Electronic and Circuits

48520
Electronics and
Circuits
Lecture Notes
2013
1 1H i (t )
vs 1F 1H
+5 V
3 7 20 k
3 k 3 k
8
REF
G=100 AD620B 6 IN
499
3 k 3 k 1 5 DIGITAL
10 k ADC DATA
2 4 OUTPUT
AD705 AGND
20 k
PMcL
i
Contents
LECTURE 1A BASIC LAWS & OP-AMP AMPLIFIERS
INTRODUCTION ....................................................................................................1A.1
CURRENT.............................................................................................................1A.2
VOLTAGE ............................................................................................................1A.3
CIRCUIT ELEMENTS AND TYPES OF CIRCUITS......................................................1A.4
ACTIVE CIRCUIT ELEMENTS ...........................................................................1A.4
PASSIVE CIRCUIT ELEMENTS ..........................................................................1A.4
TYPES OF CIRCUITS ........................................................................................1A.4
INDEPENDENT SOURCES ......................................................................................1A.5
THE INDEPENDENT VOLTAGE SOURCE ...........................................................1A.5
THE INDEPENDENT CURRENT SOURCE............................................................1A.7
THE RESISTOR AND OHMS LAW .........................................................................1A.8
THE SHORT-CIRCUIT ....................................................................................1A.11
THE OPEN-CIRCUIT ......................................................................................1A.12
CONDUCTANCE .............................................................................................1A.12
PRACTICAL RESISTORS ......................................................................................1A.13
PREFERRED VALUES AND THE DECADE PROGRESSION .................................1A.14
THE E SERIES VALUES ...............................................................................1A.14
MARKING CODES..........................................................................................1A.16
KIRCHHOFFS CURRENT LAW............................................................................1A.19
KIRCHHOFFS VOLTAGE LAW ...........................................................................1A.23
COMBINING RESISTORS .....................................................................................1A.26
SERIES RESISTORS ........................................................................................1A.26
PARALLEL RESISTORS ..................................................................................1A.27
COMBINING INDEPENDENT SOURCES ................................................................1A.30
COMBINING INDEPENDENT VOLTAGE SOURCES IN SERIES ...........................1A.30
COMBINING INDEPENDENT CURRENT SOURCES IN PARALLEL ......................1A.32
THE VOLTAGE DIVIDER RULE ...........................................................................1A.34
THE CURRENT DIVIDER RULE ...........................................................................1A.36
DEPENDENT SOURCES .......................................................................................1A.38
THE DEPENDENT VOLTAGE SOURCE ............................................................1A.38
THE DEPENDENT CURRENT SOURCE.............................................................1A.40
POWER ..............................................................................................................1A.42
POWER ABSORBED IN A RESISTOR ................................................................1A.48
AMPLIFIERS .......................................................................................................1A.49
UNITS OF GAIN .............................................................................................1A.50
AMPLIFIER POWER SUPPLIES ........................................................................1A.52
SATURATION ................................................................................................1A.53
CIRCUIT MODEL ...........................................................................................1A.54
THE OPERATIONAL AMPLIFIER..........................................................................1A.55
FEEDBACK ....................................................................................................1A.56
CIRCUIT MODEL ...........................................................................................1A.57
THE IDEAL OP-AMP ......................................................................................1A.58
OP-AMP FABRICATION AND PACKAGING ......................................................1A.60
NEGATIVE FEEDBACK .......................................................................................1A.61
NEGATIVE FEEDBACK IN ELECTRONICS ........................................................1A.62
AN AMPLIFIER WITH NEGATIVE FEEDBACK .................................................1A.63
Electronics and Circuits 2013

ii
THE NONINVERTING AMPLIFIER ....................................................................... 1A.66
THE NONINVERTING AMPLIFIER WITH AN IDEAL OP-AMP ........................... 1A.69
INPUT RESISTANCE OF THE NONINVERTING AMPLIFIER ............................... 1A.71
EQUIVALENT CIRCUIT OF THE NONINVERTING AMPLIFIER .......................... 1A.71
THE BUFFER................................................................................................. 1A.74
THE INVERTING AMPLIFIER .............................................................................. 1A.75
INPUT RESISTANCE OF THE INVERTING AMPLIFIER ...................................... 1A.77
EQUIVALENT CIRCUIT OF THE INVERTING AMPLIFIER.................................. 1A.77
SUMMARY ........................................................................................................ 1A.79
EXERCISES ........................................................................................................ 1A.85
LECTURE 1B NODAL AND MESH ANALYSIS

INTRODUCTION ................................................................................................... 1B.1
NODAL ANALYSIS .............................................................................................. 1B.2
CIRCUITS WITH RESISTORS AND INDEPENDENT CURRENT SOURCES ONLY .... 1B.5
NODAL ANALYSIS USING BRANCH ELEMENT STAMPS................................... 1B.8
CIRCUITS WITH VOLTAGE SOURCES ............................................................. 1B.11
CIRCUITS WITH DEPENDENT SOURCES ......................................................... 1B.13
SUMMARY OF NODAL ANALYSIS ................................................................. 1B.16
MESH ANALYSIS ............................................................................................... 1B.19
PLANAR CIRCUITS........................................................................................ 1B.19
PATHS, LOOPS AND MESHES ........................................................................ 1B.20
MESH CURRENT ........................................................................................... 1B.21
MESH ANALYSIS METHODOLOGY ................................................................ 1B.22
CIRCUITS WITH RESISTORS AND INDEPENDENT VOLTAGE SOURCES ONLY .. 1B.23
CIRCUITS WITH CURRENT SOURCES ............................................................. 1B.25
CIRCUITS WITH DEPENDENT SOURCES ......................................................... 1B.27
SUMMARY OF MESH ANALYSIS ................................................................... 1B.29
SUMMARY ........................................................................................................ 1B.30
EXERCISES ........................................................................................................ 1B.31
GUSTAV ROBERT KIRCHHOFF (1824-1887) ...................................................... 1B.34
LECTURE 2A CIRCUIT ANALYSIS TECHNIQUES

INTRODUCTION ................................................................................................... 2A.1
LINEARITY .......................................................................................................... 2A.2
SUPERPOSITION .................................................................................................. 2A.3
SUPERPOSITION THEOREM ............................................................................. 2A.5
SOURCE TRANSFORMATIONS .............................................................................. 2A.9
PRACTICAL VOLTAGE SOURCES..................................................................... 2A.9
PRACTICAL CURRENT SOURCES ................................................................... 2A.11
PRACTICAL SOURCE EQUIVALENCE ............................................................. 2A.13
MAXIMUM POWER TRANSFER THEOREM ..................................................... 2A.15
THVENINS AND NORTONS THEOREM ............................................................ 2A.17
SUMMARY OF FINDING THVENIN EQUIVALENT CIRCUITS .......................... 2A.25
SUMMARY ........................................................................................................ 2A.29
EXERCISES ........................................................................................................ 2A.30

iii
LECTURE 2B LINEAR OP-AMP APPLICATIONS
INTRODUCTION .................................................................................................... 2B.1
SUMMING AMPLIFIER .......................................................................................... 2B.2
DIFFERENCE AMPLIFIER ...................................................................................... 2B.5
INVERTING INTEGRATOR ..................................................................................... 2B.8
DIFFERENTIATOR............................................................................................... 2B.10
NEGATIVE IMPEDANCE CONVERTER ................................................................. 2B.11
VOLTAGE-TO-CURRENT CONVERTER ................................................................ 2B.13
NONINVERTING INTEGRATOR ............................................................................ 2B.15
SUMMARY ......................................................................................................... 2B.17
EXERCISES ........................................................................................................ 2B.21
LECTURE 3A REACTIVE COMPONENTS

INTRODUCTION ....................................................................................................3A.1
THE CAPACITOR ..................................................................................................3A.2
CAPACITOR v-i RELATIONSHIPS ......................................................................3A.4
ENERGY STORED IN A CAPACITOR ..................................................................3A.6
SUMMARY OF IMPORTANT CAPACITOR CHARACTERISTICS ............................3A.9
THE INDUCTOR ..................................................................................................3A.10
INDUCTOR v-i RELATIONSHIPS ......................................................................3A.13
ENERGY STORED IN AN INDUCTOR ...............................................................3A.18
SUMMARY OF IMPORTANT INDUCTOR CHARACTERISTICS ............................3A.21
PRACTICAL CAPACITORS AND INDUCTORS ........................................................3A.22
CAPACITORS .................................................................................................3A.22
INDUCTORS ...................................................................................................3A.24
SERIES AND PARALLEL CONNECTIONS OF INDUCTORS AND CAPACITORS .........3A.26
INDUCTORS ...................................................................................................3A.26
CAPACITORS .................................................................................................3A.29
CIRCUIT ANALYSIS WITH INDUCTORS AND CAPACITORS ...................................3A.31
DC CIRCUITS ................................................................................................3A.31
NODAL AND MESH ANALYSIS ......................................................................3A.33
DUALITY ...........................................................................................................3A.35
SUMMARY .........................................................................................................3A.39
EXERCISES ........................................................................................................3A.41
LECTURE 4A SOURCE-FREE RC AND RL CIRCUITS

INTRODUCTION ....................................................................................................4A.1
DIFFERENTIAL OPERATORS .................................................................................4A.2
PROPERTIES OF DIFFERENTIAL OPERATORS ........................................................4A.4
THE CHARACTERISTIC EQUATION .......................................................................4A.8
THE SIMPLE RC CIRCUIT ...................................................................................4A.12
PROPERTIES OF THE EXPONENTIAL RESPONSE...................................................4A.15
SINGLE TIME CONSTANT RC CIRCUITS .............................................................4A.18
THE SIMPLE RL CIRCUIT ...................................................................................4A.20
SINGLE TIME CONSTANT RL CIRCUITS..............................................................4A.23
SUMMARY .........................................................................................................4A.27
EXERCISES ........................................................................................................4A.28

iv
LECTURE 5A FIRST-ORDER STEP RESPONSE
INTRODUCTION ................................................................................................... 5A.1
THE UNIT-STEP FORCING FUNCTION .................................................................. 5A.2
THE DRIVEN RC CIRCUIT ................................................................................... 5A.6
THE FORCED AND THE NATURAL RESPONSE..................................................... 5A.10
FINDING A PARTICULAR SOLUTION USING THE INVERSE DIFFERENTIAL
OPERATOR ................................................................................................... 5A.12
FINDING A PARTICULAR SOLUTION BY INSPECTION ..................................... 5A.14
FINDING A PARTICULAR SOLUTION USING AN INTEGRATING FACTOR.......... 5A.15
STEP-RESPONSE OF RC CIRCUITS ..................................................................... 5A.18
ANALYSIS PROCEDURE FOR SINGLE TIME CONSTANT RC CIRCUITS ................ 5A.28
RL CIRCUITS ..................................................................................................... 5A.29
ANALYSIS PROCEDURE FOR SINGLE TIME CONSTANT RL CIRCUITS ................. 5A.31
SUMMARY ........................................................................................................ 5A.32
EXERCISES ........................................................................................................ 5A.33
LEONHARD EULER (1707-1783) ....................................................................... 5A.37
LECTURE 6A THE PHASOR CONCEPT

INTRODUCTION ................................................................................................... 6A.1
SINUSOIDAL SIGNALS ......................................................................................... 6A.3
SINUSOIDAL STEADY-STATE RESPONSE ............................................................. 6A.5
THE COMPLEX FORCING FUNCTION .................................................................. 6A.11
THE PHASOR ..................................................................................................... 6A.17
FORMALISATION OF THE RELATIONSHIP BETWEEN PHASOR AND SINUSOID ...... 6A.20
GRAPHICAL ILLUSTRATION OF THE RELATIONSHIP BETWEEN A PHASOR
AND ITS CORRESPONDING SINUSOID ............................................................ 6A.21
PHASOR RELATIONSHIPS FOR R, L AND C ......................................................... 6A.22
PHASOR RELATIONSHIPS FOR A RESISTOR ................................................... 6A.22
PHASOR RELATIONSHIPS FOR AN INDUCTOR ................................................ 6A.24
PHASOR RELATIONSHIPS FOR A CAPACITOR ................................................ 6A.26
SUMMARY OF PHASOR RELATIONSHIPS FOR R, L AND C .............................. 6A.28
ANALYSIS USING PHASOR RELATIONSHIPS ...................................................... 6A.29
IMPEDANCE ...................................................................................................... 6A.30
ADMITTANCE .................................................................................................... 6A.35
SUMMARY ........................................................................................................ 6A.37
EXERCISES ........................................................................................................ 6A.38
JOSEPH FOURIER (1768-1830) .......................................................................... 6A.43

v
LECTURE 6B CIRCUIT SIMULATION
INTRODUCTION .................................................................................................... 6B.1
PROJECT FLOW .................................................................................................... 6B.2
STARTING A NEW PROJECT ............................................................................. 6B.2
DRAWING THE SCHEMATIC ............................................................................. 6B.3
SIMULATION ................................................................................................... 6B.3
SCHEMATIC CAPTURE ......................................................................................... 6B.4
GROUND ......................................................................................................... 6B.4
SI UNIT PREFIXES ........................................................................................... 6B.5
ALL PARTS MUST HAVE UNIQUE NAMES ....................................................... 6B.5
LABELING NODES ........................................................................................... 6B.6
SIMULATION ........................................................................................................ 6B.7
DC BIAS ......................................................................................................... 6B.7
TIME-DOMAIN (TRANSIENT) SIMULATIONS .................................................... 6B.7
AC SWEEP / NOISE SIMULATIONS ................................................................. 6B.10
EXERCISES ........................................................................................................ 6B.15
LECTURE 7A SINUSOIDAL STEADY-STATE RESPONSE

INTRODUCTION ....................................................................................................7A.1
ANALYSIS USING PHASORS ..................................................................................7A.2
NODAL ANALYSIS ...............................................................................................7A.3
MESH ANALYSIS .................................................................................................7A.5
SUPERPOSITION ...................................................................................................7A.7
THVENINS THEOREM ........................................................................................7A.8
NORTONS THEOREM ........................................................................................7A.10
PHASOR DIAGRAMS ...........................................................................................7A.12
POWER IN THE SINUSOIDAL STEADY-STATE......................................................7A.20
INSTANTANEOUS POWER ..............................................................................7A.20
AVERAGE POWER .........................................................................................7A.21
ROOT-MEAN-SQUARE (RMS) VALUES ........................................................7A.23
RMS VALUE OF A SINUSOID .........................................................................7A.24
PHASORS AND RMS VALUES ........................................................................7A.25
AVERAGE POWER USING RMS VALUES .......................................................7A.26
APPARENT POWER ........................................................................................7A.26
POWER FACTOR ............................................................................................7A.26
COMPLEX POWER .........................................................................................7A.27
REACTIVE POWER.........................................................................................7A.28
SUMMARY OF POWER IN AC CIRCUITS .........................................................7A.30
SUMMARY .........................................................................................................7A.33
EXERCISES ........................................................................................................7A.34
OLIVER HEAVISIDE (1850-1925).......................................................................7A.40

vi
LECTURE 8B FIRST-ORDER OP-AMP FILTERS
INTRODUCTION ................................................................................................... 8B.1
BILINEAR FREQUENCY RESPONSES ..................................................................... 8B.2
FREQUENCY RESPONSE REPRESENTATION.......................................................... 8B.4
MAGNITUDE RESPONSES .................................................................................... 8B.6
PHASE RESPONSES ............................................................................................ 8B.10
SUMMARY OF BILINEAR FREQUENCY RESPONSES........................................ 8B.14
BODE PLOTS ..................................................................................................... 8B.15
FREQUENCY AND MAGNITUDE SCALING .......................................................... 8B.16
FREQUENCY SCALING (DENORMALISING) .................................................... 8B.17
MAGNITUDE SCALING.................................................................................. 8B.18
CASCADING CIRCUITS ...................................................................................... 8B.19
INVERTING BILINEAR OP-AMP CIRCUIT ........................................................... 8B.20
INVERTING OP-AMP CIRCUITS .......................................................................... 8B.22
CASCADE DESIGN ............................................................................................. 8B.23
EXERCISES ........................................................................................................ 8B.26
LECTURE 9A SECOND-ORDER STEP RESPONSE

INTRODUCTION ................................................................................................... 9A.1
SOLUTION OF THE HOMOGENEOUS LINEAR DIFFERENTIAL EQUATION ............... 9A.2
DISTINCT REAL ROOTS .................................................................................. 9A.3
REPEATED REAL ROOTS ................................................................................ 9A.4
ONLY REAL ROOTS ........................................................................................ 9A.5
DISTINCT COMPLEX ROOTS ........................................................................... 9A.7
REPEATED COMPLEX ROOTS.......................................................................... 9A.9
THE SOURCE-FREE PARALLEL RLC CIRCUIT .................................................... 9A.10
THE OVERDAMPED PARALLEL RLC CIRCUIT.................................................... 9A.13
THE CRITICALLY DAMPED PARALLEL RLC CIRCUIT ........................................ 9A.17
THE UNDERDAMPED PARALLEL RLC CIRCUIT ................................................. 9A.21
RESPONSE COMPARISON ................................................................................... 9A.25
THE SOURCE-FREE SERIES RLC CIRCUIT ......................................................... 9A.26
COMPLETE RESPONSE OF THE RLC CIRCUIT ..................................................... 9A.27
FORCED RESPONSE ...................................................................................... 9A.28
NATURAL RESPONSE.................................................................................... 9A.29
CASE I OVERDAMPED .............................................................................. 9A.31
CASE II CRITICALLY DAMPED ................................................................. 9A.33
CASE III UNDERDAMPED ......................................................................... 9A.34
MAXIMUM VALUE AND PEAK TIME ............................................................. 9A.36
SUMMARY ........................................................................................................ 9A.38
EXERCISES ........................................................................................................ 9A.39
WILLIAM THOMSON (LORD KELVIN) (1824-1907) ........................................... 9A.43

vii
LECTURE 10A SECOND-ORDER FREQUENCY RESPONSE
INTRODUCTION ..................................................................................................10A.1
RESONANCE ......................................................................................................10A.2
PARALLEL RESONANCE .....................................................................................10A.4
PHASOR DIAGRAM OF THE PARALLEL RLC CIRCUIT ....................................10A.7
QUALITY FACTOR .............................................................................................10A.8
SECOND-ORDER CIRCUIT RELATIONS .............................................................10A.11
BANDWIDTH ....................................................................................................10A.12
SERIES RESONANCE.........................................................................................10A.16
OTHER RESONANT FORMS ..............................................................................10A.18
THE SECOND-ORDER LOWPASS FREQUENCY RESPONSE .................................10A.24
PEAK FREQUENCY ......................................................................................10A.26
BANDWIDTH ...............................................................................................10A.28
CASE I RELATIVE PEAK > 2 ( Q0 1 1 2 ) ....................................10A.29
CASE II RELATIVE PEAK < 2 (1 2 Q0 1 1 2 ) .......................10A.31
CASE III NO RELATIVE PEAK ( Q0 1 2 )..............................................10A.32
BODE PLOTS ...............................................................................................10A.33
THE SECOND-ORDER HIGHPASS FREQUENCY RESPONSE ................................10A.34
PEAK FREQUENCY ......................................................................................10A.36
BANDWIDTH ...............................................................................................10A.36
STANDARD FORMS OF SECOND-ORDER FREQUENCY RESPONSES....................10A.37
SUMMARY .......................................................................................................10A.38
EXERCISES ......................................................................................................10A.39
JAMES CLERK MAXWELL (1831-1879) ...........................................................10A.43
LECTURE 10B SECOND-ORDER OP-AMP FILTERS

INTRODUCTION .................................................................................................. 10B.1
FILTER DESIGN PARAMETERS ........................................................................... 10B.1
THE LOWPASS BIQUAD CIRCUIT ....................................................................... 10B.3
THE UNIVERSAL BIQUAD CIRCUIT .................................................................... 10B.8
APPROXIMATING THE IDEAL LOWPASS FILTER ............................................... 10B.10
THE BUTTERWORTH LOWPASS FILTER ............................................................ 10B.12
SUMMARY ....................................................................................................... 10B.15
EXERCISES ...................................................................................................... 10B.16
LECTURE 11A COMPLEX FREQUENCY

INTRODUCTION ..................................................................................................11A.1
COMPLEX FREQUENCY ......................................................................................11A.2
THE DAMPED SINUSOIDAL FORCING FUNCTION ................................................11A.6
GENERALIZED IMPEDANCE AND ADMITTANCE ................................................11A.10
FREQUENCY RESPONSE AS A FUNCTION OF ................................................11A.13
FREQUENCY RESPONSE AS A FUNCTION OF ................................................11A.17
THE COMPLEX-FREQUENCY PLANE ................................................................11A.22
VISUALIZATION OF THE FREQUENCY RESPONSE FROM A POLE-ZERO PLOT ....11A.28
SUMMARY .......................................................................................................11A.31
EXERCISES ......................................................................................................11A.32

viii
LECTURE 11B SPECIALTY AMPLIFIERS
INTRODUCTION ................................................................................................. 11B.1
DIFFERENTIAL AND COMMON-MODE SIGNALS ................................................. 11B.2
DIFFERENCE AMPLIFIERS.................................................................................. 11B.3
DIFFERENCE AMPLIFIER DEFICIENCIES ........................................................ 11B.5
DIFFERENCE AMPLIFIER ICS ........................................................................ 11B.5
INSTRUMENTATION AMPLIFIERS ....................................................................... 11B.6
IN-AMP APPLICATION .................................................................................. 11B.8
PROGRAMMABLE GAIN AMPLIFIERS ................................................................. 11B.9
PGA DESIGN ISSUES .................................................................................. 11B.10
PGA EXAMPLE .......................................................................................... 11B.10
ISOLATION AMPLIFIERS .................................................................................. 11B.11
SUMMARY ...................................................................................................... 11B.14
LECTURE 12A TRANSFER FUNCTIONS

INTRODUCTION ................................................................................................. 12A.1
TRANSFER FUNCTIONS...................................................................................... 12A.2
CHARACTERISTIC EQUATION ....................................................................... 12A.3
POLE-ZERO PLOT ......................................................................................... 12A.3
TRANSFER FUNCTION FORM ........................................................................ 12A.4
RELATIONSHIP TO DIFFERENTIAL EQUATION ............................................... 12A.5
CIRCUIT ABSTRACTION ................................................................................ 12A.6
FORCED RESPONSE ........................................................................................... 12A.7
FREQUENCY RESPONSE................................................................................... 12A.11
NATURAL RESPONSE ...................................................................................... 12A.14
COMPLETE RESPONSE ..................................................................................... 12A.19
SUMMARY ...................................................................................................... 12A.22
EXERCISES ...................................................................................................... 12A.24
PIERRE SIMON DE LAPLACE (1749-1827) ....................................................... 12A.27
LECTURE 12B SENSOR SIGNAL CONDITIONING

INTRODUCTION ................................................................................................. 12B.1
SENSORS ...................................................................................................... 12B.2
PROCESS CONTROL SYSTEMS ...................................................................... 12B.3
PROGRAMMABLE LOGIC CONTROLLERS ...................................................... 12B.4
SMART TRANSDUCERS ................................................................................. 12B.5
PROGRAMMABLE AUTOMATION CONTROLLERS .......................................... 12B.6
BRIDGE CIRCUITS ............................................................................................. 12B.7
BRIDGE DESIGN ISSUES.............................................................................. 12B.11
AMPLIFYING AND LINEARIZING BRIDGE OUTPUTS .................................... 12B.12
DRIVING REMOTE BRIDGES ....................................................................... 12B.15
INTEGRATED BRIDGE TRANSDUCERS ......................................................... 12B.18
STRAIN, FORCE, PRESSURE AND FLOW MEASUREMENTS ................................ 12B.19
HIGH IMPEDANCE SENSORS ............................................................................ 12B.21
TEMPERATURE SENSORS................................................................................. 12B.22
SUMMARY ...................................................................................................... 12B.24
EXERCISES ...................................................................................................... 12B.25

ix
LECTURE 13A SYSTEM MODELLING
INTRODUCTION ..................................................................................................13A.1
DIFFERENTIAL EQUATIONS OF PHYSICAL SYSTEMS ..........................................13A.2
LINEAR APPROXIMATIONS OF PHYSICAL SYSTEMS ...........................................13A.4
THE TRANSFER FUNCTION ................................................................................13A.7
BLOCK DIAGRAMS ............................................................................................13A.8
FEEDBACK .......................................................................................................13A.15
SUMMARY .......................................................................................................13A.18
EXERCISES ......................................................................................................13A.19
LECTURE 13B REVISION
MATRICES - QUICK REFERENCE GUIDE
TEXTBOOK ANSWERS
ANSWERS

1A.1
Lecture 1A Basic Laws & Op-Amp Amplifiers

Current. Voltage. Circuit elements and types of circuits. Independent sources.
The resistor and Ohms Law. Practical resistors. Kirchhoffs Current Law
(KCL). Kirchhoffs Voltage Law (KVL). Combining resistors. Combining
independent sources. The voltage divider rule. The current divider rule.
Dependent sources. Power. Amplifiers. The operational amplifier. Negative
feedback. The noninverting amplifier. The inverting amplifier.
Introduction
Electric circuit theory and electromagnetic theory are the two fundamental
theories upon which all branches of electrical engineering are built. Many
branches of electrical engineering, such as power, electric machines, control,
electronics, communications, and instrumentation, are based on electric circuit
theory. Circuit theory is also valuable to students specializing in other branches
of the physical sciences because circuits are a good model for the study of
energy systems in general, and because of the applied mathematics, physics,
and topology involved.
Electronic circuits are used extensively in the modern world society in its
present form could not exist without them! They are used in communication
systems (such as televisions, telephones, and the Internet), digital systems (such
as personal computers, embedded microcontrollers, smart phones), and
industrial systems (such as robotic and process control systems). The study of
electronics is therefore critical to electrical engineering and related professions.
One goal in this subject is to learn various analytical techniques and computer
software applications for describing the behaviour of electric circuits. Another
goal is to study various uses and applications of electronic circuits.
We will start by revising some basic concepts, such as KVL, KCL and Ohms
Law. We will then introduce the concept of the electronic amplifier, and then
study a device called an operational amplifier (op-amp for short), which has
been used as the building block for modern analog electronic circuitry since its
invention in the 1960s.

1A.2
Current
Charge in motion represents a current. The current present in a discrete path,
such as a metallic wire, has both a magnitude and a direction associated with it
it is a measure of the rate at which charge is moving past a given reference
point in a specified direction. Current is symbolised by i and thus:
Current defined as
dq
i
the rate of change of
charge moving past (1A.1)
a reference dt
The unit of current is the ampere (A) and is equivalent to Cs -1 . In a circuit

current is represented by an arrow:
Representation of
current in a circuit
i
Figure 1A.1
The arrow does not indicate the actual direction of charge flow, but is simply
part of a convention that allows us to talk about the current in an unambiguous
manner.
The use of terms such as a current flows through the resistor is a tautology
Correct usage of the
term current and should not be used, since this is saying a a charge flow flows through the
resistor. The correct way to describe such a situation is there is a current in
the resistor.
A current which is constant is termed a direct current, or simply DC. Examples

DC and AC defined of direct currents are those that exist in circuits with a chemical battery as the
source. A sinusoidal current is often referred to as alternating current, or AC1.
Alternating current is found in the normal household electricity supply.
1
Later we shall also see that a periodic current (e.g. a square wave), with no DC term, can also
be referred to as an alternating current.

1A.3
Voltage
A voltage exists between two points in a circuit when energy is required to
move a charge between the two points. The unit of voltage is the volt (V) and
is equivalent to JC -1 . In a circuit, voltage is represented by a pair of +/- signs:
Representation of
voltage in a circuit
A
Figure 1A.2
Once again, the plus-minus pair does not indicate the actual voltage polarity.
Example
Note the voltages across the circuit elements below:
A A
v = -5 V v=5V
B B
(a) (b)
In both (a) and (b), terminal B is 5 V positive with respect to terminal A.
A A
v=5V v = -5 V
B B
(c) (d)
In both (c) and (d), terminal A is 5 V positive with respect to terminal B.

1A.4
Circuit Elements and Types of Circuits

A circuit element is an idealised mathematical model of a two-terminal
electrical device that is completely characterised by its voltage-current
Ideal circuit relationship. Although ideal circuit elements are not off-the-shelf circuit
elements are used components, their importance lies in the fact that they can be interconnected
to model real circuit
elements (on paper or on a computer) to approximate actual circuits that are composed
of nonideal elements and assorted electrical components thus allowing for the
analysis of such circuits.
Circuit elements can be categorised as either active or passive.
Active Circuit Elements
Active circuit elements can deliver a non-zero average power indefinitely.

Active circuit
element defined There are four types of active circuit element, and all of them are termed an
ideal source. They are:
the independent voltage source
the independent current source
the dependent voltage source
the dependent current source
Passive Circuit Elements
Passive circuit elements cannot deliver a non-zero average power indefinitely.

Passive circuit Some passive elements are capable of storing energy, and therefore delivering
element defined
power back into a circuit at some later time, but they cannot do so indefinitely.
There are three types of passive circuit element. They are:

the resistor
the inductor
the capacitor
Types of Circuits
The interconnection of two or more circuit elements forms an electrical

Network and circuit network. If the network contains at least one closed path, it is also an electrical
defined
circuit. A network that contains at least one active element, i.e. an independent
Active and passive or dependent source, is an active network. A network that does not contain any
circuits defined
active elements is a passive network.

1A.5
Independent Sources
Independent sources are ideal circuit elements that possess a voltage or current
value that is independent of the behaviour of the circuits to which they belong.
The Independent Voltage Source
An independent voltage source is characterised by a terminal voltage which is

completely independent of the current through it. The representation of an Independent voltage
source defined
independent voltage source is shown below:
vs
Figure 1A.3
If the value of the voltage source is constant, that is, does not change with time, An ideal battery is
then we can also represent it as an ideal battery: equivalent to an
independent voltage
source that has a
constant value
Vs Vs
Figure 1A.4
Although a real battery is not ideal, there are many circumstances under
which an ideal battery is a very good approximation.

1A.6
In general, however, the voltage produced by an ideal voltage source will be a
function of time. In this case we represent the voltage symbolically as vt .
A few typical voltage waveforms are shown below. The waveforms in (a) and
(b) are typical-looking amplitude modulation (AM) and frequency modulation
(FM) signals, respectively. Both types of signals are used in consumer radio
communications. The sinusoid shown in (c) has a wide variety of uses; for
example, this is the shape of ordinary household voltage. A pulse train, such
as that in (d), can be used to drive DC motors at a variable speed.
v v
t t
(a) (b)
v v
t
(c) (d)
Figure 1A.5
Since the voltage produced by a source is in general a function of time, then the
most general representation of an ideal voltage source is as shown below:
The most general
representation of an
ideal independent
voltage source
vs ( t ) vs ( t )
"intuitive" AS 1102
IEC 60617
Figure 1A.6

1A.7
The Independent Current Source
An independent current source establishes a current which is independent of Independent current

the voltage across it. The representation of an independent current source is source defined
shown below:
Terminal 1
is
Terminal 2
Figure 1A.7
In other words, an ideal current source is a device that, when connected to

anything, will always push i s out of terminal 1 and pull i s into terminal 2.
Since the current produced by a source is in general a function of time, then the
most general representation of an ideal current source is as shown below:
The most general
representation of an
ideal independent
current source
is ( t ) is( t )
"intuitive" AS 1102
IEC 60617
Figure 1A.8

1A.8
The Resistor and Ohms Law

In 1827 the German physicist George Ohm published a pamphlet entitled The
Galvanic Circuit Investigated Mathematically. It contained one of the first
efforts to measure currents and voltages and to describe and relate them
mathematically. One result was a statement of the fundamental relationship we
now call Ohms Law.
Consider a uniform cylinder of conducting material, to which a voltage has

been connected. The voltage will cause charge to flow, i.e. a current:
A simple resistor
i
A
v
v conductor l

Figure 3A.1
Ohm found that in many conducting materials, such as metal, the current is
always proportional to the voltage. Since voltage and current are directly
proportional, there exists a proportionality constant R, called resistance, such
that:
v Ri
Ohms Law
(1A.2)
This is Ohms Law. The unit of resistance (volts per ampere) is referred to as
the ohm, and is denoted by the capital Greek letter omega, .
We refer to a construction in which Ohms Law is obeyed as a resistor.

1A.9
The ideal resistor relationship is a straight line through the origin:

The resistor is a
linear circuit element
R
1
Figure 1A.9
Even though resistance is defined as R v i , it should be noted that R is a

purely geometric property, and depends only on the conductor shape and the
material used in the construction. For example, it can be shown for a uniform
resistor that the resistance is given by:
l The resistance of a
R (1A.3)
uniform resistor
A
where l is the length of the resistor, and A is the cross-sectional area. The
resistivity, , is a constant of the conducting material used to make the
resistor.
The circuit symbol for the resistor is shown below, together with the direction
of current and polarity of voltage that makes Ohms Law algebraically correct:
The circuit symbol

for the resistor
v R
Figure 1A.10

1A.10
Example
Consider the circuit shown below.
i1
10 V v 1 k
i2
The voltage across the 1 k resistor is, by definition of an ideal voltage source,
vt 10 V . Thus, by Ohms Law, we get:
v 10
i1 0.01 A 10 mA
R 1000
and:
v 10
i2 0.01 A 10 mA
R 1000
Note that i2 i1 , as expected.
Example
Consider the circuit shown below.
i (t )
3cos( t ) A v( t ) 50
Ohms Law yields:
vt Ri t
50 3 cost
150 cost V

1A.11
The Short-Circuit
Consider a resistor whose value is zero ohms. An equivalent representation of

such a resistance, called a short-circuit, is shown below:
The short-circuit
i( t ) i( t )
arbitrary arbitrary
v( t ) R= 0 v( t )
circuit circuit
Figure 1A.11
By Ohms Law:
v Ri
0i
0V (1A.4)
Thus, no matter what finite value it has, vt will be zero. Hence, we see that
a zero-ohm resistor is equivalent to an ideal voltage source whose value is zero
volts, provided that the current through it is finite.

1A.12
The Open-Circuit
Consider a resistor having infinite resistance. An equivalent representation of

such a resistance, called an open-circuit, is shown below:
The open-circuit
i i
arbitrary arbitrary
v R = v
circuit circuit
Figure 1A.12
By Ohms Law:
v
i
R
v

0A (1A.5)
Thus, no matter what finite value vt has, it will be zero. Thus, we may
conclude that an infinite resistance is equivalent to an ideal current source
whose value is zero amperes, provided that the voltage across it is finite.
Conductance
The reciprocal of resistance is called the conductance, G:
Conductance
1
defined
G (1A.6)
R
The unit of conductance is the siemen, and is abbreviated S. The same circuit
symbol is used to represent both resistance and conductance.

1A.13
Practical Resistors
There are many different types of resistor construction. Some are shown below:
Some types of
resistors
carbon composition carbon film metal film
wire wound wire wound array

with heat sink
chip array
chip - thick film chip - thin film
Figure 1A.13 Some types of resistors
The through-hole resistors are used by hobbyists and for prototyping real
designs. Their material and construction dictate several of their properties, such
as accuracy, stability and pulse handling capability.
The wire wound resistors are made for accuracy, stability and high power
applications. The array is used where space is a premium and is normally used
in digital logic designs where the use of pull-up resistors is required.
Modern electronics utilises surface-mount components. There are two

varieties of surface-mount chip resistor thick film and thin film.

1A.14
Preferred Values and the Decade Progression
Fundamental standardization practices require the selection of preferred values

within the ranges available. Standard values may at first sight seem to be
strangely numbered. There is, however, a beautiful logic behind them, dictated
by the tolerance ranges available.
The decade progression of preferred values is based on preferred numbers

Component values generated by a geometric progression, repeated in succeeding decades. In 1963,
have been
standardized by the the International Electrotechnical Commission (IEC) standardized the preferred
IEC
number series for resistors and capacitors (standard IEC 60063). It is based on
the fact that we can linearly space values along a logarithmic scale so a
percentage change of a value results in a linear change on the logarithmic scale.
Component values For example, if 6 values per decade are desired, the common ratio is
are spaced
equidistantly on a 6
10 1.468 . The six rounded-off values become 100, 150, 220, 330, 470, 680.
logarithmic scale
The E Series Values
The IEC set the number of values for resistors (and capacitors) per decade
based on their tolerance. These tolerances are 0.5%, 1%, 2%, 5%, 10%, 20%
and 40% and are respectively known as the E192, E96, E48, E24, E12, E6 and
The E series E3 series, the number indicating the quantity of values per decade in that
values explained
series. For example, if resistors have a tolerance of 5%, a series of 24 values
can be assigned to a single decade multiple (e.g. 100 to 999) knowing that the
possible extreme values of each resistor overlap the extreme values of adjacent
resistors in the same series.
Any of the numbers in a series can be applied to any decade multiple set. Thus,
for instance, multiplying 220 by each decade multiple (0.1, 1, 10 100, 1000
etc.) produces values of 22, 220, 2 200, 22 000, 220 000 etc.
The E series of preferred resistor and capacitor values according to

IEC 60063 are reproduced in Table 1A.1.

1A.15
0.5% 1% 2% 0.5% 1% 2% 0.5% 1% 2% 0.5% 1% 2% 0.5% 1% 2%

E192 E96 E48 E192 E96 E48 E192 E96 E48 E192 E96 E48 E192 E96 E48
100 100 100 169 169 169 287 287 287 487 487 487 825 825 825
101 172 291 493 835
102 102 174 174 294 294 499 499 845 845
104 176 298 505 856
105 105 105 178 178 178 301 301 301 511 511 511 866 866 866
106 180 305 517 876
107 107 182 182 309 309 523 523 887 887
109 184 312 530 898
110 110 110 187 187 187 316 316 316 536 536 536 909 909 909
111 189 320 542 920
113 113 191 191 324 324 549 549 931 931
114 196 328 556 942
115 115 115 196 196 196 332 332 332 562 562 562 953 953 953
117 198 336 569 965
118 118 200 200 340 340 576 576 976 976
120 203 344 583 988
121 121 121 205 205 205 348 348 348 590 590 590
123 208 352 597 5% 10% 20% 40%
124 124 210 210 357 357 604 604 E24 E12 E6 E3
126 213 361 612
127 127 127 215 215 215 365 365 365 619 619 619 100 100 100 100
129 218 370 626 110
130 130 221 221 374 374 634 634 120 120
132 223 379 642 130
133 133 133 226 226 226 383 383 383 649 649 649 150 150 150
135 229 388 657 160
137 137 232 232 392 392 665 665 180 180
138 234 397 673 200
140 140 140 237 237 237 402 402 402 681 681 681 220 220 220 220
142 240 407 690 240
143 143 243 243 412 412 698 698 270 270
145 246 417 706 300
147 147 147 249 249 249 422 422 422 715 715 715 330 330 330
149 252 427 723 360
150 150 255 255 432 432 732 732 390 390
152 259 437 741 430
154 154 154 261 261 261 442 442 442 750 750 750 470 470 470 470
156 264 448 759 510
158 158 267 267 453 453 768 768 560 560
160 271 459 777 620
162 162 162 274 274 274 464 464 464 787 787 787 680 680 680
164 277 470 796 750
165 165 280 280 475 475 806 806 820 820
167 284 481 816 910
Table 1A.1 - IEC standard E series of values in a decade

1A.16
Marking Codes
The IEC also defines how manufacturers should mark the values of resistors
and capacitors in the standard called IEC 60062. The colours used on fixed
leaded resistors are shown below:
IEC labelling for
leaded resistors
4 bands 22 , 5%
5 bands 2200 , 1%
6 bands 100 k , 0.1%, 15 ppm
silver 0.01 10%

gold 0.1 5%
black 0 0 0 0 1 200 ppm
brown 1 1 1 1 10 1% 100 ppm
red 2 2 2 2 100 2% 50 ppm
orange 3 3 3 3 1k 15 ppm
yellow 4 4 4 4 10k 25 ppm
green 5 5 5 5 100k 0.5%

blue 6 6 6 6 1M 0.25% 10 ppm
violet 7 7 7 7 10M 0.1% 5 ppm
grey 8 8 8 8 1 ppm
white 9 9 9 9 Tolerance
Multiplier Temperature
Significant Figures Coefficient
Figure 1A.14 Colour code marking of leaded resistors
The resistance colour code consists of three or four colour bands and is
followed by a band representing the tolerance. The temperature coefficient
band, if provided, is to the right of the tolerance band and is usually a wide
band positioned on the end cap.

1A.17
The resistance colour code includes the first two or three significant figures of
the resistance value (in ohms), followed by a multiplier. This is a factor by
which the significant-figure value must be multiplied to find the actual
resistance value. (i.e. the number of zeros to be added after the significant
figures).
Whether two or three significant figures are represented depends on the

tolerance: 5% and wider require two band; 2% and tighter requires three
bands. The significant figures refer to the first two or three digits of the
resistance value of the standard series of values in a decade, in accordance with
IEC 60063 as indicated in the relevant data sheets and shown in Table 1A.1.
The colours used and their basic numerical meanings are recognized
internationally for any colour coding used in electronics, not just resistors, but
some capacitors, diodes, cabling and other items.
The colours are easy to remember: Black is the absence of any colour, and
therefore represents the absence of any quantity, 0. White (light) is made up of
all colours, and so represents the largest number, 9. In between, we have the The resistor colour
code explained
colours of the rainbow: red, orange, yellow, green, blue and violet. These take
up the numbers from 2 to 7. A colour in between black and red would be
brown, which has the number 1. A colour intermediate to violet and white is
grey, which represents the number 8.

1A.18
When resistors are labelled in diagrams, such as schematics, IEC 60062 calls
for the significant figures to be printed as such, but the decimal point is
replaced with the SI prefix of the multiplier. Examples of such labelling are
shown below:
IEC labelling for

diagrams Resistor Value IEC Labelling
0.1 0R1
1 1R0
22 22R
3.3 k 3K3
100 k 100K
4.5 M 4M5
Note how the decimal point is expressed, that the ohm symbol is shown as an
R, and that 1000 is shown as a capital K. The use of a letter instead of a
decimal point solves a printing problem the decimal point in a number may
not always be printed clearly, and the alternative display method is intended to
We use a letter in
place of a decimal help misinterpretation of component values in circuit diagrams and parts lists.
point for labelling
component values
In circuit diagrams and constructional charts, a resistors numerical identity, or
designator, is usually prefixed by R. For example, R15 simply means resistor
number 15.
A portion of a schematic diagram showing designators and IEC labelling is

shown below:
Figure 1A.15 Schematic portion showing IEC labelling
Note that resistor R4 has the value 4.7 and resistor R12 has the value 330 .

1A.19
Kirchhoffs Current Law

A connection of two or more elements is called a node. An example of a node
is depicted in the partial circuit shown below:
R1 R2
v R3
R4
i
Figure 1A.16
Even if the figure is redrawn to make it appear that there may be more than one
node, as in the figure below, the connection of the six elements actually
constitutes only one node.
R2
R1
v R3
node
R4 i
Figure 1A.17

1A.20
Kirchhoffs Current Law (KCL) is essentially the law of conservation of

electric charge. If currents directed out of a node are positive in sense, and
currents directed into a node are negative in sense (or vice versa), then KCL
can be stated as follows:
KCL defined
KCL: At any node of a circuit, the
(1A.7)
currents algebraically sum to zero.
If there are n elements attached to a node then, in symbols, KCL is:
i
k 1
k 0 (1A.8)
KCL can also be stated as: The sum of the currents entering a node is equal to
the sum of the currents leaving a node.
Example
As an example of KCL, consider a portion of some circuit, shown below:
i1 R3 i6
i2 i5
i3 i4
Choosing the positive sense to be leaving, we apply KCL at the node and
obtain the equation:
i1 i2 i3 i4 i5 i6 0
Note that even if one of the elements the one which carries i3 is a short-
circuit, KCL holds. In other words, KCL applies regardless of the nature of the
elements in the circuit.

1A.21
Example
We want to find the voltage v, in the two-node circuit shown below:
i1 i2 i3
13 A 1 v 2 2A 3
The directions of i1 , i2 , i3 and the polarity of v were chosen arbitrarily (the

directions of the 13 A and 2 A sources are given). By KCL (at either of the two
nodes), we have:
13 i1 i2 2 i3 0
From this we can write:
i1 i2 i3 11
By Ohms Law:
v v v
i1 i2 i3
1 2 3
Substituting these into the previous equation yields:
v v v
11
1 2 3
v v
v 11
2 3
6v 3v 2v
11
6
11v
11
6
v 6V
Having solved for v, we can now find that:
v 6 v 6 v 6
i1 6A i2 3 A i3 2A
1 1 2 2 3 3

1A.22
Just as KCL applies to any node of a circuit, so must KCL hold for any closed
region, i.e. to satisfy the physical law of conservation of charge, the total
current leaving (or entering) a region must be zero.
Example
In the circuit shown below, three regions have been identified:
i5
i1 i2
Region 3
a i b
Region 1 i3 Region 2 i4
Applying KCL to Region 1, we get:
i0
For Region 2:
i1 i3 i4 i2
For Region 3:
i 2 i5 i 4
You may now ask, Since there is no current from point a to point b (or vice
versa) why is the connection (a short-circuit) between the points there? If the
connection between the two points is removed, two separate circuits result. The
voltages and currents within each individual circuit remain the same as before.
Having the connection present constrains points a and b to be the same node,
and hence be at the same voltage. It also indicates that the two separate
portions are physically connected (even though there is no current between
them).

1A.23
Kirchhoffs Voltage Law

Starting at any node in a circuit, we form a loop by traversing through elements
Loop defined
(open-circuits included!) and returning to the starting node, never encountering
any other node more than once.
For example, the paths fabef and fdcef are loops:
a 1 b 2 c
v1 v2
6 v6 v7 3 3V
v3
v8
f e 4 d
v4
2A
5
v5
Figure 1A.18
whereas the paths becba and fde are not:
a 1 b 2 c
v1 v2
3
6 v6 v7 3V
v3
v8
f e 4 d
v4
2A
5
v5
Figure 1A.19

1A.24
Kirchhoffs Voltage Law (KVL) is essentially the law of conservation of
energy. If voltages across elements traversed from to + are positive in sense,
and voltages across elements that are traversed from + to are negative in
sense (or vice versa), then KVL can be stated as follows:
KVL defined
KVL: Around any loop in a circuit, the
(1A.9)
voltages algebraically sum to zero.
If there are n elements in the loop then, in symbols, KVL is:
v
k 1
k 0 (1A.10)
KVL can also be stated as: In traversing a loop, the sum of the voltage rises
equals the sum of the voltage drops.
Example
In the circuit shown in Figure 1A.18, we select a traversal from to + to be

positive in sense. Then KVL around the loop abcefa gives:
v1 v 2 v3 v8 v6 0
and around loop bcdeb, we have:
v 2 3 v 4 v7 0
In this last loop, one of the elements traversed (the element between nodes b
and e) is an open-circuit; however, KVL holds regardless of the nature of the
elements in the circuit.

1A.25
Example
We want to find the current i, in the one-loop circuit shown below:
2 i
v1 34 V
10 V 4 v2
6
v3
The polarities of v1 , v2 , v3 and the direction of i were chosen arbitrarily (the

polarities of the 10 V and 34 V sources are given). Applying KVL we get:
10 v1 34 v 2 v3 0
Thus:
v1 v 2 v3 24
From Ohms Law:
v1 2i v 2 4i v3 6i
Substituting these into the previous equation yields:
2i 4i 6i 24
2i 4i 6i 24
12i 24
i 2 A
Having solved for i, we now find that:
v1 2i 2 2 4 V
v 2 4i 4 2 8 V
v3 6i 6 2 12 V

1A.26
Combining Resistors
Relatively complicated resistor combinations can be replaced by a single
equivalent resistor whenever we are not specifically interested in the current,
voltage or power associated with any of the individual resistors.
Series Resistors
Consider the series combination of N resistors shown in (a) below:
i R1 R2 RN i
arbitrary v1 v2 vN arbitrary
circuit v circuit v R eq
(a) (b)
Figure 1A.20
We apply KVL:
v v1 v 2 v N (1A.11)
and Ohms Law:
v R1i R2i RN i
R1 R2 RN i (1A.12)
and then compare this result with the simple equation applying to the
equivalent circuit shown in Figure 1A.20b:
v Req i (1A.13)
Thus, the value of the equivalent resistance for N series resistances is:
Combining series
resistors Req R1 R2 RN (series)
(1A.14)

1A.27
Parallel Resistors
A similar simplification can be applied to parallel resistors. Consider the

parallel combination of N conductances shown in (a) below:
i
i
i1 i2 iN
arbitrary
circuit v arbitrary
G1 G2 GN circuit v Geq
(a) (b)
Figure 1A.21
We apply KCL:
i i1 i2 iN (1A.15)
and Ohms Law:
i G1v G2v GN v
G1 G2 GN v (1A.16)
whereas the equivalent circuit shown in Figure 1A.21b gives:
i Geq v (1A.17)
and thus the value of the equivalent conductance for N parallel conductances is:
Geq G1 G2 GN (1A.18) Combining parallel

(parallel) conductances

1A.28
In terms of resistance instead of conductance:
1 1 1 1
Combining parallel
resistors (parallel) (1A.19)
Req R1 R2 RN
The special case of only two parallel resistors is needed often:
Combining two R1R2

resistors in Req (parallel) (1A.20)
parallel R1 R2
Note that since Geq G1 G2 then we may deduce that:
Geq G1 and Geq G2 (1A.21)
Hence:
1 1 1 1
and (1A.22)
Req R1 Req R2
or:
results in an
equivalent
resistance smaller Req R1 and Req R2 (1A.23)
than either resistor
Thus the equivalent resistance of two resistors in parallel is less than the value
of either of the two resistors.
The special case of N resistors of equal value R in parallel is:
Combining the same

R
valued resistors in
parallel
Req (parallel) (1A.24)
N

1A.29
Example
We want to find the current i in the circuit below:
i 5 1
28 V 4 3 v
In order to find i, we can replace series and parallel connections of resistors by

their equivalent resistances. We begin by noting that the 1 and 3 resistors
are in series. Combining them we obtain:
i 5
28 V 4 4
Note that it is not possible to display the original voltage v in this figure. Since
the two 4 resistors are connected in parallel, we can further simplify the
circuit as shown below:
i 5
28 V 2
Here, the 5 and 3 resistors are in series, so we may combine them into
one 7 resistor. Then, from Ohms Law, we have:
28
i 4A
7

1A.30
Combining Independent Sources

An inspection of the KVL equations for a series circuit shows that the order in
which elements are placed in a series circuit makes no difference. An
inspection of the KCL equations for a parallel circuit shows that the order in
which elements are placed in a parallel circuit makes no difference. We can use
these facts to simplify voltage sources in series and current sources in parallel.
Combining Independent Voltage Sources in Series
It is not possible to combine independent voltage sources in parallel, since this

would violate KVL. However, consider the series connection of two ideal
voltage sources shown in (a) below:
a a
v2
arbitrary arbitrary
v circuit veq v circuit
v1
b b
(a) (b)
Figure 1A.22
From KVL we know that v v1 v 2 , and by the definition of an ideal voltage

source, this must be the voltage between nodes a and b, regardless of what is
connected to them. Thus, the series connection of two ideal voltage sources is
equivalent to a single independent voltage source given by:
Combining
independent voltage
sources in series veq v1 v2 (series) (1A.25)
Clearly, the obvious generalization to N voltage sources in series holds.

