRC Press, Taylor &
Francis Group, 2006
Dorf, Richard C., Electronics, Power Electronics, Optoelectronics, Microwaves, Electromagnetics, and
Radar, 3rd edition, CRC Press, Taylor & Francis Group, 2006
Hagmann, Gerd, Grundlagen der Elektrotechnik, 13th edition, AULA-Verlag, 2008
Morris, Alan S. and Langari, Reza, Measurement and instrumentation: theory and application, 1st edi-
tion, Elsevier, 2012
Rizzoni, Giorgio, Fundamentals of Electrical Engineering, 1st edition, McGraw-Hill, 2009
Index
AC analysis 124 damping factor 171
AC power 161 DC analysis 51
acceleration sensor 147 DC/DC converter 196
active filter 160 decibel 152
active power 162 dependent source 47
alternating current 125 determinant 57
alternator 59 die 177
ampmeter 43 dielectric 10, 91
analog to digital converter 44, 87 differential capacitor 147
angular frequency 125 differentiator 95
anode 180 diode 69, 178, 180
anti-aliasing filter 155 displacement field 10
apparent power 165 donor 5
arithmetical mean 125 drain 192
band gap 2 efficiency 45
band-pass filter 173 electric circuit 15
bandwidth 173 electric field 7
base 186 electric field strength 7
bipolar transistor 49, 186 electric flux 11
Bode plot 152 electric flux density 11
body diode 194 electric potential 9
branch 16 electric relay 102
breakdown voltage 180 electrical power 44
building block 149 electrolytic capacitor 97
capacitance 91 electronic control unit 87
capacitor 91 elementary positive charge 1
cathode 180 emitter 186
ceramic capacitor 97 energy band 1
circuit simulation 199 equivalent series resistance 96
collector 186 Euler’s formula 135
comparator 83 Faraday's law 99
complex conjugate number 136 film capacitor 97
complex number 134 filter 149
conductance 21 first order circuit 103
conduction band 2, 3 forward region 189
consumer system 16 Fourier analysis 126
Coulomb’s law 7 gate 192
Cramer’s rule 57 generator system 16
critically damped case 119, 123 ground 51
current 12 HEV/EV V, 19, 98
current density 13 higher order filter 155
current divider 36 high-pass filter 150
current gain 187 homogeneous ODE 104, 114
current probe 43 independent mesh 64
current source 19 inductor 99
cut-off frequency 152, 173 inhomogeneous ODE 108
cut-off region 189, 194 Insulator 3
208 Index

internal resistance 27 RC circuit 103, 107, 133, 143, 144, 145, 154,
inverting amplifier 84 157, 160, 164
keyless entry system 175 reactive power 163
Kirchhoff’s current law 33 real inductor 101
Kirchhoff’s law 140 real operational amplifier 82
Kirchhoff’s voltage law 36 rectified value 127
leakage current 97 rectifier 182
LED 184 recuperation 98
load 15 resistance 21
low-pass filter 153 resistance strain gauge 23
lumped element model 15 resistor 21
mass action law 4 resonance frequency 173
Maxwell’s equation 9, 12, 92 resonant frequency 158
mechanical harmonic oscillator 113 reverse polarity protection 185
mesh 16 reverse region 194
mesh analysis 60 RL circuit 111, 141, 150
Metal 2 RLC circuit 112, 120, 123, 146, 169, 175
modified nodal analysis 55 RMS power 162
MOSFET 192, 202 root-mean square (RMS) 127
natural angular frequeny 169 saturation mode 189, 194
netlist 201 seat heating 191
nodal analysis 51 second order ODE 112
node 16 semiconductor 177
non-inverting amplifier 86 Semiconductor 3
Norton’s theorem 75 Shannon-Nyquist criterion 155
NTC 22 short circuit 26
ohmic region 194 SMD 191
Ohm's law 21 source 15, 192
open circuit 26 SPICE 199
operating point 39, 190 supercap 97
operational amplifier 80 superposition 65
oscilloscope 44 switching 102
overdamped case 116, 122 THD 191
parasitic resistor 97 Thevenin’s equivalent 70
peak value 125 underdamped case 117, 123
phase angle 126 unity gain buffer 88
phase difference 126 valence band 2, 3
phasor 137 vector diagram 129
pn-junction 178 voltage 8
pointer diagram 130 voltage divider 38
power factor 165 voltage source 16
power factor correction 166 voltage stabilization 184
PSPICE 200 voltmeter 43
PTC 22 wafer 177
pulse width modulation (PWM) 196 Wheatstone bridge 23, 147
quality factor 171 Wye-Delta transformation 41
Zener diode 180, 184