1A.31
Example
In a previous example we determined the current i in the one-loop circuit

shown below:
2 i
v1 34 V
10 V 4 v2
6
v3
By rearranging the order in this one loop circuit (of course this does not
affect i), we obtain the circuit shown below:
i 2
34 V v1
10 V 4 v2
6
v3
We can now combine the series independent voltage sources and the series
resistors into single equivalent elements:
-24 V v 12
By Ohms Law:
24
i 2 A
12

1A.32
Combining Independent Current Sources in Parallel
It is not possible to combine independent current sources in series, since this

would violate KCL. However, consider the parallel connection of two ideal
current sources shown in (a) below:
i a i a
i1 i2 arbitrary i eq arbitrary
circuit circuit
b b
(a) (b)
Figure 1A.23
From KCL we find that i i1 i2 , and by the definition of an ideal current

source, this must always be the current into the arbitrary circuit. Thus, the
parallel connection of two ideal current sources is equivalent to a single
independent current source given by:
Combining
independent current ieq i1 i2 (parallel) (1A.26)
sources in parallel
Clearly, the obvious generalization to N current sources in parallel holds.

1A.33
Example
In a previous example, we determined the voltage v in the two-node circuit

shown below:
13 A 1 v 2 2A 3
Combining the parallel independent current sources into a single equivalent

source, we obtain the circuit:
11 A 1 v 2 3
Since the equivalent resistance of the three resistors in parallel is given by:
1 1 1 1 6 3 2 11

Req 1 2 3 6 6
we obtain:
6
Req A
11
Then, from Ohms Law:
6
v 11 6 V
11

1A.34
The Voltage Divider Rule

It can be quite useful to determine how a voltage appearing across two series
resistors divides between them. Consider the circuit shown below:
i R1
arbitrary
v1
circuit v R2 v2
Figure 1A.24
By Ohms Law, the current in the resistors is:
v
i (1A.27)
R1 R2
By application of Ohms Law again, the voltage across R1 is:
v1 R1i (1A.28)
and therefore:
Voltage divider rule

R1
defined v1 v (1A.29)
R1 R2
Similarly, the voltage across R2 is:
R2
v2 v (1A.30)
R1 R2
These equations describe how the voltage is divided between the resistors.
Because of this, a pair of resistors in series is often called a voltage divider.

1A.35
Example
We want to find the voltage v in the circuit below:
i 5 1
28 V 4 v1 3 v
Combining the series connection of the 1 and 3 resistors, we obtain the

circuit below:
i 5
28 V 4 v1 4
Now the pair of 4 resistors in parallel can be combined as shown below:
i 5
28 V 2 v1
By voltage division:
2 56
v1 28 8V
25 7
Returning to the original circuit and applying voltage division again yields:
3 3
v v1 8 6 V
3 1 4

1A.36
The Current Divider Rule

It can be quite useful to determine how a current entering two parallel resistors
divides between them. Consider the circuit shown below:
i
i1 i2
arbitrary
circuit
v
R1 R2
Figure 1A.25
We replace the parallel connection of R1 and R2 by its equivalent resistance.

Thus, Ohms Law gives:
R1 R2
v Req i i (1A.31)
R1 R2
By application of Ohms Law again, the current in R1 is i1 v R1 and thus:
Current divider rule

R2
defined i1 i (1A.32)
R1 R2
Similarly, the current in R2 is:
R1
i2 i (1A.33)
R1 R2
These equations describe how the current is divided between the resistors.
Because of this, a pair of resistors in parallel is often called a current divider.
Note that a larger amount of current will exist in the smaller resistor thus
current tends to take the path of least resistance!

1A.37
Example
We want to find the current i in the circuit below:
i1 4
i
36 sin(100 t ) V 6 3
The total current delivered by the source is:
36 sin 100t
i1
4 63 6 3
6 sin 100t A
Therefore the desired current is:
6 2
i i1 6 sin 100t 4 sin 100t V
63 3
The current divider rule can also be derived using conductances. Referring to
Figure 1A.25, the voltage across the parallel resistors is:
i i
v Req i (1A.34)
Geq G1 G2
The current in resistor R1 is i1 G1v and thus:
G1
i1 i (1A.35)
G1 G2
A similar result obviously holds for current i2 . The advantage of this form of
the current divider rule is that it is the dual of the voltage divider rule we
replace voltages with currents, and resistors with conductances.

1A.38
Dependent Sources
An ideal source, either voltage or current, whose value depends upon some
parameter (usually a voltage or current) in the circuit to which the source
belongs is known as a dependent or controlled source.
The Dependent Voltage Source
A dependent voltage source establishes a voltage across its terminals,

independent of the current through it, which is determined by the voltage or
Dependent voltage
sources defined current at some other location in the electrical system. There are two types of
dependent voltage source the voltage-controlled voltage source (VCVS) and
the current-controlled voltage source (CCVS).
Kvx rm i x
VCVS CCVS
Figure 1A.26
Note that the dependent source is represented by a diamond-shaped symbol so

as not to confuse it with an independent source.
These sources are mathematical models that are useful in modelling real
circuits and systems, e.g. they are used in modelling operational amplifiers.

1A.39
Example
Consider the circuit shown below. This circuit contains a dependent source
whose value in this case depends on the voltage across the 4 resistor it is a
VCVS.
3v2
i 2
v1
12 V 4 v2
To analyse the circuit, we apply KVL and obtain:
v1 3v 2 v 2 12
or:
v1 2v 2 12
By Ohms Law:
v1 2i and v 2 4i
Therefore:
2i 24i 12
2i 8i 12
6i 12
i 2 A
Hence:
v 2 4i 8 V
and the value of the dependent voltage source is:
3v 2 24 V

1A.40
The Dependent Current Source
A dependent current source establishes a current, which is independent of the

Dependent current voltage across it, that is determined by the voltage or current at some other
sources defined
location in the electrical system. There are two types of dependent current
source the voltage-controlled current source (VCCS) and the current-
controlled current source (CCCS).
Ki x gm vx
CCCS VCCS
Figure 1A.27
These sources are mathematical models that are useful in modelling real
circuits and systems, e.g. they are used in modelling transistors.

1A.41
Example
Consider the circuit shown below. In this circuit the value of the dependent
current source is specified by a voltage it is a VCCS.
i1 i2
2A 3 v 4v 5
To solve for v, we apply KCL and obtain:
i1 i2 4v 2
Thus:
v v
4v 2
3 5
8v
4v 2
15
52v
2
15
30 15
v V
52 26
Consequently:
30
4v A
13
and this is the value of the dependent current source, in amperes. The other
variables in the circuit are:
v 5 v 3
i1 A and i2 A
3 26 5 26

1A.42
Power
Power is the rate at which work is done or energy is expended. Taking the
product of voltage (energy per unit charge) and current (charge per unit time)
we get a quantity that measures energy per unit time. Its for this reason that we
define p , the instantaneous power absorbed by an electrical circuit element, to
be the product of voltage and current:
Instantaneous
power defined p vi (1A.36)
The fundamental unit of power is the watt (W) and is equivalent to Js -1 . In

using the formula given in Eq. (1A.36), we need to be careful in establishing
the correct voltage polarity and current direction. Consider the circuit element:
A i
Figure 1A.28
If one terminal of the element (A) is v volts positive with respect to the other
Passive sign terminal (B), and if a current i is entering the element through terminal A, then
convention defined
it gives power a power p vi is being absorbed or delivered to the element. When the
absorbed by a
circuit element current arrow is directed into the element at the plus-marked terminal, we
satisfy the passive sign convention. If the numerical value of the power using
this convention is negative, then we say that the element is generating or
delivering power.

1A.43
Example
Note the power in the circuit elements below:
3A
2V
p = (2)(3) = 6 W absorbed
-2 V
-3 A
p = (-2)(-3) = 6 W absorbed
-5 A
4V
p = (4)(-5) = -20 W absorbed

(20 W generated)

1A.44
Example
Consider the circuit shown below:
i( t )
325cos(100 t ) V 100
By Ohms Law:
vt 325
i t cos100t A
R 100
By definition, the power absorbed by the resistor is:
p R t vt it
325 2
cos 2 100t
100
1056 cos 2 100t W
In particular, at time t 0 the power absorbed by the resistor is:
p R 0 1056 cos 2 0
1056 W
At time t 5 ms , however, since:

cos 100 5 10 3 cos 0
2
then the resistor absorbs 0 watts.

1A.45
Example
i
i1 i2 i3
5V 1 2 9A 3
We shall determine the power absorbed in each of the elements.
Note that the voltage across each of the elements is 5 V since all the elements
are in parallel. Therefore, by Ohms Law:
5 5 5
i1 5A i2 A i3 A
1 2 3
and the powers absorbed in the 1 , 2 and 3 resistors are:
p1 5i1 55 25 W
5 25
p 2 5i2 5 W
2 2
5 25
p3 5i3 5 W
3 3
respectively, for a total of:
25 25 150 75 50 275
25 W
2 3 6 6
absorbed by the resistors.

1A.46
By KCL:
5 5 5
i 9 i1 i2 i3
1 2 3
or:
30 15 10 1
i 9 A
6 6
Thus the power delivered by the voltage source is:
5
p v 5i W
6
Also, the power delivered by the current source is:
pi 59 45 W
Hence the total power delivered by the sources is:
5 270 5 275
45 W
6 6 6
We see that the total power delivered by the sources is equal to the total power
absorbed by the resistors. Since power delivered by a circuit element is equal to
the negative of the power absorbed, this is equivalent to saying that the total
power absorbed by all circuit elements is zero. Thus, the principle of
conservation of energy (and therefore power) is satisfied in this circuit (as it is
in any circuit).

1A.47
Example
Consider the circuit shown below, which is identical to the previous example
except for the value of the current source:
i
i1 i2 i3
5V 1 2 10 A 3
In this case:
25 25
p1 25 W p2 W p3 W
2 3
as before. By KCL, however:
i 10 i1 i2 i3
and thus:
55 5
i 10 A
6 6
Therefore, the powers delivered by the sources are:
5 25
p v 5 W pi 510 50 W
6 6
Hence the total power absorbed is:
275 25
p1 p 2 p3 p v pi 50 0 W
6 6
and again energy (power) is conserved. However, in this case not only do the
resistors absorb power, but so does the voltage source. It is the current source
that supplies all the power absorbed in the rest of the circuit.

1A.48
Power Absorbed in a Resistor
The power absorbed in every resistor is always a nonnegative number.

Consider the resistor shown below:
i R
Figure 1A.29
By definition, the power absorbed in the resistor is p vi . But by Ohms Law,

v Ri . Thus p Ri i , or:
(1A.37)
p Ri 2
Also, i v R , so that p vv R , or:
(1A.38)
v2
p
R
A real resistor Both formulas for calculating power absorbed in a resistor R demonstrate that p
always absorbs
power is always a nonnegative number when R is positive. Therefore a resistor always
absorbs power.
In a physical resistor, this power is dissipated as radiation (light and/or heat). In

some types of resistors (such as an incandescent bulb, a toaster, or an electric
heater), this property is desirable in that the net result may be light or warmth.
In other types of resistors, such as those found in electronic circuits, the heat
dissipated in a resistor may be a problem that cannot be ignored.
Real resistors have The physical size of a resistor determines the amount of power it can safely
a power rating that
dissipate. A power dissipation that exceeds the rating of a resistor can
must not be
exceeded physically damage the resistor. In many electronic applications, resistors need
dissipate only small amounts of power, allowing their use in integrated circuits.

1A.49
Amplifiers
A linear amplifier is a device that increases the amplitude of a signal (a voltage
Amplifier defined
or a current) whilst preserving waveform shape. The circuit symbol for an
amplifier is a triangle which clearly shows the direction of signal travel.
An example of a voltage amplifier is shown below:

The amplifiers
circuit symbol and
vo its transfer
characteristic
K
vi vo 1
K
vi
input output
(a) circuit symbol (b) transfer characteristic
Figure 1A.30
Note the use of subscript notation: i for input, and o for output. The
relationship between the input and output for the voltage amplifier is:
The input-output
vo Kvi (1A.39) relationship for an
ideal linear voltage
amplifier
The quantity K is referred to as the gain. If the gain is a positive number, then
the amplifier is said to be non-inverting. If the gain is a negative number, then
the amplifier is said to be inverting. Note that a negative number does not
imply a decrease in the signal it implies an inversion.
Since the purpose of an amplifier is to increase signal amplitude, we normally

have K 1 . Circuits with K 1 are said to attenuate, but as we shall see,
they may still be implemented with an amplifier.
Amplifiers are used in numerous places and form one of the basic building-
blocks of electronic circuits. For example, signals in telecommunications that
come from antennas are particularly weak and could be in the microvolt or
millivolt range. Reliable processing of these small signals is made easier if the
signal magnitude is much larger.

1A.50
Units of Gain
The gain K of a voltage amplifier can be expressed in two ways. The first way
is as a straight voltage ratio, with units volts per volt:
Amplifier gain vo
expressed in volts
per volt
K V/V (1A.40)
vi
Note that K must be a dimensionless quantity.
The second way comes from the historical development of amplifiers which
were first used extensively throughout telecommunication systems. In these
applications, since signals were audio in nature, it became common to compare
signal amplitudes in terms of the audio power they could deliver. Thus, we can
express the voltage gain with units of decibels:2
Amplifier gain
vo
expressed in
decibels
K 20 log10 dB (1A.41)
vi
The dB unit of voltage gain is useful when circuits are cascaded a cascade
occurs when the output of one circuit is fed into the input of another (and it has
been ensured, through careful design, that one circuit does not load the next,
i.e. each individual circuits behaviour is independent of the load placed on it).
For cascaded circuits you can add the voltage gains in dB instead of
multiplying the standard voltage gains.
2
Historically the Bel (named after Alexander Graham Bell the inventor of the first
commercially viable and practical telephone) was used to define ratios of audio loudness i.e.
ratios of power. In the metric system, a convenient unit to use is the decibel (dB):
1 decibel 10 log10 Po Pi . If electrical power is assumed to be dissipated across equal
resistors, then since P V 2 R , the power ratio is 20 log10 Vo Vi . This power ratio became a
way to express the voltage gain of amplifiers. Note that the decibel is dimensionless, so it can
be applied to any dimensionless ratio, if one wished.

1A.51
Example
A cascade of amplifiers and circuits is shown below, with the gain (or
attenuation) expressed in V/V and dB.
vi v1 v2 vo
A1 R A2
input output
amplifier circuit amplifier

v1 v2 1 vo
vi = A1 = -10 V/V v1 = K = 2 V/V v2 = A2 = 20 V/V
| A1 | = 20 dB | K | = -6 dB | A2 | = 26 dB
The overall gain can be expressed as:
vo v1 v2 vo
A1KA2 10 12 20 100 V/V
vi vi v1 v2
Note that when the gain in V/V is negative, then the signal is inverted.
The gain expressed in decibels is:
vo
20 log10 A1KA2
vi
20 log10 A1 20 log10 K 20 log10 A2
20 6 26
40 dB
As a check, we can perform the dB calculation directly on the overall V/V

gain:
vo
20 log10 100 20 2 40 dB
vi
When the gain is expressed in dB, it refers to the magnitude of the gain only
it conveys no phase information. When the gain in dB is a negative number,
then we have attenuation.

1A.52
Amplifier Power Supplies
Amplifiers require a separate and independent power supply to operate. Some

amplifiers are bipolar, meaning that they are designed to amplify both positive
and negative signals. These amplifiers require a bipolar DC power supply. For
example, it is common to power bipolar amplifiers with +15 V and -15 V. A lot
of amplifier power supply labelling is used to reflect the technology it is
implemented with, or is based on historical precedent. For example, it is
common to see bipolar power supplies represented on amplifiers as:
A bipolar amplifier
showing the power
supply connections VCC (positive supply)
vi K vo
input output
VEE (negative supply)
Figure 1A.31
where VCC 15 V and V EE 15 V , with respect to a circuit common. The

CC subscript in this case refers to the voltage at a collector of a transistor
inside the amplifier package, and the EE to an emitter of a transistor.
Some amplifiers are unipolar, which means they are designed to amplify
signals that are of one polarity only. These amplifiers only require a single DC
power supply, such as +5V.
A unipolar amplifier
showing the power
supply connections VCC (positive supply)
vi K vo
input output
Figure 1A.32

1A.53
Saturation
Real amplifiers can only output a voltage signal that is within the capabilities
of the internal circuitry and the external DC power supplies. When amplifier
outputs approach their output limitation, they are said to saturate they cannot
provide the output that is required by a linear characteristic. The resulting
transfer characteristic, with the positive and negative saturation levels denoted
L and L respectively, is shown below:
The transfer
characteristic of a
real amplifier,
vo showing that it
L+ saturates eventually
vi
Figure 1A.33
Each of the two saturation levels is usually within a volt or so of the voltage of
the corresponding power supply. Obviously, in order to avoid distorting the
output signal waveform, the input signal swing must be kept within the linear
range of operation. If we dont, then the output waveform becomes distorted
and eventually gets clipped at the output saturation levels.
The input signal and
the output signal of
a saturated amplifier
showing clipping
Figure 1A.34

1A.54
Circuit Model
For an ideal voltage amplifier, the output voltage is independent of both the
source resistance and the load resistance. Thus, to model an ideal voltage
amplifier, we would use a voltage-controlled voltage source:
The model of an
ideal voltage
amplifier
vi Kv i vo
Figure 1A.35
Real voltage amplifiers have a finite input resistance as well as a finite output
resistance. Thus, a model of a real amplifier is:
A linear model of a
real voltage
amplifier
Rout
vi Rin Kv i vo
Figure 1A.36
This model is only valid in its linear region of operation. Also note that the
amplifier is unilateral there is no path for a voltage at the output to appear
in some way at the input of the amplifier. Thus, the use of the voltage-
controlled voltage source creates a one way path for the voltage from the
input to the output.

1A.55
The Operational Amplifier

An operational amplifier (op-amp) is an integrated circuit amplifier consisting
Op-amps defined
of dozens of transistors. An op-amp amplifies the voltage difference between
its two input terminals, and produces a single-ended output voltage, i.e. the
output voltage is with respect to the power supply common. The circuit
symbol for the op-amp is shown below:
An op-amp circuit
symbol showing all
positive connections
supply
VCC
v
inputs vo output
v
VEE
negative
supply
Figure 1A.37
In many circuit diagrams it is customary to omit the power supply and common
connections (the output being understood to be taken with respect to the circuit
common), and so we normally draw:
A simplified op-amp
circuit symbol
v
vo
v
Figure 1A.38
The input labelled v is termed the noninverting terminal, and the input
labelled v is termed the inverting terminal. This naming is a result of the op-
amps ability to amplify the difference between these two voltages.

1A.56
Feedback
The op-amp is an amplifier intended for use with external feedback elements,
where these elements determine the resultant function, or operation3. As we
shall see, op-amp circuits can perform a variety of mathematical operations,
such as addition, subtraction, integration and differentiation of voltage signals.
The feedback elements are connected between the op-amps output and its
inverting terminal, thus providing what is known as negative feedback.
input
output
vi vo
Feedback
Network
Figure 1A.39
In the figure, the input is applied between the op-amp (+) input and a common,
or reference point, as denoted by the ground symbol. This reference point is
also common to the output, the feedback network and the power supply.
The feedback network can be resistive or reactive, linear or nonlinear, or any

combination of these. More detailed analysis, shown later, shows that the
circuits overall gain characteristic is predominantly determined by the
feedback network. It is critically important to note that op-amps are never
intended for use without a feedback network.
3
The naming of the operational amplifier occurred in the classic paper by John R. Ragazzini,
Robert H. Randall and Frederick A. Russell, Analysis of Problems in Dynamics by Electronic
Circuits, Proceedings of the IRE, Vol. 35, May 1947, pp. 444-452. This paper references the
op-amp circuits (feedback amplifiers) used by Bell Labs in the development of the M9 gun
director, a weapon system which was instrumental in winning WWII.

1A.57
Circuit Model
Op-amps are voltage amplifiers. They are designed to have an extremely large
input resistance, a very low output resistance, and a very large gain. Even
though they are comprised of dozens of transistors, a simple linear macro-
model of a real op-amp valid over a certain range of operating conditions is
shown below:
A simple linear
model of a real
op-amp
Rout
v vo
Rin AOL ( v - v )
Figure 1A.40
Note that, under open-circuit conditions on the output (i.e. no load is attached
to the op-amp output terminal), the op-amps output voltage is given by:
The open-loop
vo AOL v v (1A.42) output voltage of an
op-amp, under no-
load conditions
The gain of the amplifier under these conditions, AOL , is termed the open-loop
gain, hence the OL subscript. The reason for this name will become apparent
shortly.
The model parameters for a general purpose op-amp, such as the TL071, are
tabulated below:
Parameter Symbol Typical Value
Open-loop voltage gain AOL 200 000 V/V
Input resistance Rin 1 T
Output resistance Rout 200

1A.58
The Ideal Op-Amp
The ideal op-amp is a theoretical device that pushes the typical op-amp
parameters to their ideal values:
Parameter Symbol Ideal Value
Open-loop voltage gain AOL V/V
Input resistance Rin
Thus, the ideal op-amp has the circuit model:

Model of an ideal
op-amp
v vo
AOL ( v - v )
v AOL=
Figure 1A.41
There are several interesting characteristics of this model that will be useful
when analysing (and designing) circuits with ideal op-amps.

1A.59
The first characteristic is due to the infinite input resistance:

An ideal op-amp
An ideal op-amp draws no input current (1A.43) draws no input
current
The second characteristic of the model is that the output voltage is constrained
by a dependent voltage source (there is no output resistance), and thus:
An ideal op-amp has an output voltage An ideal op-amp has

an output voltage
(1A.44) that is independent
that is independent of its load of the load placed
on it
The last and most important characteristic is due to the infinite open-loop gain.
At first glance the idealisation that AOL appears problematic from a circuit
analysis viewpoint, since for a finite input voltage difference the output will be
infinite. However, the ideal op-amp can produce a finite output voltage, but
only so long as the input voltage is zero. Thus, for an ideal op-amp to produce a
finite output voltage vo , the input voltage difference must be:
vo finite
v v 0 (1A.45)
AOL
and therefore:
An ideal op-amp has
equal input voltages
v v (1A.46) if it has a finite
output voltage
Thus:
An ideal op-amp has equal input voltages (1A.47)
Since the ideal op-amp has equal input voltages (like a short-circuit), but draws
The virtual short-
no input current (like an open-circuit), we say there is a virtual short-circuit circuit defined
across its input terminals. We will use the concept of the virtual short-circuit as
the fundamental basis for the analysis and design of circuits containing ideal
op-amps.

1A.60
Op-Amp Fabrication and Packaging
There are many designs for the internal circuit of an op-amp, with each design
optimising a particular parameter (or parameters) of interest to the designer.
Such parameters may be the open-loop gain (how much the input voltage
difference is amplified), the bandwidth (the highest frequency it can amplify),
or the bias current (how much DC current it draws from the input terminals).
You will become familiar with these terms later when we look more closely at
real op-amp limitations (as opposed to the ideal op-amp).
There are several device fabrication technologies that are used to construct an
op-amp. For general-purpose op-amps, bipolar junction transistors (BJTs) are
mostly used at the input because they are easy to match and are capable of
carrying large currents. However, some operational amplifiers have a field
effect transistor (FET) input, with the rest of the circuit being made from BJTs.
Complementary metal-oxide-semiconductor (CMOS) transistors are used in
op-amps that find application in the design of analog and mixed-signal very
large scale integrated (VLSI) circuits.
It is important to be able to recognise the standard pin-outs of an op-amp IC.

All ICs conform to a standard pin numbering scheme. There is usually a notch
or mark on one end of the chip. With the notch oriented to the left, pin 1 is the
first pin on the bottom of the package. The pins are then numbered in a
counter-clockwise direction. An example is shown below for the TL071 op-
amp, which is a single op-amp housed in an 8-pin package:
Single op-amp IC
package details
TL071 package details
Figure 1A.42

1A.61
Negative Feedback
The concept of negative feedback is fundamental to life. A simple experiment
An everyday
will illustrate this point: close your eyes and then bring your index fingers example of negative
together so that they touch at the tips. You will probably miss. By closing your feedback
eyes you have broken a feedback loop which is vital to most human actions; in
order to perform an operation accurately we must be able to see what we are
doing and thus apply any small corrections as and when necessary. In effect,
we are taking the output (the action) and feeding it back to the input (the
mental instruction or intention) in such a way that the output is made equal to
the input. In other words, the action is forced to correspond exactly with the
intention.
Examples of negative feedback can also be found in the field of mechanical

engineering. One of the clearest examples is the governor which is used to
control the speed of rotating machinery. The most historically significant form
of governor was that developed by James Watt in 1788, following the
suggestion of his business partner, Matthew Boulton. The conical pendulum
governor was one of the final series of innovations that Watt made to the steam
engine that ushered in the industrial revolution and became the prime source of
motive power in the 19th century.
The conical pendulum governor is shown in basic form below:

Watts conical
pendulum governor
Figure 1A.43

1A.62
As the speed of the engine, and hence of the governor shaft, increases, the
centrifugal force causes the weights to fly outwards on their linkage. The
linkage is connected via a system of levers to the main steam valve so that, if
the speed increases, the movement of the governor weights throttles back the
main steam supply to the engine. Conversely, a tendency to slow down,
perhaps due to increased load, will allow more steam in to boost the speed back
to normal. The speed will thus settle down to a happy medium and be largely
independent of variations in the load on the engine.
In this example, as in the physiological illustration discussed first, the system is

kept under control by feeding a measure of its output back to the input.
Mechanical control systems such as the governor are often known as servo-
systems (literally, slave systems) and are fundamental to industrial automation.
Negative Feedback in Electronics
The gain of an amplifier, such as an op-amp, will vary from device to device
due to the many manufacturing variations in the transistors and resistors that
comprise it. Such component variations result in considerable uncertainty in
the overall voltage gain. For example, an op-amp datasheet may specify the
typical value for the open-loop gain as 200 000, but some specimens may
achieve a gain as low as 25 000. The open-loop gain also changes with
temperature, power supply voltage, signal frequency and signal amplitude.
Just as the steam engine needs the controlling influence of the governor, so
Negative feedback
is used to precisely most electronic amplifiers require electrical negative feedback if their gain is to
set the gain of an
amplifier be accurately predictable and remain constant with varying environmental
conditions.

1A.63
An Amplifier with Negative Feedback
The figure below shows a block diagram of an amplifier with negative

feedback:
Block diagram of an
amplifier with
amplifier negative feedback
vi ve vo
AOL
vf = vo

feedback
network
Figure 1A.44
The amplifier has a voltage gain AOL and the feedback network is an attenuator
which feeds a fixed fraction, , of the output back to the input. The feedback
signal, v f vo , is subtracted from the input signal (we thus have negative
feedback if the signal were added, we would have positive feedback).
We can now determine the effective voltage gain, ACL , of the amplifier with
feedback. This is given simply by the ratio of the output voltage to input
voltage:
vo
ACL (1A.48)
vi
The signal at the input to the basic amplifier is:
ve vi v f vi vo (1A.49)

1A.64
Also, the signal at the output of the basic amplifier is given by:
vo AOL ve (1A.50)
Therefore:
vo AOL vi vo (1A.51)
Rearranging:
vo 1 AOL AOL vi
vo AOL

vi 1 AOL (1A.52)
Hence:
The closed-loop
gain of an amplifier
AOL
with negative
ACL (1A.53)
feedback
1 AOL
Equation (1A.53) is the general equation for an amplifier with negative

feedback. The basic gain of the amplifier, AOL , is known as the open-loop gain
and the gain with feedback, ACL , as the closed-loop gain.

1A.65
Engineers design the circuit by starting with a basic amplifier with a very large
open-loop gain (e.g., the open-loop gain of an op-amp is AOL 100,000 ) and
then ensure by design that the feedback network provides an attenuation so

that:
AOL 1 (1A.54)
When this is the case, we can neglect the 1 in the denominator of Eq. (1A.53)
so that:
The closed-loop
gain of an amplifier
1
ACL (1A.55)
with negative

feedback, if the
open-loop gain is
very large
This is a very significant equation because we have designed an amplifier

with a precisely determined voltage gain. As long as the open-loop gain is
much larger than the closed-loop gain (e.g. a hundred times greater) then the
closed-loop gain is independent of the amplifier characteristics and dependent
only on the feedback network . The feedback network usually depends upon
just a handful of passive elements, such as resistors and capacitors. These
elements are the most stable components in electronics; their values can be
precisely specified to very high levels of accuracy (better than 0.1% for
resistors, and 1% for capacitors), and their value does not vary greatly with
environmental changes such as temperature and aging. Negative feedback
extends these attributes of accuracy and stability to the gain of the entire
amplifier.
Later, we will also see that negative feedback increases the frequency of
signals that we can apply to the amplifier, reduces nonlinear distortion,
increases input resistance and decreases output resistance. The price for these
benefits is a reduction in the amplifier gain a trade-off that is well worth
making.

1A.66
The Noninverting Amplifier

We seek a circuit implementation of the amplifier with negative feedback
presented in Figure 1A.44. We have already seen that an op-amp amplifies the
difference between its two input voltages. Thus, an op-amp implements the
subtracter and the amplifier in one device:
An op-amp
implements the
subtracter and the amplifier
amplifier in one
device vi vo vi
A OL vo
vf
vf
Figure 1A.45
The feedback network can consist of any combination of passive or active

elements. The simplest feedback network provides attenuation, using a
resistive voltage divider:
A simple feedback
network that
provides a fixed
attenuation
vf R2 vo
vf vo
R1
feedback
network
Figure 1A.46
Note that for this circuit the input is on the right and the output is on the left, as
we are providing a feedback path from the output of the op-amp, and back to its
inverting input terminal.

1A.67
The fixed fraction, , of the output which is fed back to the input is given by
the voltage divider rule:
vf (1A.56)
R1

vo R1 R2
Thus, a circuit that implements the amplifier with negative feedback is:
The noninverting
amplifier
vi
vo
R2
R1
Figure 1A.47
To reiterate, the closed-loop gain of this amplifier is:
vo AOL
ACL (1A.57)
vi 1 AOL
and if we have AOL 1 , then the closed-loop gain is approximately:
1 R2
ACL 1 (1A.58)
R1

1A.68
Example
An op-amp with an open-loop gain of 100 000 is connected in a noninverting

amplifier configuration to give a nominal gain of 10, as shown below:
vi
A OL vo
R2
R1 9 k
1 k
A OL = 100 000
The feedback factor is:
R1 1k 1

R1 R2 1k 9k 10
Therefore the closed-loop gain is:
AOL 10 5 10 5
ACL 9.9990 10
1 AOL 1 10 5 10 1 10001
If the op-amp open-loop gain is changed to 200 000 (e.g. a different op-amp is
used) then the closed-loop gain changes to:
AOL 2 10 5 2 10 5
ACL 9.9995 10
1 AOL 1 2 10 5 10 1 20001
Thus, the closed-loop gain changes by only 0.005%, even though the open-loop
gain changed by 100%. This is because AOL 1 , and therefore, by
Eq. (1A.57), ACL 1 10 . Thus, so long as AOL 1 is satisfied, the
closed-loop amplifier maintains a nominal gain of 10 to a very high accuracy.

1A.69
The Noninverting Amplifier with an Ideal Op-Amp
Assuming an ideal op-amp with infinite open-loop gain ( AOL ), then the
overall closed-loop gain of the amplifier is given by Eq. (1A.57):
AOL
ACL
1 AOL
1

1 AOL
1

1
1

(1A.59)
Thus, the overall closed-loop gain is:
The gain of an ideal

R
ACL 1 2 (noninverting) (1A.60)
noninverting
amplifier
R1
Thus, the approximation for the closed-loop gain that we used for a real op-
amp, ACL 1 , now turns into an exact equation, ACL 1 . Thus, we will
find it expedient to analyse op-amp circuits by assuming that ideal op-amps are
used, with an understanding that the real circuits will differ in performance by
only a tiny amount.
A key point to note in this formula is that the ratio of the resistors determines
the gain. In practice this means that a range of actual R1 and R2 values can be
used, so long as they provide the same ratio.
The amplifier in this configuration provides a gain which is always greater than
or equal to 1. The output is also in phase with the input, since the gain is
positive. Hence, this configuration is referred to as a noninverting amplifier.

1A.70
The closed-loop gain of the amplifier can also be derived using circuit analysis
and the concept of the virtual short-circuit. For such a simple circuit, the results
of the analysis can be written directly on the circuit diagram:
Analysis steps for
the ideal
noninverting
amplifier vi R2
1 0V v o = (1+ R ) v i
1
7
4 0A
5 R2
2 vi
vi
vi vi
3 R1 R2
R1 R1
6
R1
Figure 1A.48
The analysis steps are:
1. We assume an ideal op-amp, and also assume that since there is a negative
feedback path around the op-amp, then it is producing a finite output
voltage (i.e., the overall amplifier is working). Thus, the ideal op-amp
must have a virtual short-circuit (VSC) at its input terminals. We label the
voltage across the input terminals as 0 V.
2. Since there is no difference in the voltages across the VSC, the voltage at
the inverting terminal is v vi .
3. The current through resistor R1 is given by Ohms Law, i1 vi R1 .
4. Due to the infinite input resistance of the ideal op-amp, the current entering
the inverting terminal is 0 A.

1A.71
5. KCL at the inverting terminal now gives i2 i1 vi R1 .
6. The voltage drop across the resistor R2 is given by Ohms Law,

vi
v R2 R2 i2 R2 , with the polarity shown.
R1
7. KVL, from the common, across R1 , across R2 and to the output terminal
v i R2
gives vo vi R2 1 v i .
R1 R1
Once again, we see that the overall closed-loop gain is given by 1 R2 R1 .
Input Resistance of the Noninverting Amplifier
The input voltage in a noninverting amplifier is connected directly to the

The input resistance
noninverting terminal of the op-amp, which is effectively an open-circuit as far of the ideal
noninverting
as current is concerned. Thus, the input resistance of the circuit is ideally amplifier is infinite
infinite. This is an important property of the noninverting amplifier that is used
in a variety of situations.
Equivalent Circuit of the Noninverting Amplifier
The equivalent circuit of the noninverting amplifier is:

Equivalent circuit of
the ideal
noninverting
amplifier
vi vo
Kv i
R2
K = 1+
R1
Figure 1A.49
In our analysis of op-amp circuits from now on, we will assume ideal op-amps
and make frequent use of the virtual short-circuit concept.

1A.72
Example
It is required to design an amplifier with a gain of 5 V/V that presents an

almost infinite input resistance to any attached voltage source.
In this case, we decide to choose feedback resistors in the k range:
4 k
1 k
vo
vi
Notice that we have decided to draw the inverting terminal of the op-amp at the
top, and the feedback resistors pass over the top of the op-amp. The circuit is
still the same as before. Either representation can be used, and will depend on
such factors as space or clarity in the circuit schematic.
When we attach any type of source to the input of this circuit, no current will
be drawn. For example:
4 k
1 k
50 i=0A v o = 10 V
0V vi = 2 V
vs = 2 V
Source
We have therefore met the design specifications.

1A.73
Example
An ideal op-amp is connected in a noninverting amplifier configuration as

shown below:
vo
vi
We identify R1 and R2 0 . Thus, the overall closed-loop gain is:
R2 0
ACL 1 1 1
R1
Thus, the circuit provides a gain of 1, and is called a buffer or a follower.
You can also see that a buffer can be created by:
R1
vo
vi
or by:
R2
vo
vi
Both of these circuits will operate as buffers, but the circuit presented first uses
one less component.

1A.74
The Buffer
You may wonder What is the point of a buffer if it only provides a gain of 1?
The buffer is used to
couple one circuit The answer lies in the other properties of the circuit its infinite input
to another
resistance and zero output resistance.
Example
A certain function generator has an output resistance of 50 and so its

output will experience a significant internal Ri voltage drop when the attached
load draws a large current, resulting in a drop in the output terminal voltage:
50 i = 20 mA
1V
vs = 5 V vo = 4 V 200
Function Generator Load
Therefore, we need to buffer the function generator with an amplifier which

presents a high input resistance to the source and which also provides a low
output resistance to the load:
Rs Ro
vs vi Ri Avi vo RL
Source Amplifier Load

An ideal buffer amplifier with a gain of 1, when placed in between the function
generator and the load, delivers the full source voltage to the load:
i = 25 mA
50 i=0
vs = 5 V vi = 5 V vo = 5 V 200

1A.75
The Inverting Amplifier

An inverting amplifier is created by swapping the input and common
connections of the noninverting amplifier. The result is:
The inverting
amplifier
R2
R1
vi
vo
Figure 1A.50
We will analyse this circuit using the concept of the virtual short-circuit. This
can be done on the circuit schematic:
Analysis steps for
the ideal inverting
amplifier
6
5
vi
R2
R1
vi
R2
R1
3
vi 2 4
R1 R1 0V 0A
vi R2
1 0V vo = - v
R1 i
7
Figure 1A.51

1A.76
The analysis steps are:
1. We assume an ideal op-amp, and also assume that since there is a negative
feedback path around the op-amp, then it is producing a finite output
voltage (i.e., the overall amplifier is working). Thus, the ideal op-amp
must have a virtual short-circuit (VSC) at its input terminals. We label the
voltage across the input terminals as 0 V.
2. Since there is no difference in the voltages across the VSC, the voltage at
the inverting terminal is v 0 .
3. The current through resistor R1 is given by Ohms Law, i1 vi R1 .
4. Due to the infinite input resistance of the ideal op-amp, the current entering
the inverting terminal is 0 A.
5. KCL at the inverting terminal now gives i2 i1 vi R1 .
6. The voltage drop across the resistor R2 is given by Ohms Law,

vi
v R2 R2 i2 R2 , with the polarity shown.
R1
7. KVL, from the common, across the VSC, across R2 and to the output
vi R
terminal gives vo 0 R2 2 vi .
R1 R1
Thus, the overall closed-loop gain is:
The gain of an ideal

R2
inverting amplifier
ACL (inverting) (1A.61)
R1
The negative sign indicates that there is an inversion of the signal (i.e. a 180
phase change), so that a waveform will appear amplified, but upside down.

1A.77
Input Resistance of the Inverting Amplifier
The input resistance of the inverting amplifier (i.e. the resistance seen by the
input voltage source), is, by the definition of input resistance:

v
Rin vi ii vi i1 i R1 (1A.62)
of the ideal inverting
amplifier is finite
vi R1
Thus, the input resistance of the circuit is equal to R1 . This is a disadvantage

compared to the noninverting amplifier, and careful use of the inverting
amplifier is required.
Equivalent Circuit of the Inverting Amplifier
An equivalent circuit of the inverting amplifier is:

Equivalent circuit of
the ideal inverting
amplifier
vi vo
R1 Kv i
R2
K= -
R1
Figure 1A.52
One advantage of the inverting amplifier over the noninverting amplifier is that
you can achieve gain magnitudes less than one, i.e. build circuits that can
attenuate, as well as amplify.
The inverting amplifier is also the basis for many other useful circuits that we
will encounter later, such as the summer, integrator and differentiator.

1A.78
Example
It is required to design an amplifier with a gain of -5 V/V. In this case, we

decide to choose feedback resistors in the k range:
5 k
1 k
vi
vo
Suppose we now attach a source, which has an internal resistance of 1 k, to

this amplifier:
5 k
1 k i = 1 mA 1 k
1V vi = 1 V
0V
v o = -5 V
vs = 2 V
Source
We see that the 1 k input resistance of the inverting amplifier has caused the
source to deliver current, and therefore there is a significant voltage drop
across its internal resistance. A better design to suit this particular source
would use resistors in the 100s of k:
500 k
1 k i 20 A 100 k
0V
v o -9.901 V
vi 1.980 V
20 mV
vs = 2 V
Source
Now there is less than 1% error in the gain that it provides to the source.

1A.79
Summary
Current is defined as the rate of flow of charge past a certain cross-
sectional area:
dq
i
dt
Voltage is defined as the work done per unit charge in moving it from one
point to another in a circuit.
A circuit element is an idealised mathematical model of a two-terminal

electrical device that is completely characterised by its voltage-current
relationship. Active circuit elements can deliver a non-zero average power
indefinitely, whilst passive circuit elements cannot. A connection of circuit
elements is called a network. If the network contains at least one closed
path, it is also an electrical circuit.
Independent sources are ideal circuit elements that possess a voltage or

current value that is independent of the behaviour of the circuits to which
they belong. There are two types, voltage and current:
vs ( t ) is ( t )
voltage current
source source

1A.80
The resistor is a linear passive circuit element that obeys Ohms Law:
v Ri
A resistance of 0 is known as a short-circuit.
A resistance of is known as an open-circuit.
The reciprocal of resistance is called the conductance:
1
G
R
Practical resistors come in a large variety of shapes, materials and
construction which dictate several of their properties, such as accuracy,
stability, pulse handling capability, resistor value, size and cost.
Kirchhoffs Current Law (KCL) states: At any node of a circuit, the

currents algebraically sum to zero:
i
k 1
k 0
Kirchhoffs Voltage Law (KVL) states: Around any loop in a circuit, the
voltages algebraically sum to zero:
v
k 1
k 0

1A.81
Resistors in series can be combined into one equivalent resistor:
Req R1 R2 RN
Resistors in parallel can be combined into one equivalent resistor:
1 1 1 1

Req R1 R2 RN
Independent voltage sources in series can be added. Independent current

sources in parallel can be added.
Two resistors in series form a voltage divider:
i R1
arbitrary
v1
circuit v R2 v2
R2
v2 v
The voltage divider rule is:
R1 R2
Two resistors in parallel form a current divider:
i
i1 i2
arbitrary
circuit
v
R1 R2
R1
i2 i
The current divider rule is:
R1 R2

1A.82
An ideal source, either voltage or current, whose value depends upon some
parameter (usually a voltage or current) in the circuit to which the source
belongs is known as a dependent or controlled source. There are four types:
Kvx rm i x Ki x gm vx
VCVS CCVS CCCS VCCS
The instantaneous power absorbed by an electrical circuit element is the

product of voltage and current:
p vi
The power absorbed in a resistor is:
v2
p Ri 2
R
A linear amplifier is a device that increases the amplitude of a signal (a
voltage or a current) whilst preserving waveform shape. The most common
is a voltage amplifier:
vi K vo =Kvi
input output
The gain of a voltage amplifier is expressed in either V/V or in dB:
vo vo
K V/V or
K 20 log10 dB
vi vi

1A.83
The operational amplifier (op-amp) is an integrated circuit amplifier

consisting of dozens of transistors. An op-amp amplifies the voltage
difference between its two input terminals, and produces a single-ended
output voltage. It is the building block for a vast array of useful
electronic circuits.
noninverting input v
vo output
inverting input v
The op-amp is always operated with a negative feedback network.
The characteristics of the ideal op-amp are:
Parameter Symbol Ideal Value
Open-loop voltage gain AOL V/V
Input resistance Rin
If an ideal op-amp has negative feedback, then a virtual short-circuit

appears across its input terminals. This is the key to analysing and
designing op-amp circuits.
Negative feedback is a universal concept that arises in fields as diverse as

biology, mechanics and electronics. If a proportion of the output of a device
is fed back to the input and subtracted (so-called negative feedback), then
the device will exhibit certain desirable overall qualities, such as accuracy
and stability. In particular, the gain of an electronic amplifier with an open-
loop gain of AOL and a feedback factor of will have a closed-loop gain
given by:
AOL
ACL
1 AOL

1A.84
The noninverting amplifier is one of the most fundamental op-amp

arrangements:
R2
R1
vo
vi
The gain of the noninverting amplifier is:
R2
ACL 1
R1
An important special case of the noninverting amplifier is the buffer, for

which R1 , R2 0 and therefore ACL 1 .
The inverting amplifier is one of the most fundamental op-amp

arrangements:
R2
R1
vi
vo
The gain of the inverting amplifier is:
R2
ACL
R1
References
Hayt, W. & Kemmerly, J.: Engineering Circuit Analysis, 3rd Ed., McGraw-
Hill, 1984.

1A.85
Exercises
1.
A large number of electrons are moving through a conductor:
conductor
electron flow
The number varies with time t seconds.
(a) What is the direction of current?
(b) If the total charge to pass a certain point on the conductor varies according
to the equation:

q t 3 1 e 100t mC
then find the current in amperes as a function of time.
(c) When will the current be 200 mA?
(d) If the conductor has a uniform diameter of 1 mm throughout its length,

find the current density as a function of time (express as Am -2 ).
(e) Sketch charge and current as functions of time.
(f) How many electrons are moving through the conductor at time
t = 50 ms?

1A.86
2.
The charge entering the upper terminal of the element below is given by
5 sin 1000t C .
(a) How much charge enters that terminal between t 0.5 ms and
t 0.5 ms ?
(b) How much charge leaves the lower terminal in the same time interval?
(c) Find i at t 0.2 ms .
3.
For the current waveform shown below:
i( t ) (A)
300
200
100
t (s)
0 1 2 3 4 5 6 7 8
-100
-200
-300
-400
determine the total charge transferred between t 0 and t :
(a) 4 s (b) 7 s

1A.87
4.
The charging current supplied to a 12 V automotive battery enters its positive
terminal. It is given as a function of time by:
0 t0
t 10000
i 4e A 0 t 15000 s
0 t 15000 s

(a) What is the total charge delivered to the battery in the 15000 s charging
interval?
(b) What is the maximum power absorbed by the battery?
(c) What is the total energy supplied?
(d) What is the average power delivered in the 15000 s interval?
5.
The voltage v has its positive reference at terminal A of a certain circuit
element. The power absorbed by the circuit element is 4t 1 W for t 0 .
2
If v 2t 2 V for t 0 , how much charge enters terminal A between t 0

and t 2 s ?
6.
The resistance of 10 mm2 copper wire is 1.725 /km, and, with a certain type
of insulation, it can safely carry 70 A without overheating. Operating at
maximum current:
(a) What voltage exists between the ends of a one kilometre length of wire?
(b) How much power is dissipated in the conductor?
(c) What is the power dissipation per square mm of surface area?

1A.88
7.
For the circuit shown below:
20 V
10 2 A 2 i
60 V 3A v Load
find:
(a) v (b) i (c) the power absorbed by the load.
8.
With reference to the network shown below:
A
8V 6A
5V B 3A vx
ix
find:
(a) i x (b) v x (c) the power absorbed by the battery.

1A.89
9.
Find the power supplied by the 3 V source in the circuit below:
20i 1
5
3V 10
13 V
6 i1
10.
Find the power absorbed by each element in the circuit below:
-50 V
i
vx = -6i 0.2vx
10

1A.90
11.
iy
ix
30 mA v30 1 k 10 mA 2ix
(a) Find v30 , i x , and i y .
(b) Change the control on the dependent source from 2i x to 2i y and then find
v30 , i x , and i y .
12.
i
v
v 2 2 10 A
2
(a) Find v , i , and the power delivered by the independent source.
(b) Repeat if the arrow of the dependent source is reversed.

1A.91
13.
Find Req for each of the networks shown below:
(a)
4
10 6 2
20 1
R eq
5
3
(b)
8
2 7
R eq

1A.92
14.
By combining independent sources and resistances as appropriate, find:
(a) The current i in the circuit below.
30 V
8
10 20 V
2 i
(b) The voltage v in the circuit below.
30 mA v 1 k 10 mA 250
15.
Use the concepts of current division, voltage division, and resistance
combination to write expressions (by inspection) for v3 and i1 in the circuit
shown below:
R2
i1
is R1 R3 v3

1A.93
16.
Determine the necessary values of v and i in the circuit shown below:
3 v
1 1A
5 4
17.
The circuit shown below exhibits several examples of independent current and
voltage sources in series and in parallel.
12 V -3 A
3V
2A -5 V 4A
4V
(a) Find the power supplied by the -5 V source.
(b) To what value should the 4 A source be changed to reduce the power
supplied by the -5 V source to zero?

1A.94
18.
For the ideal op-amp circuit below:
i 2 10 k
i 1 1 k
io
vi v1 vo
1V iL
1 k
Determine the values of:
(a) v1 , i1 , i2 , v o , i L and io (b) the voltage gain vo vi
(c) the current gain i L ii (d) the power gain PL Pi
19.
Given the ideal op-amp circuit below:
R2
R1
vi
vo
R3
(a) Find the voltage gain vo vi .
(b) Find the resistance seen by the voltage source vi .
(c) How do the results of (a) and (b) differ from the case when R3 ?
Why?

1A.95
20.
R3
vi
vo
R2
R1
(a) Find the voltage gain vo vi .
(b) Find the resistance seen by the voltage source vi .
(c) How do the results of (a) and (b) differ from the case when R3 ?
Why?
21.
Consider the ideal op-amp circuit below:
R1
R1
vi R2
R2 100
Choose values for resistors R1 and R2 such that the 100 resistor absorbs
10 mW when vi 4 V .

1A.96
22.
R
vo
ii
(a) Find the ratio vo ii .
(b) Find the resistance seen by the current source ii .
23.
R1 R3
R2
i3
vo
ii
(c) Find the current gain i3 ii .

1A.97
24.
vo
ii Ri i2
R2
R1
(c) Find the current gain i2 ii .

1B.1
Lecture 1B Nodal and Mesh Analysis

Nodal analysis. Mesh analysis.
Introduction
After becoming familiar with Ohms Law and Kirchhoffs Laws and their
application in the analysis of simple series and parallel resistive circuits, we
must begin to analyse more complicated and practical circuits.
Physical systems that we want to analyse and design include electronic control
circuits, communication systems, energy converters such as motors and
generators, power distribution systems, mobile devices and embedded systems.
We will also be confronted with allied problems involving heat flow, fluid
flow, and the behaviour of various mechanical systems.
To cope with large and complex circuits, we need powerful and general
methods of circuit analysis. Nodal analysis is a method which can be applied to
any circuit, and mesh analysis is a method that can be applied to any planar
circuit (i.e. to circuits that are able to be laid out on a 2D surface without
crossing elements). Both of these methods are widely used in hand design and
computer simulation. A third technique, known as loop analysis, generalises
mesh analysis and can be applied to any circuit it is effectively the dual of
nodal analysis.
We will find that the judicious selection of an analysis technique can lead to a
drastic reduction in the number of equations to solve, and we should therefore
try to develop an ability to select the most convenient analysis method for a
particular circuit.

1B.2
Nodal Analysis
In general terms, nodal analysis for a circuit with N nodes proceeds as follows:
1. Select one node as the reference node, or common (all nodal voltages
are defined with respect to this node in a positive sense).
The general
principle of nodal
analysis 2. Assign a voltage to each of the remaining N 1 nodes.
3. Write KCL at each node, in terms of the nodal voltages.
4. Solve the resulting set of simultaneous equations.
As will be seen, the method outlined above becomes a little complicated if the
circuit contains voltage sources and / or controlled sources, but the principle
remains the same.
Example
We apply nodal analysis to the following 3-node circuit:
3A 2 1 -2 A
Following the steps above, we assign a reference node and then assign nodal
voltages:
v1 5 v2
3A 2 1 -2 A

1B.3
We chose the bottom node as the reference node, but either of the other two
nodes could have been selected. A little simplification in the resultant
equations is obtained if the node to which the greatest number of branches is
connected is identified as the reference node.
In many practical circuits the reference node is one end of a power supply
which is generally connected to a metallic case or chassis in which the circuit
resides; the chassis is often connected through a good conductor to the Earth.
Thus, the metallic case may be called ground, or earth, and this node
becomes the most convenient reference node.
To avoid confusion, the reference node will be called the common unless it The distinction
has been specifically connected to the Earth (such as the outside conductor on a between common
and earth
digital storage oscilloscope, function generator, etc).
Note that the voltage across any branch in a circuit may be expressed in terms
of nodal voltages. For example, in our circuit the voltage across the 5
resistor is v1 v 2 with the positive polarity reference on the left:
v1 5 v2
(v1 - v2)
We must now apply KCL to nodes 1 and 2. We do this be equating the total
current leaving a node to zero. Thus:
v1 v1 v 2
3 0
2 5
v 2 v1 v1
2 0
5 1
Simplifying, the equations can be written:
0.7v1 0.2v 2 3
0.2v1 1.2v 2 2

1B.4
Rewriting in matrix notation, we have:
0.7 0.2 v1 3
0 .2 1 .2 v 2
2
These equations may be solved by a simple process of elimination of variables,

or by Cramers rule and determinants. Using the latter method we have:
3 0.2
2 1.2 3.6 0.4 4
v1 5V
0.7 0.2 0.84 0.04 0.8
0.2 1.2
0.7 3
0.2 2 1.4 0.6 2
v2 2.5 V
0.8 0.8 0.8
Everything is now known about the circuit any voltage, current or power in
the circuit may be found in one step. For example, the voltage at node 1 with
respect to node 2 is v1 v 2 2.5 V , and the current directed downward
through the 2 resistor is v1 2 2.5 A .

1B.5
Circuits with Resistors and Independent Current Sources Only
A further example will reveal some interesting mathematical features of nodal

analysis, at least for the case of circuits containing only resistors and
independent current sources. We will find it is much easier to consider
conductance, G, rather than resistance, R, in the formulation of the equations.
Example
A circuit is shown below with a convenient reference node and nodal voltages
specified.
1/4
-3 A
v1 1/3 v2 1/2 v3
-8 A 1 1/5 25 A
We sum the currents leaving node 1:
3v1 v 2 4v1 v3 8 3 0
7v1 3v 2 4v3 11
At node 2:
3v 2 v1 1v 2 2v 2 v3 3 0
3v1 6v 2 2v3 3
At node 3:
4v3 v1 2v3 v 2 5v3 25 0

4v1 2v 2 11v3 25

1B.6
Rewriting in matrix notation, we have:
7 3 4 v1 11
3 6 2 v 3
2
4 2 11 v3 25
For circuits that contain only resistors and independent current sources, we
define the conductance matrix of the circuit as:
7 3 4
G 3 6 2
4 2 11
The conductance It should be noted that the nine elements of the matrix are the ordered array of
matrix defined
the coefficients of the KCL equations, each of which is a conductance value.
Thr first row is composed of the coefficients of the Kirchhoff current law
equation at the first node, the coefficients being given in the order of v1 , v 2
and v3 . The second row applies to the second node, and so on.
The major diagonal (upper left to lower right) has elements that are positive.
The conductance matrix is symmetrical about the major diagonal, and all
elements not on this diagonal are negative. This is a general consequence of the
systematic way in which we ordered the equations, and in circuits consisting of
only resistors and independent current sources it provides a check against
errors committed in writing the circuit equations.
We also define the voltage and current source vectors as:
v1 11
v v 2 i 3
v3 25

1B.7
Our KCL equations can therefore be written succinctly in matrix notation as:
Nodal analysis
Gv i expressed in matrix
notation
The solution of the matrix equation is just:
v G 1i
Computer programs that do nodal analysis use sophisticated numerical

methods to efficiently invert the G matrix and solve for v . When solving the
equations by hand we resort to matrix reduction techniques, or use Cramers
rule (up to 3 x 3). Thus:
11 3 4
3 6 2
25 2 11
v1
7 3 4
3 6 2
4 2 11
To reduce work, we expand the numerator and denominator determinants by

minors along their first columns to get:
6 2 3 4 3 4
11 3 25
2 11 2 11 6 2
v1
6 2 3 4 3 4
7 3 4
2 11 2 11 6 2
1162 3 41 2530 682 123 750

762 3 41 430 434 123 120
191
1 V
191
Similarly:
7 11 4 7 3 11
3 3 2 3 6 3
4 25 11 4 2 25
v2 2 V v3 3 V
191 191

1B.8
Nodal Analysis Using Branch Element Stamps
The previous example shows that nodal analysis leads to the equation Gv i .
We will now develop a method whereby the equation Gv i can be built up
on an element-by-element basis by inspection of each branch in the circuit.
Consider a resistive element connected between nodes i and j:
vi G vj vi G vj
(vi - vj) (vj - vi )
(a) (b)
Figure 1B.1
Suppose that we are writing the ith KCL equation because we are considering
the current leaving node i (see Figure 1B.1a). The term that we would write in
this equation to take into account the branch connecting nodes i and j is:
Gvi v j 0 (1B.1)
This term appears in the ith row when writing out the matrix equation.
If we are dealing with the jth KCL equation because we are considering the
current leaving node j (see Figure 1B.1b) then the term that we would write in
this equation to take into account the branch connecting nodes j and i is:
Gv j vi 0 (1B.2)
This term appears in the jth row when writing out the matrix equation.

1B.9
Thus, the branch between nodes i and j contributes the following element
stamp to the conductance matrix, G :
i j
The element stamp
i G G (1B.3)
for a conductance

j G G
If node i or node j is the reference node, then the corresponding row and
column are eliminated from the element stamp shown above.
For any circuit containing only resistors and independent current sources, the
conductance matrix can now be built up by inspection. The result will be a G
matrix where each diagonal element g ii is the sum of conductances connected
to node i, and each off-diagonal element g ij is the total conductance between
nodes i and j but with a negative sign.
Now consider a current source connected between nodes i and j:
I
vi vj
Figure 1B.2
In writing out the ith KCL equation we would introduce the term:
I 0 (1B.4)
In writing out the jth KCL equation we would introduce the term:
I 0 (1B.5)

1B.10
Thus, a current source contributes to the right-hand side (rhs) of the matrix
equation the terms:
The element stamp i I

for an independent
current source
j I (1B.6)
Thus, the i vector can also be built up by inspection each row is the addition
of all current sources entering a particular node. This makes sense since
Gv i is the mathematical expression for KCL in the form of current leaving
a node = current entering a node.
Example
We will analyse the previous circuit but use the formal approach to nodal
analysis.
1/4
-3 A
v1 1/3 v2 1/2 v3
-8 A 1 1/5 25 A
By inspection of each branch, we build the matrix equation:
1 2 3
1 3
4 3 4 v1 8 3

2 3 1 3 2 2 v 2 3

3
4 2 5 4 2 v3 25
This is the same equation as derived previously.

1B.11
Circuits with Voltage Sources
Voltage sources present a problem in undertaking nodal analysis, since by

definition the voltage across a voltage source is independent of the current
through it. Thus, when we consider a branch with a voltage source when
writing a nodal equation, there is no way by which we can express the current
through the branch as a function of the nodal voltages across the branch.
There are two ways around this problem. The more difficult is to assign an
unknown current to each branch with a voltage source, proceed to apply KCL
at each node, and then apply KVL across each branch with a voltage source.
The result is a set of equations with an increased number of unknown variables.
The easier method is to introduce the concept of a supernode. A supernode

The concept of a
encapsulates the voltage source, and we apply KCL to both end nodes at the supernode
same time. The result is that the number of nodes at which we must apply KCL
is reduced by the number of voltage sources in the circuit.
Example
Consider the circuit shown below, which is the same as the previous circuit
except the 1 2 resistor between nodes 2 and 3 has been replaced by a 22 V
voltage source:
1/4
-3 A
supernode
22 V
v1 1/3 v2 v3
-8 A 1 1/5 25 A

1B.12
KCL at node 1 remains unchanged:
3v1 v 2 4v1 v3 8 3 0
7v1 3v 2 4v3 11
We find six branches connected to the supernode around the 22 V source

(suggested by a broken line in the figure). Beginning with the 1 3 resistor
branch and working clockwise, we sum the six currents leaving this supernode:
3v 2 v1 3 4v3 v1 25 5v3 1v 2 0
7v1 4v 2 9v3 28
We need one additional equation since we have three unknowns, and this is
provided by KVL between nodes 2 and 3 inside the supernode:
v3 v 2 22
Rewriting these last three equations in matrix form, we have:
7 3 4 v1 11
7 4 9 v 2 28

0 1 1 v3 22
The solution turns out to be v1 4.5 V , v 2 15.5 V and v3 6.5 V .
Note the lack of symmetry about the major diagonal in the G matrix as well as
the fact that not all of the off-diagonal elements are negative. This is the result
The presence of a
dependent source of the presence of the voltage source. Note also that it does not make sense to
destroys the
symmetry in the G call the G matrix the conductance matrix, for the bottom row comes from the
matrix
equation v 2 v3 22 , and this equation does not depend on any
conductances in any way.
A supernode can contain any number of independent or dependent voltage

sources. In general, the analysis procedure is the same as the example above
one KCL equation is written for currents leaving the supernode, and then a
KVL equation is written for each voltage source inside the supernode.

1B.13
Circuits with Dependent Sources
Dependent current sources are fairly easy to include into nodal analysis we
just need to express the dependent current in terms of nodal voltages.
Dependent voltages sources are dealt with using the concept of the supernode.
Of the two types of dependent voltage source, the current controlled voltage
source (CCVS) requires the most effort to incorporate into nodal analysis. We
will analyse this case before summarizing the method of nodal analysis for any
resistive circuit.
Example
Consider the circuit shown below, which is the same as the previous circuit
except now the 22 V voltage source has been replaced by a current controlled
voltage source:
1/4
-3 A ix
supernode
ix / 8
v1 1/3 v2 v3
-8 A 1 1/5 25 A
KCL at node 1 remains unchanged:
3v1 v 2 4v1 v3 8 3 0
7v1 3v 2 4v3 11

1B.14
KCL at the supernode, formed by considering nodes 2 and 3 together, remains

unchanged:
3v 2 v1 3 4v3 v1 25 5v3 1v 2 0
7v1 4v 2 9v3 28
Finally, we turn our attention to the dependent source. We rewrite the

dependent current in terms of nodal voltages:
i x 4v1 v3
and then we write KVL for the dependent source as:
i x 4v1 v3
v3 v 2
8 8
v1 2v 2 3v3 0
In matrix form we now have:
7 3 4 v1 11
7 4
9 v 2 28
1 2 3 v3 0
The nodal voltages turn out to be v1 1.431 V , v 2 3.373 V and

v3 2.726 V .
In the previous example, the source dependency ( i x ) was easily expressed
using Ohms Law across the 1 4 resistor. In other cases, the dependency
may need to be found using KCL.

1B.15
Example
2ix
supernode
v1 1 v2 ix
1/2 1/3 18 A
KCL at the supernode, formed by considering nodes 1 and 2 together, gives:
2v1 1v1 v 2 1v 2 v1 3v 2 18 0
2v1 3v 2 18
Note how the 1 resistor contributes nothing to the KCL equation. Next, we
turn our attention to the dependent source inside the supernode. We rewrite the
dependent current in terms of nodal voltages using KCL at node 2:
v 2 v1
3v 2 i x 0
1
i x v1 4v 2
and then we write KVL for the dependent source as:
v 2 v1 2i x 2v1 4v 2
3v1 9v 2 0
2 3 v1 18
3 9 v 0
2
The nodal voltages turn out to be v1 6 V , v 2 2 V .

1B.16
Summary of Nodal Analysis
The general We perform nodal analysis for any resistive circuit with N nodes by the
procedure to follow
when undertaking following method:
nodal analysis
1. Make a neat, simple, circuit diagram. Indicate all element and source
values. Each source should have its reference symbol.
2. Select one node as the reference node, or common. Then write the node
voltages v1 , v 2 , , v N 1 at their respective nodes, remembering that
each node voltage is understood to be measured with respect to the
chosen reference.
3. If the circuit contains dependent sources, express those sources in terms

of the variables v1 , v 2 , , v N 1 , if they are not already in that form.
4. If the circuit contains voltage sources, temporarily modify the given

circuit by replacing each voltage source by a short-circuit to form
supernodes, thus reducing the number of nodes by one for each voltage
source that is present. The assigned nodal voltages should not be
changed. Relate each supernodes source voltage to the nodal voltages.
5. Apply KCL at each of the nodes or supernodes. If the circuit has only
resistors and independent current sources, then the equations may be
built using the element stamp approach.

1B.17
Example
Consider the circuit shown below, which contains all four types of sources and
has five nodes.
v2
vx 1
1/2
6A
3V
v4
v1 v3
2vx
1
1/2 4vy
vy
We select the bottom node as the reference and assign v1 to v4 in a clockwise

direction starting from the left node.
We next relate the controlled sources to the nodal voltages:
2vx 2v2 v1
4v y 4v1
We form supernodes around the two voltage sources, and write relations for
them in terms of the nodal voltages:
v 4 v1 3
v3 4v y 4v1
Thanks to the supernodes, we see that we only need to write KCL equations at
node 2 and the supernode containing both nodes 1 and 4. At node 2:
2v 2 v1 6 1v 2 v3 0
2v1 3v 2 v3 6

1B.18
while at the supernode:
2v1 v 2 6 2v 2 v1 2v 4 1v1 0

2v x
v1 2v 4 6
Thus, we obtain four equations in the four node voltages:
1 0 01 v1 3
4
0 1 0 v 2 0

2 3 1 0 v3 6

1 0 0 2 v 4 6
The solution is:
v1 4 V
v 2 14 3 V
v3 16 V
v 4 1 V
The technique of nodal analysis described here is completely general and can
always be applied to any electrical circuit.

1B.19
Mesh Analysis
Before we embark on mesh analysis, we need to define the concept of: a planar
circuit, a path through a circuit, a loop and a mesh. We can then outline the
analysis strategy using these terms.
Planar Circuits
A planar circuit is one where it is possible to draw the circuit on a plane

Planar circuits
surface in such a way that no branch passes over or under any other branch. defined
(a) (b) (c)
Figure 1B.3
In the figure above, circuit (a) is planar, circuit (b) is nonplanar and circuit (c)
is planar, but drawn so that it appears nonplanar.

1B.20
Paths, Loops and Meshes
A path is made through a circuit when we start on one node and traverse
elements and nodes without encountering any nodes previously visited. A loop
is any closed path i.e. the last node visited is the same as the starting node. A
mesh is a loop which does not contain any other loops within it. A mesh is a
property of a planar circuit and is not defined for a nonplanar circuit.
Examples of loops
and meshes
(a) (b) (c)
(d) (e) (f)
Figure 1B.4
In the figure above, the set of branches in (a) identified by the heavy lines is
neither a path nor a loop. In (b) the set of branches is not a path since it can be
traversed only by passing through the central node twice. In (c) the closed path
is a loop but not a mesh. In (d) the closed path is also a loop but not a mesh. In
(e) and (f) each of the closed paths is both a loop and a mesh. This circuit
contains four meshes.

1B.21
Mesh Current
We define a mesh current as a mathematical (or imaginary) current in which

charge flows only around the perimeter of a mesh. A mesh current is indicated
by a curved arrow that almost closes on itself and is drawn inside the
appropriate mesh.
Mesh currents
defined
i1 i2
Figure 1B.5
Although the direction of mesh currents is arbitrary, we draw the mesh currents
in a clockwise direction so that a symmetry in the equations results when
performing mesh analysis. One of the great advantages of mesh currents is that
KCL is automatically satisfied, and no branch can appear in more than two
meshes.
We no longer have a current or current arrow shown on each branch in the

Branch currents can
circuit. The current through any branch may be determined by superimposing be expressed in
terms of mesh
each mesh current that exists in it. For example, the branch current heading currents
down in the middle resistor in the circuit above is given by i1 i2 .

1B.22
Mesh Analysis Methodology
In general terms, mesh analysis for a planar circuit with M meshes proceeds as
follows:
The general 1. Assign a clockwise mesh current in each of the M meshes.

principle of mesh
analysis
2. Write KVL around each mesh, in terms of the mesh currents.
As will be seen, the method outlined above becomes a little complicated if the
circuit contains current sources and / or controlled sources, but the principle
remains the same.
Example
A two-mesh circuit is shown below.
6 4
42 V i1 3 i2 10 V
We apply KVL to each mesh. For the left-hand mesh:
42 6i1 3i1 i2 0
9i1 3i2 42
For the right-hand mesh:
3i2 i1 4i2 10 0
3i1 7i2 10
The solution is obtained by solving simultaneously: i1 6 A and i2 4 A .

1B.23
Circuits with Resistors and Independent Voltage Sources Only
When the circuit contains only resistors and voltages sources, the KVL
equations have a certain symmetrical form and we can define a resistance
matrix with the circuit. We will find again that the matrix equation can be
formulated by inspection of the circuit.
Example
Consider the five-node, three-mesh circuit shown below.
1
i2 2
i1 3
7V
6V
i3 1
2
The three required mesh currents are assigned as indicated, and we

methodically apply KVL about each mesh:
7 1i1 i2 6 2i1 i3 0
1i2 i1 2i2 3i2 i3 0
2i1 i3 6 3i3 i2 1i3 0
Simplifying and writing as a matrix equation:
3 1 2 i1 1
1 6 3 i 0
2
2 3 6 i3 6

1B.24
For circuits that contain only resistors and independent voltage sources, we
The resistance
matrix defined define the resistance matrix of the circuit as:
3 1 2
R 1 6 3
2 3 6
Once again we note the symmetry about the major diagonal. This occurs only
for circuits with resistors and independent voltage sources when we order the
equations correctly (rows correspond to meshes).
We also define the current and voltage source vectors as:
i1 1
i i2 v 0
i3 6
Our KVL equations can therefore be written succinctly in matrix notation as:
Mesh analysis
expressed in matrix Ri v
notation
Applying Cramers rule to the formulation for i1 gives:
1 1 2
0 6 3
6 3 6 27 0 90 117
i1 3A
3 1 2 81 12 30 39
1 6 3
2 3 6
The other mesh currents are: i2 2 A and i3 3 A .

1B.25
Circuits with Current Sources
When a mesh has a current source in it, we must modify the procedure for
forming the circuit equations. There are two possible methods. In the first
method, we can relate the source current to the assigned mesh currents, assign
an arbitrary voltage across it (thereby increasing the number of variables by
one) and write KVL equations using this voltage. Alternately, a better method
is to take a lead from nodal analysis and formulate the dual of a supernode - a
supermesh.
To create a supermesh, we open-circuit or remove current sources, thereby The concept of a

reducing the total number of meshes. We apply KVL only to those meshes in supermesh
the modified circuit.
Example
Consider the circuit shown below in which a 7 A independent current source is

in the common boundary of two meshes.
1
i2 2
i1 3
7V
7A
i3 1
2
For the independent current source, we relate the source current to the mesh
currents:
i1 i3 7

1B.26
We then mentally open-circuit the current source, and form a supermesh whose
interior is that of meshes 1 and 3:
1
i2 2
i1 3
7V
i3 1
supermesh 2
Applying KVL about the supermesh:
7 1i1 i2 3i3 i2 1i3 0
Around mesh 2 we have:
1i2 i1 2i2 3i2 i3 0
Rewriting these last three equations in matrix form, we have:
1 0 1 i1 7
1 4 4 i 7
2
1 6 3 i3 0
Notice that we have lost all symmetry in the matrix equation Ri v , and we
can no longer call R the resistance matrix. Applying Cramers rule for i1 :
7 0 1
7 4 4
0 6 3 84 42 126
i1 9A
1 0 1 12 2 14
1 4 4
1 6 3
The other mesh currents are: i2 2.5 A and i3 2 A .

1B.27
Circuits with Dependent Sources
Dependent voltage sources are fairly easy to include into mesh analysis we
just need to express the dependent voltage in terms of mesh currents.
Dependent current sources are dealt with using the concept of the supermesh.
Of the two types of dependent current source, the voltage controlled current
source (VCCS) requires the most effort to incorporate into mesh analysis. We
will analyse this case before summarizing the method of mesh analysis for any
resistive circuit.
Example
In the circuit shown below we have both a dependent and an independent

current source:
1
i2 2
i1 3
15 A
vx
1v
9
x
i3 1
2
For the independent current source, we relate the source current to the mesh
currents:
i1 15
Turning our attention to the dependent source, we describe the dependency in

terms of the mesh currents:
v x 3i3 i2
i3 i1
9 9
3i1 i2 2i3 0

1B.28
Since the current sources appear in meshes 1 and 3, when they are open-
circuited, only mesh 2 remains. Around mesh 2 we have:
1i2 i1 2i2 3i2 i3 0
1 0 0 i1 15
3 1 2 i 0
2
1 6 3 i3 0
The mesh currents turn out to be i1 15 A , i2 11 A and i3 17 A . We note
that we wasted a little time in assigning a mesh current i1 to the left mesh we
should simply have indicated a mesh current and labelled it 15 A.

1B.29
Summary of Mesh Analysis
We perform mesh analysis for any resistive circuit with M meshes by the The general
procedure to follow
following method: when undertaking
mesh analysis
1. Make certain that the circuit is a planar circuit. If it is nonplanar, then
mesh analysis is not applicable.
2. Make a neat, simple, circuit diagram. Indicate all element and source
values. Each source should have its reference symbol.
3. Assuming that the circuit has M meshes, assign a clockwise mesh

current in each mesh, i1 , i2 , , i M .
4. If the circuit contains dependent sources, express those sources in terms

of the variables i1 , i2 , , i M , if they are not already in that form.
5. If the circuit contains current sources, temporarily modify the given

circuit by replacing each current source by an open-circuit to form
supermeshes, thus reducing the number of meshes by one for each
current source that is present. The assigned mesh currents should not be
changed. Relate each source current to the mesh currents.
6. Apply KVL around each of the meshes or supermeshes. If the circuit

has only resistors and independent voltage sources, then the equations
may be formed by inspection.

1B.30
Summary
Nodal analysis can be applied to any circuit. Apart from relating source
voltages to nodal voltages, the equations of nodal analysis are formed from
application of Kirchhoffs current law.
In nodal analysis, a supernode is formed by short-circuiting a voltage

source and treating the two ends as a single node.
Mesh analysis can only be applied to planar circuits. Apart from relating
source currents to mesh currents, the equations of mesh analysis are formed
from application of Kirchhoffs voltage law.
In mesh analysis, a supermesh is formed by open-circuiting a current

source and treating the perimeter of the original two meshes as a single
mesh.
References
Hill, 1984.

1B.31
Exercises
1.
(a) Find the value of the determinant:
2 1 0 3
1 1 0 1

4 0 3 2
3 0 0 1
(b) Use Cramers rule to find v1 , v 2 and v3 if:
2v1 35 v 2 3v3 0
2v3 3v 2 4v1 56
v 2 3v1 28 v3 0
2.
Use nodal techniques to determine ik in the circuit shown below:
20
3ik
4A 10 5 -3 A
ik

1B.32
3.
Set up nodal equations for the circuit shown below and then find the power
furnished by the 5 V source.
1/2
5V
1/4
vx
5A 1/3 1
4vx
4.
Write mesh equations and then determine i x in each of the circuits shown
below:
60 V
10 ix
10 V 4 2A 5
(a)
4V
10
8ix
6V 4
2
ix
(b)

1B.33
5.
Assign mesh currents in the circuit below, write a set of mesh equations, and
determine i.
2 k 2 k
1 k 5V
i
vx
10 V vx 1 k
6.
With reference to the circuit shown below, use mesh equations to find i A and
the power supplied by the dependent source.
2iA
100
iA
14 V 200
300
100 2V

1B.34
Gustav Robert Kirchhoff (1824-1887)
Kirchhoff was born in Russia, and showed an early interest in mathematics. He

studied at the University of Knigsberg, and in 1845, while still a student, he
pronounced Kirchhoffs Laws, which allow the calculation of current and
voltage for any circuit. They are the Laws electrical engineers apply on a
routine basis they even apply to non-linear circuits such as those containing
semiconductors, or distributed parameter circuits such as microwave striplines.
He graduated from university in 1847 and received a scholarship to study in

Paris, but the revolutions of 1848 intervened. Instead, he moved to Berlin
where he met and formed a close friendship with Robert Bunsen, the inorganic
chemist and physicist who popularized use of the Bunsen burner.
In 1857 Kirchhoff extended the work done by the German physicist Georg
Simon Ohm, by describing charge flow in three dimensions. He also analysed
circuits using topology. In further studies, he offered a general theory of how
electricity is conducted. He based his calculations on experimental results
which determine a constant for the speed of the propagation of electric charge.
Kirchhoff noted that this constant is approximately the speed of light but the
greater implications of this fact escaped him. It remained for James Clerk
Maxwell to propose that light belongs to the electromagnetic spectrum.
Kirchhoffs most significant work, from 1859 to 1862, involved his close
collaboration with Bunsen. Bunsen was in his laboratory, analysing various
salts that impart specific colours to a flame when burned. Bunsen was using
coloured glasses to view the flame. When Kirchhoff visited the laboratory, he
suggested that a better analysis might be achieved by passing the light from the
flame through a prism. The value of spectroscopy became immediately clear.
Each element and compound showed a spectrum as unique as any fingerprint,
which could be viewed, measured, recorded and compared.
Spectral analysis, Kirchhoff and Bunsen wrote not long afterward, promises
the chemical exploration of a domain which up till now has been completely
closed. They not only analysed the known elements, they discovered new

1B.35
ones. Analyzing salts from evaporated mineral water, Kirchhoff and Bunsen
detected a blue spectral line it belonged to an element they christened
caesium (from the Latin caesius, sky blue). Studying lepidolite (a lithium-
based mica) in 1861, Bunsen found an alkali metal he called rubidium (from
the Latin rubidius, deepest red). Both of these elements are used today in
atomic clocks. Using spectroscopy, ten more new elements were discovered
before the end of the century, and the field had expanded enormously
between 1900 and 1912 a handbook of spectroscopy was published by
Kayser in six volumes comprising five thousand pages!
Kirchhoffs work on spectrum analysis led on to a study of the composition of [Kirchhoff is] a
perfect example of
light from the Sun. He was the first to explain the dark lines (Fraunhofer lines) the true German
investigator. To
in the Sun's spectrum as caused by absorption of particular wavelengths as the search after truth in
light passes through a gas. Kirchhoff wrote It is plausible that spectroscopy is its purest shape and
to give utterance
also applicable to the solar atmosphere and the brighter fixed stars. We can with almost an
abstract self-
now analyse the collective light of a hundred billion stars in a remote galaxy forgetfulness, was
the religion and
billions of light-years away we can tell its composition, its age, and even how purpose of his life.
Robert von
fast the galaxy is receding from us simply by looking at its spectrum! Helmholtz, 1890.
As a consequence of his work with Fraunhofers lines, Kirchhoff developed a

general theory of emission and radiation in terms of thermodynamics. It stated
that a substances capacity to emit light is equivalent to its ability to absorb it at
the same temperature. One of the problems that this new theory created was the
blackbody problem, which was to plague physics for forty years. This
fundamental quandary arose because heating a black body such as a metal
bar causes it to give off heat and light. The spectral radiation, which depends
only on the temperature and not on the material, could not be predicted by
classical physics. In 1900 Max Planck solved the problem by discovering
quanta, which had enormous implications for twentieth-century science.
In 1875 he was appointed to the chair of mathematical physics at Berlin and he

ceased his experimental work. An accident-related disability meant he had to
spend much of his life on crutches or in a wheelchair. He remained at the
University of Berlin until he retired in 1886, shortly before his death in 1887.

2A.1
Lecture 2A Circuit Analysis Techniques

Linearity. Superposition. Source transformations. Thvenin and Norton
equivalent circuits.
Introduction
Many of the circuits that we analyse and design are linear circuits. Linear
circuits posses the property that outputs are proportional to inputs, and that
a sum of inputs leads to a sum of corresponding outputs. This is the principle
of superposition and is a very important consequence of linearity. As will be
seen later, this principle will enable us to analyse circuits with multiple sources
in an easy way.
Nonlinear circuits can be analysed and designed with graphical methods or

numerical methods (with a computer) the mathematics that describe them can
only be performed by hand in the simplest of cases. Examples of nonlinear
circuits are those that contain diodes, transistors, and ferromagnetic material.
In reality all circuits are nonlinear, since there must be physical limits to the
linear operation of devices, e.g. voltages will eventually break down across
insulation, resistors will burn because they cant dissipate heat to their
surroundings, etc. Therefore, when we draw, analyse and design a linear
circuit, we keep in mind that it is a model of the real physical circuit, and it is
only valid under a defined range of operating conditions.
In modelling real physical circuit elements, we need to consider practical

sources as opposed to ideal sources. A practical source gives a more realistic
representation of a physical device. We will study methods whereby practical
current and voltage sources may be interchanged without affecting the
remainder of the circuit. Such sources will be called equivalent sources.
Finally, through the use of Thvenins theorem and Nortons theorem, we will
see that we can replace a large portion of a complex circuit (often a
complicated and uninteresting part) with a very simple equivalent circuit, thus
enabling analysis and focus on one particular element of the circuit.

2A.2
Linearity
A linear circuit A linear circuit is one that contains linear elements, independent sources, and
defined
linear dependent sources.
A linear element is one that possesses a linear relationship between a cause and
A linear element
defined an effect. For example, when a voltage is impressed across a resistor, a current
results, and the amount of current (the effect) is proportional to the voltage (the
cause). This is expressed by Ohms Law, v Ri . Notice that a linear element
means simply that if the cause is increased by some multiplicative constant K,
then the effect is also increased by the same constant K.
If a linear elements relationship is graphed as cause vs. effect, the result is

a straight line through the origin. For example, the resistor relationship is:
A linear relationship
is defined by a
straight line through
the origin v
R
1
Figure 2A.1
A linear dependent source is one whose output voltage or current is

A linear dependent
source defined proportional only to the first power of some current or voltage variable in the
circuit (or a sum of such quantities). For example, a dependent voltage source
given by v s 0.6i1 14v 2 is linear, but v s 0.6i12 and v s 0.6i1v 2 are not.
From the definition of a linear circuit, it is possible to show that the response
Output is
proportional to input
is proportional to the source, or that multiplication of all independent sources
for a linear circuit by a constant K increases all the current and voltage responses by the same
factor K (including the dependent source outputs).

2A.3
Superposition
The linearity property of a circuit leads directly to the principle of
superposition. To develop the idea, consider the following example:
Example
We have a 3-node circuit:
v1 5 v2
ia 2 1 ib
There are two independent current sources which force the currents ia and ib
into the circuit. Sources are often called forcing functions for this reason, and
the voltages they produce at each node in this circuit may be termed response
functions, or simply responses.
The two nodal equations for this circuit are:
0.7v1 0.2v 2 ia
0.2v1 1.2v 2 ib
Now we perform experiment x. We change the two current sources to iax and
ibx ; the two unknown node voltages will now be different, and we let them be
v1x and v 2 x . Thus:
0.7v1x 0.2v 2 x iax

0.2v1x 1.2v 2 x ibx

2A.4
If we now perform experiment y by changing the current sources again, we get:
0.7v1 y 0.2v 2 y iay

0.2v1 y 1.2v 2 y iby
We now add or superpose the two results of the experiments:
0.7v1x v1 y 0.2v 2 x v 2 y iax iay

0.2v1x v1 y 1.2v 2 x v 2 y ibx iby
Compare this with the original set of equations:
0.7v1 0.2v 2 ia
0.2v1 1.2v 2 ib
We can draw an interesting conclusion. If we let iax iay ia , ibx iby ib ,
then the desired responses are given by v1 v1x v1 y and v 2 v 2 x v 2 y . That
is, we may perform experiment x and note the responses, perform experiment y
and note the responses, and finally add the corresponding responses. These are
the responses of the original circuit to independent sources which are the sums
of the independent sources used in experiments x and y.
Superposition allows This is the fundamental concept involved in the superposition principle. It is
us to treat inputs
evident that we may break an independent source into as many pieces as we
separately, then
combine individual wish, so long as the algebraic sum of the pieces is equal to the original source.
responses to obtain
the total response
In practical applications of the superposition principle, we usually set each
independent source to zero, so that we can analyse the circuit one source at a
time.

2A.5
Superposition Theorem
We can now state the superposition theorem as it is mostly applied to circuits:

The superposition
In any linear network containing several sources, we can theorem
calculate any response by adding algebraically all the

(2A.1)
individual responses caused by each independent source
acting alone, with all other independent sources set to zero.
When we set the value of an independent voltage source to zero, we create a

short-circuit by definition. When we set the value of an independent current
source to zero, we create an open-circuit by definition.
Setting a voltage
source to zero
creates a short-
circuit. Setting a
current source to
0V S.C. 0A O.C. zero creates an
open-circuit.
Figure 2A.2
Note that dependent sources cannot be arbitrarily set to zero, and are generally
active when considering every individual independent source.
The theorem as stated above can be made much stronger a group of

independent sources may be made active and inactive collectively. For
example, sometimes it is handy to consider all voltage sources together, so that
mesh analysis can be applied easily, and then all current sources together so
that nodal analysis may be applied easily.
There is also no reason that an independent source must assume only its given
value or zero it is only necessary that the sum of the several values be equal
to the original value. However, an inactive source almost always leads to the
simplest circuit.

2A.6
Example
We use superposition in the following circuit to write an expression for the

unknown branch current i x .
6
ix
vs = 3 V 9 is = 2 A
We first set the current source equal to zero (an open-circuit) and obtain the
portion of i x due to the voltage source as 0.2 A. Then if we let the voltage
source be zero (a short-circuit) and apply the current divider rule, the
remaining portion of i x is seen to be 0.8 A.
We may write the answer in detail as:
3 6
ix ix ix 2 0 .2 0 .8 1 A
is 0 vs 0
69 69

2A.7
Example
The circuit below contains a dependent source.
2 ix 1
10 V v 3A 2ix
We seek i x , and we first open-circuit the 3 A source. The single mesh equation
is:
10 2i x 1i x 2i x 0
so that:
i x 2
Next, we short-circuit the 10 V source and write the single node equation:
v v 2i x
3 0
2 1
and relate the dependent-source-controlling quantity to v :
v 2i x
We find:
i x 0.6
and thus:
i x i x i x 2 0.6 1.4 A
It usually turns out that little, if any, time is saved in analysing a circuit
containing dependent sources by use of the superposition principle, because
there are at least two sources in operation: one independent source and all the
dependent sources.

2A.8
Superposition is We must be aware of the limitations of superposition. It is applicable only to
often misapplied to
power in a circuit linear responses, and thus the most common nonlinear response power is
element
not subject to superposition.
Example
The circuit below contains two 1 V batteries in series.
1V 1V
If we apply superposition, then each voltage source alone delivers 1 A and

furnishes 1 W. We might then mistakenly calculate the total power delivered to
the resistor as 2 W. This is incorrect.
Each source provides 1 A, making the total current in the resistor 2 A. The
power delivered to the resistor is therefore 4 W.

2A.9
Source Transformations
Practical Voltage Sources
The ideal voltage source is defined as a device whose terminal voltage is

independent of the current through it. Graphically, its characteristic is:
An ideal voltage
source, and its
terminal
i v characteristic
vs
vs v
ideal source terminal characteristic i
Figure 2A.3
The ideal voltage source can provide any amount of current, and an unlimited
amount of power. No such device exists practically. All practical voltage
sources suffer from a voltage drop when they deliver current the larger the
current, the larger the voltage drop. Such behaviour can be modelled by the
inclusion of a resistor in series with an ideal voltage source:
A practical voltage
source, and its
terminal
R sv i characteristic
v
vs
vs v
R sv
1
practical source terminal characteristic i
Figure 2A.4

2A.10
The terminal characteristics of the practical voltage source are given by KVL:
The terminal
characteristic of a
practical voltage v vs Rsvi (2A.2)
source
The resistance Rsv is known as the internal resistance or output resistance.

shows the effect This resistor (in most cases) is not a real physical resistor that is connected in
of the internal
resistance series with a voltage source it merely serves to account for a terminal voltage
which decreases as the load current increases.
The applicability of this model to a practical source depends on the device and
the operating conditions. For example, a DC power supply such as found in a
laboratory will maintain a linear relationship in its terminal characteristics over
a larger range of currents than a chemical battery.
When we attach a load to a practical voltage source:

A load attached to a
practical voltage
source will always R sv i
exhibit less voltage L
and current than the
ideal case
vs vL RL
Figure 2A.5
we get a load voltage which is always less than the open-circuit voltage, and
given by the voltage divider rule:
RL
vL vs vs (2A.3)
Rsv RL
The load current will also be less than we expect from an ideal source:
vs v
iL s (2A.4)
Rsv RL RL

2A.11
Practical Current Sources
The ideal current source is defined as a device whose current is independent of

the voltage across it. Graphically, its characteristic is:
An ideal current
source, and its
terminal
v characteristic
i
is v is
i
ideal source terminal characteristic
Figure 2A.6
The ideal current source can support any terminal voltage regardless of the load
resistance to which it is connected, and an unlimited amount of power. An
ideal current source is nonexistent in the real world. For example, transistor
circuits and op-amp circuits can deliver a constant current to a wide range of
load resistances, but the load resistance can always be made sufficiently large
so that the current through it becomes very small. Such behaviour can be
modelled by the inclusion of a resistor in parallel with an ideal current source:
A practical current
source, and its
terminal
i v characteristic
R si i s
is R si v
R si
1
is
practical source terminal characteristic
i
Figure 2A.7

2A.12
The terminal characteristics of the practical current source are given by KCL:
The terminal
characteristic of a v
practical current i is
source Rsi
or v Rsiis Rsii (2A.5)
When we attach a load to a practical current source:

A load attached to a
practical current
source will always
exhibit less voltage iL
and current than the
ideal case
is R si v L RL
Figure 2A.8
we get a load current which is always less than the short-circuit current, and
given by the current divider rule:
Rsi
iL is is (2A.6)
Rsi RL
The load voltage will also be less than we expect from an ideal source:
Rsi RLis
vL RLis (2A.7)
Rsi RL

2A.13
Practical Source Equivalence
We define two sources as being equivalent if each produces identical current

and identical voltage for any load which is placed across it terminals. With
reference to the practical voltage and current source terminal characteristics:
v v
vs R si i s
R sv R si
1 1
is
practical voltage source
i practical current source
i
Figure 2A.9
we can easily establish the conditions for equivalence. We must have:

If practical sources
Rsv Rsi Rs (2A.8) are equivalent then

they have the same
internal resistance
so that the slopes of the two terminal characteristics are equal. We now let Rs
represent the internal resistance of either practical source. To achieve the same
axes intercepts, we must have:
The relationship
between a practical
vs Rsis (2A.9)
voltage source and
a practical current
source

2A.14
We can now transform between practical voltage and current sources:

The equivalence of
practical sources
Rs i i
vs v is Rs v
= Rs is v
= s
Rs
practical voltage source practical current source
Figure 2A.10
Example
Consider the practical current source shown below:
3A 2
Since its internal resistance is 2 , the internal resistance of the equivalent

practical voltage source is also 2 . The voltage of the ideal voltage source
contained within the practical voltage source is v s Rs i s 2 3 6 V . The
equivalent practical voltage source is shown below:
6V

2A.15
Maximum Power Transfer Theorem
Consider a practical DC voltage source:
Rs IL
Vs VL RL
Figure 2A.11
The power delivered to the load RL is:
RLVs2
PL R I 2
Rs RL 2
L L (2A.10)
A graph of the load power vs. load resistance is shown below:
A graph of the
power delivered to a
PL load versus the load
resistance shows
clearly that a peak
occurs at a certain
PLmax resistance
R L max RL
Figure 2A.12

2A.16
To find the value of RL that absorbs maximum power from the practical
source, we differentiate with respect to RL :
PL Rs RL Vs2 Vs2 RL 2Rs RL

2
(2A.11)
RL Rs RL 4
and equate the derivative to zero to obtain the relative maximum:
Rs RL 2 2RL Rs RL 0 (2A.12)
or:
The load resistance
which maximizes
power delivered
from a practical
RL Rs (2A.13)
source
Since the values RL 0 and RL both give a minimum ( PL 0 ), then this

value is the absolute maximum.
Since we have already proved the equivalence between practical voltage and
current sources, we have proved the following maximum power transfer
theorem:
The maximum
power transfer An independent voltage source in series with a resistance Rs ,
theorem
or an independent current source in parallel with a resistance Rs ,
(2A.14)
delivers a maximum power to that load resistance R L when
R L Rs .
We can only apply the maximum power transfer theorem when we have
only applies to a
control over the load resistance, i.e. if we know the source resistance, then we
choice of load
resistor can choose RL Rs to maximize power transfer. On the other hand, if we are
given a load resistance and we are free to design or choose a source resistance,
we do not choose Rs RL to maximize power transfer by examining
Eq. (2A.10), we see that in this case we should choose Rs 0 .

2A.17
Thvenins and Nortons Theorem

Thvenins and Nortons theorems greatly simplify the analysis of many linear
circuits. Lon Thvenin was a French engineer working in telegraphy who first
published a statement of the theorem in 1883. Edward Norton was a scientist
with the Bell Telephone Laboratories who mentioned his theorem in a
technical memorandum in 1926, but never published it.
Thvenins theorem tells us that it is possible to replace a large portion of a

linear circuit (often a complicated and / or uninteresting portion) by an
equivalent circuit containing only an independent voltage source in series with
a resistor:
Thvenins theorem
allows us to replace
part of a circuit with
R Th a practical voltage
source
large rest rest
portion
of vTh of
of
linear circuit circuit
circuit
Thvenin equivalent circuit
Figure 2A.13
Nortons theorem is the dual of Thvenins theorem, and uses a current source:
Nortons theorem
allows us to replace
part of a circuit with
a practical current
large rest rest source
portion
of iN RN of
of
linear circuit circuit
circuit
Norton equivalent circuit
Figure 2A.14

2A.18
We can state Thvenins theorem formally as:
Thvenins theorem
Given any linear circuit, rearrange it in the form of two circuits
A and B that are connected together at two terminals.
If either circuit contains a dependent source, its control variable
must be in the same circuit.
Define a voltage voc as the open-circuit voltage which would
(2A.15)
appear across the terminals of A if B were disconnected so that
no current is drawn from A. Then all the currents and voltages in
B will remain unchanged if A has all its independent sources set
to zero and an independent voltage source voc is connected in
series with the inactive A network.
The inactive circuit A will always reduce to a single resistor, which we call the
Thvenin resistance, RTh . Also, since voc appears as an independent voltage
source in the Thvenin equivalent circuit, it is also denoted as vTh .
We can state Nortons theorem formally as:

Nortons theorem
Given any linear circuit, rearrange it in the form of two circuits
A and B that are connected together at two terminals.
If either circuit contains a dependent source, its control variable
must be in the same circuit.
Define a current i sc as the short-circuit current which would
(2A.16)
appear across the terminals of A if B were short-circuited so that
no voltage is provided by A. Then all the voltages and currents in
B will remain unchanged if A has all its independent sources set
to zero and an independent current source i sc is connected in
parallel with the inactive A network.
The inactive circuit A will always reduce to a single resistor, which we call the
Norton resistance, R N . Also, since i sc appears as an independent current
source in the Norton equivalent circuit, it is also denoted as i N .

2A.19
The Thvenin or Norton equivalent circuit allows us to draw a new simpler
circuit and make rapid calculations of the voltage, current, and power. It also
allows us to easily choose a load resistance for maximum power transfer.
Example
3 7
12 V 6 RL
Circuit A Circuit B
The broken lines separate the original circuit into circuits A and B. We shall
assume that our main interest is in circuit B, which consists only of a load
resistor R L . To form the Thvenin equivalent circuit, we disconnect circuit B
and use voltage division to determine that voc 8 V . When we set all
independent sources in circuit A to zero, we replace the 12 V source with a
short-circuit. Looking back into the inactive A circuit, we see a 7
resistor connected in series with the parallel combination of 6 and 3 .
Thus, the inactive A circuit can be represented by a 9 resistor. If we now
replace circuit A by its Thvenin equivalent circuit, we have:
8V RL
Thvenin equivalent of Circuit B

Circuit A

2A.20
Note that the Thvenin equivalent circuit we have obtained for circuit A is
completely independent of circuit B an equivalent for A may be obtained no
matter what arrangement of elements is connected to the A circuit, even if
circuit B is nonlinear!
From the viewpoint of the load resistor R L , the Thvenin equivalent circuit is
identical to the original; from our viewpoint, the circuit is much simpler and we
can now easily compute various quantities. For example, the power delivered
to the load is:
2
8
p L R L
9 RL
Furthermore, we can now easily see that the maximum voltage which can be
obtained across R L is 8 V when R L . A quick transformation of the
Thvenin equivalent circuit to a practical current source (the Norton
equivalent) indicates that the maximum current which may be delivered to the
load is 8 9 A for R L 0 . The maximum power transfer theorem shows that a
maximum power is delivered to R L when R L 9 . None of these facts is
readily apparent from the original circuit.

2A.21
To form the Norton equivalent circuit, we short-circuit the B circuit and use the
current divider rule to discover:

6 12 72 72 8
i sc A
67 6 7 39 42 81 9
3
67
When we set all independent sources in circuit A to zero, we get the same
results as for the Thvenin circuit, and so R N 9 . The Norton equivalent
circuit is therefore:
8/9 A 9 RL
Norton equivalent of Circuit B

Circuit A
It should be apparent from the previous example that we can easily find the
Norton equivalent circuit from the Thvenin equivalent circuit, and vice versa,
by a simple source transformation. Using our previous results, we must have:
The Thvenin and
RTh RN (2A.17) Norton equivalent
resistances are the
same
Because of this result, we usually just refer to the resistor in either equivalent
circuit as the Thvenin resistance, RTh .
We also have:
The relationship
between the
voc RThisc (2A.18) Thvenin and
Norton equivalent
circuits

2A.22
Example
2 3
4V 2A 1
Circuit A Circuit B
The Thvenin and Norton equivalent circuits are desired from the perspective
of the 1 resistor. We determine RTh for the inactive network, and then find
either v oc or i sc . Making the independent sources inactive, we have:
2 3
We thus have RTh 5 . The open-circuit voltage can be determined using

superposition. With only the 4 V source operating , the open-circuit voltage is
4 V. When only the 2 A source is on, the open-circuit voltage is also 4 V. Thus,
with both sources operating, we have voc 8 V . This determines the Thvenin
equivalent, and from it the Norton equivalent, shown below:
8V 1 1.6 A 5 1

2A.23
We will often find it convenient to determine either the Thvenin or Norton
equivalent by finding both the open-circuit voltage and the short-circuit
current, and then determining the Thvenin resistance by:
The Thvenin
equivalent
voc resistance can be
RTh (2A.19) obtained from the
isc open-circuit voltage
and short-circuit
current
This is especially true for circuits that contain dependent sources.
Example

2 3
vx vx
4V 4
To find v oc we note that v x voc , and that the dependent source current must
pass through the 2 resistor since there is an open circuit to the right. KCL at
the top of the dependent source gives:
voc 4 voc
0
2 4
voc 8 V
Upon short-circuiting the output terminals, it is apparent that v x 0 and the
dependent current source is zero. Hence i sc 4 2 3 0.8 A . Thus
RTh v oc i sc 8 0.8 10 and the Thvenin equivalent is:
10
8V

2A.24
If a circuit contains only dependent sources and no independent sources, then
If a circuit does not
contain independent the circuit qualifies as the inactive A circuit and voc 0 and i sc 0 . We seek
sources, then one
must be applied to the value of RTh represented by the circuit, but in this case RTh v oc i sc is
obtain the Thvenin
equivalent undefined, and we must use a different approach. We apply either an
resistance
independent voltage source or an independent current source externally to the
A circuit and measure the resultant current or voltage.
Example

3 i
1.5 i 2
If we apply a 1 A source externally, and measure the resultant voltage, then

RTh v 1 :
3 i
1.5 i 2 v 1A
We see that i 1 A and KCL gives:
v 1.5 1 v
1 0
3 2
so that v 0.6 V and RTh 0.6 . The Thvenin equivalent is:
0.6

2A.25
Summary of Finding Thvenin Equivalent Circuits
We have seen three approaches to finding the Thvenin equivalent circuit. The
first example contained only independent sources and resistors, and we could
use several different methods on it. One involved finding v oc for the active
circuit, and then RTh for the inactive circuit. We could also have found i sc and
RTh , or v oc and i sc . In the second example, both independent and dependent
sources were present, and the method we used required us to find v oc and i sc .
The last example did not contain any independent sources, and we found RTh
by applying a 1 A source and finding RTh v 1 . We could also have applied a
1 V source and determined RTh 1 i .
These important techniques and the types of circuits to which they may be
applied most readily are indicated in the table below:
Suitable methods to
Circuit contains obtain the Thvenin
equivalent circuit
Methods
RTh and v oc or i sc
v oc and i sc Possible
i 1 A or v 1 V
Table 2A.1

2A.26
All possible methods do not appear in the table. Another method has a certain
A methods to obtain
the Thvenin appeal because it can be used for any of the three types of circuit tabulated.
equivalent circuit
that works for all
Simply label the terminals of the A circuit as v, define the current leaving the
circuits positive polarity as i, then analyse the A circuit to obtain an equation in the
form v v oc RTh i .
Example
2 v1 3 i
4V 2A v
KCL at the middle node gives:

v1 4
2i 0
2
v1 8 2i
KVL at the output gives:
v v1 3i
8 5i
v oc RTh i
from which voc 8 V and RTh 5 , as before:
5 i
8V v

2A.27
Example

2 v1 3 i
v v
4V 4
KCL at the middle node gives:
v1 4 v
i 0
2 4
v
v1 4 2i
2
KVL at the output gives:
v v1 3i
v
4 5i
2
8 10i
voc RTh i
from which voc 8 V and RTh 10 , as before:
10
8V

2A.28
Example

3 i
1.5 i 2 v
KCL at the top right node gives:
v 1.5i v
i 0
3 2
5v 3i 0
v 0 0.6i
voc RTh i
from which voc 0 V and RTh 0.6 , as before:
0.6
This procedure is universally applicable, but one of the other methods is

usually easier and quicker to perform.

2A.29
Summary
A linear circuit is one that contains linear elements, independent sources,
and linear dependent sources. For a linear circuit, it is possible to show that
the response is proportional to the source.
The superposition theorem states that, in evaluating the response in a

linear circuit due to several sources, we are free to treat each independent
source separately, collectively, or in any number of parts, and then
superpose the response caused by each part.
A practical voltage source consists of an ideal voltage source v s in series
with a resistance Rsv . A practical current source consists of an ideal current
source i s in parallel with a resistance Rsi . The practical sources can be
made equivalent by setting Rsv Rsi Rs and vs Rsis .
The maximum power transfer theorem states that if we know the source
resistance Rs of a practical source, then to maximize power transfer to a
load RL , we set RL Rs .
Thvenins theorem tells us that it is possible to replace a large portion of a

linear circuit by an equivalent circuit containing only an independent
voltage source in series with a resistor.
Nortons theorem tells us that it is possible to replace a large portion of a

linear circuit by an equivalent circuit containing only an independent
current source in parallel with a resistor.
There are several methods that can be applied to determine a Thvenin or

Norton equivalent circuit. Some methods are only applicable to certain
circuits, and the most convenient analysis method should be chosen.
References
Hill, 1984.

2A.30
Exercises
1.
Find the power dissipated in the 20 resistor of the circuit shown below by
each of the following methods:
(a) nodal equations
(b) mesh equations
(c) source transformations to eliminate all current sources, then a method of

your choice
(d) source transformations to eliminate all voltage sources, followed by any

method you wish.
1 20
27 V 4 5 6A
2.
The circuit shown below contains a dependent source. Use superposition to
find I .
-4 V
3
1
Vx 2
5Vx
2A
I

2A.31
3.
Consider the linear circuit shown below.
vx
Linear
i s1 i s2
Circuit
(a) The circuit contains only resistors. If i s1 8 A and i s 2 12 A , v x is found
to be 80 V. However, if is1 8 A and is 2 4 A , then vx 0 V . Find v x
when i s1 i s 2 20 A .
(b) The circuit now contains a source such that vx 40 V when
is1 is 2 0 A . All data in part (a) are still correct. Find v x when
is1 is 2 20 A .
4.
1 6
a
18 A 2 12 18 V
b
(a) Find the Norton equivalent of the circuit.
(b) If a variable resistor R were placed between terminals a and b, what value
would result in maximum power being drawn from the terminals?
(c) Find the maximum power that could be drawn from terminals a and b.

2A.32
5.
Find the maximum power that can be delivered to a variable R in the circuit
below:
1 2
20 V
R
3 4
6.
In the circuit below, what value of resistance should be connected between
terminals a-b to draw maximum power?
vx 1
2
a
5vx 3 4 9 sin(120 t ) V
b

2A.33
7.
Consider the circuit below:
rm iA
iA R2
R1 IS R3
(a) Determine an expression for the current i A .
(b) If rm R2 , what is the current i A ?
(c) What would happen if you tried to build such a circuit with rm R2 ?

2B.1
Lecture 2B Linear Op-Amp Applications

Summing amplifier. Difference amplifier. Inverting integrator. Differentiator.
Negative impedance converter. Voltage-to-current converter. Noninverting
integrator.
Introduction
One of the reasons for the popularity of the op-amp is its versatility. As we
shall see shortly, you can do almost anything with op-amps! More importantly,
the IC op-amp has characteristics that closely approach the assumed ideal. This
implies that it is quite easy to design circuits using the IC op-amp. It also
means that a real op-amp circuit will work in a manner that is very close to the
predicted theoretical performance.
Op-amp circuits which exhibit a linear characteristic (i.e. the response is

proportional to the source) are known as linear op-amp circuits. They are
chiefly, but not necessarily, composed of linear circuit elements, such as
resistors and capacitors, Many important op-amp circuits can be created with
linear circuit elements. For example, the weighted summer is able to output a
voltage which is a weighted sum of the input signals. The difference amplifier
is able to subtract two signals. Integrators, of both the inverting and
noninverting variety, are used in many control and signal processing
applications.
Also, by exploiting the active nature of op-amp circuits, we can develop

circuits which have no passive equivalent, such as the negative impedance
converter, and the voltage-to-current converter.

2B.2
Summing Amplifier
An inverting
summing amplifier
Rn
vn
R2 Rf
v2
R1
v1
vo
Figure 2B.1 An Inverting Summing Amplifier
The basic configuration is that of an inverting amplifier. We have a resistance

R f in the negative feedback path, but there are a number of input signals,
v1 , v 2 , , v n each applied to a corresponding resistor R1 , R2 , , Rn which
are connected to the inverting terminal of the op-amp. Just like in the inverting
amplifier configuration, the op-amp and the negative feedback will maintain a
virtual short-circuit across the op-amp input terminals, and therefore maintain
the inverting terminal at 0 V. Ohms Law then tells us that the currents
i1 , i2 , , in are given by:
v1 v2 vn
i1 ,
i2 , ,
in (2B.1)
R1 R2 Rn

2B.3
All these currents sum together at the inverting terminal, also known as the
summing junction, to produce the current i:
i i1 i2 in (2B.2)
Remembering that the ideal op-amp input terminals are open-circuits,

application of KCL at the inverting terminal shows that this current is forced
through R f . The output voltage v o may now be determined by an application
of Ohms Law and KVL:
vo 0 R f i R f i (2B.3)
Thus:
R R R
vo f v1 f v2 f vn
(2B.4)
R1 R2 Rn
The output voltage is a weighted sum of the input signals v1 , v 2 , , v n . This
circuit is therefore called a weighted summer. Note that each summing

coefficient may be independently adjusted by varying the corresponding feed-
in resistor ( R1 to Rn ). This nice property is a consequence of the virtual
common that exists at the inverting op-amp terminal.

2B.4
Example
We want to design a weighted summer such that vo 2v1 v 2 4v3 .
We can firstly form the sum vo1 2v1 v 2 by using the circuit:
20 k 20 k
v2
10 k
v1
vo 1
Then we can form vo vo 1 4v3 using:
2.5k 10 k
v3
10 k
vo 1
vo
Finally, we can put the two circuits together, in a cascade, to create:
20 k 20 k
v2 2.5k 10 k
v3
10 k
v1 10 k
vo

2B.5
Difference Amplifier
A simple difference amplifier can be constructed with four resistors and an
op-amp, as shown below:
The difference
amplifier
R2
R1
v i1
R3
v i2
vo
R4
Figure 2B.2 A Difference Amplifier
There are a number of ways to find the output voltage, but the easiest uses the
principle of superposition (since the circuit is linear). To apply superposition
we first reduce vi 2 to zero that is, connect the terminal to which vi 2 is applied
to the common and then find the corresponding output voltage, which will be
due entirely to vi1 . We denote this output v o1 , as shown in (a) below:
Analyzing a
difference amplifier
R2 R2 using superposition
R1 R1
v i1
vo1 vo2
R3
v i2
R3 R4 R4
(a) (b)

2B.6
Figure 2B.3 Analyzing a Difference Amplifier
We recognize the (a) circuit as that of an inverting amplifier. The existence of

R3 and R4 does not affect the gain expression, since there is no current in
either of them. Thus:
R2
vo1 vi1
R1 (2B.5)
Next, we reduce vi1 to zero and evaluate the corresponding output voltage vo 2 .
The circuit will now take the form shown in Figure 2B.3(b), which we
recognize as the noninverting configuration with an additional voltage divider,
made up of R3 and R4 , connected across the input vi 2 . The output voltage vo 2
is therefore given by:
R R4
vo 2 1 2 vi 2
R1 3R R4
(2B.6)
The superposition principle tells us that the output voltage v o is equal to the
sum of v o1 and vo 2 . Thus we have:
R2 1 R2 R1
vo vi1 vi 2
R1 1 R3 R4 (2B.7)
To act as a difference amplifier, the output must produce a signal which is

proportional to vi 2 vi1 . That is, the coefficients of vi1 and vi 2 in Eq. (2B.7)
must be equal in magnitude but opposite in sign. This requirement leads to the
condition:
R2 1 R2 R1

R1 1 R3 R4 (2B.8)

2B.7
Simplifying we get:
R2 R4

R1 R3 (2B.9)
Substituting in Eq. (2B.7) results in the output voltage:
The output of a
R2
vi 2 vi1
vo (2B.10)
R1
Thus, if we choose R4 R3 R2 R1 then we have produced a difference
amplifier with a gain of R2 R1 .
Example
A difference amplifier with an input resistance of 20 k and a gain of 10 is

shown below:
100 k
10 k
vi 10 k v o =10 v i
100 k
Note that input resistance is defined as the resistance seen between the two
input terminals. Thanks to the virtual short-circuit at the op-amp input
terminals, KVL around the input resistors gives Rin 20 k .

2B.8
Inverting Integrator
An inverting integrator, also known as the Miller integrator after its inventor, is
shown below:
The inverting
integrator
R
vi
vo
Figure 2B.4 The Miller Integrator
We use the virtual short circuit concept to analyze the gain (it is essentially the
same analysis as for the inverting amplifier). KCL at the inverting terminal,
which is held at 0 V by the op-amp and negative feedback, gives:
vi dv
C o (2B.11)
R dt
and therefore:
1 t
vo t vi t dt vo 0 (2B.12)
RC 0
Thus, vo t is the time integral of vi t , and vo 0 is the initial condition of
this integration process. The product RC has units of time and is called the
integration time constant. This integrator circuit is inverting because of the
minus sign in front of the integral.

2B.9
Unfortunately, for DC input voltages, the op-amp circuit is operating as an
open-loop, which is easily seen when we recall that capacitors behave as open
circuits with DC. Thus, the above analysis does not apply, since there is no
virtual short circuit. With an ideal op-amp and a DC input voltage, there is no
current and vi appears at the inverting terminal the resulting output of the
ideal op-amp will be infinite. In practice the output of a real op-amp will
saturate close to one of the supply rail voltages (depending on the sign of the
DC input voltage).
A practical circuit that alleviates this problem is shown below:
Rf
R
vi
vo
Figure 2B.5 A Practical Approximation to the Miller Integrator
This circuit provides a feedback path for DC voltages (i.e. the op-amp circuit is
operating in a closed-loop) and prevents the output from saturating. To keep
the DC offset at the output of the integrator low, we should select a small R f .
Unfortunately, however, the lower the value of R f , the less ideal the integrator
becomes. Thus selecting a value for R f is a trade-off between DC
performance and integrator performance.

2B.10
Differentiator
An ideal differentiator is shown below:
The differentiator
C
vi
vo
Figure 2B.6 The Differentiator
We use the virtual short circuit concept again to analyze the circuit. KCL at the
inverting terminal, which is held at 0 V by the op-amp and negative feedback,
gives:
dvi v
C o (2B.13)
dt R
and therefore:
vo t RC
dvi
(2B.14)
dt
Thus, vo t is the time derivative of vi t .
The very nature of a differentiator circuit causes it to be a noise magnifier.

This is due to spikes introduced at the output every time there is a sharp change
in the input. For this reason they are rarely used in practice.

2B.11
Negative Impedance Converter

A negative impedance1 converter (NIC) is shown below:
The negative
impedance
converter
vi
R1 R2
vi
R1 R1
vi vo
ii R
vi
R2 1
vi
R1 R
R in
Figure 2B.7 The Negative Impedance Converter
To investigate the operation of this circuit, we will evaluate the input resistance
Rin of the circuit. To find Rin we apply an input voltage vi and evaluate the
input current ii . Then, by definition, Rin vi ii .
Owing to the virtual short circuit between the op-amp input terminals, the
voltage at the inverting terminal will be equal to vi . The current through R1
will therefore be vi R1 . Since the input resistance of the ideal op-amp is
1
Impedance generalises the concept of resistance as will be seen later with the introduction
of sinusoidal steady-state analysis, phasors, and reactance.

2B.12
infinite, the current through R2 will also be vi R1 . Thus, the voltage at the op-
amp output will be:
vi R2
vo vi R2 1 vi (2B.15)
R1 R1
which we recognise as the output of a normal noninverting amplifier. We now

apply Ohms Law to R and obtain the current through it as:
v1 R2 R1 v R 1
v 2 (2B.16)
R R1 R
Since there is no current into the positive input terminal of the op-amp, KCL
gives:
v R2
ii (2B.17)
R R1
Thus:
R1
Rin R (2B.18)
R2
That is, the input resistance is negative with a magnitude equal to R, the
resistance in the positive-feedback path, multiplied by the ratio R1 R2 . We
now see why the circuit is called a negative impedance converter (NIC), where
R may in general be replaced by an arbitrary circuit element, such as a
capacitor or inductor.

2B.13
Voltage-to-Current Converter
To investigate the NIC further, consider the case R1 R2 r , where r is an
arbitrary value. It follows that Rin R . Let the input be fed with a voltage
source Vs having a source resistance equal to R, as shown below:
A voltage-to-current
converter, invented
by Prof. Bradford
r Howland, MIT,
around 1962
R R
Vs
IL
VL
RL
Figure 2B.8 Application of the Negative Impedance Converter
Utilizing the information gained previously, we can replace the circuit in the
dashed box by a resistance equal to R :
R
Vs
IL
RL -R
Figure 2B.9

2B.14
The figure below illustrates the conversion of the voltage source to its Norton
equivalent:
IL
Vs
R R RL -R
Figure 2B.10
Finally, the two parallel resistances R and R are combined to produce an

infinite resistance, resulting in the equivalent circuit:
I L = Vs
R
Vs
RL
R
Figure 2B.11
We see that the load current is given by:
Vs
IL (2B.19)
R
independent of the value of RL !. This is an interesting result; it tells us that the
circuit of Figure 2B.8 acts as a voltage-to-current converter, providing a
current I L that is directly proportional to Vs and is independent of the value of
the load resistance. That is, the output terminal acts as a current-source output,
with the impedance looking back into the output terminal equal to infinity.
Note that this infinite resistance is obtained via the cancellation of the positive
source resistance R with the negative input resistance R .

2B.15
Noninverting Integrator
A specific application of the voltage-to-current converter, where a capacitor C
is used as a load, is illustrated below:
A noninverting
integrator, invented
by Gordon Deboo in
r 1966 whilst working
at the NASA Ames
Research Centre
r
vo
R R
vi
vi
i=
R vL
Figure 2B.12
From the previous analysis of the voltage-to-current converter we conclude that

capacitor C will be supplied by a current i vi R . From the branch
relationship for a capacitor we conclude:
dvL vi
C (2B.20)
dt R
and thus:
1 t
vL t vi t dt vL 0
RC 0
(2B.21)
Thus, v L t is the time integral of vi t .

2B.16
This circuit has some interesting properties. The integral relationship does not
have an associated negative sign, as is the case with the Miller integrator.
Noninverting, or positive, integrators are required in many applications.
Another useful property is the fact that one terminal of the capacitor is
connected to common. This would simplify the initial charging of the capacitor
to v L 0 , as may be necessary to set an initial condition.
The output of the circuit cannot be taken at the terminal labelled v L since the
connection of a load there will change the preceding analysis. Fortunately, a
voltage source output is available that is proportional to v L at the output of
the op-amp where you can easily verify that vo 2v L . Thus the output of the
circuit is:
2 t
vo t vi t dt vo 0
RC 0
(2B.22)

2B.17
Summary
The op-amp is a versatile electronic building block. Real op-amps perform
close to the ideal, making circuit design and verification relatively easy.
The inverting summing amplifier:
Rn
vn
R2 Rf
v2
R1
v1
vo
has an output equal to:
R R R
vo f v1 f v2 f vn
R1 R2 Rn
The difference amplifier:
R2
R1
v i1
R1
v i2
vo
R2
vo
R2
vi 2 vi1
R1
2B.18
The inverting integrator:
R
vi
vo
is a poor implementation of an integrator since the op-amp is operating in

an open-loop for DC input voltages, causing the output to saturate. A
practical approximation that alleviates this problem is:
Rf
R
vi
vo
The differentiator:
C
vi
vo
vo t RC
dvi
dt
It is rarely used in practice, because it tends to act as a noise magnifier.

2B.19
A negative impedance converter (NIC):
R2
R1
R
vi
R in
has an input resistance given by:
R1
Rin R
R2
The Howland voltage-to-current converter:
R R
Vs
IL
VL
RL
creates a voltage-controlled current source where:
Vs
IL
R

2B.20
The Deboo noninverting integrator:
r
vo
R R
vi
has an output:
2 t
vo t vi t dt vo 0
RC 0
It is a very practical and useful integrator circuit.
References
Deboo, Gordon: A Novel Integrator, NASA-TM-X-57906, NASA Ames
Research Center, 1966.
Sedra, A. and Smith, K.: Microelectronic Circuits, Saunders College

Publishing, New York, 1991.
Sheingold, D.: Impedance & Admittance Transformations, The Lightning

Empiricist, vol. 12, no. 1, Philbrick Researches, Inc., Boston, 1964.

2B.21
Exercises
1.
Design an op-amp circuit with a 10 k input resistance which converts a
symmetrical square wave at 1 kHz having 2 V peak-to-peak amplitude and zero
average value into a triangle wave of 2 V peak-to-peak amplitude.
2.
Design a two op-amp circuit with inputs v1 and v 2 and input resistances of
100 k whose output is vo v1 10v 2 .
3.
Determine the output voltage of the following circuit:
60 k
30 k
v3
20 k vo
v2
20 k 20 k
v1
4.
Design a negative impedance converter having an input resistance of -1 k.
This circuit is connected to the output terminal of a source whose open-circuit
voltage is 1 mV and whose output resistance is 900 . What voltage is then
measured at the output of the source?

2B.22
5.
Express io as a function of vi for the circuit below:
R1
R1
vi
R1 R1
R
io
RL

3A.1
Lecture 3A Reactive Components

The capacitor. The inductor. Practical capacitors and inductors. Series and
parallel connections of inductors and capacitors. Circuit analysis with
inductors and capacitors. Duality.
Introduction
The capacitor is a circuit element whose voltage-current relationship involves
the rate of change of voltage. Physically, a capacitor consists of two conducting
surfaces on which a charge may be stored, separated by a thin insulating layer
which has a very large resistance. Energy is stored in the electric field that
exists between the capacitors two conducting surfaces. In addition, the
insulating layer may be made of a high permittivity material (such as ceramic)
which will dramatically increase the capacitance (compared to air).
The inductor is a circuit element whose voltage-current relationship involves

the rate of change of current. Physically, an inductor may be constructed by
winding a length of wire into a coil. Energy is stored in the magnetic field that
exists around the wire in the coil, and the geometry of a coil increases this field
compared to a straight wire. In addition, the centre of the coil may be made of a
ferromagnetic material (such as iron) which will dramatically increase the
inductance (compared to air).
Both the capacitor and the inductor are capable of storing and delivering finite
amounts of energy, but they cannot deliver non-zero average power over an
infinite time interval. They are therefore passive circuit elements, like the
resistor.
The capacitor and inductor are linear circuit elements. Therefore all the circuit
methods previously studied, such as nodal analysis, superposition, Thvenins
theorem, etc., can be applied to circuits containing capacitors and inductors.
Lastly, in dealing with the capacitor and inductor in a circuit, we will note that
the equations describing their behaviour bear a similar resemblance they are
the duals of each other. It will be shown that the concept of duality is a
recurring theme in circuit analysis, and can be readily applied to many simple
circuits, saving both time and effort.

3A.2
The Capacitor
The simplest capacitor is formed by two conductive plates separated by a
dielectric layer:
A parallel plate
capacitor
A
q
conductor
d
insulator
v
Figure 3A.1
One of the plates carries a positive charge, q, whilst the other carries an equal
but opposite charge, -q. Therefore, the capacitor stores charge. There is a
potential difference, v, between the plates. Ideally, the amount of charge q
deposited on the plates is proportional to the voltage v impressed across them.
We define a constant1 called the capacitance, C, of the structure by the linear
relationship:
q Cv
The definition of
(3A.1)
capacitance
The unit of capacitance is the farad, with symbol F.
There is an electric field between the capacitor plates which is established by

the charges present on the plates. Energy is stored in the capacitor by virtue of
this electric field.
1
The constant only models the behaviour of the structure under certain operating conditions.
The capacitance of a structure in the real world will vary with temperature, voltage, pressure,
frequency, chemical aging, etc,

3A.3
The ideal capacitance relationship is a straight line through the origin:

The capacitor is a
C
1
Figure 3A.2
Even though capacitance is defined as C q v , it should be noted that C is a

purely geometric property, and depends only on the conductor arrangements
and the materials used in the construction. For example, it can be shown for a
thin parallel plate capacitor that the capacitance is approximately:
A The capacitance of
C (3A.2)
a parallel plate
capacitor
d
where A is the area of either of the two parallel plates, and d is the distance
between them. The permittivity, , is a constant of the insulating material
between the plates. The permittivity is usually expressed in terms of relative
permittivity, r :
Relative permittivity
r 0 (3A.3) defined
where 0 8.854 pFm -1 is the permittivity of free space (and, for all practical
purposes, air).

3A.4
Capacitor v-i Relationships
We now seek a v-i relationship for the capacitor. From the definition of current:
dq
i (3A.4)
dt
we substitute q Cv and obtain:
The capacitors v-i

dv
branch derivative
relationship
iC (3A.5)
dt
The circuit symbol for the capacitor is based on the construction of the physical
device, and is shown below together with the passive sign convention for the
voltage and current:
The circuit symbol

for the capacitor
v C
Figure 3A.3
The capacitor voltage may be expressed in terms of the current by integrating

Eq. (3A.5):
The capacitors v-i

1 t
vt id vt0
branch integral
relationship (3A.6)
C t0

3A.5
Example
The voltage waveform across a 3 F capacitor is shown below:
v(t) (V)
1
-1 0 1 2 3 t (s)
Since the voltage is zero and constant for t 1 , the current is zero in this
interval. The voltage then begins to increase at the linear rate dv dt 1 Vs -1 ,
and thus a constant current of i C dv dt 3 A is produced. During the
following 2 second interval, the voltage is constant and the current is therefore
zero. The final decrease of the voltage causes a negative 3 A and no response
thereafter. The current waveform is sketched below:
i (t) (A)
3
-1 0 1 2 3 t (s)
-3

3A.6
Energy Stored in a Capacitor
The power delivered to a capacitor is:
dv
p vi Cv (3A.7)
dt
and the energy stored in its electric field is therefore:
wC t pdt wC t 0
t
t0
dt wC t 0
t dv
C v
t0 dt
v t
C vdv wC t 0
v t 0
1

C v 2 t v 2 t 0 wC t 0
2
(3A.8)
If the capacitor voltage is zero at t0 , then the electric field, and hence the
stored capacitor energy, wC t 0 , is also zero at that instant. We then have:
The stored energy in

1
wC t Cv 2 t
a capacitor
(3A.9)
2
We can see that the energy stored in a capacitor depends only on the
capacitance and the voltage. Therefore, a finite amount of energy can be stored
in a capacitor even if the current through the capacitor is zero.
Whenever the voltage is not zero, and regardless of its polarity, energy is
stored in the capacitor. It follows, therefore, that power must be delivered to
the capacitor for a part of the time and recovered from the capacitor later.

3A.7
Example
A sinusoidal voltage source is connected in parallel with an 80 k resistor and

a 2 F capacitor, as shown below:
vs 80 k 2 F
The 80 k resistor in the circuit represents the dielectric losses present in a

typical 2 F capacitor. Let v s 325 sin 100t V , which corresponds to a
typical low voltage 230 V RMS supply.
The current through the resistor is:
4.063 sin 100t mA

v
iR
R
and the current through the capacitor is:
2 10 6 325 sin 100t 204.2 cos100t mA

dv d
iC C
dt dt
The power delivered to the capacitor is:
p C viC 325 sin 100t 0.2042 cos100t 33.18 sin 200t W
A graph of the power delivered to the capacitor versus time is shown below:
pC (W)
33.18
-10 -5 0 5 10 15 20 t (ms)
-33.18

3A.8
The instantaneous power delivered to the capacitor is sinusoidal, and varies at
twice the frequency of the voltage source. Negative power delivered to the
capacitor means that power is sourced from the capacitor. Note that the
average power dissipated in the capacitor is zero.
The energy stored in the capacitor is:
1 2 1
Cv 2 10 6 325 sin 100t 105.6 sin 2 100t mJ
2
wC
2 2
A graph of the stored energy versus time is shown below:
wC (mJ)
105.6
-10 -5 0 5 10 15 20 t (ms)
The energy stored is sinusoidal and varies at twice the frequency of the voltage
source, but it also has a finite average component of 52.81 mJ.
We see that the energy increases from zero at t 0 to a maximum of 105.6 mJ

at t 5 ms and then decreases to zero in another 5 ms. During this 10 ms
interval, the energy dissipated in the resistor as heat is:
1.320 sin 2 100t dt 0.66021 cos200t dt 6.602 mJ

0.01 0.01 0.01
wR p R dt
0 0 0
Thus, an energy equal to 6.25% of the maximum stored energy is lost as heat in
the process of storing and removing the energy in the physical capacitor. Later
we will formalise this concept by defining a quality factor Q that is
proportional to the ratio of the maximum energy stored to the energy lost per
period.

3A.9
Summary of Important Capacitor Characteristics
Several important characteristics of the ideal capacitor are summarised below.
1. If a constant voltage is held across a capacitor, then dv dt 0 and

The capacitor is an
subsequently no current enters (or leaves) it. A capacitor is thus an open-circuit to DC
open-circuit to DC. This fact is represented by the capacitor symbol.
2. A capacitor voltage cannot change instantaneously, for this implies

A capacitor voltage
dv dt , and the capacitor would require infinite current. Thus, cannot change
instantaneously
capacitor voltage is smooth and continuous. This fact will be used
frequently when undertaking transient analysis of circuits.
v(t)
V
0 t0 t1 t
v(t)
V
0 t0 t1 t2 t3 t
Figure 3A.4
3. For an ideal capacitor, current can change instantaneously.
v
5
0 1 2 3 4 t (ms)
dv
i = C dt
20
0 1 2 3 4 t (ms)
Figure 3A.5
4. An ideal capacitor never dissipates energy, but stores it and releases it

using its electric field.

3A.10
The Inductor
On 21 April 1820, the Danish scientist rsted noticed a compass needle
deflected from magnetic north when an electric current from a battery was
An electric current switched on and off, confirming a direct relationship between electricity and
produces a magnetism. Three months later he began more intensive investigations and
magnetic field
soon thereafter published his findings, showing that an electric current
produces a circular magnetic field.
An inductor is a two-terminal device, usually constructed as a coil of wire, that

is designed to store energy in a magnetic field. The coil effectively increases
the magnetic field by the number of turns of wire, and as we shall see, also
increases the number of circuits which link the magnetic field.
An ideal inductor produces a magnetic field which is wholly confined within it.
The closest approximation to an ideal inductor that we can physically produce
is a toroid, which has an almost uniform magnetic field confined within it:
A toroidal inductor
i
core
magnetic
field
B
winding
Figure 3A.6

3A.11
To explain magnetism, 19th century scientists invoked an analogy with fluids

and postulated the existence of a magnetic fluid, known as magnetic flux, ,
which streamed throughout space and manifested itself as magnetism.
Magnetic flux
Magnetic flux always streams out of north poles and into south poles, and defined
forms a closed loop. We still use this concept of flux today, as the theory has
been spectacularly successful.
If a tube of magnetic flux (which is a closed loop) streams through a closed

loop of wire (a circuit), then it is said to link with the circuit. The amount of
flux linkage, , is given by:
N (3A.10) Magnetic flux

linkage defined
where N is the number of loops of wire, or turns, in the circuit, and is the
average amount of magnetic flux streaming through each loop.
Magnetic flux
linkage shown
graphically
N turns
1 2 3 ... N
= 1+ 2 + 3+ ...+ N = N av
Figure 3A.7
An ideal inductor is a structure where the flux linkage (with itself) is directly
proportional to the current through it. We define a constant, called the self
inductance, L, of the structure by the linear relationship:
Li (3A.11) The definition of

inductance
The unit of inductance is the henry, with symbol H.

3A.12
The ideal inductance relationship is a straight line through the origin:
The inductor is a
L
1
Figure 3A.8
Even though inductance is defined as L i , it should be noted that L is a

purely geometric property, and depends only on the conductor arrangements
and the materials used in the construction.
For example, it can be shown for a closely wound toroid that the inductance is
approximately:
The inductance of a
A
toroidal inductor
L N2 (3A.12)
l
where A is the cross-sectional area of the toroid material, and l is the mean path
length around the toroid. The permeability, , is a constant of the material
used in making the toroid. The permeability is usually expressed in terms of
relative permeability, r :
Relative
permeability defined r 0 (3A.13)
where 0 400 nHm -1 is the permeability of free space (and, for all practical
purposes, air).

3A.13
Inductor v-i Relationships
We now seek a v-i relationship for the inductor. In 1840, the great British
experimentalist Michael Faraday2 discovered that a changing magnetic field
could induce a voltage in a neighbouring circuit, or indeed the circuit that was
producing the magnetic field. Faradays Law states that the induced voltage is
equal to the rate-of-change of magnetic flux linkage:
d Faradays Law
v (3A.14)
dt
The minus sign comes from the fact that the polarity of the induced voltage is
such as to oppose the change in flux. For an inductor, we can figure out that the
polarity of the induced voltage must be positive at the terminal where the
current enters the inductor. If we know this, then we can mark the polarity on a
circuit diagram and deal with the magnitude of the induced voltage by
dropping the minus sign (the determination of the voltage polarity is called
Lenzs Law).
If we allow the polarity to be set by Lenzs Law, and substitute Li into the
previous equation, then we get:
The inductors v-i

di
vL (3A.15)
branch derivative
relationship
dt
2
The American inventor Joseph Henry discovered this phenomenon independently, but
Faraday was the first to publish.

3A.14
The circuit symbol for the inductor is based on the construction of the physical
device, and is shown below together with the passive sign convention for the
voltage and current:
The circuit symbol

for the inductor
v L
Figure 3A.9
The inductor current may be expressed in terms of the voltage by integrating

Eq. (3A.15):
The inductors v-i

1 t
branch integral
relationship
it vd it0 (3A.16)
L t0

3A.15
Example
The variation of current through a 25 mH inductor as a function of time is

shown below:
i(t) (mA)
20
10
t (ms)
-1 0 1 2 3 4 5
-10
Since the current is constant for t 0 , the voltage is zero in this interval. The
current then begins to increase at the linear rate di dt 10 As -1 , and thus a
constant voltage of v L di dt 250 mV is produced. During the following 2
millisecond interval, the current decreases at the linear rate di dt 15 As -1 ,

and so the voltage is v L di dt 375 mV . The final increase of the current
causes a positive 125 mV and no response thereafter. The voltage waveform is
sketched below:
v(t) (mV)
400
200
-1 0 1 2 3 4 5
t (ms)
-200
-400

3A.16
Example
Consider the simple circuit below:
i v 2H
Let the current it which is produced by the source be described by the

function of time shown below:
i(t) (A)
1
t (s)
0 1 2
Since v L di dt 2 di dt then the voltage across the inductor is as shown

below:
v(t) (V)
2
t (s)
0 1 2
-1
-2

3A.17
We know that a resistor always absorbs power and the energy absorbed is
dissipated as heat but how about an inductor? For the inductor in this
example, the instantaneous power p t vt i t absorbed by the inductor is
given by the graph below:
p(t) (W)
2
t (s)
0 1 2
-1
-2
We see that the power absorbed by the inductor is zero for t 0 and
2 t . For 0 t 1 , since p t is a positive quantity, the inductor is
absorbing power (which is produced by the source). However, for 1 t 2 ,
since p t is a negative quantity, the inductor is actually supplying power (to
the source).
To get the energy absorbed by the inductor, we simply integrate the power
absorbed over time. For this example, the energy absorbed increases from 0 to
1
2
12 1 J as time goes from t 0 to t 1 s . However, from t 1 to t 2 s ,
the inductor supplies energy such that at time t 2 s and thereafter, the net
energy absorbed by the inductor is zero. Since all of the energy absorbed by the
inductor is not dissipated but is eventually returned, we say that the inductor
stores energy. The energy is stored in the magnetic field that surrounds the
inductor.

3A.18
Energy Stored in an Inductor
The power delivered to an inductor is:
di
p vi L i (3A.17)
dt
and the energy stored in its magnetic field is therefore:
wL t pdt wL t0
t
t0
dt wL t0
t di
L i
t0 dt
i t
L idi wL t0
i t0

1 2
2

L i t i 2 t0 wL t0 (3A.18)
If the inductor current is zero at t0 , then the magnetic field, and hence the
stored inductor energy, wL t 0 , is also zero at that instant. We then have:
1 2
wL t Li t
The stored energy in
an inductor (3A.19)
2
We can see that the energy stored in an inductor depends only on the
inductance and the current. Therefore, a finite amount of energy can be stored
in an inductor even if the voltage across the inductor is zero.
Whenever the current is not zero, and regardless of its direction, energy is
stored in the inductor. It follows, therefore, that power must be delivered to the
inductor for a part of the time and recovered from the inductor later.

3A.19
Example
A sinusoidal current source is connected in series with a 100 m resistor and a

3 mH inductor, as shown below:
100 m
vR
is 3 mH vL
The 100 m resistor in the circuit represents the resistance of the wire which
must be associated with the physical coil. Let i s 12 sin 100t A .
The voltage across the resistor is:
v R Ri 1.2 sin 100t V
and the voltage across the inductor is:
3 10 3 12 sin 100t 11.31cos100t V

di d
vL L
dt dt
The power delivered to the inductor is:
p L v L i 11.31 cos100t 12 sin 100t 67.86 sin 200t W
A graph of the power delivered to the inductor versus time is shown below:
pL (W)
67.86
-10 -5 0 5 10 15 20 t (ms)
-67.86

3A.20
The instantaneous power delivered to the inductor is sinusoidal, and varies at
twice the frequency of the current source. Negative power delivered to the
inductor means that power is sourced from the inductor. Note that the average
power dissipated in the inductor is zero.
The energy stored in the inductor is:
1 2 1
Li 3 10 3 12 sin 100t 216 sin 2 100t mJ
2
wL
2 2
A graph of the stored energy versus time is shown below:
wL (mJ)
216
-10 -5 0 5 10 15 20 t (ms)
The energy stored is sinusoidal and varies at twice the frequency of the current
source, but it also has a finite average component of 108 mJ.
We see that the energy increases from zero at t 0 to a maximum of 216 mJ at

t 5 ms and then decreases to zero in another 5 ms. During this 10 ms interval,
the energy dissipated in the resistor as heat is:
14.4 sin 2 100t dt 7.21 cos200t dt 72 mJ

0.01 0.01 0.01
wR p R dt
0 0 0
Thus, an energy equal to 33.33% of the maximum stored energy is lost as heat
in the process of storing and removing the energy in the physical inductor.

3A.21
Summary of Important Inductor Characteristics
Several important characteristics of the ideal inductor are summarised below.
1. If a constant current is passed through an inductor, then di dt 0 and

The inductor is a
subsequently no voltage appears across it. An inductor is thus a short- short-circuit to DC
circuit to DC.
2. An inductor current cannot change instantaneously, for this implies

An inductor current
di dt , and the inductor would require infinite voltage. Thus, cannot change
instantaneously
inductor current is smooth and continuous. This fact will be used
frequently when undertaking transient analysis of circuits.
i (t)
I
0 t0 t1 t
i (t)
I
0 t0 t1 t2 t3 t
Figure 3A.10
3. For an ideal inductor, voltage can change instantaneously.
i
5
0 1 2 3 4 t (ms)
di
v= L dt
20
0 1 2 3 4 t (ms)
Figure 3A.11
4. An ideal inductor never dissipates energy, but stores it and releases it

using its magnetic field.

3A.22
Practical Capacitors and Inductors

Practical capacitors and inductors are manufactured for different values,
voltage and current ratings, accuracy, volumetric efficiency, temperature
Practical
components have stability, etc. As such, their construction will play a role in their electrical
non-ideal and
parasitic effects behaviour. To represent these physical components in a circuit we need to
model their non-ideal and parasitic effects.
Capacitors
There are many different types of capacitor construction. Some are shown
below, labelled by the type of dielectric:
Some types of
capacitors
polystyrene polypropylene polycarbonate
PTFE (Teflon) polyester paper
ceramic tantalum electrolytic aluminium electrolytic
Figure 3A.12 Some types of capacitors

3A.23
A reasonably accurate circuit model of a real capacitor is shown below:
Model of a real
capacitor
Rp
C
Rs Ls
Figure 3A.13
In the model, the series resistance Rs takes into account the finite resistance of
the plates used to make the capacitor. The series inductance Ls is used to
model the fact that a current is required to charge and discharge the plates, and
this current must have a magnetic field. Finally, no practical material is a
perfect insulator, and the resistance R p represents conduction through the
dielectric.
We call Rs , Ls and R p parasitic elements. Capacitors are designed to
minimize the effects of the parasitic elements consistent with other

Parasitic elements
requirements such as physical size and voltage rating. However, parasitics are defined
always present, and when designing circuits care must be taken to select
components for which the parasitic effects do not compromise the proper
operation of the circuit.
Electrolytic Capacitors
One side (the anode) of an electrolytic capacitor is formed from a foil of

Electrolytic
aluminium which is anodised to produce an oxide layer which is the dielectric. capacitors have a
The oxide-coated foil is immersed in an electrolytic solution which forms the large capacitance
per unit volume, but
cathode. A large capacitance per unit volume is achieved. However, exposure are polarised
to a reverse voltage for a long time leads to rapid heating of the capacitor and
to breakdown. Thus, electrolytic capacitors are polarised and it must be
ensured that the correct voltage polarity is applied to avoid failure.

3A.24
Inductors
There are many different types of inductor construction. Some are shown
below, labelled by the type of core:
Some types of
inductors
ferrite core laminated steel core air core
Figure 3A.14 Some types of inductors
The construction of the core of the inductor is especially important in

determining its properties. Laminated steel cores are common for low
frequencies, and ferrites (non-conducting oxides of iron) are used for high
The core of an
inductor is generally frequencies. Since the core is subjected to a changing magnetic field, the
a magnetic material
which exhibits induced voltages in the core create what are known as eddy currents. Steel
energy losses
laminations and ferrites reduce the energy losses caused by these eddy currents.
There are also energy loss mechanisms associated with reversing the magnetic
domains in a core, and so the losses due to the magnetic characteristic and the
induced eddy currents are combined into a term called the core loss.
Air-cored inductors Air-cored inductors are linear and do not exhibit core losses (since there is no
do not exhibit core
losses
conductive core). They can be made by winding a coil on a non-magnetic
former, such as plastic, or may be self-supporting if made large enough. Air-
cored inductors have lower inductance than ferromagnetic-core inductors, but
are often used at high frequencies because they are free of core losses.

3A.25
A reasonably accurate circuit model of a real inductor is shown below:

Model of a real
inductor
Rp
Rs L
Cp
Figure 3A.15
The series resistance Rs takes into account the finite resistance of the wire
used to create the coil. The parallel capacitance C p is associated with the
electric field in the insulation surrounding the wire, and is called interwinding
capacitance. The parallel resistance R p represents the core losses.
The following inductor model, showing just the predominant non-ideal effect
of finite winding resistance, is often used at low frequencies:
Low frequency
model of a real
Rs L inductor showing the
winding resistance
Figure 3A.16

3A.26
Series and Parallel Connections of Inductors and Capacitors

Resistors connected in series and parallel can be combined into an equivalent
resistor. Inductors and capacitors can be combined in the same manner.
Inductors
A series connection of inductors is shown below:
v1 L1
v
v2 L2
Figure 3A.17
We find from KVL that:
v v1 v2
di di
L1 L2
dt dt
L1 L2
di
dt
di
L
dt (3A.20)
where:
Combining inductors
in series L L1 L2 (series)
(3A.21)

3A.27
We depict this below:
L1 L2 L = L 1+L 2
Figure 3A.18
For the case of two inductors in parallel, as shown below:
i
i1 i2
v L1 L2
Figure 3A.19
we have, by KCL:
i i1 i2
1 t 1
vdt i t
t
L1 t0
vdt i1 t 0 2 0
L2 t0
1 1 t
vdt i1 t0 i2 t0
L1 L2 t0
1 t
vdt it0 (3A.22)
L t0

3A.28
where:
1 1 1
Combining inductors
in parallel (parallel) (3A.23)
L L1 L2
and:
it0 i1 t0 i2 t0 (3A.24)
L1
i1(t0) 1 1 1
L = +
L L1 L2
L2
i (t0) = i1(t0) + i2(t0)
i2(t0)
Figure 3A.20

3A.29
Capacitors
Consider two capacitors in parallel as shown below:
i
i1 i2
v C1 C2
Figure 3A.21
By KCL, we have:
i i1 i2
dv dv
C1 C2
dt dt
C1 C2
dv
dt
dv
C
dt (3A.25)
so that we obtain:
Combining
C C1 C2 (parallel)
(3A.26)
capacitors in parallel

3A.30
C1
C = C1 + C2
C2
Figure 3A.22
For capacitors in series, analogously to inductors in parallel, we get the result:
1 1 1

Combining
capacitors in series
C C1 C2 (series) (3A.27)
which is shown below:
1=1 + 1
C1 C2 C C C1 C2
v1(t0) v2(t0) v (t0) = v1(t0) + v2(t0)
Figure 3A.23
In summary, inductors in series and parallel are treated like resistors, whereas
capacitors in series and parallel are treated like conductances.

3A.31
Circuit Analysis with Inductors and Capacitors

For the ideal inductor and ideal capacitor, all of the relationships between
voltage and current are linear relationships (apart from the integral Linear circuit
analysis techniques
relationships that have an initial condition term). Consequently, circuit analysis can be applied to
techniques which rely on the linearity property (such as nodal analysis, mesh circuits with
inductors and
analysis, superposition, Thvenins theorem and Nortons theorem) can be capacitors
applied to circuits containing inductors and capacitors.
DC Circuits
An ideal inductor behaves as a short-circuit to DC, and the capacitor behaves

A DC circuit is
as an open-circuit to DC. We can use these facts to determine voltages and treated as a purely
resistive circuit in
currents in DC circuits that contain both inductors and capacitors. Other the steady-state
sources, such as sinusoidal sources, will be treated later.
Example
Determine the current i in the circuit below:
2F
1 4H
3F
6A 1F 2 4
3
i
The circuit has one independent current source whose value is constant. For a
resistive circuit we would naturally anticipate that all voltages and currents are
constant. However, this is not a resistive circuit. Yet, our intuition suggests that
the constant-valued current source produces constant-valued responses. This
fact will be confirmed more rigorously later. In the meantime, we shall use the
result that a circuit containing only constant-valued sources is a DC circuit.

3A.32
Since for DC all inductors behave like short-circuits and all capacitors behave
like open-circuits, we can replace the original circuit with an equivalent
resistive circuit:
6A 2 4
3
i
By current division we find that:
4
i 6 4 A
42

3A.33
Nodal and Mesh Analysis
Just as we analysed resistive circuits with the use of node and mesh equations,
we can write a set of equations for circuits that contain inductors and capacitors Nodal and mesh
analysis can be
in addition to resistors and sources. The procedure is similar to that described applied to circuits
with inductors and
for the resistive case the difference being that for inductors and capacitors the capacitors
appropriate relationship between voltage and current is used in place of Ohms
Law.
Example
6F
5
i4 4H i3
1F 3H
1A i1 2 i2 3V
By mesh analysis, for mesh i1 :
i1 1
For mesh i2 :
2i2 i1 3 i2 i3 3
d
dt
For mesh i3 :
d
i3 i2 4 d i3 i4 1 i3 dt 0
t
3
dt dt 6

3A.34
For mesh i4 :
1 t
i4 i1 dt 5i4 4 i4 i3 0
d
1 dt
The equation obtained for mesh i2 is called a differential equation since it

contains variables and their derivatives. The equations obtained for meshes i3
and i4 are called integrodifferential equations since they contain integrals as

well as derivatives.
Writing the equations for a circuit, as in the preceding example, is not difficult.
Finding the solution of equations like these, however, is another matter it is
no simple task. Thus, with the exception of some very simple circuits, we shall
have to resort to additional concepts and techniques to be introduced later.

3A.35
Duality
Duality is a concept which arises frequently in circuit analysis. To illustrate,
consider the two circuits shown below:
Dual circuits
R v
v1
vs v2 is G C
i
i1 i2
Circuit A Circuit B
Figure 3A.24
Using mesh analysis for circuit A and nodal analysis for circuit B, we get the
following results:
Circuit A Circuit B
v s v1 v 2 i s i1 i2
di dv
v s Ri L i s Gv C
dt dt
In other words, both circuits are described by the same equations:
xs x1 x2
dy
xs a1 y a2 (3A.28)
dt
except that a variable that is a current in one circuit is a voltage in the other and
vice versa. For these two circuits this result is not a coincidence, but rather is
due to a concept known as duality, which has its roots in the subject of graph
theory.

3A.36
In actuality, circuits A and B are dual circuits since the following roles are
interchanged:
Dual circuit
vi
RG
relationships
LC
series parallel (3A.29)
The usefulness of duality lies in the fact that once a circuit is analysed, its dual
is in essence analysed also. Note that if circuit B is the dual of circuit A, then
taking the dual of circuit B in essence results in circuit A. Not every circuit,
however, has a dual. With the aid of graph theory it can be shown that a circuit
has a dual if and only if it is a planar network.
The technique for obtaining the dual of a planar network is:

The technique for
obtaining the dual of
a planar circuit 1. Inside of each mesh, and in the infinite region surrounding the
circuit, place a node.
2. Suppose two of these nodes, say nodes a and b, are in adjacent

meshes. Then there is at least one element in the boundary common
to these two meshes. Place the dual of each common element
between nodes a and b.
When using this procedure to obtain the dual circuit, in order to get the mesh
equations of the original circuit to correspond to the node equations of the dual
circuit, place clockwise mesh currents i1 , i2 , ..., in in the finite regions. The
corresponding nodes in the dual are labelled with the voltages v1 , v 2 , ..., v n
respectively. The reference node of the dual circuit corresponds to the infinite
region of the original circuit.

3A.37
Example
For the circuit given in the previous example (shown in blue), the dual is
obtained as (shown in red):
6F
5
v4 v3
4F 4H 3F
5S 6H
1H
3H
1F
1A 2 3V
v1 2 S v2
1V 3A
The dual circuit is redrawn for simplicity:
4F
1H 3F
v1 2S v2
v4 v3
5S 1V 3A 6H

3A.38
By nodal analysis, at node v1 :
v1 1
At node v 2 :
2v 2 v1 3 v2 v3 3
d
dt
At node v3 :
d
v3 v2 4 d v3 v4 1 v3 dt 0
t
3
dt dt 6
At node v 4 :
1 t
v 4 v1 dt 5v 4 4 v 4 v3 0
d
1 dt
Note that these are the duals of the mesh equations that we obtained earlier for
the original circuit.

3A.39
Summary
The v-i relationship for a capacitor is:
dv
iC
dt
A capacitor behaves as an open-circuit to direct current.
The voltage across a capacitor cannot change instantaneously.
The energy stored in a capacitor is:
1
wC Cv 2
2
The v-i relationship for an inductor is:
di
vL
dt
An inductor behaves as a short-circuit to direct current.
The current through an inductor cannot change instantaneously.
The energy stored in an inductor is:
1 2
wL Li
2
Inductors in series and parallel are combined in the same way as are
resistances. Capacitors in series and parallel are combined in the same way
as are conductances.
Writing node and mesh equations for circuits containing inductors and
capacitors is done in the same manner as for resistive circuits. Obtaining
solutions of equations in this form will be avoided, except for simple
circuits.
A planar circuit and its dual are in essence described by the same equations.

3A.40
References
Bobrow, L.: Elementary Linear Circuit Analysis, Holt-Saunders, 1981.
Hill, 1984.

3A.41
Exercises
1.
There is a current i 5 sin 10t A through an inductance L 2 H . What is the
first instant of time after t 0 when the power entering the inductor is exactly:
(a) 100 W
(b) -100 W
2.
The energy stored in a certain 10 mH inductor is zero at t 1 ms and increases
linearly by 20 mJ each second thereafter. Find the inductor current and voltage
for t 1 ms if neither is ever negative.
3.
A 25 F capacitor having no voltage across it at t 0 is subjected to the single
pulse of current shown below.
i(t) (mA)
20
10
t (ms)
0 10 20 30
Determine the voltage across, the power entering, and the energy stored in C at
t :
(a) 17 ms
(b) 40 ms

3A.42
4.
Find C eq for the lattice network shown below if terminals a and b are:
(a) open-circuited as shown
(b) short-circuited
1 nF
a
3 nF
Ceq 4 nF
b
2 nF
5.
The series combination of a 4 F and a 3 F capacitor is in series with the
parallel combination of a 2 F , a 1 F and a C F capacitor.
(a) What is the maximum possible value for the equivalent capacitance of
the five capacitors?
(b) Repeat for the minimum value.
(c) Find C if C eq 1.5 F .

3A.43
6.
At t 0 , i 5 A in the circuit shown below:
8 sin 10t A 1H 3H v
(a) Find vt for all t.
(b) Find i t for t 0 .
7.
(a) Write the single nodal equation for the circuit (a) below:
(b) Write the single mesh equation for the circuit (b) below:
is
vC (0) 40 mH
5 k
i L (0)
200 F
vC (0) 200 F vs
5 k
40 mH
i L (0)
(a) (b)

3A.44
8.
Using a reference node at the bottom of the circuit, assign node-to-reference
voltages in the circuit below and write nodal equations. Let i L 0 0.5 A and
vC 0 12 V .
3v1
v2
0.1 H
iL 50
0.8e -100 t A
v1 20 200 F vC

4A.1
Lecture 4A Source-Free RC and RL Circuits

Differential operators. Properties of differential operators. The characteristic
equation. The simple RC circuit. Properties of the exponential response. Single
time constant RC circuits. The simple RL circuit. Single time constant RL
circuits.
Introduction
The analysis of a linear circuit that has storage elements (capacitors and
inductors) inevitably gives rise to a linear differential equation. The solution of
the differential equation always consists of two parts one part leads to the so-
called forced response, the other is the natural response. The forced response
is due to the application of a source to the circuit. The natural response is due
entirely to the circuits configuration, its initial energy, and the amplitude of
the applied source at the instant of application.
Initially, we will study the theory of differential operators, which leads to a

rather simple method of solving linear differential equations. For simple RC
and RL circuits, the differential operator seems like overkill a powerful
method applied to a simple situation. However, some of the insights resulting
from such a method will put us in good stead when we look at more
complicated circuits that have both capacitors and inductors in them.
We shall then study the natural response of some simple source-free RC and
RL circuits. This study will reveal some surprising results, such as the fact that
there is only ever one style of natural response an exponential response.
We will become familiar with the exponential response, and we will give some
special names to the algebraic terms involved in it, such as initial condition and
time constant.
With practice, we will also see that we can write down the natural response for
simple circuits by inspection. This will lead to intuition of circuit behaviour,
and we will get a feel for the way a circuit behaves by simply looking at it.

4A.2
Differential Operators
Let D denote differentiation with respect to t:
D is just shorthand
d
for d dt D
dt (4A.1)
Then D 2 denotes differentiation twice with respect to t, and so on. That is, for
positive integer k:
dk y
D y k
k
dt (4A.2)
The expression:
(4A.3)
A an D n an1D n1 ... a1D a0
is called a differential operator of order n. The coefficients ak in the operator A
may be functions of t, but most of the time they will be constant coefficients.
The product AB of two operators A and B is defined as that operator which

produces the same result as is obtained by using the operator B followed by the
operator A. Thus:
ABy ABy (4A.4)
Differential
operators commute The product of two differential operators always exists and is a differential
if they have constant operator. For operators with constant coefficients it is true that AB BA .
coefficients

4A.3
Example
Let A D 2 and B 3D 1 .
Then:
By 3D 1 y 3
dy
y
dt
and:
dy
ABy D 2 3 y
dt
2
d y dy dy
3 2 6 2y
dt dt dt
2
d y dy
3 2 5 2y
dt dt

3D 5 D 2 y
2

Hence AB D 23D 1 3D 2 5 D 2 .
The sum of two differential operators is obtained by expressing each in the

form:
(4A.5)
an D n an1D n1 ... a1D a0
and adding corresponding coefficients.
Example
Let A 3D 3 D 2 and B D 2 4 D 7 .
Hence A B 3D 3 D 2 3D 9 .
Differential operators are linear operators. Therefore, differential operators Differential

operators are linear
with constant coefficients satisfy all the laws of the algebra of polynomials operators that obey
the laws of algebra
with respect to the operations of addition and multiplication.

4A.4
Properties of Differential Operators

Since for constant s and positive integer k:
(4A.6)
D k e st s k e st
it is easy to find the effect that a differential operator has upon e st . Let f D
be a polynomial in D:
f D an D n an 1 D n 1 ... a1 D a0 (4A.7)
Then:
f D e st an s ne st an1s n1e st ... a1sest a0e st (4A.8)
so:
Operating with D
and multiplication by
s are equivalent for f D e st f s e st (4A.9)
the function e st
This equation does not mean that f D f s . f D is an operator, and it
means to take a linear sum of derivatives of a function that it operates on.
f s is a standard polynomial in s. The equation means that the effect of the D
operator on e st is the same as multiplication by s. The function e st is the only

function that enjoys this property a deeper insight into the special function
e st will be developed later.
If s is a root of the equation f s 0 , then in view of the above result:
f D e st 0, if f s 0 (4A.10)

4A.5
Example
Let f D 2 D 2 5D 12 . Then the equation f s 0 is:
2 s 2 5s 12 0
or:
s 42s 3 0
of which the roots are s1 4 and s2 32 .
With the aid of Eq. (4A.10), it can be seen that:
2 D 2
5D 12e 4t 0
and that:
2 D 2
5D 12e3t 2 0
In other words, y1 e 4t and y2 e3t 2 are solutions of:
2 D 2
5 D 12y 0 .

4A.6
Next consider the effect of the operator D s on the product of e st and a

function y. We have:
D s e st y De st y sest y (4A.11)
e st Dy
and:
D s 2 e st y D s e st Dy (4A.12)
e st D 2 y
Repeating the operation, we are led to:
D s n e st y e st D n y (4A.13)
Using the linearity of differential operators, we conclude that when f D is a
Shifting an polynomial in D with constant coefficients, then:

exponential factor
e st f D y f D s e st y
from the left of a
differential operator (4A.14)
to the right
This relation shows us how to shift an exponential factor from the left of an
operator to the right of an operator.

4A.7
Example
Let us solve the differential equation:
D 34 y 0
First, we multiply the equation by e3t to obtain:
e3t D 3 y 0
4
Applying the exponential shift as in Eq. (4A.14) leads to:
D 4 e3t y 0
Integrating four times gives us:
e3t y c0 c1t c2t 2 c3t 3
and finally:

y c0 c1t c2t 2 c3t 3 e 3t
It can be shown that the four functions e 3t , te 3t , t 2e 3t and t 3e 3t are linearly

independent thus the solution given is the general solution of the differential
equation.
In Eq. (4A.14), if we let f D D n and y t k , then:
D s n t k e st e st D nt k (4A.15)
But D nt k 0 for k 0, 1, 2, ..., n 1 , and so:
D s t 0,
The solution to a
k 0, 1, ..., n 1
n k st (4A.16) special class of
e differential equation

4A.8
The Characteristic Equation

The general linear differential equation with constant coefficients of order n is
an equation that can be written:
dny d n 1 y
an n an1 n1 a1 a0 y r t
dy
dt dt dt (4A.17)
If r t is identically zero (i.e. zero for all time, not just a specific time), then
we have an equation that is said to be homogeneous:
Homogeneous
differential equation dny d n 1 y dy
defined an n an 1 n 1 a1 a0 y 0
dt dt dt (4A.18)
Any linear combination of solutions of a linear homogeneous differential

equation is also a solution. If yi , with i 1, 2, ..., k , are solutions of Eq.
(4A.18) and if ci , with i 1, 2, ..., k , are constants, then:
y c1 y1 c2 y2 ck yk (4A.19)
is a solution of Eq. (4A.18).
If the solutions yi are linearly independent (i.e. they cannot be expressed as

linear combinations of one another), then there are n of them and we say that
the general solution of the differential equation is:
y c1 y1 c2 y2 cn yn (4A.20)

4A.9
The linear homogeneous differential equation with constant coefficients may

also be written in the form:
f D y 0 (4A.21)
where f D is a linear differential operator. As we saw in the preceding

A solution to a linear
section, if s is any root of the algebraic equation f s 0 , then: homogeneous
differential equation
with constant
f D e st 0
coefficients is
(4A.22)
y e st
which means simply that y e st is a solution of Eq. (4A.21). The equation:
f s 0 (4A.23) Characteristic
equation defined
is called the characteristic equation associated with Eq. (4A.21).
The characteristic equation for Eq. (4A.21) is of degree n. Let its roots be
s1 , s2 , , sn . If these roots are all real and distinct, then the n solutions
y1 e s1t , y2 e s2t , , yn e snt are linearly independent and the general

solution of Eq. (4A.21) can be written at once. It is: General solution of
a linear
homogeneous
(4A.24)
y c1e s1t c2e s2t cne snt differential equation
with constant
coefficients
in which c1 , c2 , , cn are arbitrary constants. Repeated roots of the

characteristic equation will be treated later.

4A.10
Example
Solve the equation:
d3y d 2 y dy
4 6y 0
dt 3 dt 2 dt
First write the characteristic equation:
s 3 4s 2 s 6 0
s 1s 2s 3 0
whose roots are s 1, 2, 3 . Then the general solution is seen to be:
y c1e t c2e 2t c3e3t
Example
Solve the equation:
3D 3
5 D 2 2 D y 0
The characteristic equation is:
3s 3 5s 2 2s 0
s s 23s 1 0
whose roots are s 0, 2, 13 . Using the fact that eot 1 , the desired solution
may be written:
y c1 c2e 2t c3et 3

4A.11
Example
Solve the equation:
d2y
4y 0
dt 2
with the conditions that when t 0 , y 0 and dy dt 3 .
s2 4 0
with roots s 2, 2 . Hence the general solution of the differential equation is:
y c1e 2t c2e 2t
It remains to enforce the conditions at t 0 .
The condition that y 0 when t 0 requires that:
0 c1 c2
Now:
dy
2c1e 2t 2c2e 2t
dt
The condition that dy dt 3 when t 0 requires that:
3 2c1 2c2
From the simultaneous equations for c1 and c2 we conclude that c1 3

4
and
c2 34 . Therefore:
y 3
4 e 2t
e 2t

4A.12
The Simple RC Circuit

Consider the parallel RC circuit shown below.
The simple RC
circuit
i
C v R
Figure 4A.1
We designate the time-varying voltage by vt , and we shall let the value of

vt at t 0 be prescribed as V0 .
Kirchhoffs Current Law (KCL) applied to the top node gives us:
dv v
C 0
dt R (4A.25)
Division by C gives us:
The governing dv v
of the simple RC
0
circuit
dt RC (4A.26)
and we must determine an expression for vt which satisfies this equation and
also has the value V0 at t 0 .

4A.13
Written in operator notation, Eq. (4A.26) is:
1
D v 0
RC (4A.27)
for which the characteristic equation is:
1
s 0
RC (4A.28)
which has a root at s 1 RC . Therefore, the solution to the differential

equation is:
(4A.29)
v c1e t RC
It remains to enforce the condition that v0 V0 . Thus:
V0 c1 (4A.30)
and the final form of the solution is:

The solution to the
(4A.31) governing
v V0e t RC differential equation
of the simple RC
circuit
This solution is known as the natural response of the circuit its form depends
only on the initial energy in the circuit (in this case the initial capacitor
voltage), and the circuit element configuration and values.

4A.14
Let us check the power and energy relationships in this circuit. The power
being dissipated in the resistor is:
(4A.32)
v 2 V02e 2t RC
pR
R R
Energy and power and the total energy turned into heat in the resistor is found by integrating the
dissipated in the
simple RC circuit instantaneous power from zero time to infinite time:
V02
ER pR dt e 2t RC dt
0 R 0

V02 RC 2t RC
2 e
R 0
1
2 CV02 (4A.33)
This is the result we expect, because the total energy stored initially in the
capacitor is 1
2 CV02 , and there is no energy stored in the capacitor at infinite
time. All the energy is accounted for by dissipation in the resistor.

4A.15
Properties of the Exponential Response

We will now consider the nature of the response in the parallel RC circuit. We
found that the voltage is represented by:
(4A.34)
v V0e t RC
At zero time, the voltage is the assumed value V0 and as time increases, the
voltage decreases and approaches zero. The shape of this decaying exponential
is seen by a plot of v V0 versus t, as shown below:
The decaying
exponential
v response
V0
1
Figure 4A.2
Since the function we are plotting is e t RC , the curve will not change if RC
does not change. Thus, the same curve must be obtained for every RC circuit
having the same product of R and C. Lets see how this product affects the
shape of the curve.
If we double the product RC, then the exponent will be unchanged if t is also
doubled. In other words, the original response will occur at a later time, and the
new curve is obtained by moving each point on the original curve twice as far
to the right. With this larger RC product, the voltage takes longer to decay to
any given fraction of its original value.

4A.16
To get a handle on the rate at which the curve decays, lets consider the time
that would be required for the voltage to drop to zero if it continued to drop at
its initial rate.
The initial rate of decay is found by evaluating the derivative at zero time:
(4A.35)
d v 1 t RC 1
e
dt V0 t 0
RC t 0 RC
We designate the value of time it takes for v V0 to drop from unity to zero,
assuming a constant rate of decay, by T (or sometimes the Greek letter ).

Thus T RC 1 , or:
Time constant
defined for the
simple RC circuit T RC (4A.36)
The product RC has the units of seconds since the exponent t RC must be
dimensionless. The value of time T is called the time constant. It is shown
below:
Time constant
shown graphically
on the natural
response curve
v
V0
1
T t
Figure 4A.3

4A.17
An equally important interpretation of the time constant T is obtained by

determining the value of v V0 at t T . We have:
v (4A.37)
e 1 0.3679 or v 0.3679V0
V0
Thus, in one time constant the response has dropped to 36.8 percent of its
initial value. The value of T may be determined graphically from this fact
from the display on an oscilloscope, as indicated below:
Response curve
values at integer
v multiples of the time
constant
V0
1
0.3679
0.1353
0.0498
0
t
T 2T 3T 4T
Figure 4A.4
At some point three to five time constants after zero time, most of us would
agree that the voltage is a negligible fraction of its former self.
Why does a larger value of the time constant RC produce a response curve
which decays more slowly? An increase in C allows a greater energy storage
for the same initial voltage, and this larger energy requires a longer time to be
dissipated in the resistor. For an increase in R, the power flowing into the
resistor is less for the same initial voltage; again, a greater time is required to
dissipate the stored energy.
In terms of the time constant T , the response of the parallel RC circuit may be
written simply as:
(4A.38)
v V0et T

4A.18
Single Time Constant RC Circuits

Many of the RC circuits for which we would like to find the natural response
contain more than a single resistor and capacitor. We first consider those cases
Single time constant in which the given circuit may be reduced to an equivalent circuit consisting of
RC circuit defined
only one resistor and one capacitor. Such a circuit is known as a single time
constant circuit.
Firstly, consider a circuit with any number of resistors and one capacitor. We
fix our attention on the two terminals of the capacitor and determine the
equivalent resistance across these terminals. The circuit is thus reduced to the
simple parallel case.
Example
The circuit shown in (a) below may be simplified to that of (b):
i1
R2
R1 R3 C v R eq
C v
(a) (b)
enabling us to write:
t ReqC
v V0e
where:
v0 V0
R1R3
and Req R2
R1 R3
Every current and voltage in the resistive portion of the circuit must have the
t ReqC
form Ae , where A is the initial value of that current or voltage.

4A.19
Since the current in a resistor may change instantaneously, we shall indicate

the instant after any change that may have occurred at t 0 by use of the
symbol 0 .
Thus, the current in R1 , for example, may be expressed as:
i1 i1 0 e t T
where:
RR
T R2 1 3 C
R1 R3
and i1 0 remains to be determined from some initial condition. Suppose that

v0 is given. Since v cannot change instantaneously, we may think of the
capacitor as being replaced by an independent DC source, v0 , for the instant
of time t 0 . Thus:
v0

i1 0
R3
R2 R1R3 R1 R3 R1 R3
The solution is obtained by collecting all these results:
v0
t
R3
i1 e R2 R1R3 R1 R3 C
R2 R1R3 R1 R3 R1 R3
Another special case includes those circuits containing one resistor and any
number of capacitors. The resistor voltage is easily obtained by establishing the
value of the equivalent capacitance and determining the time constant.
Some circuits containing a number of both resistors and capacitors may be

replaced by an equivalent circuit containing only one resistor and one Not all RC circuits
can be converted
capacitor; it is necessary that the original circuit be one which can be broken into single time
constant RC circuits
into two parts, one containing all resistors and the other containing all
capacitors. This is not possible in general.

4A.20
The Simple RL Circuit

Consider the series RL circuit shown below.
The simple RL
circuit
i
L v R
Figure 4A.5
We designate the time-varying current by i t , and we shall let the value of i t

at t 0 be prescribed as I 0 .
Kirchhoffs Voltage Law (KVL) applied around the loop gives us:
di
L Ri 0
dt (4A.39)
Division by L gives us:
The governing
differential equation di R
of the simple RL i0
circuit dt L (4A.40)
This equation has a familiar form; comparison with:
dv v
0
dt RC (4A.41)
shows that the replacement of v by i and RC by L R produces the identical

equation Eq. (4A.40). It should, for the RL circuit we are now analysing is the
dual of the RC circuit we considered first.

4A.21
This duality forces i t for the RL circuit and vt for the RC circuit to have
identical expressions if the resistance of one circuit is equal to the conductance
of the other and if L is numerically equal to C. That is, we will obtain the dual
circuit (and equation) if we make the substitution:
Duality relations for

v i (4A.42) all dual circuits
R G
CL
Thus, the response of the RC circuit:
vt v0et RC V0et RC (4A.43)
enables us to write immediately:

The solution to the
governing
it i0etR L I 0etR L (4A.44) differential equation
of the simple RL
circuit
for the RL circuit.
Lets examine the physical nature of the response of the RL circuit as expressed
by Eq. (4A.44). At t 0 we obtain the correct initial condition, and as t
becomes infinite the current approaches zero. This latter result agrees with our
thinking that if there were any current remaining through the inductor, then
energy would continue to flow into the resistor and be dissipated as heat. Thus
a final current of zero is necessary.

4A.22
The time constant of the RL circuit may be found by using the duality
relationships on the expression for the time constant of the RC circuit, or it may
be found by simply noting the time at which the response has dropped to 36.8
percent of its initial value:
Time constant L
defined for the
simple RL circuit
T (4A.45)
R
Our familiarity with the negative exponential and the significance of the time
constant T enables us to sketch the response curve readily:
i
I0
0.3679 I0
0
t
T
Figure 4A.6
An increase in L allows a greater energy storage for the same initial current,
and this larger energy requires a longer time to be dissipated in the resistor. If
we reduce R, the power flowing into the resistor is less for the same initial
current; again, a greater time is required to dissipate the stored energy.

4A.23
Single Time Constant RL Circuits

It is not difficult to extend the results obtained for the series RL circuit to a
circuit containing any number of resistors and inductors. We first consider
Single time constant
those cases in which the given circuit may be reduced to an equivalent circuit RL circuit defined
consisting of only one resistor and one inductor; that is, single time constant
(STC) circuits.
Example
The circuit shown in (a) below may be simplified to that of (b):
R3 L i i
R1 R2 R4 L R eq
i1 i2
(a) (b)
The equivalent resistance which the inductor faces is:
R1R2
Req R3 R4
R1 R2
and the time constant is therefore:
L
T
Req
The inductor current is:
i i 0e t T
and represents what we might call the basic solution to the problem. It is quite
possible that some current or voltage other than i is needed, such as the current

4A.24
i2 in R2 . We can always apply Kirchhoffs laws and Ohms law to the resistive
portion of the circuit without any difficulty, but current division provides the
quickest answer in this circuit:
i 0e t T
R1
i2
R1 R2
It may also happen that we know the initial value of some current other than

the inductor current. Thus, if we are given the initial value of i1 as i1 0 , then
it is apparent that the initial value of i2 is:
RR
i2 0 i1 0 1
From these values, we obtain the necessary initial value of i 0 :

i 0 i1 0 i2 0
R1 R2
R2

i1 0
and the expression for i2 becomes:
i2
R1
R2

i1 0 e t T
We can obtain this last expression more directly. Every current and voltage in
the resistive portion of the circuit must have the form Ae t T , where A is the
initial value of that current or voltage. We therefore express i2 as:
i2 Ae t T
and A must be determined from a knowledge of the initial value of i2 . We

found this initial value previously. We therefore have:
i2
R1
R2

i1 0 e t T

4A.25
A similar sequence of steps provides a solution to a large number of problems.
This technique is also applicable to a circuit which contains one resistor and
any number of inductors, as well as to those special circuits containing two or
more inductors and also two or more resistors that may be simplified by
resistance or inductance combination until the simplified circuits have only one
inductance or one resistance.
Example
120
60
iL
t=0 1 mH 50
18 V 90 2 mH 3 mH
i1
After t 0 , when the voltage source is disconnected, we easily calculate an

equivalent inductance,
23
Leq 1 2.2 mH
23
an equivalent resistance,
9060 120
Req 50 110
90 180
and the time constant:
Leq 2.2 10 3
T 20 s
Req 110

4A.26
Thus, the form of the natural response is Ae 50000t . With the independent source
connected t 0 , iL is 18
50 , or 0.36 A, while i1 is 18
90 , or 0.2 A. At t 0 , iL
must still be 0.36 A, but i1 will jump to a new value determined by iL 0 .

Thus, by the current divider rule:

i1 0 iL 0 180
270 0.24 A
Hence:
iL 0.36 t0
0.36e 50000 t
t0
and:
i1 0.2 t0
0.24e 50000 t
t0
The responses are sketched below:
iL i1
0.4 0.4
0.3 0.3
0.2 0.2
0.1 0.1
t ( s) -20 0 20 40 60 80
t ( s)
-20 0 20 40 60 80
-0.1
-0.2
-0.3
A circuit containing several resistances and several inductances does not in

general possess a form which allows either the resistances or inductances to be
Not all RL circuits combined into single equivalent elements. There is no single negative
can be converted
into single time exponential term or single time constant associated with the circuit. Rather, in
constant RL circuits
general, there will be several negative exponential terms, the number of terms
being equal to the number of inductances that remain after all possible inductor
combinations have been made. The natural response of these more complex
circuits is obtained using mathematical techniques that will appear later.

4A.27
Summary
The solution to the linear homogeneous differential equation f D y 0 is
y c1e s1t c2e s2t cn e snt where the si s are the roots of the characteristic
equation f s 0 and the ci s are arbitrary constants.
The natural response of any voltage or current in a single time constant

circuit always takes the form of y Ae t T , where T is the time constant
and A is determined from the initial conditions.
For single time constant RC circuits, the time constant is T ReqCeq .
Leq
For single time constant RL circuits, the time constant is T .
Req
Not all RC and RL circuits can be reduced to single time constant circuits.
References
Bedient, P. & Rainville, E.: Elementary Differential Equations, 6th Ed.
Macmillan Publishing Co., 1981.
Hayt, W. & Kemmerly, J.: Engineering Circuit Analysis, 3rd Ed., Mcgraw-Hill,
1984.

4A.28
Exercises
1.
The current in a simple source-free series RC circuit is given by
i t 20e 5000t mA , and the capacitor voltage is 2 V in magnitude at t 0 .
Find R and C.
2.
200
t=0
1 kV 600 50 F
(a) Find the charge in coulombs present on the capacitor at t 0 .
(b) What is the charge at t 0.01 s ?
3.
Let v and i be the voltage and current variables for a capacitor, assuming the
passive sign convention. The capacitor is the only energy-storage element

present in a source-free resistive circuit. If v 0 80 V , i 0 0.1 A , and

q 0 for the capacitor is 20 mC, find v0.01 .

4A.29
4.
100
ix
5 ix 400 2 F

If ix 0 3 mA , find ix t for t 0 .
5.
fuse
vs 10 M 2 F
The 10 M resistor in the circuit represents the leakage resistance present in a

typical 2 F high-voltage capacitor. The fuse blows (opens) at t 0 . Let
v s 23000 cos100t V . Assuming 50 V is relatively nonlethal, at what time is
it safe to get your hands across the capacitor?

4A.30
6.
t=0
2 M 9V
i2
4 M 1 M 20 F 30 F
The switch closes at t 0 . Find i2 t for t 0 .
7.
The magnitude of the current in a series RL circuit decreases at a rate of
2000 As-1 at t 0 and 100 As-1 at t 0.2 s . At what time has the energy
stored in the inductor decreased to 1 per cent of its initial value?
8.
v( t )
2 8
i1 i2
0.4
1H 2H
t=0
24 V
Find:
(a) i1 0 (b) i2 0 (c) i1 t , t 0 (d) i2 t , t 0 (e) vt , t 0

4A.31
9.
What is the time constant of a series RL circuit if:
(a) the current decreases by a factor of 1000 in 0.2 s?
(b) the time required for the current to drop to half of its initial value is 0.1 s
less than the time required for it to drop to one-quarter of its initial value?
10.
The switch in the circuit shown below has been open for a long time.
1.5 H 2H
i
3
30 20
t=0
12 V
Find i at t :
(a) -0.08 s (b) +0.08 s
11.
The voltage across the resistor in a simple source-free series RL circuit is
vR t 50e 400t V for t 0 . If the value of resistance changes from 200 to
40 at t 0 when a second resistor is placed in parallel with it, find vR t
for t 0 .

4A.32
12.
The switch in the circuit shown below has been open for a long time.
200
1 k 100
t=0
12 V
4 F 0.5 H
i
It closes at t 0 . Find i t .
Hint: Superposition can be very useful.

5A.1
Lecture 5A First-Order Step Response

The unit-step forcing function. The driven RC circuit. Forced and natural
response. Step-response of RC circuits. Analysis procedure for single time
constant RC circuits. Step-response of RL circuits. Analysis procedure for
single time constant RL circuits.
Introduction
The determination of the natural response of a source-free circuit relies solely
on the configuration of the circuit elements and on any initial energy storage
present in the system (capacitor voltages and inductor currents).
If we apply a DC source to a circuit, it will eventually settle to a state where all

voltages and currents are DC. Such a DC source is termed a forcing function
it forces the form of the response in the circuit voltages and currents.
If we apply a sinusoidal source to a circuit, it will eventually settle to a state

where all voltages and currents are sinusoidal. Thus the concept of a forcing
function is applicable to other styles of sources. We shall see later that the DC
source and the sinusoidal source are special cases of a general forcing function.
The application of a forcing function to a circuit will result in a response

consisting of two parts. The first part is termed the forced response, or steady-
state response. The second part is the natural response. As we know, the
natural response dies out after a period of time, leaving only the forced
response. We can think of the natural response of a circuit as a necessary
transient response to move the circuit from one state to another.
We will consider circuits that are initially in a known state any sources have
either been off for a very long time, or on for a very long time. Any forcing
functions are switched on at t 0 .
The forced response can be obtained by considering the response of the circuit
after a very long time. The form of the natural response will be the same as that
obtained for the source-free circuit.
The complete response will be obtained by adding the forced response to the
natural response.

5A.2
The Unit-Step Forcing Function

In an analysis of a circuit we often find the need to turn source voltages or
currents on at a specified time. To do this, we define the unit-step function to
be zero for all values of time less than zero, and unity for all values of time
greater than zero:
0, t 0
u t
The unit-step
function defined
1, t 0 (5A.1)
At t 0 , u t changes abruptly from 0 to 1. Its value at t 0 is undefined, but

its value is known for all points arbitrarily close to t 0 . We often indicate

this by writing u 0 0 and u 0 1 .
Graphically, the unit-step function looks like:
and graphed
u(t )
1
0 t
Figure 5A.1
Note that a vertical line of unit length is shown at t 0 . This riser is of

course impossible (how can a function take on many values simultaneously?)
but is usually shown in drawings.
We cannot always arrange for the switching of a source to occur at t 0 . Since

the unit-step function provides us with a discontinuity when the value of its
argument is zero, then we need to consider a delayed version of the unit-step.

5A.3
We will now make a very important observation: it is the argument of the The argument of a
function which determines the position of the function along the t-axis. We function determines
its position
therefore have the delayed unit-step function:
0, t t0
u t t0
1, t t0 (5A.2)
We obtain the conditions on the values of the function by the simple

substitution t t t 0 in Eq. (5A.1). Graphically, we have:
u(t- t0 )
1
0 t0 t
Figure 5A.2
We see that the argument t t0 simply shifts the origin of the original
function to t 0 . A positive value of t 0 shifts the function to the right

corresponding to a delay in time. A negative value shifts the function to the left
an advance in time.
To represent a constant voltage V being turned on in a circuit, we represent it as

a source which is zero before t t0 and a constant V after t t0 :
General
V u(t - t0) network
Figure 5A.3

5A.4
The utility of the unit-step function is that it can be used as a switch to turn
another function (not necessarily a constant) on or off at some point. For
example, the product given by u t 1cos2t is shown below:
The unit-step
function as a
switch u ( t- 1) cos 2 t
1
-2 -1 0 1 2 3 4 t
Figure 5A.4
Another useful forcing function may be obtained by manipulating the unit-step

forcing function. Let us define a rectangular voltage pulse by the following
conditions:
The rectangular
pulse
0, t t0

vt V , t0 t t1 (5A.3)
0, t1 t

The pulse is shown below:
v(t)
V
0 t0 t1 t
Figure 5A.5
This rectangular pulse is the difference of two step functions:
vt V u t t0 u t t1 (5A.4)

5A.5
Example
The unit-step finds application in mathematically representing the application

of a source via a switch. Consider the circuit shown below:
t=0
General
(a) V network
It is tempting to replace the voltage source and switch by a step function:
General
(b) V u(t - t0) network
However, this is incorrect, because the circuit with the step function actually
represents:
t=0 The circuit

equivalent of a step-
voltage function
General
(c) V network
That is, the step function applies v 0 for t 0 , and then v V for t 0 . The
first circuit applies an open circuit for t 0 , and then v V for t 0 .
However, circuit (b) can often be used if we establish that all initial currents
and voltages in the original network and in circuit (b) for t 0 are equivalent.
This is always the case for circuits that start out with zero initial conditions (no
stored energy) at t 0 .

5A.6
The Driven RC Circuit

Consider a circuit consisting of a battery V in series with a switch, a resistance
R, and a capacitance C. The switch is closed at t 0 , as shown in the circuit
diagram (a) below:
A simple RC circuit
driven by a step-
voltage
t=0 R R
V C v( t ) Vu ( t ) C v( t )
(a) (b)
Figure 5A.6
We will assume that there is no stored energy in the capacitor before t 0 , and
we are therefore able to replace the battery and switch by a voltage-step forcing
function Vu t , which also produces no response prior to t 0 . Hence, we will
analyse the equivalent circuit shown in circuit diagram (b).
We shall find vt by writing down the appropriate differential equation that

describes the circuit, and then solve it using Eulers integrating factor.
Applying KCL to the top capacitor node, we have:
v Vu t dv (5A.5)
C 0
R dt
which can be rewritten as:
The governing
Vu t
(5A.6)
for the RC circuit dv v
driven by a step-
voltage dt RC RC

5A.7
Since the unit-step function is discontinuous at t 0 , we first consider the

solution for t 0 and then for t 0 . It is obvious that the application of zero
voltage since t has not produced any response, and therefore:
vt 0, t 0 (5A.7)
For positive time u t is unity, and we must solve the equation:
dv v V
, t 0 (5A.8)
dt RC RC
To solve, first multiply both sides by an integrating factor equal to et RC . This
gives:
dv t RC v V
et RC e et RC (5A.9)
dt RC RC
Thus, recognising that the left hand-side is the derivative of vet RC , we have:
d
dt

vet RC et RC
V
RC
(5A.10)
Integrating both sides with respect to time gives:
Vet RC
ve t RC
dt A (5A.11)
RC
where A is a constant of integration. Dividing both sides by the integrating
factor gives:
Vet RC
RC dt Ae
t RC t RC
ve (5A.12)

5A.8
Noting that V is a constant, we perform the integration and obtain:
vt V Ae t RC (5A.13)

Prior to t 0 , vt 0 , and thus v 0 0 . Since the voltage across a capacitor
cannot change by a finite amount in zero time without being associated with an

infinite current, we thus have v 0 0 . We thus invoke the initial condition

that v 0 0 and get:
vt V Ve t RC , t 0 (5A.14)
Thus, an expression for the response valid for all t would be:
The complete

response of an RC
circuit to a step-
voltage vt V 1 e t RC u t (5A.15)
This is the desired solution, but it has not been obtained in the simplest manner.
The complete
response is In order to establish a more direct procedure, we will interpret the two terms
composed of two
parts appearing in Eq. (5A.15).
The exponential term has the functional form of the natural response of the RC
circuit it is a negative exponential that approaches zero as time increases, and
the natural response
and it is characterized by the time constant RC. The functional form of this part of
the response is identical with that which is obtained in the source-free circuit.
However, the amplitude of this exponential term depends on V, the forcing
function.
Eq. (5A.15) also contains a constant term, V. Why is it present? The natural
response approaches zero as the energy stored in the capacitor gradually
the forced response reaches its limit. Eventually the capacitor will be fully charged and it will
appear as an open circuit the current will be zero, and the battery voltage V
will appear directly across the capacitor terminals. This voltage is a part of the
response which is directly attributable to the forcing function, and we call it the
forced response. It is the response which is present a long time after the switch
is closed.

5A.9
We see that the complete response is composed of two parts:
The complete
complete forced
natural
response is the sum
of the forced
response response response (5A.16) response and the
natural response
The forced response has the characteristics of the forcing function; it is found
The forced response
by pretending that all switches have been thrown a long time ago. For circuits is determined by
forcing function
with only switches and DC sources, the forced response is just the solution of a
simple DC circuit problem (all capacitors are open-circuits, all inductors are
short-circuits).
The natural response is a characteristic of the circuit, and not of the sources. Its
The natural
form may be found by considering the source-free circuit, and it has an response is
determined by the
amplitude which depends on the initial amplitude of the source and the initial circuit
energy storage.
The reason for the two responses, forced and natural, may also be seen from
physical arguments. We know that our circuit will eventually assume the
forced response. However, at the instant the switches are thrown, the initial
capacitor voltages (or the currents through the inductors in other circuits) will The natural
have values which depend only on the energy stored in these elements. These response provides
the link between the
voltages or currents cannot be expected to be the same as the voltages and initial state and final
state of a circuit
currents demanded by the forced response. Hence, there must be a transient
period during which the voltages and currents change from their given initial
values to their required final values. The portion of the response which
provides the transition from initial to final values is the natural response (often
called the transient response).

5A.10
The Forced and the Natural Response

There is an excellent mathematical reason for considering the complete
The solution to a
general differential
response to be composed of two parts, the forced response and the natural
equation consists of response. The reason is based on the fact that the solution of a linear
two parts
differential equation may be expressed as the solution of two parts, the
particular solution (forced response) and the complementary solution (natural
response).
Recall that a general linear differential equation with constant coefficients of

order n is an equation that can be written:
dny d n 1 y
an n an1 n1 a1 a0 y r t
dy
dt dt dt (5A.17)
This linear nonhomogeneous differential equation with constant coefficients

may also be written in the form:
f D y r t (5A.18)
We may identify r t as the forcing function.
We have already seen that the solution to the homogeneous differential

equation with constant coefficients is the solution of the equation:
f D yc 0
the complementary
solution (natural (5A.19)
response) and
Such a solution, yc , is called the complementary solution. We know that the

complementary solution has a form given by:
(5A.20)
yc c1e s1t c2e s2t cn e snt
where the ci s are arbitrary constants and the si s are the roots of the
characteristic equation.

5A.11
The particular solution, y p , is any solution we happen to come up with that
satisfies the original differential equation. That is, the particular solution
satisfies:
f D y p r t
the particular
(5A.21) solution (forced
response)
To see that the general solution of Eq. (5A.18) is composed of two parts, let the
complete solution be written as the sum of the particular solution and the
complementary solution:
The general solution
is the sum of the
y y p yc (5A.22)
particular solution
and the
complementary
solution
Substitution into Eq. (5A.18) results in:
f D y p yc f D y p f D yc
r t 0
r t (5A.23)
That is, we can safely add yc to any particular solution, since it contributes
nothing to the right-hand side.
We now have to find a way to obtain the particular solutions of differential

equations. We will look at three methods one where we utilise the concept of
an inverse differential operator, one where the solution can be obtained by
inspection, and one where the differential equation is first-order.

5A.12
Finding a Particular Solution using the Inverse Differential Operator
The concept of the differential operator can be used to find particular solutions
for nonhomogeneous differential equations.
In seeking a particular solution of:
f D y r t (5A.24)
it is natural to write:
(5A.25)
The inverse
1
differential operator
yp r t
is introduced
notationally f D
and then try to find an operator 1 f D so that the function y p will have
meaning and satisfy Eq. (5A.24). All we require is that:
with the only

(5A.26)
1
f D r t r t
requirement that it
f D
gives a solution to
the original
We proved before that:
f D e st f s e st (5A.27)
This suggests that we define:

The effect of the
1 st 1 st
f s 0
inverse differential
operator on an e e ,
exponential function
f D f s (5A.28)
Now from Eq. (5A.27) it follows that:
f s e st
(5A.29)
e st
f D e st
f s f s
and thus, with the requirement of Eq. (5A.26), Eq. (5A.28) is verified.

5A.13
Example
Solve the equation:
D 2

D y e 2t
Here the roots of the characteristic equation are s 0, 1 . Further,
f D D 2 D and f 2 0 . Hence, using Eq. (5A.28):
1 e 2t 1
yp 2 e 2
2t
e 2t
D D 2 2 6
So the solution is:
y c1 c2e t 16 e 2t
Example
Find a particular solution of:
D 2
9y 5 3et
Here the roots of the characteristic equation are s 3, 3 and f D D 2 9 .

We also have f 0 0 and f 1 0 . Hence, using superposition, we have:
1 5 5
y1 5 2
D 9
2
0 9 9
and:
1 3et 3
y2 3e t
et
D 9
2
1 9
2
8
Hence:
y p y1 y2 95 83 et
is a particular solution.

5A.14
Finding a Particular Solution by Inspection
For the second method of finding a particular solution, we note that it is

frequently easy to obtain a particular solution of a nonhomogeneous equation:
a D
n
n
an1D n1 a1D a0 y r t (5A.30)
by inspection.
For example, if r t is a constant R0 and a0 0 , then:
The particular
(5A.31)
solution for a R
constant forcing yp 0
function is a a0
constant
is a solution because all derivatives of y p are zero.
Example
Solve the equation:
D 2

3D 2 y 16
We obtain the complementary function:
yc c1e t c2e 2t
By inspection a particular solution of the original equation is:
y p 162 8
Hence the general solution is:
y c1e t c2e 2t 8

5A.15
Finding a Particular Solution using an Integrating Factor
For the third method of finding a particular solution, let us restrict ourselves to
first-order differential equations. The general equation of the type encountered
in analysing the RC circuit of the previous section can be written as:
dv
Pv Q (5A.32)
dt
P and Q can, in general, be functions of time. We identify Q as the forcing
function, and P as a quantity due solely to the circuit configuration. Following
the steps as before, we first multiply both sides by an integrating factor equal
to e Pt . This gives:
dv Pt
e Pt e Pv e Pt Q (5A.33)
dt
Thus, recognising that the left hand-side is the derivative of ve Pt , we have:
d
dt

ve Pt Qe Pt (5A.34)
Integrating both sides with respect to time gives:
ve Pt Qe Pt dt A
(5A.35)
where A is a constant of integration. Since the constant is explicitly shown, we

should remember that no integration constant needs to be added to the
remaining integral when it is evaluated.
Dividing both sides by the integrating factor gives:

The general solution
v e Pt Qe Pt dt Ae Pt
of a first-order
differential
(5A.36) equation

5A.16
If Qt , the forcing function, is known, then it remains only to evaluate the

integral to obtain the complete response. However, we shall not evaluate such
an integral for each problem. Instead we are interested in using Eq. (5A.36) to
draw several general conclusions.
We should note first that, for a source-free circuit, Q must be zero, and the
solution is the natural response:
has a natural
(5A.37)
response vn Ae Pt
In linear time-invariant circuits, P is always a positive constant that depends

only on the passive circuit elements and their interconnection in the circuit (if a
circuit contains a dependent source or a negative resistance, it is possible for P
to be negative). The natural response therefore approaches zero as time
increases without limit.
We therefore find that one of the two terms making up the complete response
has the form of the natural response. It has an amplitude which will depend on
the initial energy of the circuit as well as the initial value of the forcing
function.
We next observe that the first term of Eq. (5A.36) depends on the functional
and a forced form of Qt , the forcing function. Whenever we have a circuit in which the
response
natural response dies out as t becomes infinite, then this first term must
describe the response completely after the natural response has disappeared.
Thus, this term is the forced response.

5A.17
We have only considered those problems involving the sudden application of

DC sources, and Qt is therefore a constant for all values of time after the
switch has been closed. We can now evaluate the integral in Eq. (5A.36),
obtaining the forced response:
(5A.38)
Q The forced response
vf for DC excitation
P
and we can write the complete response:
Q The complete
v v f vn Ae Pt (5A.39) response for DC
P excitation
For the RC series circuit, Q P is the constant voltage V and 1 P is the time
constant T. We can see that the forced response might have been obtained
without evaluating the integral, because it must be the complete response at
infinite time. The forced response is thus obtained by inspection using DC
circuit analysis.
In the following section we shall attempt to find the complete response for
several RC circuits by obtaining the forced and natural responses and adding
them.

5A.18
Step-Response of RC Circuits
We will use the simple RC series circuit to illustrate how to determine the
complete response by the addition of the forced and natural response. This
circuit, shown below, has been analysed earlier, but by a longer method.
The simple RC
circuit driven by a
step-voltage R
Vu ( t ) C v( t )
Figure 5A.7
The desired response is the voltage across the capacitor, vt , and we first
express this voltage as the sum of the forced and natural voltage:
The complete
response expressed
as the sum of the v v f vn (5A.40)
forced response and
natural response
The functional form of the natural response must be the same as that obtained
without any sources. We therefore replace the step-voltage source by a short-
circuit and recognize the resulting parallel source-free RC circuit. Thus:
The form of the
(5A.41)
vn Ae t RC
natural response
that results from the
source-free circuit
where the amplitude A is yet to be determined.

5A.19
We next consider the forced response, that part of the response which depends
upon the nature of the forcing function itself. In this particular problem the
forced response must be constant because the source is a constant V for all
positive values of time.
After the natural response has died out the capacitor must be fully charged and
the forced response is simply:
The forced response

vf V (5A.42) obtained from DC
circuit analysis
Note that the forced response is determined completely there is no unknown

amplitude. We next combine the two responses:
(5A.43)
v V Ae t RC
and apply the initial condition to evaluate A. The voltage is zero prior to t 0 ,
Determining the
and it cannot change value instantaneously since it is the voltage across a amplitude of the
decaying
capacitor. Thus, the voltage is zero immediately after t 0 , and: exponential term
0 V A (5A.44)
and:
The complete

v V 1 e t RC (5A.45) response obtained
without solving a
differential equation!

5A.20
The complete response is plotted below, and we can see the manner in which
the voltage builds up from its initial value of zero to its final value of V.
A graph of the step-

response of the RC
circuit v
V
0.6321V
0
t
T 2T 3T 4T
Figure 5A.8
The transition is effectively accomplished in a time 4T . In one time constant,

the voltage has attained 63.2% of its final value.

5A.21
Example
a t=0
b
10
60 50 v( t )
200 50 mF
120 V
50 V i( t )
The switch is assumed to have been in position a for a long time, or, in other
words, the natural response which resulted from the original excitation of the
circuit has decayed to a negligible amplitude, leaving only a forced response
caused by the 120 V source.
We begin by finding the forced response when the switch is in position a. The
voltages throughout the circuit are all constant, and there is thus no current
through the capacitor (which is treated like an open-circuit). Simple voltage
division determines the forced response prior to t 0 :
50
vf 120 100, t 0
50 10
and we thus have the initial condition:
v0 100
Since the capacitor voltage cannot change instantaneously, this voltage is

equally valid at t 0 and t 0 .

5A.22
The switch is now thrown to b, and the complete response is:
v v f vn
The form of the natural response is obtained by replacing the 50 V source by a

short-circuit and evaluating the equivalent resistance:
1
Req 24
1 50 1 200 1 60
t ReqC
vn Ae
or
vn Ae t 1.2
In order to evaluate the forced response with the switch at b, we wait until all
the voltages and currents have stopped changing, thus treating the capacitor as
an open circuit, and use voltage division once more:
vf
50200 50 200 50 20
60 50200 50 200
Thus:
v 20 Ae t 1.2
and from the initial condition already obtained:
100 20 A
Therefore, the complete response is:
v 20 80e t 1.2 , t 0

5A.23
This response is sketched below:
v
100
20
0
t
-1 1T 2 3 4
The natural response is seen to form a transition from the initial to the final
response.
Finally, let us calculate some response that need not remain constant during the
instant of switching, such as i t in the circuit diagram. With the contact at a, it
is evident that i 50 260 0.192 A . When the switch is in position b, the
forced response for this current now becomes:
50 50
if 0.1
60 50200 50 200 50 200
The form of the natural response is the same as that which we already
determined for the capacitor voltage:
in Ae t 1.2
Combining the forced and natural responses, we obtain:
i 0.1 Ae t 1.2

5A.24

To evaluate A, we need to know i 0 . This is found by fixing our attention on
the energy-storage element, here the capacitor, for the fact that v must remain
100 V during the switching interval is the governing condition establishing

other currents and voltages at t 0 . Since v 0 100 V , and since the

capacitor is in parallel with the 200 resistor, we find i 0 0.5 , A 0.4 ,
and thus:
i t 0.192 t0
i t 0.1 0.4e t 1.2
t0
or:

i t 0.192u t 0.1 0.4e t 1.2 u t
This response is sketched below:
0.5
0.192
0.1
0
t
-1 1T 2 3 4

5A.25
Example
Let us apply a rectangular voltage pulse of amplitude V and duration t0 to the

simple RC series circuit.
R
v( t )
V -V u ( t - t0 )
C v( t )
Vu ( t )
0 t0 t
(a) (b)
We represent the forcing function as the sum of two step-voltage sources Vu t

and Vu t t0 , and plan to obtain the response by using the superposition
principle. We will designate that part of vt which is due to the lower source
Vu t acting alone by the symbol v1 t and then let v2 t represent that part due
to Vu t t0 acting alone. Then:
vt v1 t v2 t
We now write each of the partial responses v1 and v2 as the sum of a forced
and a natural response. The response v1 t is familiar:

v1 t V 1 e t RC , t 0
Note that the range of t , t 0 , in which the solution is valid, is indicated.

5A.26
We now consider the upper source and its response v2 t . Only the polarity of
the source and the time of its application are different. There is thus no need to
determine the form of the natural response and the forced response, the
solution for v1 t enables us to write:

v2 t V 1 e t t0 RC , t t0
where the applicable range of t , t t0 , must again be indicated.
We now add the two solutions carefully, since each is valid over a different
interval of time: Thus:

vt V 1 e t RC 0 t t0
vt V 1 e t RC
V 1 e t t0 RC
t t0
or, simplifying the second equation:

vt V 1 e t RC 0 t t0
vt V e t0 RC
1e t RC
t t0

5A.27
The solution is completed by stating that vt is zero for negative t and

sketching the response as a function of time. The type of curve obtained
depends upon the relative values of t0 and the time constant T. Two possible
curves are shown below:
v v
V V
0
t t
1
2 t0 t0 2t0 0 t0 2t0 3t0 4t0
(T) (T)
(a) (b)
The left curve is drawn for the case where the time constant is only one-half as
large as the length of the applied pulse the rising portion of the exponential
has therefore almost reached V before the decaying exponential begins.
The opposite situation is shown to the right there, the time constant is twice
t0 and the response never has a chance to reach the larger amplitudes.

5A.28
Analysis Procedure for Single Time Constant RC Circuits

The procedure we have been using to find the response of an RC circuit after
Step-by-step guide
DC sources have been switched on or off or in or out of the circuit at some
to solving STC RC instant of time, say t 0 , is summarized in the following. We assume that the
circuits with step-
sources circuit is reducible to a single equivalent resistance Req in parallel with a single
equivalent capacitance Ceq when all independent sources are set equal to zero,
i.e. we have a single time constant circuit. The response we seek is represented
by f t .
1. With all independent sources set to zero, simplify the circuit to

determine Req , Ceq , and the time constant T ReqCeq .
2. Viewing Ceq as an open circuit, use DC-analysis methods to find

vC 0 , the capacitor voltage just prior to the discontinuity.
3. Again viewing Ceq as an open circuit, use DC-analysis methods to
find the forced response. This is the value approached by f t as

t ; we represent it by f .
4. Write the total response as the sum of the forced and natural
responses: f t f Ae t T .
5. Find f 0 by using the condition that vC 0 vC 0 . If desired, Ceq
may be replaced by a voltage source vC 0 [a short circuit if

vC 0 0 ] for this calculation. With the exception of capacitor
voltages, other voltages and currents in the circuit may change
abruptly.
6.
Then f 0 f A and f t f f 0 f e t T .

5A.29
RL Circuits
The complete response of any RL circuit may also be obtained as the sum of
Solving RL circuits
the forced response and natural response. uses the same
techniques as for
RC circuits
Example
i (t )
50u(t )
6 3H
50 V
The circuit contains a DC voltage source as well as a step-voltage source. Let

us determine i t for all values of time. We might choose to replace everything
to the left of the inductor by the Thvenin equivalent, but instead let us merely
recognize the form of that equivalent as a resistor in series with some voltage
source. The circuit contains only one energy-storage element, the inductor, and
the natural response is therefore a negative exponential:
L 3
T 2
Req 1.5
and:
in Ae t 2
The forced response must be that produced by a constant voltage of 100 V. The
forced response is constant, and no voltage is present across the inductor (it
behaves like a short-circuit) and therefore:
100
if 50
2

5A.30
Thus:
i i f in 50 Ae t 2
In order to evaluate A, we must establish the initial value of the inductor

current. Prior to t 0 , this current is 25 A, and it cannot change
instantaneously. Thus:
25 50 A or A 25
Hence:
i 50 25e t 2 , t 0
We complete the solution by also stating:
i 25, t 0
or by writing a single expression valid for all t:
i 25 251 e t 2 u t A
The complete response is sketched below:
50
25
0
t
-2 2 4 6 8
Note how the natural response serves to connect the response for t 0 with the
constant forced response.

5A.31
Analysis Procedure for Single Time Constant RL Circuits

This is the dual of the statements given for the analysis procedure for single
time constant RC circuits.
The procedure we have been using to find the response of an RL circuit after
DC sources have been switched on or off or in or out of the circuit at some
instant of time, say t 0 , is summarized in the following. We assume that the Step-by-step guide
to solving STC RL
circuit is reducible to a single equivalent resistance Req in parallel with a single circuits with step-
sources
equivalent inductance Leq when all independent sources are set equal to zero,
i.e. we have a single time constant circuit. The response we seek is represented
by f t .
1. With all independent sources set to zero, simplify the circuit to

determine Req , Leq , and the time constant T Leq Req .
2. Viewing Leq as a short circuit, use DC-analysis methods to find
iL 0 , the inductor current just prior to the discontinuity.
3. Again viewing Leq as a short circuit, use DC-analysis methods to find
the forced response. This is the value approached by f t as t ;

we represent it by f .
4. Write the total response as the sum of the forced and natural
responses: f t f Ae t T .
5. Find f 0 by using the condition that iL 0 iL 0 . If desired, Leq
may be replaced by a current source iL 0 [an open circuit if
iL 0 0 ] for this calculation. With the exception of inductor

currents, other voltages and currents in the circuit may change
abruptly.
6.
Then f 0 f A and f t f f 0 f e t T .

5A.32
Summary
The unit-step function can be used to simulate the opening and closing of
switches under certain conditions.
The complete response of a circuit to a forcing function consists of two

parts: the forced response, and the natural response.
Mathematically, the forced response is the particular solution of a

nonhomogeneous linear differential equation with constant coefficients.
The natural response is the complementary solution to the corresponding
homogeneous equation.
For circuits driven by constant voltages or currents, the forced response can
be obtained by undertaking DC circuit analysis.
For single time constant circuits, the natural response is a decaying

exponential, y Ae t T , where T is the time constant.
The complete response of single time constant circuits to DC excitation

always takes the form of f t f f 0 f e t T . We just need to
determine the time constant of the circuit T, the forced response f , and
the initial response f 0 .
References
Hayt, W. & Kemmerly, J.: Engineering Circuit Analysis, 3rd Ed., Mcgraw-Hill,
1984.

5A.33
Exercises
1.
A current source of 5 A, a 4 resistor, and a closed switch are in parallel. The
switch opens at t 0 , closes at t 0.2 s , opens at t 0.4 s , and continues in
this periodic pattern. Express the voltage across the switch as an infinite
summation of step functions.
2.
30 u (t ) V
2 k
vC ( t ) 25 F v1 ( t ) 2 k 10 u (t ) mA
Find vC t and v1 t .
3.
10 u (t ) A vC ( t ) 1 F 1 10 u (t ) V
2 i1
i1
Find vC t .

5A.34
4.
After being open for several minutes, the switch in the circuit below closes
at t 0 .
t=0 4 k 1 k
Vs vC ( t ) 0.2F 100 V
Find vC t for all t if Vs :
(a) -200 V
(b) +100 V
5.
100
100 V R C
Specify values for R and C in the circuit so that the capacitor voltage will reach
80 V, 10 ms after the switch is closed, but will not drop below 90 V until 0.5 s
after the switch is opened, assuming that it has been closed for a very long
time.

5A.35
6.
10 V
ix t=0
10 v (t ) 2 F 40 V
0.2ix
Find vt for t 0 .
7.
A 250 resistor and a source, 12u t 12u t 103 V , are in series with a
0.2 H inductor. Find the inductor current magnitude at t 0.5 and 2 ms.
8.
The general solution to the driven series RL circuit is given by Eq. (5A.36).
Use it to find i t for t 0 if R 250 , L 0.2 H , and the source voltage is:
(a) 100u t V
(b) 100e 100t u t V
(c) 100 cos1250t u t V

5A.36
9.
20 4 mH
6 u (t ) A v 40 60 u (t ) V
Find v as a function of time.
10.
40
100 u (t ) V 2A 2H
Find the power being absorbed by the inductor at t :
(a) 0 (b) 0 (c) 0.05 (d)

5A.37
Leonhard Euler (1707-1783) (Len ard Oy ler)
The work of Euler built upon that of Newton and made mathematics the tool of
analysis. Astronomy, the geometry of surfaces, optics, electricity and
magnetism, artillery and ballistics, and hydrostatics are only some of Eulers
fields. He put Newtons laws, calculus, trigonometry, and algebra into a
recognizably modern form.
Euler was born in Switzerland, and before he was an adolescent it was

recognized that he had a prodigious memory and an obvious mathematical gift.
He received both his bachelors and his masters degrees at the age of 15, and
at the age of 23 he was appointed professor of physics, and at age 26 professor
of mathematics, at the Academy of Sciences in Russia.
Among the symbols that Euler initiated are the sigma () for summation
(1755), e to represent the constant 2.71828(1727), i for the imaginary 1
(1777), and even a, b, and c for the sides of a triangle and A, B, and C for the
opposite angles. He transformed the trigonometric ratios into functions and
abbreviated them sin , cos and tan , and treated logarithms and exponents as
functions instead of merely aids to calculation. He also standardised the use of
for 3.14159
His 1736 treatise, Mechanica, brought a universality to Newton mechanics

with the guidance of a rigorous mathematical notation. An introduction to pure
mathematics, Introductio in analysin infinitorum, appeared in 1748 which
treated algebra, the theory of equations, trigonometry and analytical geometry.
In this work Euler gave the formula e ix cos x i sin x . It did for calculus what
Euclid had done for geometry. Euler also published the first two complete
works on calculus: Institutiones calculi differentialis, from 1755, and
Institutiones calculi integralis, from 1768.
Euler's work in mathematics is vast. He was the most prolific writer of

mathematics of all time. After his death in 1783 the St Petersburg Academy
continued to publish Euler's unpublished work for nearly 50 more years!

5A.38
Some of his phenomenal output includes: books on the calculus of variations;
on the calculation of planetary orbits; on artillery and ballistics; on analysis; on
shipbuilding and navigation; on the motion of the moon; lectures on the
differential calculus. He made decisive and formative contributions to
geometry, calculus and number theory. He integrated Leibniz's differential
calculus and Newton's method of fluxions into mathematical analysis. He
introduced beta and gamma functions, and integrating factors for differential
equations. He studied continuum mechanics, lunar theory, the three body
problem, elasticity, acoustics, the wave theory of light, hydraulics, and music.
He laid the foundation of analytical mechanics. He proved many of Fermats
assertions including Fermats Last Theorem for the case n 3 . He published a
full theory of logarithms of complex numbers. Analytic functions of a complex
variable were investigated by Euler in a number of different contexts, including
the study of orthogonal trajectories and cartography. He discovered the
Cauchy-Riemann equations used in complex variable theory.
Euler made a thorough investigation of integrals which can be expressed in

terms of elementary functions. He also studied beta and gamma functions. As
well as investigating double integrals, Euler considered ordinary and partial
differential equations. The calculus of variations is another area in which Euler
made fundamental discoveries.
He considered linear equations with constant coefficients, second order

differential equations with variable coefficients, power series solutions of
differential equations, a method of variation of constants, integrating factors, a
method of approximating solutions, and many others. When considering
vibrating membranes, Euler was led to the Bessel equation which he solved by
introducing Bessel functions.
Euler made substantial contributions to differential geometry, investigating the

theory of surfaces and curvature of surfaces. Many unpublished results by
Euler in this area were rediscovered by Gauss.

5A.39
Euler considered the motion of a point mass both in a vacuum and in a resisting
medium. He analysed the motion of a point mass under a central force and also
considered the motion of a point mass on a surface. In this latter topic he had to
solve various problems of differential geometry and geodesics.
He wrote a two volume work on naval science. He decomposed the motion of a

solid into a rectilinear motion and a rotational motion. He studied rotational
problems which were motivated by the problem of the precession of the
equinoxes.
He set up the main formulas for the topic of fluid mechanics, the continuity
equation, the Laplace velocity potential equation, and the Euler equations for
the motion of an inviscid incompressible fluid.
He did important work in astronomy including: the determination of the orbits

of comets and planets by a few observations; methods of calculation of the
parallax of the sun; the theory of refraction; consideration of the physical
nature of comets.
Euler also published on the theory of music...
Euler did not stop working in old age, despite his eyesight failing. He
eventually went blind and employed his sons to help him write down long
equations. Euler died of a stroke after a day spent: giving a mathematics lesson
to one of his grandchildren; doing some calculations on the motion of balloons;
and discussing the calculation of the orbit of the planet Uranus, recently
discovered by William Herschel.

6A.1
Lecture 6A The Phasor Concept

Sinusoidal signals. Sinusoidal steady-state response. The complex forcing
function. Phasors. Phasor relationships for R, L and C. Impedance.
Admittance.
Introduction
The sinusoid is the most important function in electrical engineering. There are
several reasons for this.
It was the great Swiss mathematician Euler (1707-1783) who first identified
the fact that:
The special
For circuits described by linear differential equations a (6A.1) relationship enjoyed
by sinusoids and
sinusoidal source yields a sinusoidal response. linear circuits
The response sinusoid has the same frequency as the source, it is however
altered in amplitude and phase.
Only sinusoids have this property with respect to linear systems. For example,
applying a square wave input does not produce a square wave output.
The source / response form-invariance of the sinusoid can be attributed to the

fact that derivatives and integrals of a sinusoid yield sinusoids. Euler
recognised that this was because the sinusoid is really composed of exponential
components, and it is the exponential function which enjoys the rather peculiar
property that:
dt

d t
e et (6A.2)
That is, the derivative of the exponential function is itself an exponential

function. We have already seen that this relationship plays a vital role in
determining the solution to differential equations, where the form of each term
has to be the same for the differential equation to work, i.e. so that each term,
such as 3 dy dt or 5 d 2 y dt 2 , can actually be added together.

6A.2
The second major reason for the importance of the sinusoid can be attributed to
Joseph Fourier (1768-1830), who in 1807 recognised that:
Periodic signals are
made up of Any periodic function can be represented as the weighted (6A.3)
sinusoids - Fourier
Series sum of a family of sinusoids.
With this observation, we can analyse the behaviour of a linear circuit for any
periodic forcing function by determining the response to the individual
sinusoidal (or exponential) components and adding them up (superposition).
This decomposition of a periodic forcing function into a number of

appropriately chosen sinusoidal forcing functions is a very powerful analytical
method. It is called frequency-domain analysis, and we shall use it extensively.
The third reason is that the sinusoid is easy to generate, transmit and utilise.
Sinusoids are
Most of our electric power is generated as a sinusoid, and the functional form
utilised in a lot of
practical of the sinusoid is needed to make most of our motors turn (it doesnt really
applications
matter about the lights or heaters). It is therefore of great practical importance
to engineers who specialise in power and machines.
It is also extremely important to engineers who specialise in communications.

Our modern-day communications are carried through the air (and the
vacuum of space) on sinusoidal carriers. In fact, the basis of analog
communication is to modulate some aspect of a sinusoid such as its
amplitude (AM), its frequency (FM), or its phase (PM).
With digital communications, we encode the binary ones and zeros into some
aspect of the sinusoid, e.g. amplitude shift keying (ASK), frequency shift
keying (FSK), phase shift keying (PSK), and Gaussian minimum-shift keying
(GMSK) which is used in the Groupe Spcial Mobile (GSM) mobile telephone
network.
Thus, the sinusoid plays a prominent role in electrical engineering due to both
its theoretical and practical importance.

6A.3
Sinusoidal Signals
Consider the sinusoid:
vt Vm cost (6A.4)
which is shown graphically below:
The cosinusoid
graphed
v( t )
Vm
2
t (rad)
3
2 2
-Vm
(a)
v( t )
Vm
t (s)
T T 3T T
4 2 4
-Vm
(b)
Figure 6A.1
The amplitude of the cosine wave is Vm , and the argument is t . The radian
frequency or angular frequency is . In Figure 6A.1 (a) the function repeats

itself every 2 radians, and its period is therefore 2 radians. In Figure 6A.1
(b), the function is plotted against t and the period is now T where T 2 .
A cosine wave having a period T seconds per period must execute 1 T periods
each second its frequency is therefore:
1 The relationship
f (6A.5)
between frequency
and period
T

6A.4
Since T 2 , we have the relationship between frequency and radian

frequency:
The relationship
between frequency
and radian 2f (6A.6)
frequency
A more general form of the sinusoid is:
vt Vm cost (6A.7)
which now includes a phase angle in its argument t . Eq. (6A.7) is

plotted below as a function of t , and the phase angle appears as the number
of radians by which the original cosine wave, shown as a dotted line, is shifted
to the left or earlier in time.
The phase angle
defined
v( t )
Vm
Vm cos ( t )
2 t (rad)
-Vm
Vm cos ( t + )
Figure 6A.2
Since corresponding points on the cosinusoid cost occur rad earlier

compared to cost , we say that cost leads cost by rad.
Conversely, cost lags cost by rad. In either case, leading or
lagging, we say that the cosinusoids are out of phase. If the phase angles are
equal, they are said to be in phase.
In electrical engineering the phase angle is commonly given in degrees rather

than radians. Thus, instead of writing v 100 cos100t 6 we write
v 100 cos100t 30 .

6A.5
Sinusoidal Steady-State Response

We are now ready to apply a sinusoidal forcing function to a simple circuit and
obtain the forced response. We shall first write the differential equation which
applies to the given circuit. We know that the complete solution is composed of
the natural response and the forced response.
The form of the natural response is independent of the mathematical form of The natural
the forcing function and depends only upon the type of circuit and the element response is
determined by the
values. We have already determined the natural response of simple RC and RL circuit
circuits.
The forced response has the mathematical form of the forcing function. We
therefore expect the forced response to be sinusoidal. The term steady-state is The forced response
for a sinusoid is
used synonymously with forced response, and the circuits we are about to termed the
sinusoidal steady-
analyse are commonly said to be in the sinusoidal steady-state. state response
Unfortunately, steady-state implies not changing with time, but this is not
correct the sinusoidal forced response definitely changes with time. The
steady-state simply refers to the condition which is reached after the natural
response has died out.
Consider the series RL circuit below:
R i( t )
vs ( t ) =
L
Vm cos( t )
Figure 6A.3
The sinusoidal source voltage vs Vm cost has been switched into the
circuit at some remote time in the past, and the natural response has died out
completely.

6A.6
We seek the forced response, or steady-state response, or particular solution,
and it must satisfy the differential equation:
Ri Vm cost
di
L (6A.8)
dt
We now invoke the inverse differential operator to find a particular solution to
the differential equation and write:
LD R i Vm cost
1
i Vm cost
LD R (6A.9)
We only know the effect of the inverse differential operator on exponentials.

We therefore use Eulers formula:
e j cos j sin (6A.10)

Eulers identity
and note also that:
to write:
A cosinusoid
expressed as a sum e j e j
of complex cos (6A.12)
exponentials 2
We shall also later use:
A sinusoid
expressed as a sum
e j e j
of complex
sin (6A.13)
exponentials
j2

6A.7
Then Eq. (6A.9) can be written:
Vm e jt e jt
i
LD R 2 (6A.14)
Now remembering that:
1 st 1 st
e e , f s 0
f D f s
(6A.15)
we can use superposition and rewrite Eq. (6A.14) as:
Vm e jt Vm e jt
i
R jL 2 R jL 2 (6A.16)
Realizing the denominators, we get:
R jL e j t
i 2 Vm
R 2 L2 2
R j L e j t
2 V (6A.17)
R L 2 2 m
2
Collecting terms gives:
RVm e jt e jt
i 2
R 2 L2 2
LVm e jt e jt
2
R 2 L2 j2 (6A.18)

6A.8
Using Eulers identities for cos and sin , the forced response is obtained:
it cost
RVm
R 2 2 L2
LVm
2 sin t
R L 2 2
(6A.19)
This expression is quite unwieldy, and a clearer picture of the response can be
obtained by expressing the response as a single sinusoid with a phase angle.
We will therefore let:
it A cost (6A.20)
After expanding the function cost , we have:
(6A.21)
cost
RVm
A cos cos t A sin sin t
R 2 L2
2
LVm
2 sin t
R 2 L2
Equating like coefficients of cost and sin t , we find:
RVm
A cos
R 2 2 L2
LVm
A sin 2
R 2 L2 (6A.22)

6A.9
To find A and , we square both equations and add the results:
A2 cos 2 A2 sin 2 A2
R 2Vm
2
2 L2Vm 2

R 2
2 L2
2
R 2 2 L2 2
2
V
2 m 2 2 (6A.23)
R L
and also divide one equation by the other:
A sin L
tan (6A.24)
A cos R
Hence:
Vm
A (6A.25)
R L
2 2 2
and:
L (6A.26)
tan 1
R
The alternative form of the forced response therefore becomes:
The forced current

1 L
it
Vm response for the
cos t tan (6A.27) series RL circuit to a
R 2 2 L2 R sinusoidal voltage
source
We can see that the amplitude of the response is proportional to the amplitude
of the forcing function thus linearity between input and output holds (e.g.
doubling the input leads to a doubling of the output). We can also see that the
current decreases for any increase in R , or L , but not proportionately.

6A.10
Also, the current lags the applied voltage by tan 1 L R , an angle between
0 and 90 . When 0 or L 0 , the current must be in phase with the
voltage since the former situation is DC and the inductor appears as a short-
circuit, and the latter situation is a resistive circuit. If R 0 then the current
lags the voltage by 90 .
Note also that the frequency of the response is the same as the forcing function.
The applied voltage and the resultant current are shown below:
The forced response
graphed, showing
only an amplitude
and phase change v, i

i( t )
0 2 t
v( t )
Figure 6A.4
The fact that current lags the voltage in this simple RL circuit is now visually
apparent.
The method by which we found the sinusoidal steady-state response for the
simple RL circuit is quite intricate. It would be impractical to analyse every
circuit by this method. We shall see in the next section that there is a way to
simplify the analysis. It involves the formulation of complex algebraic
equations instead of differential equations, but the advantage is that we can
produce a set of complex algebraic equations for a circuit of any complexity.
Sinusoidal steady-state analysis becomes almost as easy as the analysis of
resistive circuits.

6A.11
The Complex Forcing Function

It seems strange at first, but the use of complex quantities in sinusoidal steady-
state analysis leads to methods which are simpler than those involving only real
quantities.
Consider a sinusoidal source:
Vm cost (6A.28)
which is connected to a general, passive, linear, time-invariant (LTI) circuit as

shown below:
Excitation of a
passive LTI circuit
by a real sinusoid
produces a real
sinusoidal response
passive
Vm cos ( t+ ) LTI Im cos ( t+ )
circuit
Figure 6A.5
A current response in some other branch of the circuit is to be determined, and

we know for a sinusoidal forcing function that the forced response is
sinusoidal. Let the sinusoidal forced response be represented by:
I m cost (6A.29)
Note that the frequency stays the same only the amplitude and phase are
unknown.

6A.12
If we delay the forcing function by 90 , then since the system is time-

invariant, the corresponding forced response must be delayed by 90 also
(because the frequencies are the same). Thus, the forcing function:
Vm cost 90 Vm sin t (6A.30)
will produce a response:
I m cost 90 I m sin t (6A.31)
Since the circuit is linear, if we double the source, we double the response. In
fact, if we multiply the source by any constant k, we achieve a response which
is k times bigger. We now construct an imaginary source we multiply the
source by j 1 . We thus apply:
jVm sin t (6A.32)
and the response is:
jI m sin t (6A.33)
The imaginary source and response are shown below:

Excitation of a
passive LTI circuit
by an imaginary
sinusoid produces
an imaginary
sinusoidal response passive
j Vm sin ( t+ ) LTI j Im sin ( t+ )
circuit
Figure 6A.6
We have applied a real source and obtained a real response, and we have
applied an imaginary source and obtained an imaginary response. We can now

6A.13
use the superposition theorem (the circuit is linear) to find the response to a
complex forcing function which is the sum of the real and imaginary forcing
functions. Thus, the sum of the forcing functions of Eqs. (6A.28) and (6A.32)
is:
Vm cost jVm sin t (6A.34)
and it produces a response which is the sum of Eqs. (6A.29) and (6A.33):
I m cost jI m sin t (6A.35)
The complex source and response may be represented more simply by applying
Eulers identity. Thus, the forcing function:
Vm e j t (6A.36)
produces:
I m e j t (6A.37)
The complex source and response are illustrated below:
Excitation of a
passive LTI circuit
by a complex source
produces a complex
passive response
Vm e j ( t+ ) LTI Im e j ( t+ )
circuit
Figure 6A.7

6A.14
We are now ready to see how this helps with sinusoidal analysis. We first note
that the real part of the complex response is produced by the real part of the
complex forcing function, and the imaginary part of the complex response is
produced by the imaginary part of the complex forcing function.
We analyse circuits Our strategy for sinusoidal analysis will be to apply a complex forcing function
in the sinusoidal
steady-state by whose real part is the given real forcing function we should then obtain a
using a complex
forcing function complex response whose real part is the desired real response.
whose real part is
the given real
forcing function We will try this strategy on the previous RL circuit:
R i
vs = Vm cos( t ) L
Figure 6A.8
The real source Vm cost is applied, and the real response i t is desired.
We first construct the complex forcing function by adding an appropriate

imaginary component to the given real forcing function. The necessary
complex source is:
Vm cost jVm sin t Vm e jt (6A.38)
The complex response which results is expressed in terms of an unknown

amplitude I m and an unknown phase angle :
I m e j t (6A.39)

6A.15
Writing the differential equation for this circuit:
di
Ri L vs (6A.40)
dt
we insert our complex expressions for vs and i :
RI m e j t L
d
dt

I m e j t Vm e jt (6A.41)
Taking the indicated derivative gives:

Using complex
j t j t jt (6A.42)
RI m e jLI m e Vm e sources and
responses reduces
the original
which is a complex algebraic equation. This is a considerable advantage we to a complex
algebraic equation
have turned a differential equation into an algebraic equation. The only
penalty is that the algebraic equation uses complex numbers. It will be seen
later that this is not a significant disadvantage.
In order to determine the value of I m and , we divide through by the common
factor e jt :
RI m e j jLI m e j Vm (6A.43)
Factoring the left side gives:
R jL I me j Vm (6A.44)
Rearranging, we have:
The complex
jVm
I me (6A.45)
response expressed
R jL in rectangular form

6A.16
This is the complex response, and it was obtained in a few easy steps. If we
express the response in exponential or polar form, we have:
The complex
(6A.46)
e
response expressed
j Vm j tan 1 L R
in polar form
I me
R 2 2 L2
Thus, by comparison:
(6A.47)
Vm
Im
R 2 2 L2
and:
tan 1 L R (6A.48)
We said that the complex response was:
I m e j t I m e j e jt (6A.49)
and that the real response was just the real part of the complex response.
Therefore, the real response i t is obtained by multiplying both sides of Eq.
(6A.46) by e jt and taking the real part. Thus:
it I m cost
The real sinusoidal
steady-state
response
Vm L
cos t tan 1
R L
2 2 2
R (6A.50)
This agrees with the response derived before using the D operator.
Although the analysis was straightforward, we have not yet taken advantage of
the full power of the complex representation. In order to do so, we must
introduce the concept of the phasor.

6A.17
The Phasor
A sinusoidal voltage or current at a given frequency is characterized by only
two parameters, an amplitude and a phase. The complex representation of the
voltage or current is characterized by a magnitude and an angle. For example,
the assumed sinusoidal form of the current response in the previous example
was:
I m cost (6A.51)
and the corresponding representation of this current in complex form is:
I m e j t (6A.52)
Once I m and are specified, the current is exactly defined. Throughout any
linear circuit operating in the sinusoidal steady-state at a single frequency ,

every voltage and current may be characterized completely by a knowledge of
its amplitude and phase angle.
A sinusoid of a
All sinusoidal responses in a linear circuit have a frequency of . given frequency is
Therefore, instead of writing I m cost , we could just say amplitude I m amplitude and
specified by an
phase
and phase .
All complex responses in a linear circuit have the factor e jt . of

A complex response
a given frequency
j t is specified by a
Therefore, instead of writing I m e , we could just say magnitude I m and
magnitude and
angle
angle .
Thus, we can simplify the voltage source and current response of the example
by representing them concisely as complex numbers:
Vm e j 0 and I m e j (6A.53)

6A.18
We usually write the complex representation in polar form. Thus, the source
voltage:
vt Vm cost (6A.54)
is represented in complex form as:
V Vm 0 (6A.55)
and the current response:
i t I m cost
A general (6A.56)
sinusoid
as:
and its phasor

representation I I m (6A.57)
The abbreviated complex representation is called a phasor. Phasors are printed

in boldface because they are effectively like a vector, they have a magnitude
and direction (angle). In hand writing, we normally place a tilde underneath:
V Vm0 and I I m (6A.58)

~ ~
Capital letters are used to represent phasors because they are constants they
are not functions of time.
Introducing the In general, we refer to xt as a time-domain representation and the

frequency-domain
corresponding phasor X as a frequency-domain representation.

6A.19
We can see that the magnitude of the complex representation is the amplitude
of the sinusoid and the angle of the complex representation is the phase of the
sinusoid.
It is a simple matter to convert a signal from the time-domain to the frequency-

domain it is achieved by inspection:
The time-domain
xt A cost X Ae j and frequency-
domain relationships
for a sinusoid
amplitude magnitude
phase angle (6A.59)
Example
If x t 3sint 30 then we have to convert to our cos notation:

x t 3cost 120 . Therefore X 3 120 .
Note carefully that X 3 cost 120 . All we can say is that

x t 3cost 120 is represented by X 3 120 .
The convenience of complex numbers extends beyond their compact

Phasors make
representation of the amplitude and phase. The sum of two phasors corresponds manipulating
sinusoids of the
to the sinusoid which is the sum of the two component sinusoids represented same frequency
easy
by the phasors. That is, if x 3 t x1 t x 2 t where x1 t , x 2 t and x 3 t are
sinusoids with the same frequency, then X3 X1 X 2 .
Example
If x 3 t cost 2sint then X3 10 2 90 1 j 2 2.2463 which
corresponds to x 3 t 2.24cost 63 .

6A.20
Formalisation of the Relationship between Phasor and Sinusoid
Using Eulers identity:
we have:
Ae j e jt Ae j t A cos t jA sin t (6A.61)
We can see that the sinusoid A cos t represented by the phasor
X Ae j is equal to the real part of Xe jt . Therefore:
The phasor / time-

domain relationship
xt Re Xe jt (6A.62)
This can be visualised as:
Graphical
interpretation of
rotating phasor /
time-domain Im
relationship
Xe j t = Ae je j t
A
Re
x( t )
complex plane
Figure 6A.9

6A.21
Graphical Illustration of the Relationship between a Phasor and its
Corresponding Sinusoid
Consider the representation of a sinusoid by its phasor: xt ReXe jt .

Graphically, x t can be generated by taking the projection of the rotating
phasor formed by multiplying X by e jt , onto the real axis:
A sinusoid can be
generated by taking
complex plane time-domain the real part of a
rotating complex
number
X
Figure 6A.10
Phasor Representations
Phasors can be represented in four different ways:
X X m polar form
X X m e j exponential form
X X m cos j sin trigonometric form
X a jb rectangular form

6A.22
Phasor Relationships for R, L and C

Now that we can transform into and out of the frequency-domain, we can
Phasor relationships
for the passive
derive the phasor relationships for each of the three passive circuit elements.
elements This will lead to a great simplification of sinusoidal steady-state analysis.
We will begin with the defining time-domain equation for each of the
elements, and then let both the voltage and current become complex quantities.
After suppressing e jt throughout the equation, the desired relationship
between the phasor voltage and phasor current will become apparent.
Phasor Relationships for a Resistor
The resistor provides the simplest case. The defining time-domain equation is:
vt Rit (6A.63)
If we apply a complex voltage Vm e j t and assume a complex current
I m e j t , we obtain:
Vm e j t RI m e j t (6A.64)
By dividing throughout by e jt , we find:
Vm e j RI m e j (6A.65)
or in polar form:
Vm RI m (6A.66)
But Vm and I m are just the voltage and current phasors V and I . Thus:
Phasor V-I
relationship for a
resistor V RI (6A.67)

6A.23
Equality of the angles and is apparent, and the current and voltage are
thus in phase.
The voltage-current relationship in phasor form for a resistor has the same form
as the relationship between the time-domain voltage and current as illustrated
below:
i I
v=Ri R V=R I R
time-domain frequency-domain
Figure 6A.11
Example
Assume a voltage of 8 cos100t 50 across a 4 resistor. Working in the

time-domain, the current is:
vt
i t 2 cos100t 50
R
The phasor form of the same voltage is 8 50 , and therefore:
V
I 2 50
R
If we transform back to the time-domain, we get the same expression for the
current.
No work is saved for a resistor by analysing in the frequency-domain because

the resistor has a linear relationship between voltage and current.

6A.24
Phasor Relationships for an Inductor
The defining time-domain equation is:
di t
vt L (6A.68)
dt
After applying the complex voltage and current equations, we obtain:
Vm e j t L
d
dt

I m e j t (6A.69)
Taking the indicated derivative:
Vm e j t jLI m e j t (6A.70)
Suppressing e jt , we find:
Vm e j jLI m e j (6A.71)
Thus the desired phasor relationship is:

Phasor V-I
relationship for an
inductor
V j L I (6A.72)
The time-domain equation Eq. (6A.68) has become an algebraic equation in the
frequency-domain. The angle of jL is exactly 90 and you can see from
Eq. (6A.71) that 90 . I must therefore lag V by 90 in an inductor.

6A.25
The phasor relationship for an inductor is indicated below:
i I
di
v=L L V=j LI L
dt
Figure 6A.12
Example
Assume a voltage of 8 cos100t 50 across a 4 H inductor. Working in the

vt
i t dt
L
2 cos100t 50dt
0.02 sin 100t 50
0.02 cos100t 140
V 8 50
I 0.02 140
jL 1004 90
current.

6A.26
Phasor Relationships for a Capacitor
The defining time-domain equation is:
dvt
it C (6A.73)
dt
I m e j t C
d
dt

Vm e j t (6A.74)
I m e j t jCVm e j t (6A.75)
Suppressing e jt , we find:
I m e j jCVm e j (6A.76)
Thus the desired phasor relationship is:

Phasor V-I
relationship for a
capacitor I j C V (6A.77)
Thus I leads V by 90 in a capacitor.

6A.27
The time-domain and frequency-domain representations are compared below:
dv
i=C dt I =j C V
v V
C C
Figure 6A.13
Example
Assume a voltage of 8 cos100t 50 across a 4 F capacitor. Working in the

dvt
i t C
dt
4 8 cos100t 50
d
dt
3200 sin 100t 50
3200 cos100t 40
I jCV 1004 90 8 50 320040
current.

6A.28
Summary of Phasor Relationships for R, L and C
We have now obtained the phasor V I relationships for the three passive
elements. These results are summarized in the table below:
Summary of phasor
V-I relationships for Time-domain Frequency-domain
the passive
elements
v V
i I
v Ri V RI
R R
v V
i I
di V j L I
vL
dt
L j L
v V
i 1 1 I
C
v idt V I
jC
C 1 j C
All the phasor equations are algebraic. Each is also linear, and the equations
relating to inductance and capacitance bear a great similarity to Ohms Law.
Before we embark on using the phasor relationships in circuit analysis, we

need to verify that KVL and KCL work for phasors. KVL in the time-domain
is:
v1 t v2 t vn t 0 (6A.78)
If all voltages are sinusoidal, we can now use Eulers identity to replace each
real sinusoidal voltage by the complex voltage having the same real part,
suppress e jt throughout, and obtain:
KVL and KCL are

obeyed by phasors V1 V2 Vn 0 (6A.79)
Thus KVL holds. KCL also holds by a similar argument.

6A.29
Analysis Using Phasor Relationships
We now return to the series RL circuit that we considered several times before,
shown as (a) in the figure below. We draw the circuit in the frequency-domain,
as shown in (b):
A circuit and its
frequency-domain
R R equivalent
i( t ) I
VR
vs ( t ) = Vs VL j L
L
Vm cos( t )
(a) (b)
Figure 6A.14
From KVL in the frequency-domain:
VR VL Vs (6A.80)
We now insert the recently obtained V I relationships for the elements:
RI jLI Vs (6A.81)
The phasor current is then found:
Vs
I (6A.82)
R j L
The source has a magnitude of Vm and a phase of 0 (it is the reference by

which all other phase angles are measured). Thus:
V 0 The response of the

I m (6A.83)
circuit in the
R j L
frequency-domain

6A.30
The current may be transformed to the time-domain by first writing it in polar

form:
tan 1 L R
Vm
I
R 2 2 L2
I m (6A.84)
Transforming back to the time-domain we get:
The response of the it I m cost

circuit in the time-
domain
Vm L
cos t tan 1 (6A.85)
R L
2 2 2
R
which is the same result as we obtained before the hard way.
Impedance
The voltage-current relationships for the three passive elements in the
frequency-domain are:
I
V RI V j L I V (6A.86)
j C
If these equations are written as phasor-voltage phasor-current ratios, we get:
Phasor V-I
V V V 1
relationships for the
R j L (6A.87)
j C
passive elements
I I I
These ratios are simple functions of the element values, and in the case of the
inductor and capacitor, frequency. We treat these ratios in the same manner we
treat resistances, with the exception that they are complex quantities and all
algebraic manipulations must be those appropriate for complex numbers.

6A.31
We define the ratio of the phasor voltage to the phasor current as impedance,
symbolized by the letter Z :
V
Z
Impedance defined
(6A.88)
I
The impedance is a complex quantity having the dimensions of ohms.
Impedance is not a phasor and cannot be transformed to the time-domain by

multiplying by e jt and taking the real part.
In the table below, we show how we can represent a resistor, inductor or

capacitor in the time-domain with its frequency-domain impedance:
Time-domain Frequency-domain Impedances of the

three passive
elements
R R
L j L
C 1 j C
Impedances may be combined in series and parallel by the same rules we use
for resistances.
In a circuit diagram, a general impedance is represented by a rectangle:
I Z
V = ZI
Figure 6A.15

6A.32
Example
We have an inductor and capacitor in series:
100 F
5 mH
At 104 rads-1 , the impedance of the inductor is Z L jL j 50 and the

impedance of the capacitor is Z C 1 jC j1 . Thus the series
combination is equivalent to Z eq Z L Z C j 50 j1 j 49 :
j 49
The impedance of inductors and capacitors is a function of frequency, and this

equivalent impedance is only valid at 10 4 rads-1 . For example, if
5000 rads-1 , then the impedance would be Z eq j 23 .
Impedance may be expressed in either polar or rectangular form.
In polar form an impedance is represented by:
Z Z
(6A.89)
No special names or symbols are assigned to the magnitude and angle. For
example, an impedance of 100 60 is described as having an impedance
magnitude of 100 and an angle of 60 .

6A.33
In rectangular form an impedance is represented by:

Impedance is
composed of a
Z R jX (6A.90) resistance (real part)
and a reactance
(imaginary part)
The real part, R, is termed the resistive component, or resistance. The
imaginary component, X, including sign, but excluding j, is termed the reactive
component, or reactance. The impedance 100 60 in rectangular form is
50 j86.6 . Thus, its resistance is 50 and its reactance is 86.6 .
It is important to note that the resistive component of the impedance is not

necessarily equal to the resistance of the resistor which is present in the circuit.
Example
Consider a resistor and an inductor in series:
20 5H
At 4 rads-1 , the equivalent impedance is Z eq 20 j 20 . In this case the
resistive component of the impedance is equal to the resistance of the resistor

because the network is a simple series network. Now consider the same
elements placed in parallel:
20 5H
The equivalent impedance is:
20 j 20
Z eq 10 j10
20 j 20
The resistive component of the impedance is now 10 .

6A.34
Example
We will use the impedance concept to analyse the RLC circuit shown below:
i ( t ) 1.5 k 1 k
1 1
40 sin (3000t ) V 3 H 6 F
The circuit is shown in the time-domain, and a time-domain response is

required. However, analysis should be carried out in the frequency-domain. We
therefore begin by drawing the frequency-domain circuit the source is
transformed to the frequency-domain, becoming 40 90 , the response is
transformed to the frequency-domain, being represented as I , and the
impedances of the inductor and capacitor, determined at 3000 rads-1 , are
j1 k and j 2 k respectively. The frequency-domain circuit is shown
below:
I 1.5 k 1 k
40 -90 V j1 k -j2 k
The equivalent impedance offered to the source is:
Z eq 1.5
j11 j 2 1.5 2 j
j1 1 j 2 1 j
2 j 1 j 1 j3
1 .5 1 .5
1 j 1 j 2
2 j1.5 2.536.9 k

6A.35
The phasor current is thus:
Vs 40 90
I 16 126.9 mA
Z eq 2.536.9
Upon transforming the current to the time-domain, the desired response is

obtained:
i t 16 cos3000t 126.9 mA
Admittance
The reciprocal of impedance can offer some convenience in the sinusoidal
steady-state analysis of circuits. We define admittance as the ratio of phasor
current to phasor voltage:
Admittance defined
I
Y (6A.91)
V
and thus:
1 Admittance is the
Y (6A.92)
reciprocal of
impedance
Z
The real part of the admittance is the conductance G , and the imaginary part
of the admittance is the susceptance B . Thus:
Admittance is
composed of a
1 1
Y G jB
conductance (real
(6A.93) part) and a
Z R jX susceptance
(imaginary part)
This formula should be scrutinized carefully. It does not mean that G 1 R

(unless Z R , a pure resistance), nor does it mean B 1 X .
Admittance, conductance and susceptance are all measured in siemens (S).

6A.36
Example
Consider a resistor and a capacitor in series:
0.1F
1
At 5 Mrads-1 , the equivalent impedance is Z eq 1 j 2 . Its admittance
is:
1 1 1 1 j2
Y
Z 1 j2 1 j2 1 j2
0.2 j 0.4 S
It should be apparent that the equivalent admittance of a circuit consisting of a

number of parallel branches is the sum of the admittances of the individual
branches. Thus, the admittance obtained above is equivalent to:
5 0.8 F
only at 5 Mrads-1 . As a check , the equivalent impedance of the parallel

network at 5 Mrads-1 is:
5 j 2.5
Z eq 1 j2
5 j 2.5
as before.

6A.37
Summary
Sinusoids are important theoretically and practically. A sinusoidal source
yields a sinusoidal response.
The sinusoidal forced response is also known as the sinusoidal steady-state

the condition which is reached after the natural response has died out.
A complex forcing function produces a complex response the real part of

the forcing function creates the real part of the response.
The application of a complex forcing function to a linear circuit turns the

describing differential equation into a complex algebraic equation.
The phasor representation of a sinusoid captures the amplitude and phase

information in a complex number amplitude corresponds to magnitude,
and phase corresponds to angle.
The phasor corresponding to xt A cost is X Ae j A .
Phasor V I relationships for the three passive elements lead to the

concept of frequency-domain impedance. The impedances of the three
passive elements are: Z R R , Z L jL , ZC 1 jC . Impedances can
be combined and manipulated like resistors except we use complex algebra.
Impedance consists of a real resistive component and an imaginary reactive

component: Z R jX .
Admittance is defined as the inverse of impedance: Y 1 Z G jB .
References
Hill, 1984.

6A.38
Exercises
1.
A sinusoidal voltage is zero and increasing at t 1.6 ms . The next zero
crossing occurs at t 4.65 ms .
(a) Calculate T , f and .
(b) If v0 20 V , find vt .
(c) By what angle does vt lead the current i 5 cost 110 A ?
2.
For the sinusoidal waveform shown below:
f (t )
8
0 t (ms)
1.2
-5
-8
Find:
(a) A and if f t A sin t (b) T (c) f (d)
(e) A and if f t A cost (f) f 0.0014

6A.39
3.
200 i ( t )
vs ( t ) 0.1 H
Let i 2 sin 500t 40 A . Determine the source voltage vs t .
4.
300 200
iL(t)
60cos(2000t ) V 600 150 mH
Find iL t .
5.
10 i( t )
vs ( t ) 2H
Let vs 20 cos 5t 30 cos10t V . Find i t .

6A.40
6.
Give in polar form:
(a) 20110 840 (b) 6.1 j 3.82 1.17 j 0.541
Give in rectangular form:
(c) j 6.3 j 9.71 (d) e j 2.152
7.
A box contains a voltage source vs1 and a current source is 2 . The voltage
between an available pair of terminals is labelled v AB .
If vs1 5 cos1000t 40 V and is 2 0.1cos500t 20 A , then
v AB 2 cos1000t 10 3 cos500t 30 V.
However, if vs1 5 cos500t 40 V while is 2 0.1cos1000t 20 A, then
v AB 3 cos1000t 20 2 cos500t 10 V.
Find v AB if vs1 20 cos1000t 10 cos500t V and
is 2 0.3 sin 1000t 0.2 sin 500t A .
8.
Assume that only three currents, i1 , i2 and i3 , enter a certain node.
(a) Find i1 t if I 2 65 110 mA and I 3 45 50 mA
(b) Find I 2 if i1 t 55 cos400t 40 mA and i3 t 35 sin 400t 70 mA .

6A.41
9.
At the input of an RLC circuit it is found that V 41 j18 V and
I 3.5 j 7.5 A . Assuming V and I satisfy the passive sign convention,
determine the power entering the network at:
(a) t 0 (b) t 2
10.
In the figure below, the voltage v is given as the phasor, V 60 j 25 V .
i
circuit
v
element
If 2000 rads-1 , find the power being delivered to the element at

t 0.1 ms if the element is:
(a) a 100 resistor (b) a 50 mH inductor (c) a 5 F capacitor
11.
The circuit below is operated at 100 rads-1 .
I1
IR
40 F 1H 400
(a) If I R 0.0120 A , find I1 .
(b) If I1 2 30 A , find I R .

6A.42
12.
Using a 1 H inductor and a 1 F capacitor, at what frequency (in hertz) may an
impedance be obtained having a magnitude of 2000 if the two elements are
combined in:
(a) series (b) parallel
13.
A 10 F capacitor and a 25 resistor are in parallel. What size inductor
should be placed in series with this parallel combination so that the impedance
of the final series network has zero reactance at 8 krads -1 ?
14.
What size capacitor should be placed in series with the series combination of
800 and 20 mH to give an admittance whose magnitude is 1 mS at
10 krads-1 ?
15.
If the input admittance and impedance of the network shown below are equal at
every frequency, find R and L.
L
Zin
R 1F
Yin

6A.43
Joseph Fourier (1768-1830) (Jo sef Foor yay)
Fourier is famous for his study of the flow of heat in metallic plates and rods.
The theory that he developed now has applications in industry and in the study
of the temperature of the Earths interior. He is also famous for the discovery
that many functions could be expressed as infinite sums of sine and cosine
terms, now called a trigonometric series, or Fourier series.
Fourier first showed talent in literature, but by the age of thirteen, mathematics
became his real interest. By fourteen, he had completed a study of six volumes
of a course on mathematics. Fourier studied for the priesthood but did not end
up taking his vows. Instead he became a teacher of mathematics. In 1793 he
became involved in politics and joined the local Revolutionary Committee. As
he wrote:-
As the natural ideas of equality developed it was possible to conceive the

sublime hope of establishing among us a free government exempt from kings
and priests, and to free from this double yoke the long-usurped soil of
Europe. I readily became enamoured of this cause, in my opinion the
greatest and most beautiful which any nation has ever undertaken.
Fourier became entangled in the French Revolution, and in 1794 he was

arrested and imprisoned. He feared he would go to the guillotine but political
changes allowed him to be freed. In 1795, he attended the Ecole Normal and
was taught by, among others, Lagrange and Laplace. He started teaching again,
and began further mathematical research. In 1797, after another brief period in
prison, he succeeded Lagrange in being appointed to the chair of analysis and
mechanics. He was renowned as an outstanding lecturer but did not undertake
original research at this time.
In 1798 Fourier joined Napoleon on his invasion of Egypt as scientific adviser.

The expedition was a great success (from the French point of view) until
August 1798 when Nelsons fleet completely destroyed the French fleet in the
Battle of the Nile, so that Napoleon found himself confined to the land he was
occupying. Fourier acted as an administrator as French type political

6A.44
institutions and administrations were set up. In particular he helped establish
educational facilities in Egypt and carried out archaeological explorations.
The Institute
d'gypte was While in Cairo, Fourier helped found the Institute d'gypte and was put in
responsible for the
completely charge of collating the scientific and literary discoveries made during the time
serendipitous
discovery of the in Egypt. Napoleon abandoned his army and returned to Paris in 1799 and soon
Rosetta Stone in
1799. The three held absolute power in France. Fourier returned to France in 1801 with the
inscriptions on this remains of the expeditionary force and resumed his post as Professor of
stone in two
languages and three Analysis at the Ecole Polytechnique.
scripts (hieroglyphic,
demotic and Greek)
enabled Thomas Napoleon appointed Fourier to be Prefect at Grenoble where his duties were
Young and Jean-
Franois many and varied they included draining swamps and building highways. It
Champollion, a
protg of Fourier, was during his time in Grenoble that Fourier did his important mathematical
to invent a method
work on the theory of heat. His work on the topic began around 1804 and by
of translating
hieroglyphic writings 1807 he had completed his important memoir On the Propagation of Heat in
of ancient Egypt in
1822. Solid Bodies. It caused controversy both Lagrange and Laplace objected to
Fouriers expansion of functions as trigonometric series.
This extract is from it was in attempting to verify a third theorem that I employed the
a letter found among procedure which consists of multiplying by cos x dx the two sides of the
Fouriers papers, equation
and unfortunately
lacks the name of
the addressee, but
x a0 a1 cos x a2 cos 2 x ...
was probably
intended for and integrating between x 0 and x . I am sorry not to have known the
Lagrange. name of the mathematician who first made use of this method because I
would have cited him. Regarding the researches of dAlembert and Euler
could one not add that if they knew this expansion they made but a very
imperfect use of it. They were both persuaded that an arbitraryfunction
could never be resolved in a series of this kind, and it does not seem that
any one had developed a constant in cosines of multiple arcs
[i.e. found a1 , a 2 ,, with 1 a1 cos x a 2 cos 2 x ... for 2 x 2 ]
the first problem which I had to solve in the theory of heat.
Other people before Fourier had used expansions of the form
f x ~ r a r expirt but Fouriers work extended this idea in two totally

new ways. One was the Fourier integral (the formula for the Fourier series
coefficients) and the other marked the birth of Sturm-Liouville theory (Sturm
and Liouville were nineteenth century mathematicians who found solutions to

6A.45
many classes of partial differential equations arising in physics that were
analogous to Fourier series).
Napoleon was defeated in 1815 and Fourier returned to Paris. Fourier was
elected to the Acadmie des Sciences in 1817 and became Secretary in 1822.
Shortly after, the Academy published his prize winning essay Thorie
analytique de la chaleur (Analytical Theory of Heat). In this he obtains for the
first time the equation of heat conduction, which is a partial differential
equation in three dimensions. As an application he considered the temperature
of the ground at a certain depth due to the suns heating. The solution consists
of a yearly component and a daily component. Both effects die off
exponentially with depth but the high frequency daily effect dies off much
more rapidly than the low frequency yearly effect. There is also a phase lag for
the daily and yearly effects so that at certain depths the temperature will be
completely out of step with the surface temperature.
All these predictions are confirmed by measurements which show that annual
variations in temperature are imperceptible at quite small depths (this accounts
for the permafrost, i.e. permanently frozen subsoil, at high latitudes) and that
daily variations are imperceptible at depths measured in tenths of metres. A
reasonable value of soil thermal conductivity leads to a prediction that annual
temperature changes will lag by six months at about 23 metres depth. Again
this is confirmed by observation and, as Fourier remarked, gives a good depth
for the construction of cellars.
As Fourier grew older, he developed at least one peculiar notion. Whether

influenced by his stay in the heat of Egypt or by his own studies of the flow of
heat in metals, he became obsessed with the idea that extreme heat was the
natural condition for the human body. He was always bundled in woollen
clothing, and kept his rooms at high temperatures. He died in his sixty-third
year, thoroughly cooked.
References
Krner, T.W.: Fourier Analysis, Cambridge University Press, 1988.

6B.1
Lecture 6B Circuit Simulation

Project Flow. Schematic capture. Simulation.
Introduction
The most popular circuit simulation package is known as PSpice. SPICE is
an acronym for Simulation Program with Integrated Circuit Emphasis a
program developed at the University of California, Berkeley in the 1970s that
was used to simulate analog electronic circuits. It quickly became an industry
standard versions became available in the public domain and large companies
developed their own versions and continue to do so. The P prefix came
about in 1984 when the first version for a PC became available. Most
professional computer-aided electronic design tools now incorporate PSpice
compatible simulators.
There are many software packages which use PSpice (or PSpice compatible)
simulators, and they are updated frequently. It is not the purpose of this
document to outline how each version of the various software packages are
used. There are many other resources that are available that do this online
tutorials, textbook appendices, student and demo versions of software, etc.
This document will give an overview of the procedure of circuit simulation,

and then highlight particular topics that are important for us in generating time-
domain and frequency-domain analyses of simple analog circuits.
This document will be based on OrCAD an integrated software package

used for electronic design automation (EDA). It captures schematic designs,
simulates them, and allows a printed circuit board to be designed for
manufacture. Demo versions are freely available.
Australians tend to use Altium Designer EDA software originally developed

in Hobart at the University of Tasmania. Altium is now a large international
company listed on the ASX, and has a large share of the EDA market.
PSpice and OrCAD are registered trademarks of Cadence Design Systems, Inc.

6B.2
Project Flow
The basic steps to running a circuit simulation are:
Start a new project
Draw the schematic
Set up and run simulation profiles
Starting a New Project
1. Open OrCAD Capture CIS - Demo Edition (freely available).
2. Go to File New Project.
3. Enter a name for the project.
4. Choose Analog or Mixed A/D.
5. Set the location. You should create a new directory for your project since
PSpice will generate a lot of project files in this folder.
6. Click OK.
7. Choose Create a blank project and click OK.
8. You should see a window where you can draw the schematic (i.e. your
circuit diagram).

6B.3
Drawing the Schematic
To add parts for your circuit (i.e. resistors, etc.)
1. Go to Place Parts.
2. Click on the library you want to use, or select multiple libraries by

holding Ctrl or dragging the mouse. In the part window you should see
at least the ANALOG, BIPOLAR, EVAL, SOURCE and SPECIAL
libraries.
3. Find the part you want to add and press OK.
4. Click where you want to place the part on your schematic. Press R to
rotate the part by 90 degrees.
5. Use wires to connect the part to complete your circuit.
Simulation
Simulations run via simulation profiles. You need to set up a simulation

profile for each of the types of simulation you want to run.
1. Go to PSpice New Simulation Profile.
2. Enter a name for the simulation profile, e.g. Step Response or

Frequency Response.
3. Press Create.
4. In the Analysis Type drop-down list, choose the simulation type, e.g.
Time Domain (Transient) for a step response simulation, AC Sweep /
Noise for a frequency response simulation.
5. Set up the parameters associated with the type of simulation.
6. To run the simulation, press F11 or choose PSpice Run.

6B.4
Schematic Capture
Here are some tips for laying out the schematic for simulation purposes.
Ground
There are many types of grounds (common points in the circuit) and there may
be more than one common point, e.g. an analog ground and a digital
ground in a mixed circuit. PSpice uses nodal analysis for circuit simulation
and therefore needs a reference node with zero voltage. This is provided by a
special ground symbol called 0/SOURCE. You need to have it in your circuit!
It looks like a ground symbol with a zero. If you don't have it, PSpice may
complain of "floating nodes" even if you have a ground.
To place the ground on the circuit Go to Place Ground and choose 0/source
(if you don't see source in the Libraries section, you will need to add the
source library). Alternatively, click on the ground button in the schematic

toolbar and choose 0/source. Or just push G.

6B.5
SI Unit Prefixes
PSpice uses letters to represent the most common SI unit prefixes:
SI Unit Prefix PSpice letter
giga G
mega Meg
kilo k
milli m
micro u
nano n
pico p
femto f
The two differences are mega and micro.
The mega prefix is written Meg (case does not matter). M is NOT
mega, it is milli.
Example: For 6.5 MHz, enter "6.5 Meg", for 3 mV, enter "3 m".
The micro prefix is not the Greek letter mu () as in the SI system of units,
because it is not supported on a standard keyboard. We use the letter u
instead.
Example: For 680 H , enter "680u".
All Parts Must Have Unique Names
You can't have two parts named "R1" in your circuit. If you are copying and
pasting parts or circuits, OrCAD usually increments the part number
automatically. However, you may accidentally name two parts the same, which
will cause a netlist error.

6B.6
Labeling Nodes
You should use aliases to label your input and output nodes. This makes your
nodes easier to find when dealing with simulation output. V(Vout) is simpler
than finding V(R1:1).
1. Go to Place Net Alias.
2. Enter a name, e.g., Vout or Vin.
3. Place the label close to a node.
The example below shows a simple circuit with aliases:

6B.7
Simulation
The following sections detail the procedures used in setting up and running
simulations.
DC Bias
The response of the circuit to DC sources is always calculated. To display DC

bias voltages, currents and power on your circuit after you run the simulation,
click on the Enable Bias Display buttons:
Time-Domain (Transient) Simulations
For sinusoids, use VSIN for your voltage source instead of VAC (VOFF is the
DC offset, VAMPL is the amplitude, and FREQ is the frequency).
For square and triangular waves, use VPULSE:
V2
PW
V1
TR TF
t
TD TD+PER
Parameter Meaning
V1 Initial value (V)
V2 Pulsed value (V)
TD Delay time (s)
TR Rise time (s)
TF Fall time (s)
PW Pulse width (s)
PER Period (s)
We normally set the delay time, TD, to zero.

6B.8
Square Wave
A square wave is the VPULSE function in the limit of TR=TF=0 and

PW=0.5*PER (PER is the period of the wave). This limiting case, however,
causes numerical difficulties in calculations. In any case, we can never make
such a square function in practice. In reality, square waves have very small TR
and TF. Typically, we use a symmetric function, i.e., we set TR=TF and
PW=0.5*PER-2*TR. Thus, for a given frequency we can set up the square
function if we choose TR. If we choose TR too large, the function does not look
like a square wave. If we choose TR too small, the program will take a long
time to simulate the circuit and for TR smaller than a certain value, the
simulation will not converge numerically. A good choice for TR is to set it to
be 1% of the PER (a period): TR=TF=0.01*PER, PW=0.48*PER. This
usually results in a nice signal without a huge amount of computational need.
Note that TR does not have to be exactly 1% of PER. You can choose nice
round numbers for TR, TF, and PW.
Triangular Wave
A triangular wave is the VPULSE function in the limit of TR=TF=0.5*PER

and PW=0 (convince yourself that this is the case). As before, the limiting case
of PW=0 causes numerical difficulties in calculations. So we have to choose PW
to be a reasonably small value. A good choice for PW is to be set at 1% of the
PER (period): PW=0.01*PER, TR=TF=0.495*PER. This usually results in
a nice signal without a huge amount of computational need. Again, note that
PW does not have to be exactly 1% of PER. You can choose nice round
numbers for TR, TF, and PW.

6B.9
Simulation Settings
1. Go to PSpice Edit Simulation Profile.
2. Select the "Time Domain (Transient)" Analysis type.
3. Enter a Run to time: so that a few periods will be displayed.

Remember that the period (seconds) = 1/frequency (Hz).
Example: If you are using a 1 kHz sine wave, it has a 1/1 kHz = 1 ms
period, so use a Run to time of 5 ms for 5 periods.
4. Set the Maximum step size to be much smaller than the period.
Example: For a 1 kHz sine wave, it has a 1 ms period, so set a

Maximum step size of approximately 0.01ms. This works out to 100
data points per period.
5. If you don't set the Maximum step size, PSpice may choose one which
is too big, making your waveforms look angular and ugly (because it
plots straight lines between data points).

6B.10
AC Sweep / Noise Simulations
Use VAC for your voltage source.
Simulation Settings
1. Go to PSpice Edit Simulation Profile.
2. Select the "AC Sweep / Noise" Analysis type.
3. Select the frequency range of interest. Don't start frequency sweeps at 0!
4. Set the Points/Decade to be at least 20.
Bode Plots
1. Use a logarithmic x-axis for the frequency.
2. The magnitude should be measured in decibels. Use the PSpice DB()

function to convert to decibels.
Example: Use DB(V(Vout)/V(Vin)), assuming you have labelled

your output and input nodes with Vout and Vin aliases.
3. Remember you also need a phase response (unless instructed otherwise).

Use the PSpice P() function to get the phase angle.
Example: Use P(V(Vout)/V(Vin)), assuming you have labelled

your output and input nodes with Vout and Vin aliases.

6B.11
4. Be sure to mark the cutoff frequency or other points of interest on your

Bode plots (on both magnitude AND phase graphs). Cutoff frequency is
defined as 3dB below the peak of the magnitude response (NOT always
at -3dB).
a. Click the "Toggle Cursor" button (or go through the menu,

Trace Cursor Display). You will now be able to move the cursor
along your plot.
b. Click the "Cursor Max" button to find the highest point (or go
through the menu, Trace Cursor Max).
c. Click the "Mark Label" button to label that point (or go

through the menu, Plot Label Mark).
d. Click the "Cursor Search" button (or go through the menu,

Trace Cursor Search Commands).
e. Select 1 for Cursor To Move to search along the y-axis of trace 1.
f. Enter "search forward level (max-3)" (don't enter the quotation

marks) to move the cursor to the right, to the point which is 3
below the max.

6B.12
g. Or enter "search back level (max-3)" (don't enter the quotation

marks) to move the cursor to the left.
h. Click the "Mark Label" button to label that cutoff point.
Unclick the Toggle Cursor button to disable the cursor so

you can move the label.
Double-click on the label to edit the text (to add units, or to
name the point).
i. It may help to increase the width of the traces in the plot.
Right click on a trace. Make sure the selection list has

Information, Properties, Cursor 1, and Cursor 2 (if it lists
Settings and Properties, you clicked on the background, not
on the trace).
Select Properties.
You can change the width and other properties of that trace.

6B.13
Example
We wish to simulate the following circuit to determine its magnitude response:
5 k
10 nF
1 k
vi
vo
We enter the following schematic into OrCAD.
Note the use of a real op-amp the LF411. You can search for it in the parts
library by Place Part then Part Search. The op-amp requires DC voltage
sources on its power pins.

6B.14
In this schematic, we have used the power symbol which simply ties
nets together with a common name in this case the name is +15V. One
tie is placed on the DC voltage source, the other on the op-amp power pin.
This avoids cluttering wires on the schematic.
We set up a frequency response simulation using AC Sweep as shown below:
for which the output is:

6B.15
Exercises
1.
Simulate the step response and frequency response of the following circuit:
10 nF
10 k
2.5 k 2
vi
TL071 6 vo
3
Use the following axes for graphing:
Response X Y
Step 0 s to 1 ms -5 V to 0 V
Magnitude 10 Hz to 100 kHz -30 dB to 20 dB
Phase 10 Hz to 100 kHz 90 to 180

7A.1
Lecture 7A Sinusoidal Steady-State Response

Analysis using phasors. Nodal analysis. Mesh analysis. Superposition. Source
transformations. Thvenins theorem. Nortons theorem. Phasor diagrams.
Power in the sinusoidal steady-state.
Introduction
In the analysis of resistive circuits of arbitrary complexity, we are able to
employ many different circuit analysis techniques to determine the response
nodal analysis, mesh analysis, superposition, source transformations,
Thvenins and Nortons theorems.
Sometimes one method is sufficient, but more often we find it convenient to

combine several methods to obtain the response in the most direct manner.
We now want to extend these techniques to the analysis of circuits in the

sinusoidal steady-state. We have already seen that impedances combine in the
same manner as do resistances. We have seen that KVL and KCL are obeyed
by phasors, and we also have an Ohm-like law for the passive elements,
V ZI . We can therefore extend our resistive circuit analysis techniques to
the frequency-domain to determine the phasor response, and therefore the
sinusoidal steady-state response.

7A.2
Analysis using Phasors

Phasors can only be used for sinusoidal steady-state analysis. They cannot be
used to determine transients in a circuit.
Phasor analysis is a transform method of analysis. In phasor analysis we

transform a problem from the time-domain to the frequency-domain. To find a
response in the frequency-domain, we solve equations using complex algebra.
Once the response is found, we transform the solutions back to the time-
domain. This is illustrated conceptually below:
differential
time-domain equations time-domain
problem solution
time-domain transform transform

to to
frequency-domain
complex
frequency- algebra frequency-
domain domain
problem solution
Figure 7A.1
Transform methods are a common method of analysis in branches of

engineering, and you will be introduced to more powerful transform methods
in more advanced subjects.

7A.3
Nodal Analysis
As an example of nodal analysis, consider the circuit shown below:
40 mF
v1 v2
2H
cos(5 t ) A 5 1H 10 0.5 sin (5 t ) A
20 mF
Figure 7A.2
Noting from the sources that 5 rads-1 , we draw the frequency-domain

circuit and assign nodal voltages V1 and V2 :
Nodal analysis in
the frequency-
domain
-j5
V1 V2
j10
1 0 A 5 j5 10 0.5 -90 A
-j10
Figure 7A.3
Each passive element is specified by its impedance, which has been determined
by knowing the frequency of the sources (which are the same) and the element
values. Two current sources are given as phasors, and phasor node voltages V1
and V2 are indicated.
At the left node, we apply KCL and I V Z :
V1 V V V2 V1 V2
1 1 1 j0 (7A.1)
5 j10 j5 j10

7A.4
At the right node:
V2 V1 V2 V1 V2 V2
j 0.5 (7A.2)
j5 j10 j 5 10
Combining terms we have the two equations:
0.2 j 0.2V1 j 0.1V2 1

(7A.3)
j 0.1V1 0.1 j 0.1V2 j 0.5
Using Cramers Rule to solve, we obtain:
1 j 0.1
j 0.5 0.1 j 0.1 0.1 j 0.1 0.05
V1
0.2 j 0.2 j 0.1 0.02 j 0.02 j 0.02 0.02 0.01
j 0.1 0.1 j 0.1
0.05 j 0.1
1 j2 V
0.05
0.2 j 0.2 1
j 0.1 j 0.5 0.1 j 0.1 j 0.1 (7A.4)
V2 2 j 4 V
0.05 0.05
The time-domain solutions are best obtained by representing the phasors in

polar form:
V1 5 63.4 V
(7A.5)
V2 2 5116.6 V
and passing to the time-domain:
v1 t 5 cos5t 63.4 V
v2 t 2 5 cos5t 116.6 V (7A.6)
Note how simple phasor analysis is compared to the work involved if we

stayed in the time-domain solving differential equations!

7A.5
Mesh Analysis
As an example of mesh analysis, consider the circuit shown below:
500 F
3 i1
3
10cos(10 t ) V 4 mH 2 i1
Figure 7A.4
Noting from the left source that 103 rads-1 , we draw the frequency-domain
circuit and assign mesh currents I1 and I 2 :
Mesh analysis in the
frequency-domain
-j2
3
10 0 V I1 I2 2I 1
j4
Figure 7A.5
Around mesh 1:
3I1 j 4I1 I 2 100 (7A.7)
while mesh 2 leads to:
j 4I 2 I1 j 2I 2 2I1 0 (7A.8)

7A.6
Combining terms we have the two equations:
3 j 4I1 j 4I 2 10
(7A.9)
2 j 4I1 j 2I 2 0
Solving:
14 j8
I1 1.24029.7 A
13
20 j 30
I2 2.77456.3 A (7A.10)
13
or:

i1 t 1.240 cos 103 t 29.7 A
i t 2.774 cos10 t 56.3 A
2
3 (7A.11)
The solution above could be checked by working entirely in the time-domain,

but it would be quite an undertaking!

7A.7
Superposition
Linear circuits are those that consist of any of the following: idealised linear
passive circuit elements (R, L and C), ideal independent voltage and current
sources and linearly dependent voltage and current sources. Such circuits are
amenable to the superposition principle.
We can analyse linear circuits with phasors and the principle of superposition.
(You may remember that linearity and superposition were invoked when we
combined real and imaginary sources to obtain a complex source).
Lets look again at the circuit of Figure 7A.3, redrawn below with each pair of
parallel impedances replaced by a single equivalent impedance (for example, 5
and j10 in parallel yield 4 j 2 ):
Superposition in the
frequency-domain
V1 -j10 V2
1 0 A 4 -j2 2+j4 0.5 -90 A
Figure 7A.6
To find V1 we first activate only the left source and find the partial response:
V1L 10
4 j 2 j10 2 j 4 4 j 28 2 j 2
4 j 2 j10 2 j 4 6 j8 (7A.12)
With only the right source active, current division helps us to obtain:
2 j4 6 j8
V1R 0.5 90 4 j 2 1
4 j 2 j10 2 j 4 6 j8 (7A.13)

7A.8
Summing, we get:
V1 2 j 2 1 1 j 2 V (7A.14)
which agrees with our previous result.
Thvenins Theorem
We will use the same circuit to see whether Thvenins Theorem can help us:
Thvenins Theorem
in the frequency-
domain
V1 -j10 V2
1 0 A 4 -j2 2+j4 0.5 -90 A
Figure 7A.7
Suppose we determine the Thvenin equivalent faced by the j10

impedance. The open circuit voltage (+ reference to the left) is:
Voc 104 j 2 0.5 902 j 4

4 j 2 2 j1 6 j 3 (7A.15)
The impedance of the inactive circuit, as viewed from the load terminals, is
simply the sum of the two remaining impedances (because the current sources
are set to zero open circuits). Hence:
Z th 6 j 2 (7A.16)

7A.9
Thus, when we reconnect the circuit, the current directed from node 1 toward
node 2 through the j10 load is:
6 j3 (7A.17)
I12 0.6 j 0.3
6 j 2 j10
Subtracting this from the left source current, the downward current through the
4 j 2 branch is found:
I1 1 0.6 j 0.3 0.4 j 0.3 (7A.18)
and , thus:
V1 Z1I1 4 j 20.4 j 0.3 1 j 2 V (7A.19)

7A.10
Nortons Theorem
Again using the same circuit, if our chief interest is in V1 we could use
Nortons Theorem on the three right elements:
Nortons Theorem in
the frequency-
domain
-j10
2+j4 0.5 -90 A
Figure 7A.8
The short circuit current is obtained using current division:
2 j4
I sc 0.5 90
2 j 4 j10
2 j 1 j
A (7A.20)
2 j6 4
and the Norton impedance (equal to the Thvenin impedance) is simply:
Z th 2 j 6 (7A.21)

7A.11
We thus need to analyse the circuit:
V1
1A 4 -j2 2-j6 1+j

A
4
Figure 7A.9
The voltage V1 is therefore:
V1
4 j 22 j 6 1 0.25 j 0.25
4 j 2 2 j 6
4 j 28
0.75 j 0.25 2 j 20.75 j 0.25 1 j 2 V (7A.22)
6 j8
It should now be clear that all methods available for linear circuit analysis can
be applied to the frequency-domain. The slight additional complexity that is
apparent now arises from the necessity of using complex numbers and not from
any more involved theoretical considerations.

7A.12
Phasor Diagrams
The phasor diagram is a sketch in the complex plane of the phasor voltages and
A phasor diagram is
a graphical sketch of currents throughout a specific circuit. It provides a graphical method for
phasors in the
complex plane solving problems which may be used to check more exact analytical methods.
Since phasor voltages and currents are complex numbers, they may be
identified as points in a complex plane. For example, the phasor voltage
V1 6 j8 1053.1 is identified on the complex voltage plane shown
below:
A simple phasor
diagram
Imaginary
axis (V)
j8 V1
10
53.1
Real axis (V)
6
Figure 7A.10
The axes are the real voltage axis and the imaginary voltage axis. The voltage
V1 is located by an arrow drawn from the origin. Since addition and
subtraction are particularly easy to perform and display on a complex plane, it
is apparent that phasors may be easily added and subtracted in a phasor
diagram. Multiplication and addition result in a change in magnitude and the
addition and subtraction of angles.

7A.13
Figure 7A.11 shows the sum of V1 and a second phasor voltage

V2 3 j 4 5 53.1 :
Phasor diagram
showing addition
V1
V1 + V2
V2
Figure 7A.11
Figure 7A.12 shows the current I1 , which is the product of V1 and the
admittance Y 1 j1 :
Phasor diagram
showing
multiplication
V1
I1 = 45
(1+ j 1) V1
Figure 7A.12
This last phasor diagram shows both current and voltage phasors on the same
complex plane it is understood that each will have its own amplitude scale,
but a common angle scale.

7A.14
The phasor diagram can also show the connection between the frequency-
domain and the time-domain. For example, let us show the phasor V Vm
on the phasor diagram, as in (a) below:
Phasor diagram
showing
transformation to the
time-domain by
rotating the phasor
and taking the real
part Vm t+
t
Vm V= Vm t+ Vm

(a) (b)
Figure 7A.13
In order to transform V to the time-domain, we first need to multiply by e jt .

We now have the complex voltage V Vme j e jt Vmt . This voltage
may be interpreted as a phasor which possesses a phase angle that increases
linearly with time. On a phasor diagram it therefore represents a rotating line
segment, the instantaneous position being t rad ahead (counterclockwise) of
Vm . Both Vm and Vmt are shown on the phasor diagram in (b).
The transformation to the time-domain is completed by taking the real part of

See the Phasors Vmt , which is the projection of the phasor onto the real axis. It is helpful
PC program
to think of the arrow representing the phasor V on the phasor diagram as the
snapshot, taken at t 0 , of a rotating arrow whose projection onto the real axis
is the instantaneous voltage vt .

7A.15
Example
The series RLC circuit shown below has several different voltages associated
with it, but only a single current:
3 j6
I
VR VL
Vs VC -j 10
The phasor diagram is constructed most easily by employing the single current
as the reference phasor all other phasors with have their angles measured
with respect to the reference. Let us arbitrarily set I I m0 and place it along
the real axis of the phasor diagram, as shown below:
VL VR + VL
j6
j5
VR I
3
-j 4
Vs = VR + VL + VC
-j 5
-j10 VR + VC
VC

7A.16
The resistor, inductor and capacitor voltages may next be calculated and placed
on the diagram, where the 90 phase relationships stand out clearly. The sum
of these three voltages is the source voltage for this circuit. The total voltage
across the resistance and inductance or resistance and capacitance or
inductance and capacitance is easily obtained from the phasor diagram.
We can design using the phasor diagram quite easily instead of embarking on
complex algebraic manipulation. For example, suppose we would like to
determine a single extra passive element that can be added in series with the
circuit without changing the peak value of the current. This additional circuit
element will contribute to an additional voltage drop, but we still must have
KVL satisfied so that the total voltage drop magnitude equals the source
voltage magnitude. Therefore, the addition of the voltage drop due to the new
element must keep the source voltage on a circle of radius Vs . From the
phasor diagram, we can see that we can only add an inductor with an
impedance Z L j8 .
We know that an increase in frequency will cause the voltage across the
inductor to increase, whilst simultaneously decreasing the voltage across the
capacitor. In fact, if we increase the frequency by 29.1%, the inductor voltage
and capacitor voltage will exactly cancel one another, and we have a condition
known as resonance.

7A.17
VL
j7.75 VR + VL
Vs = VR I
3
-j 7.75 VR + VC
VC

7A.18
Example
The figure below shows a simple parallel circuit in which it is logical to use the
single voltage between the two nodes as a reference phasor:
IR IC
Is V 5 50 F
=
2000 rad/s
Let us arbitrarily set V Vm0 and place it along the real axis of the phasor
diagram. The resistor current is in phase with this voltage, I R 0.2Vm A , and
the capacitor current leads the reference voltage by 90 , I C j 0.1Vm A . After

these two currents are added to the phasor diagram, shown below, they may be
summed to obtain the source current. The result is I s 0.2 j 0.1Vm A .
I C = j 0.1Vm I s = (0.2+ j 0.1) Vm
I R=0.2Vm V= Vm 0

7A.19
If the source current were specified initially as, for example, I s 10 A , and
the node voltage is not initially known, it is still convenient to begin
construction of the phasor diagram by assuming, say V 10 . The source
current, as a result of the assumed node voltage, is now I s 0.2 j 0.1 A . The
true source current is 10 A , however, and thus the true node voltage is
greater by the factor 1 0.2 j 0.1 the true node voltage is therefore 4 j 2 V .
The assumed voltage leads to a phasor diagram which differs from the true
phasor diagram by a change of scale (the assumed diagram is smaller by a
factor of 1 20 ) and an angular rotation (the assumed diagram is rotated
counterclockwise through 26.6 ). The real phasor diagram in this case is
shown below:
IC
I s = 1 0 A
IR
V= 20 -26.6 V
Phasor diagrams are usually very simple to construct, and most sinusoidal
steady-state analyses will be more meaningful if such a diagram is included.

7A.20
Power in the Sinusoidal Steady-State

Consider a single sinusoidal source supplying a network as shown below:
i v
v network t
Figure 7A.14
If the network contains reactive components, then in general there will be a

phase shift between the voltage and current. Let:
v Vm cost
i I m cost (7A.23)
and define:
(7A.24)
Thus is the angle by which the voltage leads the current.
Instantaneous Power
The instantaneous power delivered to the network is:
p vi
Vm I m cost cost

Vm I m
cos cos2t
2 (7A.25)
Notice that the first term is a constant, and the second term oscillates with time
at double the supply frequency.

7A.21
Average Power
Average power is just the average value of the instantaneous power. We define
this average in the normal way (the mean value theorem) as:
1 T2
pt dt
T T T 2
P lim (7A.26)
If the instantaneous power is periodic with period T0 , we have the special case:
1
pt dt
T0 2
P
T0 T0 2
(7A.27)
That is, for a periodic instantaneous power, we can integrate over one period,
and divide by the period. A graph of the instantaneous power in a network
operating in the sinusoidal steady-state (Eq. (7A.25)) is shown below:
Vm Im /2
Vm Im
cos
2
Vm Im /2 t
T0 T0
2
Figure 7A.15
From this graph it is easily seen that the average power is the constant part of
the instantaneous power (the oscillating part averages to zero) and we have:
P is the average
V I
P m m cos (W) (7A.28)
value of p

7A.22
Example
A graph of the instantaneous power is shown below for V 40 and

I 2 60 :
6
p
4 v
2 P
i
t
0 T0
-2
Figure 7A.16
Note that, on occasion, the power delivered to the network is negative, which
implies that the network is actually sourcing power back to the voltage supply.
The average power is calculated to be:
P
Vm I m
cos 1
2 42cos 60 2 W
2
Both the 2 W average power and its period, one-half the period of either the
current or the voltage, are evident in the graph. The zero value of the
instantaneous power at each instant when either the voltage or current is zero is
also apparent.

7A.23
Root-Mean-Square (RMS) Values
It is customary, when dealing with AC power, to refer to voltage and current

using a root-mean-square, or RMS value. As we shall see, this leads to some
simplification for many power formulas. The concept of an RMS value for a
voltage or current comes about by considering the average power dissipated in
a resistor when it carries a periodic current:
1 T0 2
P Ri 2 dt RI RMS
2
T0 T0 2 (7A.29)
That is, the RMS value of any periodic current is equal to the value of the
direct current which delivers the same average power. Removing R from the
above formula, we thus have:
RMS value defined

1 T0 2
I RMS
T0 T0 2
i 2 dt (7A.30)
The operation involved in finding this value is the root of the mean of the
square, hence the name root-mean-square value, or RMS value for short. A
similar expression is obtained for voltage, VRMS (or for any other signal for that
matter).

7A.24
RMS Value of a Sinusoid
A sinusoid is the most important special case of a periodic signal. Consider a

sinusoidal current given by:
i t I m cost (7A.31)
The easiest way to find its RMS value is by performing the mean-square
operations in Eq. (7A.30) graphically. For the arbitrary sinusoid given, we can
graph the square of the current, i 2 t I m2 cos 2 t , as shown below:
2
i (t )
I 2m
equal areas
2
I m
t
T0
Figure 7A.17
Note that in drawing the graph we dont really need to know the identity
cos 2 1 cos 2 2 all we need to know is that if we start off with a
sinusoid uniformly oscillating between I m and I m , then after squaring we
obtain a sinusoid that oscillates (at twice the frequency) uniformly between I m2
and 0. We can now see that the average value of the resulting waveform is
I m2 2 , because there are equal areas above and below this value. This is the
mean of the square, and so we now just take the root and get:
The RMS value of

Im
an arbitrary sinusoid
I RMS (7A.32)
2

7A.25
Note that the RMS value depends on the magnitude of the sinusoid only the
frequency and phase is irrelevant!
In the power industry, it is tacitly assumed that values of voltage and current
will be measured using their RMS value. For example, in Australia the
electricity delivered to your home has a frequency of 50 Hz and an RMS value
of 230 V. This means the voltage available at a general power outlet is a 50 Hz
sinusoid with a peak value of approximately 325 V.
It should be noted that this formula can only be applied to a sinusoid for other
waveforms, you will obtain a different ratio between the peak and the RMS
value. For example, the RMS value of a triangle waveform is I m 3 whilst for
a square wave it is simply I m .
Phasors and RMS Values
We defined a phasor corresponding to:
it I m cost (7A.33)
as:
I I m (7A.34)
We could have just as easily defined it to be:
Phasor magnitudes
(7A.35)
I can be defined as
I m I RMS RMS values
2
If we use this definition, then all relationships involving phasors, such as
V ZI , KCL, KVL, etc. must also use this definition. When working with
power and machines, it is customary to use the RMS value for the phasor
magnitude. In other fields, such as telecommunications and electronics, we use
the amplitude for the phasor magnitude. You need to be aware of this usage.

7A.26
Average Power Using RMS Values
Returning to the formula for average power, we can now rewrite it using RMS
values. We have:
The average power

P using RMS values
of voltage and
P VRMS I RMS cos (W)
current
(7A.36)
Apparent Power
The average power in a DC network is simply P VI . In sinusoidal steady-

state analysis, we define:
Apparent power
defined apparent power VRMS I RMS (VA) (7A.37)
Dimensionally, average power and apparent power have the same units, since
cos is dimensionless. However, to avoid confusion, the term volt-amperes, or
VA, is applied to apparent power.
Power Factor
The ratio of average power to the apparent power is called the power factor,
symbolized by PF:
P
Power factor defined
PF (7A.38)
VRMS I RMS
In the sinusoidal case, the power factor is also equal to:
PF cos (7A.39)
We usually refer to leading PF or lagging PF when referring to loads to resolve

the ambiguity in taking the cos. The terms leading and lagging refer to the
phase of the current with respect to the voltage. Thus, inductive loads have a
lagging power factor, capacitive loads have a leading power factor.

7A.27
Complex Power
Using RMS phasors of v Vm cost and i I m cost :
Vm
V VRMS
2 (7A.40)
I
I m I RMS
2
we know that the average power is:
P VRMS I RMS cos (7A.41)
In the same way that we introduced the concept of the phasor, we now
associate the average power with the real part of a complex power:
P ReVRMS I RMS e j (7A.42)
ReVRMS e j I RMS e j
Thus, we can define average power as:
P Re VI* (7A.43)
Note the use of I * because of the way we defined . It is an accident

of history that was defined this way, as it just as easily could have been
defined as .
We therefore define complex power as:
Complex power
S VI *
(complex VA) (7A.44) defined

7A.28
In polar form the complex power is:
(7A.45)
S VI* VRMS I RMS
so we can now see that the apparent power is:
(7A.46)
S V I VRMS I RMS
Written in rectangular form, the complex power is:
S VRMS I RMS cos j sin P jQ (7A.47)
Reactive Power
From the rectangular form, we can see that the average power is also the real
power. It can also be seen that the imaginary power, which we call the
reactive power, is:
Reactive power
defined Q VRMS I RMS sin ( var) (7A.48)
It has the same dimensions as the real power P, the complex power S and the
apparent power S . In order to avoid confusion, the unit of Q is defined as the
volt-ampere-reactive, or var for short.
Reactive power is a measure of the energy flow rate into or out of the reactive
components of a network. It is positive for inductive loads, and negative for
capacitive loads.
The physical interpretation of reactive power causes a lot of confusion. Even

though it is the imaginary component of complex power, it has a physical and
real interpretation, and must be generated by a power system. (A voltage
phasor, such as V 3 j 4 , has an imaginary component of 4 which
contributes to the amplitude and phase of the real voltage sinusoid so we
expect Q to also be physically real).

7A.29
To see how Q manifests itself in the real world, we return to the formula for
instantaneous power. Using cos A B cos A cos B sin A sin B , and noting
that , we get:
p VRMS I RMS cos cos2t

VRMS I RMS cos cos2t 2
VRMS I RMS cos cos2t 2 cos sin 2t 2 sin
(7A.49)
VRMS I RMS cos 1 cos2t 2 VRMS I RMS sin sin 2t 2
Then:
p P1 cos2t 2 Q sin 2t 2 (1.1)
The instantaneous power associated with the real and reactive power
components is shown below:
p associated with P
total p
P
Q
t
0
p associated with Q
Figure 7A.18
Note that the instantaneous power associated with P follows a cos, and the
instantaneous power associated with Q follows a sin. Thus, the two
waveforms are 90 apart and are said to be in quadrature. You can see that
reactive power does not transfer energy instantaneous power is both
delivered to, and received from, the network in a cyclic fashion, with an
average of zero. In contrast, real power does transfer energy instantaneous
power is always delivered to the network in a cyclic fashion, but it has a non-
zero average.

7A.30
Summary of Power in AC Circuits
In summary we have:
S VI* S P jQ (7A.50)
where:
Components of
complex power S V I apparent power (VA)
Real power is also
known as active P S cos real power (W)
power and average
Q S sin reactive power (var)

power
(7A.51)
= angle by which the current I lags the voltage V

2
S P2 Q2
These relationships are illustrated below:
jQ S
V
|S|
I

P
Figure 7A.19
It can be shown that the total complex power S P jQ consumed by a

network is the sum of the complex powers consumed by all the component
parts of the network. This conservation property is not true of the apparent
power S .

7A.31
Example
Consider three loads connected in parallel across a 230 V (RMS) 50 Hz line as

shown below:
V S1 S2 S3
Load 1 absorbs 10 kW and 7.5 kvar. Load 2 absorbs 3.84 kW at 0.96 PF

leading. Load 3 absorbs 5 kW at unity power factor. Find the overall power
factor.
The first load is given in rectangular form:
S1 10 j 7.5 kVA
The complex power supplied to the second load must have a real part of
3.84 kW and an angle (refer to the power triangle) of cos 1 0.96 16.26 .
Hence,
P 3.84
S2 16.26 4 16.26 3.84 j1.12 kVA
cos 0.96
The third load is simply:
S 3 5 j 0 kVA
The total complex power is:
S total S1 S 2 S 3 10 j 7.5 3.84 j1.12 5 j 0 18.84 j 6.38 kVA
Thus, the combined load is operating at a power factor equal to:
P 18.84
PF 0.9472
S 19.89

7A.32
The magnitude of the line current drawn by the combined load is:
S 19890
I 86.48 A
V 230
Electricity supply authorities do all they can to improve the PF of their loads by
installing capacitors or special machines called synchronous condensers which
supply vars to the system. They also impose tariffs which encourage consumers
to correct their PF.
If we now seek to raise the PF to 0.98 lagging, without affecting the existing
real power, the total complex power must become:
18.84
S new cos-1 0.98 19.2211.48 18.84 j 3.826 kVA
0.98
We would therefore need to add a corrective load of:
S 4 S new S total j 2.554 kVA
Now since:
2
VV* V
S 4 VI * *
*
4
Z4 Z4
then:
2
V 2302
Z4 * j 20.71
S4 j 2554
Thus, the corrective load is a capacitor of value:
1 1
C 153.7 F
X C 2 50 20.71
The magnitude of the line current drawn by the new combined load is:
S 19220
I 83.57 A
V 230

7A.33
Summary
A linear circuit can be converted to the frequency-domain where we use the
concept of phasors and impedances, and in particular the branch phasor
relationship V ZI . The circuit is then amenable to normal circuit analysis
techniques: nodal analysis, mesh analysis, superposition, source
transformations, Thvenins theorem, Nortons theorem, etc. Time-domain
responses are obtained by transforming phasor responses back to the time-
domain.
A phasor diagram is a sketch in the complex plane of the phasor voltages

and currents throughout a circuit and is a useful graphical tool to illustrate,
analyse and design the sinusoidal steady-state response of the circuit.
The RMS value of a sinusoid xt X m cos(t ) is X RMS X m 2.
There are many power terms in AC circuits:
- instantaneous power, p t vt i t (W)

- average power, P VRMS I RMS cos (W)
- reactive power, Q VRMS I RMS sin , (var)
- complex power, S VI* P jQ , (VA)

- apparent power, S V I , (VA)
- power factor, PF cos
The average power delivered to the resistive component of a load is

nonzero. The average power delivered to the reactive component of a load
is zero.
Capacitors are commonly used to improve the PF of industrial loads to

minimize the reactive power required from the electricity utility.
References
Hill, 1984.

7A.34
Exercises
1.
4 4 4
50 0 V -j4 -j4 -j4 V2
(a) Find V2 .
(b) To what identical value should each of the 4 resistors be changed so

that V2 is 180 out of phase with the source voltage?
2.
400 F
1 1
4
10cos (10 t +30) A
1 400 H i3 2 v3
(a) Find v3 t in the steady-state by using nodal analysis.
(b) Find i3 t in the steady-state by using mesh analysis

7A.35
3.
2 F 20 4 mH
i2
v1 1 mH v2 3v1
10cos (104 t ) A
(a) Find i2 t .
(b) Change the control voltage from 3v1 to 3v2 and again find i2 t .
4.
2A
1
4
F
vR 2 vL 2 H
6
1cos (10 t ) A
(a) Find vR t .
(b) Find vL t .

7A.36
5.
Find the output impedance of the network shown below:
12 j 20
1
Vx
0.1Vx -j 3
6.
Use superposition to find I x in the circuit below:
1 0 A
1 j1
Ix 1 0 V
1 0 V
1
-j1
1 0 A
7.
For the circuit below, find the phase angles of Vx and Vy graphically if
Vx 90 V and Vy 150 V .
Vx Vy
120 0 V 50 80 V

7A.37
8.
If vs 20 cos1000t 30 V in the circuit below, find the power being
absorbed by each passive element at t 0 .
2.5 k
R1
vs 1H 1 F R2 10 k
9.
Determine the average power delivered to each resistor in the network shown
below if: (a) k 0 (b) k 1
I1 20 j60
10
1000 0 V k I1
-j50
10.
The series combination of a 1000 resistor and a 2 H inductor must not
absorb more than 100 mW of power at any instant. Assuming a sinusoidal
current with 400 rads-1 , what is the largest RMS current that can be
tolerated?

7A.38
11.
Consider the following circuit:
0.1 H
100 0 V RMS R 0.3 H
= 1000 rads-1
(a) What value of R will cause the RMS voltages across the inductors to be
equal?
(b) What is the value of that RMS voltage?
12.
A composite load consists of three loads connected in parallel.
One draws 100 W at 0.9 lagging PF, another takes 200 W at 0.8 lagging PF,
and the third requires 150 W at unity PF. The composite load is supplied by a
source Vs in series with a 10 resistor. If the loads are all to operate at 110 V
RMS, determine:
(a) the RMS current through the source
(b) the PF of the composite load

7A.39
13.
A load operating at 2300 V RMS draws 25 A RMS at a power factor of
0.815 lagging. Find:
(a) the real power taken by the load
(b) the reactive power
(c) the complex power
(d) the apparent power drawn by the load
(e) the impedance of the load
14.
In the circuit shown below, load A receives S A 80 j 40 VA , while load B
absorbs S B 100 j 200 VA . Find the complex power supplied by each
source.
100 0 V RMS B 2 0 A RMS

7A.40
Oliver Heaviside (1850-1925)
The mid-Victorian age was a time when the divide between the rich and the
poor was immense (and almost insurmountable), a time of unimaginable
disease and lack of sanitation, a time of steam engines belching forth a steady
rain of coal dust, a time of horses clattering along cobblestoned streets, a time
when social services were the fantasy of utopian dreamers. It was into this
smelly, noisy, unhealthy and class-conscious world that Oliver Heaviside was
born the son of a poor man on 18 May, 1850.
A lucky marriage made Charles Wheatstone (of Wheatstone Bridge fame)

Heavisides uncle. This enabled Heaviside to be reasonably well educated, and
at the age of sixteen he obtained his first (and last) job as a telegraph operator
with the Danish-Norwegian-English Telegraph Company. It was during this
job that he developed an interest in the physical operation of the telegraph
cable. At that time, telegraph cable theory was in a state left by Professor
William Thomson (later Lord Kelvin) a diffusion theory modelling the
passage of electricity through a cable with the same mathematics that describes
I remember my first heat flow.
look at the great
treatise of
MaxwellsI saw By the early 1870s, Heaviside was contributing technical papers to various
that it was great, publications he had taught himself calculus, differential equations, solid
greater and
greatest, with geometry and partial differential equations. But the greatest impact on
prodigious
possibilities in its Heaviside was Maxwells treatise on electricity and magnetism Heaviside
power. Oliver
Heaviside was swept up by its power.
In 1874 Heaviside resigned from his job as telegraph operator and went back to
live with his parents. He was to live off his parents, and other relatives, for the
rest of his life. He dedicated his life to writing technical papers on telegraphy
and electrical theory much of his work forms the basis of modern circuit
theory and field theory.
In 1876 he published a paper entitled On the extra current which made it clear
that Heaviside (a 26-year-old unemployed nobody) was a brilliant talent. He
had extended the mathematical understanding of telegraphy far beyond

7A.41
Thomsons submarine cable theory. It showed that inductance was needed to
permit finite-velocity wave propagation, and would be the key to solving the
problems of long distance telephony. Unfortunately, although Heavisides
paper was correct, it was also unreadable by all except a few this was a trait
of Heaviside that would last all his life, and led to his eventual isolation from
the academic world. In 1878, he wrote a paper On electromagnets, etc.
which introduced the expressions for the AC impedances of resistors,
capacitors and inductors. In 1879, his paper On the theory of faults showed that
by faulting a long telegraph line with an inductance, it would actually
improve the signalling rate of the line thus was born the idea of inductive
loading, which allowed transcontinental telegraphy and long-distance
telephony to be developed in the USA.
When Maxwell died in 1879 he left his electromagnetic theory as twenty Now all has been
blended into one
equations in twenty variables! It was Heaviside (and independently, Hertz) theory, the main
who recast the equations in modern form, using a symmetrical vector calculus equations of which
can be written on a
notation (also championed by Josiah Willard Gibbs (1839-1903)). From these page of a pocket
notebook. That we
equations, he was able to solve an enormous amount of problems involving have got so far is
due in the first place
field theory, as well as contributing to the ideas behind field theory, such as to Maxwell, and next
to him to Heaviside
energy being carried by fields, and not electric charges. and Hertz. H.A.
Lorentz
A major portion of Heavisides work was devoted to operational calculus.
This caused a controversy with the mathematicians of the day because although
it seemed to solve physical problems, its mathematical rigor was not at all Rigorous
clear. His knowledge of the physics of problems guided him correctly in many mathematics is
narrow, physical
instances to the development of suitable mathematical processes. In 1887 mathematics bold
and broad. Oliver
Heaviside introduced the concept of a resistance operator, which in modern Heaviside
terms would be called impedance, and Heaviside introduced the symbol Z for
it. He let p be equal to time-differentiation, and thus the resistance operator for
an inductor would be written as pL. He would then treat p just like an algebraic
quantity, and solve for voltage and current in terms of a power series in p. In
other words, Heavisides operators allowed the reduction of the differential
equations of a physical system to equivalent algebraic equations.

7A.42
Paul Dirac derived Heaviside was fond of using the unit-step as an input to electrical circuits,
the modern notion of
the impulse, when especially since it was a very practical matter to send such pulses down a
he used it in 1927,
at age 25, in a paper telegraph line. The unit-step was even called the Heaviside step, and given the
symbol H t , but Heaviside simply used the notation 1. He was tantalizingly
on quantum
mechanics. He did
his undergraduate
work in electrical
close to discovering the impulse by stating p 1 means a function of t
engineering and
was both familiar
which is wholly concentrated at the moment t 0 , of total amount 1. It is an
with all of impulsive function, so to speak[it] involves only ordinary ideas of
Heavisides work
and a great admirer differentiation and integration pushed to their limit.
of his.
Heaviside also played a role in the debate raging at the end of the 19th century
about the age of the Earth, with obvious implications for Darwins theory of
evolution. In 1862 Thomson wrote his famous paper On the secular cooling of
the Earth, in which he imagined the Earth to be a uniformly heated ball of
The practice of
eliminating the molten rock, modelled as a semi-infinite mass. Based on experimentally
physics by reducing
a problem to a derived thermal conductivity of rock, sand and sandstone, he then
purely mathematical
exercise should be
mathematically allowed the globe to cool according to the physical law of
avoided as much as thermodynamics embedded in Fouriers famous partial differential equation for
possible. The
physics should be heat flow. The resulting age of the Earth (100 million years) fell short of that
carried on right
through, to give life needed by Darwins theory, and also went against geologic and palaeontologic
and reality to the
problem, and to evidence. John Perry (a professor of mechanical engineering) redid Thomsons
obtain the great
assistance which
analysis using discontinuous diffusivity, and arrived at approximate results that
the physics gives to could (based on the conductivity and specific heat of marble and quartz) put
the mathematics.
Oliver Heaviside, the age of the Earth into the billions of years. But Heaviside, using his
Collected Works,
Vol II, p.4 operational calculus, was able to solve the diffusion equation for a finite
spherical Earth. We now know that such a simple model is based on faulty
premises radioactive decay within the Earth maintains the thermal gradient
without a continual cooling of the planet. But the power of Heavisides
methods to solve remarkably complex problems became readily apparent.
Throughout his career, Heaviside released 3 volumes of work entitled

Electromagnetic Theory, which was really just a collection of his papers.
Heaviside shunned all honours, brushing aside his honorary doctorate from the
University of Gttingen and even refusing to accept the medal associated with
his election as a Fellow of the Royal Society, in 1891.

7A.43
In 1902, Heaviside wrote an article for the Encyclopedia Britannica entitled
The theory of electric telegraphy. Apart from developing the wave propagation
theory of telegraphy, he extended his essay to include wireless telegraphy,
and explained how the remarkable success of Marconi transmitting from
Ireland to Newfoundland might be due to the presence of a permanently
conducting upper layer in the atmosphere. This supposed layer was referred to
as the Heaviside layer, which was directly detected by Edward Appleton and
M.A.F. Barnett in the mid-1920s. Today we merely call it the ionosphere.
Heaviside spent much of his life being bitter at those who didnt recognise his
genius he had disdain for those that could not accept his mathematics without
formal proof, and he felt betrayed and cheated by the scientific community
who often ignored his results or used them later without recognising his prior
work. It was with much bitterness that he eventually retired and lived out the
rest of his life in Torquay on a government pension. He withdrew from public
and private life, and was taunted by insolently rude imbeciles. Objects were
thrown at his windows and doors and numerous practical tricks were played on
him.
Today, the historical obscurity of Heavisides work is evident in the fact that Heaviside should be
remembered for his
his vector analysis and vector formulation of Maxwells theory have become vectors, his field
theory analyses, his
basic knowledge. His operational calculus was made obsolete with the 1937 brilliant discovery of
publication of a book by the German mathematician Gustav Doetsch it the distortionless
circuit, his
showed how, with the Laplace transform, Heavisides operators could be pioneering applied
mathematics, and
replaced with a systematic, routine method. for his wit and
humor. P.J. Nahin
The last five years of Heavisides life, with both hearing and sight failing, were
years of great privation and mystery. He died on 3rd February, 1925.
References
Nahin, P.: Oliver Heaviside: Sage in Solitude, IEEE Press, 1988.

8B.1
Lecture 8B First-Order Op-Amp Filters

Introduction. Bilinear frequency responses. Frequency response
representation. Magnitude responses. Phase responses. Bode plots. Frequency
and magnitude scaling. Cascading circuits. Inverting bilinear op-amp circuit.
Inverting op-amp circuits. Cascade design.
Introduction
Filters are essential to electrical engineering. They are used in all modern
electronic systems. In communications, filters are essential for the generation
Filters are essential
and detection of analog and digital signals, whether via cable, optic fibre, air or to all modern
electronic systems
satellite. In instrumentation, filters are essential in cleaning up noisy signals,
or to recover some special part of a complicated signal. In control, feedback
through a filter is used to achieve a desired response. In power, filters are used
to inject high frequency signals on the power line for control purposes, or for
removing harmonic components of a current. In machines, filters are used to
suppress the generation of harmonics, or for controlling switching transients.
The design of filters is therefore a useful skill to possess.
Filters can be of two types: analog and digital. In this subject, we will
concentrate on analog filters. There are two reasons for this: analog filters are
necessary components in digital systems, and analog filter theory serves as a
precursor to digital filter design. The analog filters we will be looking at will
also be of two types: passive and active. Active filters represent the most
common, and use electronic components (such as op-amps) for their
implementation. This is opposed to passive filters, which use the ordinary
circuit elements: resistors, capacitors, inductors.
Filter has the commonly accepted meaning of something retained, something

rejected. For us, a filter is very simple: it is an electric circuit designed to A filter is a circuit
that implements a
implement a specific frequency response. Given a filter, obtaining the specific frequency
frequency response is just a matter of applying circuit theory. This is analysis. response
The choice of a frequency response and the choice of an implementation for a
filter, however, are never unique. This is called design.

8B.2
Bilinear Frequency Responses

Filter design and analysis is predominantly carried out in the frequency-
domain. The circuits we design and analyse will be assumed to be operating
with sinusoidal sources and be in the steady-state. This means we can use
phasors and complex numbers to describe a circuits response.
We define a circuits frequency response as the ratio of the output voltage

phasor to the input voltage phasor, as a function of frequency:
T j
Frequency response Vo
defined (8B.1)
Vi
Example
Consider a simple RC circuit:
1
j C
Vi Vo
Figure 8B.1
The frequency response is:
Vo 1 j C
T j
1 1 1
, 0
Vi R 1 jC 1 jRC 1 j 0 RC
This particular frequency response is one member of a larger class of frequency

response functions.

8B.3
The variable of the frequency response, T j , is j . A first-order

polynomial in this variable would be written as P j ja1 a 2 since the
power of j is one. A first-order polynomial in real numbers could be written
as y mx b , which is the equation of a straight line. Thus a first-order
polynomial such as ja1 a 2 is also known as a linear polynomial. When a
frequency response is the quotient of two linear polynomials it is said to be
bilinear. Thus a bilinear frequency response is of the form:
ja1 a2
T j
Bilinear frequency
response defined
jb1 b2 (8B.2)
where the a and b constants are real numbers. If Eq. (8B.2) is written in a
standard form:
a2 1 ja1 a2 1 j z
T j
and rewritten in
K terms of poles and
b2 1 jb1 b2 1 j p (8B.3) zeros
then -z is the zero of T j and -p is the pole of T j (the reason for

these names will become apparent later). For real circuits, p will always be a
positive real number, while z may be either a positive or negative real number.
The constant K depends on the circuit used to create the bilinear frequency
response for active circuits (e.g. those with op-amps) it can be positive or
negative.
Thus for the RC circuit in the example, if we were to write the frequency
response in the standard form, we would identify:
K 1
(8B.4)
z
p 0 1 RC

8B.4
Frequency Response Representation

The complex function T j can also be written using a complex exponential
in terms of magnitude and phase:
T j T j e j (8B.5a)
which is normally written in polar coordinates:

The frequency
T j T j
response in terms of (8B.7b)
magnitude and
phase
We also normally plot the magnitude and phase of T j as a function of

or f . We use both linear and logarithmic scales.
If the logarithm (base 10) of the magnitude is multiplied by 20, then we have
the gain of the frequency response in decibels (dB):
The magnitude of
the frequency
response in dB
A 20 log T j dB (8B.6)
A negative gain in decibels is referred to as attenuation. For example, -3 dB

gain is the same as 3 dB attenuation.
The phase function is usually plotted in degrees.

8B.5
Example
For the RC circuit, let 0 1 RC so that the frequency response can be written
as:
T j
1
1 j 0
The magnitude function is found directly as:
T j
1
1 0
2
The phase is:

tan 1
0
These are graphed below, using a normalised log scale for :

8B.6
Magnitude Responses
A magnitude response is the magnitude of the frequency response, plotted
The magnitude against the frequency of the input. Magnitude responses can be classified
response is the
magnitude of the
according to their particular properties. To look at these properties, we will use
frequency response linear magnitude versus linear frequency plots. For the RC circuit used in the
examples, the magnitude function T j 1 1 j 0 has three frequencies
of special interest corresponding to these values of T j :
T j 0 1
T j0
1
0.707
2
T j 0 (8B.7)
The frequency 0 is known as the half-power frequency. The plot below

shows the complete magnitude response of T j as a function of , and the
circuit that produces it:
A simple lowpass
filter
|T|
1
R
1 2
vi C vo
0
0 0 2 0
Figure 8B.2

8B.7
An idealisation of the response in Figure 8B.2, known as a brick wall, and the
circuit that produces it are shown below:
An ideal lowpass
filter
|T| Cutoff
1
ideal
vi vo
Pass Stop filter
0
0 0
Figure 8B.3
For the ideal filter, the output voltage remains fixed in amplitude until a critical
frequency is reached, called the cutoff frequency, 0 . At that frequency, and
for all higher frequencies, the output is zero. The range of frequencies with
output is called the passband; the range with no output is called the stopband. Pass and stop
bands defined
The obvious classification of the filter is a lowpass filter.
Even though the response shown in the plot of Figure 8B.2 differs from the Half-power
frequency and cutoff
ideal, it is still known as a lowpass filter, and, by convention, the half-power frequency are the
same thing for a real
frequency is taken as the cutoff frequency.
filter

8B.8
If the positions of the resistor and capacitor in the circuit of Figure 8B.2 are
interchanged, then the resulting circuit is:
vi R vo
Figure 8B.4
Show that the frequency response is:
j
T j
1 RC j (8B.8)
Letting 0 1 RC again, we can write it in the standard form:
j 0
T j
1 j 0 (8B.9)
The magnitude function of this equation, at the three frequencies given in Eq.
(8B.7), is:
T j 0 0
T j0
1
0.707
2
T j 1 (8B.10)

8B.9
The plot below shows the complete magnitude response of T j as a function

of , and the circuit that produces it:
A simple highpass
filter
|T|
1
C
1 2
vi R vo
0
0 0 2 0 3 0
Figure 8B.5
This filter is classified as a highpass filter. The ideal brick wall highpass filter
is shown below:
An ideal highpass
filter
|T| Cutoff
1
ideal
vi vo
Stop Pass filter
0
0 0
Figure 8B.6
The cutoff frequency is 0 , as it was for the lowpass filter.

8B.10
Phase Responses
Like magnitude responses, phase responses are only meaningful when we look
at sinusoidal steady-state circuits. From Eq. (8B.1), a frequency response can
be expressed in polar form as:
Phase response is
Vo Vo
T j
obtained in the
sinusoidal steady T
state
Vi Vi 0 (8B.11)
where the input is taken as reference (zero phase).
For the bilinear frequency response:
1 j z
T j K
1 j p (8B.12)
the phase is:
The phase of the

bilinear frequency K tan 1 tan 1
response z p (8B.13)
If K is positive, its phase is 0 , and if negative it is 180 . Assuming that it is

positive, we have:

tan 1 tan 1
z p (8B.14)
z p
Lead and lag circuits We use the sign of this phase angle to classify circuits. Those giving positive
defined
are known as lead circuits, those giving negative as lag circuits.

8B.11
Example
For the simple RC circuit of Figure 8B.2, we have already seen that:
tan 1 0
Since is negative for all , the circuit is a lag circuit. When 0 ,
tan 1 1 45 . A complete plot of the phase response is shown below:
0 0 2 0
0
Lagging phase
response for a
simple RC lowpass
-45 filter
-90
Example
For the simple RC circuit of Figure 8B.4, we can show that the phase is given
by:
90 tan 1 0
The phase response is shown below:
90
Leading phase
response for a
45 simple RC highpass
filter
0
0 0 2 0
The angle is positive for all , and so the circuit is a lead circuit.

8B.12
Example
Consider the RL circuit shown below:
RL
Vi Vo
jL
Using the voltage divider rule, we obtain:
R L j L
T j
R R L jL
If we write this in the standard form:
1 j z
T j K
1 j p
then:
RL R R RL
K , z L and p
R RL L L
so that p z . We can see that:
RL
T j 0 K
R RL
T j K
1z
1
1 p

8B.13
The magnitude response is:
|T( j) |
1
RL
R+ R L
0
log 10
z p
This does not approximate the ideal of Figure 8B.6 very well, but it is still
known as a highpass filter.
From (8B.14) we see that is characterized by the difference of two angles,

the first a function of the zero numerator term, the second a function of the pole
denominator term: z p . For p z the phase function z reaches
45 at a low frequency, while p reaches 45 for a higher frequency.
Therefore, the phase response looks like:

90
z
45
z + p
z
0 p log 10
p
-45
-90
Thus, the circuit provides phase lead.

8B.14
Summary of Bilinear Frequency Responses
We can summarize the magnitude and phase responses of T j for various

values of z and p , where K is assumed to be positive, in the table below:
T j Magnitude Response Phase Response

|T |
1 90
K j 0 K
0 log 10
log 10
0 0
|T |
2 log 10
0
K
j 0 K
0
log 10 -90
0
|T |
3 90
45
K 1 j z
2K
K
0 log 10
log 10 0
z z
|T | p
4 K log 10
0
K/ 2
1 -45
K
1 j p 0
log 10 -90
p
|T |
5 K 90
j p K/ 2
45
K
1 j p 0
log 10 log 10
p 0 p
|T | p z
6 log 10
0
1 j z
K
1 j p 0
log 10 90
z p
p z
|T |
7 90
1 j z
K
1 j p 0
log 10 log 10
0 z p
pz
z p
Table 8B.1

8B.15
Bode* Plots
If we plot the magnitude A as in Eq. (8B.6) as a function of with logarithmic
coordinates, then the plot is known as a magnitude Bode plot. If we plot as a Bode plots defined
function of with logarithmic coordinates it is known as a phase Bode plot.
The advantages of Bode plots in the design of filters are:
1. Using a logarithmic scale for the magnitude of T makes it possible to add

and subtract rather than multiply and divide. This applies to phase also.
2. The slope of all asymptotic lines in magnitude plots for bilinear functions is
0 or 20 dB per decade. The advantages of
using Bode plots
3. Asymptotic plots serve to sketch out ideas in design a kind of shorthand.
4. Because decades occur at equal linear distances, the shape of a Bode

magnitude plot is maintained when frequency is scaled (see the next
section).
The following figure shows how simple it is to design a particular response:
Composing a Bode
plot from first-order
A1 A2 A3 A
parts
2 3
1 + + = 1 2 3
A
Scale
1 2 3 '1 '2 '3

shape maintained
Figure 8B.7
*
Dr. Hendrik Bode grew up in Urbana, Illinois, USA, where his name is pronounced boh-dee.
Purists insist on the original Dutch boh-dah. No one uses bohd.

8B.16
Frequency and Magnitude Scaling

In filter design, it is common practice to normalise equations so that they have
the same form. For example, we have seen the bilinear frequency response:
T j
1 RC
1 RC j (8B.15)
expressed as:
T j
1
1 j 0 (8B.16)
Normalising We normalise the equation by setting 0 1 :

frequency means
setting it to 1
T j
1
1 j (8B.17)
Every equation in filter design will be normalised so that 0 1 . This is

helpful, since every equation will be able to be compared on the same base.
The difficulty that now arises is denormalising the resulting equations, values
or circuit designs.

8B.17
Frequency Scaling (Denormalising)
Frequency scaling, or denormalising, means we want to change 0 from 1

back to its original value. To do this, we must change all frequency dependent and denormalising
means setting the
terms in the frequency response, which also means frequency dependent frequency back to its
elements in a circuit. Furthermore, the frequency scaling should not affect the original value
magnitude of any impedance in the frequency response.
To scale the frequency by a factor k f , for a capacitor, we must have:
1 1 1
ZC
C k f 1 k f C k f Cnew (8B.18)
We must decrease the capacitance by the amount 1 k f , while increasing the Frequency scaling
must keep the
frequency by the amount k f if the magnitude of the impedance is to remain the magnitude of the
impedance the
same. same
For an inductor, we must have:
Z L L k f L k f Lnew
1
kf (8B.19)
Therefore, new element values may be expressed in terms of old values as

follows:
Rnew Rold (8B.20a)
1
Lnew Lold (8B.24b)
kf
The frequency
scaling equations
1
Cnew Cold (8B.24c)
kf

8B.18
Magnitude Scaling
Since a frequency response is always a ratio, if we increase the impedances in

Magnitude scale to
get realistic element the numerator and denominator by the same amount, it changes nothing. We do
values
this to obtain realistic values for the circuit elements in the implementation.
If the impedance magnitudes are normally:
1
Z R R, Z L L, ZC (8B.21)
C
then after magnitude scaling with a constant k m they will be:
k m Z R k m R,
km Z L kmL,
1
km ZC
C km (8B.22)
Therefore, new element values may be expressed in terms of old values as

follows:
Rnew k m Rold (8B.23a)
The magnitude Lnew km Lold

scaling equations
(8B.27b)
1
Cnew Cold
km
(8B.27c)

8B.19
Cascading Circuits
How can we create circuits with higher than first-order frequency responses by
cascading first-order circuits? Consider the following circuit:
R R
vi C C vo
Figure 8B.8
Show that the frequency reponse for the above circuit is:
Vo

1 RC 2
Vi 1 RC 2 2 j 3 RC (8B.24)
Compare with the following circuit:
R R
vi C C vo
Figure 8B.9

8B.20
which has the frequency response:
Vo 1 RC 1 RC

Vi 1 RC j 1 RC j

1 RC 2
1 RC 2 2 j 2 RC (8B.25)
Cascading buffered Therefore, when we cascade circuits, if the output of each circuit presents a
circuits is highly low impedance to the next stage, so that each successive circuit does not load
desitable from a
design perspective the previous circuit, then we can simply multiply the individual frequency
responses of the individual circuits to achieve an overall frequency response.
Op-amp circuits of both the inverting and non-inverting type are ideal for
cascading.
Inverting Bilinear Op-Amp Circuit

We seek an implementation of the bilinear frequency response that can be
cascaded. The frequency response of the inverting op-amp circuit is:
T j
Z2
Z1 (8B.26)
Therefore, we would like to have:
The inverting op-

amp circuit is one Z2 1 j z
way to implement a T K
bilinear frequency
response
Z1 1 j p (8B.27)
The specifications of the design problem are the values K, z and p. These may
be found from a Bode plot the break frequencies and the gain at some
frequency or obtained in any other way. The solution to the design problem
involves finding a circuit and the values of the elements in that circuit. Since
we are using an active device the op-amp inductors are excluded from our
circuits. Therefore, we want to find values for the Rs and the Cs. Once found,

8B.21
these values can be adjusted by any necessary frequency scaling, and then by
magnitude scaling to obtain convenient element values.
For the general bilinear frequency response, we can make the following
identification:
1 j z K 1 j p Z 2
K
1 j p 1 1 j z
(8B.28)
Z1
Therefore:
1 1
Z2 The impedances for
1 K j pK 1 R2 jC2 (8B.29a) the inverting op-amp
circuit to implement
1 1 a bilinear frequency
Z1 response
1 j z 1 R1 jC1 (8B.33b)
The design equations become:
Element values for

1 1 the inverting op-amp
R1 1, C1 , R2 K , C2 circuit in terms of
the pole, zero and
z pK (8B.30) gain
One realisation of the bilinear frequency response is then:
An inverting op-amp
K circuit that
implements a
bilinear frequency
response
1
1 pK
vi
vo
1 z
Figure 8B.10

8B.22
Inverting Op-Amp Circuits

The approach we took in obtaining a circuit to implement the bilinear
frequency response can be applied to other forms of T j to give the entries
in the table below:
Frequency Response Circuit

T j
1 K
1
K 1 j z vi
vo
1 z
2 K
1
1
K 1 pK
1 j p vi
vo
3 K
1 p
j p 1
K vi
1 j p vo
4 K
1
1 j z 1
K pK
1 j p vi
vo
1 z
5 K p2
j p1 p 2 1 p1
1 K
K 1
1 j p1 1 j p 2 vi
vo
Table 8B.2

8B.23
The last entry in the table illustrates an important point. If we use a series RC
connection for Z 1 but a parallel RC connection for Z 2 , then the frequency
response becomes one of second-order. So the manner in which the capacitors
are connected in the circuit determines the order of the circuit.
Cascade Design
We can make use of cascaded modules, each of first-order, to satisfy
specifications that are more complicated than the bilinear function.
Example
The asymptotic Bode plot shown below is for a band-enhancement filter:
A , dB 20 dB
A band-
0 dB enhancement filter
10
2
10
3
10
4
10
5
rad/s (log scale)
We wish to provide additional gain over a narrow band of frequencies, leaving

the gain at higher and lower frequencies unchanged. We wish to design a filter
to these specifications and the additional requirement that all capacitors have
the value C 10 nF .

8B.24
The composite plot may be decomposed into four first-order factors as shown
below:
A , dB 20 dB
0 dB
10 rad/s (log scale)
2 3 4 5
10 10 10
Decomposing a
A , dB
Bode plot into first- 1 4
order factors
rad/s (log scale)
2 3
Those marked 1 and 4 represent zero factors, while those marked 2 and 3 are
pole factors.
From the break frequencies given, we have:
T j
14 1 j 102 1 j 105
23 1 j 103 1 j 104
We next write T j as a product of bilinear functions. The choice is arbitrary,

but one of the two possibilities is:
The frequency T j T1 j T2 j
response as a 1 j 102 1 j 105
cascade of bilinear
functions 1 j 103 1 j 104

8B.25
For a circuit realisation of T1 and T2 we decide to use the inverting op-amp

circuit of Figure 8B.10. Using the formulas for element values given there, we
obtain the realisation shown below:
-2 -3 -5 -4
10 10 10 10
vi vo A realisation of the
specifications
1 1 1 1
Frequency scaling is not required since we have worked directly with specified
frequencies (we did not normalize the initial Bode plot). The magnitude scaling
of the circuit is accomplished with the equations:
Magnitude scaling is
1
Cnew Cold and Rnew km Rold required to get
km realistic element
values
Since the capacitors are to have the value 10 nF, this means km 108 . The
element values that result are shown below and the design is complete:
1 M 100 k 1 k 10 k
vi vo
A realistic
10 nF 10 nF 10 nF 10 nF implementation of
the specifications
References
Van Valkenburg, M. E.: Analog Filter Design, Holt-Saunders, Tokyo, 1982.

8B.26
Exercises
1.
For the circuit shown below, prepare the asymptotic Bode plot for the
magnitude of T j . Carefully identify all slopes and low and high frequency
asymptotes.
10 k 10 k
10 k 1 nF 10 k 10 pF
vi
vo
2.
Design an RC op-amp filter to realise the bandpass response shown below.
A, dB
20 dB
+20 dB/decade -20 dB/decade
0 dB
10 10 4 rad/s (log scale)
Use a minimum number of op-amps in your design, and scale so that the
elements are in a practical range.

8B.27
3.
The asymptotic Bode plot shown below represents a lowpass filter-amplifier
with a break frequency of 0 1000 rads-1 .
A, dB
20 dB
-40 dB/decade
1000 rad/s
Design a circuit to be connected in cascade with the amplifier such that the
break frequency is extended to 0 5000 rads-1 :
A, dB
20 dB
-40 dB/decade
5000 rad/s

9A.1
Lecture 9A Second-Order Step Response

Solution of the homogeneous linear differential equation. Source-free parallel
RLC circuit. Overdamped parallel RLC circuit. Critically damped parallel
RLC circuit damping. Underdamped parallel RLC circuit. Source-free series
RLC circuit. Complete response of the RLC circuit.
Introduction
Circuits that contain resistors, inductors and capacitors are known as RLC
circuits. We will examine the simplest RLC circuits the parallel RLC circuit
where the resistor, inductor and capacitor are in parallel, and the series RLC
circuit in which they are connected in series.
Many circuits can be reduced to one of these equivalent circuits. For example,
transistor amplifiers in analog radios, crystal oscillators driving the clocks of
digital circuits and microprocessors, and power system power factor
correction circuits, can each be reduced to a parallel RLC circuit.
The inclusion of both inductance and capacitance in a circuit leads to at least a

second-order system that is, one that is characterized by a linear differential
equation including a second-order derivative, or by a set of two first-order
differential equations. We shall see that such systems are much more
complicated than first-order systems there will be two arbitrary constants to
find, we will need initial conditions for variables and their derivatives, and
there are three different functional forms of the natural response that depend
upon element values.
However, there is a side benefit to studying second-order systems: we will see

a unifying underlying structure emerge for all linear circuits, regardless of their
complexity. We will also discover a way to exploit this underlying structure.

9A.2
Solution of the Homogeneous Linear Differential Equation

We have seen that any linear homogeneous differential equation with constant
coefficients of order n is an equation that can be written:
dny d n 1 y dy
an n an 1 n 1 a1 a0 y 0
dt dt dt (9A.1)
or more simply using the D operator as:
f D y 0 (9A.2)
We have seen that the solution to such an equation was found by using the
property that:
f D e st f s e st (9A.3)
then letting:
f s 0 (9A.4)
which leads to y e st as a solution. To find the general solution, we find the n

roots of the nth-order characteristic equation f s 0 , and then form a linear
sum of each individual solution.

9A.3
Distinct Real Roots
The characteristic equation is of degree n. Let its roots be s1 , s2 , , sn . If
these roots are all real and distinct, then the n solutions:
(9A.5)
y1 e s1t , y2 e s2t , , yn e snt
are linearly independent and the general solution can be written at once. It is:
(9A.6)
y c1e s1t c2e s2t cne snt
in which c1 , c2 , , cn are arbitrary constants.
Example
Solve the equation:
d2y dy
2
7 6y 0
dt dt
First write the characteristic equation:
s 2 7s 6 0
s 1s 6 0
whose roots are s 1, 6 . Then the general solution is seen to be:
y c1e t c2e 6t

9A.4
Repeated Real Roots
Suppose that f s 0 has repeated roots. Then Eq. (9A.6) does not yield the
general solution. To see this, let the characteristic equation have three equal
roots s1 b , s2 b and s3 b . The corresponding part of the solution yielded
by Eq. (9A.6) is:
y c1ebt c2ebt c3ebt

c1 c2 c3 ebt
c4ebt (9A.7)
Thus, corresponding to the three roots under consideration, this method yields
only one solution.
What is needed is a method for obtaining m linearly independent solutions

corresponding to m equal roots of the characteristic equation. Suppose that the
characteristic equation f s 0 has the m equal roots:
s1 s2 sm b (9A.8)
Then the operator f D must have a factor D b . We wish to find m

m
linearly independent ys for which:
D bm y 0 (9A.9)
Turning to Eq. (4A.16) in Lecture 4A, and writing s b , we find that:
D bm t k ebt 0, k 0, 1, ..., m 1
(9A.10)
The functions yk t k ebt where k 0, 1, ..., m 1 are linearly independent
because, aside from the common factor ebt , they contain only the respective
powers t 0 , t 1 , t 2 , ..., t m1 .

9A.5
The general solution of Eq. (9A.9) is therefore:
(9A.11)
y c1ebt c2tebt cmt m1ebt
Furthermore, if f D contains the factor D b m , then the equation
f D y 0 can be written g D D b y 0 where g D contains all the

m
factors of f D except D b . Then any solution of D b y 0 is also a

m m
solution of g D D b y 0 and therefore of f D y 0 .

m
Only Real Roots
Now we are in a position to write the solution of f D y 0 whenever the

characteristic equation has only real roots. Each root of the characteristic
equation is either distinct from all the other roots or it is one of a set of equal
roots.
Corresponding to a root si distinct from all others, there is the solution:
(9A.12)
yi ci e sit
and corresponding to m equal roots s1 , s2 , , sm , each equal to b, there are

solutions:
(9A.13)
c1ebt , c2tebt , , cmt m1ebt

9A.6
Example
Solve the equation:
D 4
7 D 3 18D 2 20 D 8y 0
The characteristic equation:
s 4 7 s 3 18s 2 20s 8 0
has the roots s 1, 2, 2, 2 . Then the general solution is:
y c1e t c2e 2t c3te 2t c4t 2e 2t
or:

y c1e t c2 c3t c4t 2 e 2t
Example
Solve the equation:
d4y d3y d2y

2 0
dt 4 dt 3 dt 2
s 4 2s 3 s 2 0
with roots s 0, 0, 1, 1 . Hence the desired solution is:
y c1 c2t c3e t c4te t

9A.7
Distinct Complex Roots
In all real systems, the describing differential equation f D y 0 has a

characteristic equation f s 0 that has real coefficients. We know from
algebra that if the characteristic equation has any complex roots then those
roots must occur in conjugate pairs. Thus, if:
s1 a jb (9A.14)
is a root of the equation f s 0 with a and b real and b 0 , then:
s2 a jb (9A.15)
is also a root of f s 0 . According to the preceding section, f D y 0 is

satisfied by:
y c1ea jb t c2ea jb t (9A.16)
where for y to be real, we must have c1 c2* . Since t is real along with a and b,
Eulers identity gives us the result:
y c1e at cos bt j sin bt c2e at cos bt j sin bt (9A.17)
This may be written:
y c1 c2 e at cos bt j c1 c2 e at sin bt (9A.18)
Finally, let c1 c2 2 Rec1 c3 , and j c1 c2 2 Imc1 c4 , where c3 and
c4 are real arbitrary constants. Then f D y 0 is seen to have the solution:
(9A.19)
y c3e at cos bt c4e at sin bt
corresponding to the two roots s1 a jb and s2 a jb ( b 0 ) of the

characteristic equation.

9A.8
Example
Solve the equation:
D 3
5 D 2 17 D 13y 0
For the characteristic equation:
s 3 5s 2 17 s 13 0
one root, s1 1 , is easily found. When the factor s 1 is removed, it is seen

that the other two roots are solutions of the quadratic equation:
s 2 4 s 13 0
Those roots are found to be s2 2 j 3 and s3 2 j 3 . The characteristic
equation has the roots s 1, 2 j 3 . Hence the general solution of the

differential equation is:
y c1e t c2e 2t cos 3t c3e 2t sin 3t

9A.9
Repeated Complex Roots
Repeated complex roots lead to solutions analogous to those brought in by

repeated real roots. For instance, if the roots s a jb occur three times, then
the corresponding six linearly independent solutions of the differential equation
are those appearing in the expression:
c c t c t e
1 2 3
2 at

cos bt c4 c5t c6t 2 e at sin bt (9A.20)
Example
Solve the equation:
D 4
8 D 2 16y 0
The characteristic equation s 4 8s 2 16 0 may be written:
s 2
4 0
2
so its roots are seen to be s j 2, j 2 . The roots, s1 j 2 and s2 j 2
occur twice each. Thinking of j 2 as 0 j 2 and recalling that e0t 1 , we write

the solution of the differential equation as:
y c1 c 2 t cos 2t c3 c 4 t sin 2t

9A.10
The Source-Free Parallel RLC Circuit

A suitable model for portions of many practical circuits is given by the parallel
RLC circuit:
iL
R L v C
Figure 9A.1
We will assume that energy may be stored initially in both the inductor and the
capacitor, and thus nonzero values of both inductor current and capacitor
voltage are initially present. With reference to the circuit above, we may then
write the single nodal equation:
v 1 t
vdt iL t0 C
dv
0 (9A.21)
R L t0 dt
When both sides are differentiated once with respect to time and divided by C
the result is the linear second-order homogeneous differential equation:
d 2v 1 dv 1
v0 (9A.22)
dt 2 RC dt LC
We must solve this equation subject to the initial conditions:

iL 0 I 0
v0 V
(9A.23)

0

9A.11
With our theory of the D operator behind us, we now embark on solving
Eq. (9A.22). We write the characteristic equation:
1 1
f s s 2 s 0 (9A.24)
RC LC
and identify the two roots:
2
1 1 1
s1
2 RC 2 RC LC
2
1 1 1
s2 (9A.25)
2 RC 2 RC LC
These roots can be either real and distinct, real and repeated, or complex and
distinct depending on the values of R, L and C.
Case Graph of y f s Solution of f s 0
Real and distinct y Im s
Re s
s
s -plane
Real and repeated y Im s
Re s
s
s -plane
Complex and distinct y Im s
Re s
s
s -plane
Table 9A.1 Three types of roots to the 2nd-order characteristic equation

9A.12
Therefore, there will be three different forms of the natural response,
corresponding to each of the cases in Table 9A.1.
It will be helpful to make some simplifying substitutions into Eq. (9A.25) for
the sake of conceptual clarity. Let us define the undamped natural frequency:
1
0 (9A.26)
LC
We will also define the exponential damping coefficient:
1
(parallel) (9A.27)
2 RC
This description is used because is a measure of how rapidly the natural

response decays or damps out to its steady final value (usually zero).
Finally, s , s1 and s2 are called complex frequencies and will form the basis for
some of our later work.
Thus, our characteristic equation becomes:
s 2 2s 02 0
(9A.28)
and has the two roots:
s1 2 02
s2 2 02 (9A.29)
It is now apparent that the nature of the response depends upon the relative
magnitudes of and 0 . The square root appearing in the expressions for s1
and s2 will be real when 0 , zero when 0 , and imaginary when
0 . Each of these cases will be considered separately.

9A.13
The Overdamped Parallel RLC Circuit

When 0 , both s1 and s2 will be real and distinct:
s1 2 02
s2 2 02 (9A.30)
Thus, the natural response is of the form:
(9A.31)
vn K1e s1t K 2e s2t
We also know that:
2 02 (9A.32)
and therefore:

2 02 2 02 0 (9A.33)
s2 s1 0
This shows that both s1 and s2 are negative real numbers. Thus, the response
vt is the algebraic sum of two decreasing exponential terms, both of which
approach zero as time increases without limit. In fact, since the absolute value
of s2 is larger than that of s1 , the term containing s2 has the more rapid rate of
decrease.
It only remains to find the arbitrary constants K1 and K 2 using the initial
conditions, and we have the solution.

9A.14
Example
Consider the circuit:
iR iL iC
4
2 9
H v 10 mF
in which v0 0 and i L 0 8 . We may easily determine the values of the

several parameters:
25 0 15
s1 5 s 2 45
and since 0 immediately write the general form of the natural response:
vt K 1e 5t K 2 e 45t
Only evaluation of the two constants K1 and K 2 remains. Using the initial
value of vt :
v0 0
and therefore:
0 K1 K 2
A second relation between K1 and K 2 must be obtained by taking the

derivative of vt with respect to time, determining the initial condition of this
derivative through the use of the remaining initial condition i L 0 8 , and
equating the results.

9A.15
Taking the derivative of both sides of the response, we get:
dv
5K 1e 5t 45 K 2 e 45t
dt
Evaluating the derivative at t 0 :
dv
5 K 1 45 K 2
dt t 0
We next pause to consider how the initial value of the derivative can be found
numerically. This next step is always suggested by the derivative itself, dv dt
suggests capacitor current, for:
dv
iC C
dt
Thus:
dv iC 0 i L 0 i R 0 i L 0
800 Vs -1
dt t 0 C C C
We thus have our second equation:
800 5 K 1 45 K 2
and we can solve our two equations in K1 and K 2 simultaneously to get

K1 20 and K 2 20 . Thus, the final numerical solution for the natural
response is:
vt 20e 5t e 45t
We can interpret this result. We note that vt is zero at t 0 , as required. We

also interpret the first exponential term as having a time constant of 1 5 s and
the other exponential, a time constant of 1 45 s . Each starts with unity
amplitude, but the second decays more rapidly vt is thus never negative.
We thus have a response curve which is zero at t 0 , zero at t , and is

9A.16
never negative. Since it is not everywhere zero, it must have at least one
maximum, and this can be found easily.
We differentiate the response:
dv
dt

20 5e 5t 45e 45t
Then set the derivative to zero to determine the time t p at which the voltage
reaches its peak:

5t p 45t p
0 5e 45e
40 t p
e 45 5
ln 9
tp
40
54.93 ms
and obtain:
vt p 13.51 V
A reasonable sketch of the response may be obtained by plotting the two

exponential terms 20e 5t and 20e 45t , taking their difference, and noting the
peak value obtained above:
20
20 e -5 t
16
v( t ) 12
(V)
8
4
20 e -45 t
0 t (s)
0.2 0.4 0.6 0.8 1
You can see that the dominant term is 20e 5t for large t since the other term
has effectively decayed to zero.

9A.17
The Critically Damped Parallel RLC Circuit

The very special case of 0 is known as critical damping. Any one of the
three elements R, L or C may be changed to obtain critical damping however
it is usual to select R to obtain critical damping in a circuit, and thus leave 0
unchanged.
For critical damping, we have real repeated roots :
s1 0
(9A.34)
s2 0
vn K1 K 2t et (9A.35)
In a real physical system it is impossible to obtain the exact conditions

necessary for critical damping, since even an infinitesimally small change in
the circuits conditions will cause to differ from 0 . However, it can always
be a design goal to obtain a critically damped response.

9A.18
Example
We will use the same circuit as before as an example, but this time set
R 3 13 to obtain critical damping:
iR iL iC
4
33
1
9
H v 10 mF
The initial conditions are again v0 0 and i L 0 8 . In this case:
0 15
s1 s 2 15
and since 0 immediately write the general form of the natural response:
vt K 1 K 2 t e 15t
We establish the values of K1 and K 2 by first imposing the initial condition on

vt itself, v0 0 . Thus, K1 0 . This simple result occurs in this example
because the initial value of the response was selected as zero.
The second initial condition must be applied to the derivative dv dt just as in

the overdamped case. We therefore differentiate, remembering that K1 0 :
K 2 t 15e 15t K 2 e 15t

dv
dt
Evaluating the derivative at t 0 :
dv
K2
dt t 0

9A.19
We next express the derivative in terms of the initial capacitor current:
dv iC 0 i L 0 i R 0 i L 0
800 Vs -1
dt t 0 C C C
and thus:
K 2 800
The natural response is therefore:
vt 800te 15t
Before plotting the response, we try to anticipate its form by qualitative

reasoning. We note that vt is zero at t 0 , as required. It is not immediately
apparent that the response also approaches zero as t becomes infinitely large.
Using LHpitals rule:
1
lim vt 800 lim
t
15t
800 lim 0
t t e t 15e15t
and once again we have a response which begins and ends at zero and has
positive values at all other times.
A maximum value v p again occurs at time t p :
t p 1 15 s and v p 19.62 V
This maximum is larger than that obtained in the overdamped case and is a
result of the smaller losses that occur in the larger resistor. The time of the
maximum response is slightly larger than it was with overdamping.

9A.20
The natural response curve for critical damping is shown below:
v ( t ) (V)
20
15
10
0 t (s)
0.2 0.4 0.6 0.8 1.0

9A.21
The Underdamped Parallel RLC Circuit

If we further increase the resistance R from the value we had at critical
damping, whilst leaving L and C unchanged, the damping coefficient
decreases while 0 remains constant. We thus have 0 , and the roots of
the characteristic equation become:
s1 2 02 j 02 2
s2 2 02 j 02 2 (9A.36)
We now take the new square root, which is real for the underdamped case, and
call it d , the damped natural frequency:
d 02 2 (9A.37)
Thus, the roots are distinct complex conjugates and are located at:
s1 jd
(9A.38)
s2 jd
vn e t K1 cos d t K 2 sin d t (9A.39)

9A.22
Example
We will use the same circuit as before as an example, but this time set
R 5 95 , L and C are unchanged:
iR iL iC
4
59
5
9
H v 10 mF
Again we have v0 0 and i L 0 8 . In this case:
1
9
2 RC
1
0 15
LC
and since 0 we identify:
d 02 2 12
and immediately write the general form of the natural response:
vt e 9t K 1 cos12t K 2 sin 12t
The determination of the two constant proceeds as before. Again, if v0 0

then K1 0 . Hence:
vt K 2 e 9t sin 12t
The derivative is:
dv
12 K 2 e 9t cos12t 9 K 2 e 9t sin 12t
dt

9A.23
and at t 0 :
dv i L 0
12 K 2 800
dt t 0 C
The natural response is therefore:
vt 200 3 e 9t sin 12t
Notice that, as before, vt is zero at t 0 , as required. The response also has a

final value of zero because the exponential term vanishes for large values of t.
As t increases from zero through small positive values, vt increases because

the exponential term remains unable to damp the increase due to the sinusoidal
term. But at a time t p , the exponential function begins to decrease more
rapidly than sin 12t is increasing, so vt reaches a maximum v p and begins to
decrease. We should note that t p is not the value of t for which sin 12t is a
maximum, but must occur before sin 12t reaches its maximum value.
When t d 12 , vt 0 . For the interval d t 2 d , the
response is negative, becoming zero again at t 2 d .
Thus vt is an oscillatory function of time and crosses the time axis an infinite
number of times at t n d where n is any positive integer.
The oscillatory nature of the response becomes more noticeable as

decreases. If 0 , which corresponds to an infinitely large resistance, then
vt is an undamped sinusoid which oscillates with constant amplitude. We
have merely assumed an initial energy in the circuit and have not provided any
means to dissipate this energy. It is transferred from its initial location in the
inductor to the capacitor, then returns to the inductor, and so on, forever.

9A.24
Differentiation of the natural response locates the first maximum of vt ,
t p1 77.27 ms and v p1 26.60 V
the succeeding minimum,
t p 2 339.1 ms and v p 2 2.522 V
and so on.
The natural response curve for the underdamped case is shown below:
v ( t ) (V)
25
20
15
10 200 -9 t
e
3
5
0 t (s)
0.2 0.6 0.8 1.0
-5
-10 200 -9 t
- 3 e
-15
-20
Notice that the envelope of the damped sinusoid is given by Ke t .

9A.25
Response Comparison
The overdamped, critically damped and underdamped responses for the
example circuit are shown on the same graph below:
v ( t ) (V)
Underdamped iL
4
25
R 9
H v 10 mF
Critically damped
20
15 Overdamped
10
0 0.2 0.6 0.8 1.0 t (s)
The table below shows the possibilities and names associated with the
second-order natural response.
Condition Criteria Natural Response Example
0
v ( t ) (V)
Overdamped f n K 1e s1t K 2 e s2t 25
20
15
s1, 2 2 02 10
0 0.2 0.6 0.8 1.0 t (s)
0 f n K 1 K 2 t e t
v ( t ) (V)
Critically 25
damped
20
s1, 2 15
10
0 0.2 0.6 0.8 1.0 t (s)
0 f n e t K1 cos d t K 2 sin d t
v ( t ) (V)
Underdamped 25
20
s1, 2 j d 15
10
d
5
2 2
0 0 0.2 0.6 0.8 1.0 t (s)
Table 9A.2 Second-Order Natural Responses

9A.26
The Source-Free Series RLC Circuit

The series RLC circuit is the dual of the parallel RLC circuit, and this makes
the analysis fairly simple. The figure below shows the source-free series RLC
circuit:
L
i
R C vC
Figure 9A.2
KVL around the circuit gives:
di 1 t
Ri L idt vC t0 0 (9A.40)
dt C t0
When both sides are differentiated once with respect to time and divided by L
the result is the linear second-order homogeneous differential equation:
d 2i R di 1
i0 (9A.41)
dt 2 L dt LC
This is the dual of Eq. (9A.22). Thus, if we define:
R
(series) (9A.42)
2L
then we get the same characteristic equation as for the parallel RLC circuit,
Eq. (9A.28). It is now apparent that our discussion of the parallel RLC circuit is
directly applicable to the series RLC circuit.

9A.27
Complete Response of the RLC Circuit

Consider a series RLC circuit with a DC source that is connected by a switch
that closes at t 0 :
t=0 i (t ) R L
1.3 mH
Vs C v (t )
1V 1.1F
Figure 9A.3
Assuming zero initial conditions, we would like to calculate the capacitor

voltage vt at t 120 s for the following values of resistance R:
1. R 330
2. R 68.7552
3. R 33

9A.28
Forced Response
First, we find the forced response. Since we have a DC source, we can find this
part of the solution by replacing the inductance by a short circuit and the
capacitance by an open circuit, i.e. we analyse the circuit under DC conditions.
This is shown below:
if R
Vs vf
1V
Figure 9A.4
The current is zero, the voltage drop across the resistance is zero, and the
voltage across the capacitance (an open circuit) is equal to the DC source
voltage. Therefore, the forced response is:
v f Vs 1 V (9A.43)
Notice that in this circuit the forced response for vt is the same for all three
values of resistance.

9A.29
Natural Response
First, we can write an expression for the current in terms of the voltage across
the capacitance:
dv
i C (9A.44)
dt
Then, we write a KVL equation for the circuit:
Ri v Vsu t
di
L (9A.45)
dt
Using Eq. (9A.44) to substitute for i , we get, for t 0 :
d 2v dv
LC 2 RC v Vs (9A.46)
dt dt
Dividing through by LC, we have:
d 2v R dv 1 1
2
v Vs (9A.47)
dt L dt LC LC
In D operator notation, the equation is:
2 R 1 1
D D v Vs (9A.48)
L LC LC
Therefore, the characteristic equation is:
R 1
s2 s 0 (9A.49)
L LC

9A.30
If we let:
R 1
and 0 (9A.50)
2L LC
then the characteristic equation can be written as:
s 2 2s 02 0 (9A.51)
which has general solutions:
s1 2 02
(9A.52)
s2 2 2
0
Next, we find the natural response and complete response for each value of R.
For all three cases we have:
1
0 26444 (9A.53)
LC

9A.31
Case I Overdamped ( R 330 )
In this case, we get:
R
126923
2L (9A.54)
Since we have 0 , this is the overdamped case. The roots of the

characteristic equation are given by:
s1 2 02
(9A.55)
126923 126923 26444 2 2
2785
and
s2 2 02
(9A.56)
251060
For the overdamped case, the natural response has the form:
(9A.57)
vn K1e s1t K 2 e s2t
Adding the forced response given by Eq. (9A.43) to the natural response, we
obtain the complete response:
vt 1 K1e s1t K 2e s2t (9A.58)

9A.32
Now, we must find values of K1 and K 2 so the solution matches the known
initial conditions in the circuit. It was given that the initial voltage on the
capacitance is zero, hence v0 0 . Evaluating Eq. (9A.58) at t 0 , we obtain:
0 1 K1 K 2 (9A.59)
Furthermore, the initial current was given as i 0 0 . Since the current through
the capacitance is given by:
dv
i C (9A.60)
dt
we conclude that:
dv
0 (9A.61)
dt
Taking the derivative of Eq. (9A.58) and evaluating at t 0 , we have:
s1 K1 s 2 K 2 0 (9A.62)
Now, we can solve Eqs. (9A.59) and (9A.62) for the values of K1 and K 2 . The
results are K1 1.0112 and K 2 0.01122 . Substituting these values into
Eq. (9A.58), we have the solution:
vt 1 1.0112e 2785t 0.01122e 251060t (9A.63)
Evaluating this expression at t 120 s , we get:
v120s 0.27607 V (9A.64)

9A.33
Case II Critically Damped ( R 68.7552 )
In this case, we get:
R
26444
2L (9A.65)
Since 0 , this is the critically damped case. The roots of the characteristic
equation are given by:
s1 s2 2 02 26444 (9A.66)
The natural response has the form of:
vn K1 K 2t e s1t (9A.67)
Adding the forced response to the natural response, we find that:
vt 1 K1 K 2t e s1t (9A.68)
As in Case I, the initial conditions require v0 0 and dv0 dt 0 . Thus,

substituting t 0 into Eq. (9A.68), we get:
0 1 K1 (9A.69)
Differentiating Eq. (9A.68) and substituting t 0 yields:
s1 K1 K 2 0 (9A.70)
Solving Eqs. (9A.69) and (9A.70) yields K1 1 and K 2 26444 . Thus the
complete response is:
vt 1 e 26444 t 26444te 26444 t (9A.71)
v120s 0.82528 V (9A.72)

9A.34
Case III Underdamped ( R 33 )
For this value of resistance, we have:
R
12692 (9A.73)
2L
Since 0 , this is the underdamped case. Using:
d 02 2 (9A.74)
we compute the damped natural frequency:
d 23199 (9A.75)
The natural response has the form:
vn e t K1 cos d t K 2 sin d t (9A.76)
Adding the forced response found earlier to the natural response, we obtain the
complete response:
vt 1 e t K1 cos d t K 2 sin d t (9A.77)
As in the previous cases, the initial conditions are v0 0 and dv0 dt 0 .

Evaluating Eq. (9A.77) at t 0 , we obtain:
0 1 K1 (9A.78)
Differentiating Eq. (9A.77), we get:
dvt (9A.79)
e t d K1 sin d t d K 2 cos d t
dt
e t K1 cos d t K 2 sin d t

9A.35
Evaluating at t 0 , we have:
d K 2 K1 0 (9A.80)
Solving Eqs. (9A.78) and (9A.80), we obtain K1 1 and K 2 0.5471 .

Thus, the complete solution is:
vt 1 e 12692t cos 23199t

0.5471e 12692t sin 23199t (9A.81)
v120s 1.16248 V (9A.82)
For this underdamped case, the response will look like:
v (t )
vp
1.00
0
t
tp
Figure 9A.5

9A.36
Maximum Value and Peak Time
We can evaluate the maximum value, and the time to reach the maximum value
(called the peak time), by observing that the derivative of vt is zero at
relative maxima and minima.
We already found the derivative of vt :
dvt (9A.83)
K1e t cos d t d K1e t sin d t
dt
K 2 e t sin d t d K 2 e t cos d t
Grouping terms, this can be written:
dvt (9A.84)
e t d K 2 K 1 cos d t d K1 K 2 sin d t
dt
By applying the dv0 dt 0 initial condition, we can see that:
d K 2 K 1 0 (9A.85)
Therefore, the derivative is:
dvt (9A.86)
e t d K1 K 2 sin d t
dt
Equating this to zero, we get:
sin d t 0 (9A.87)
Therefore, the times of relative maxima and minima are:
n
tm (9A.88)
d

9A.37
The case of n 1 corresponds to the peak time. Thus:

tp (9A.89)
d
Substituting this value into the output expression:
vt 1 e t K1 cos d t K 2 sin d t (9A.90)
we get:
v p vt p 1 K1e d (9A.91)
Putting values of this particular case into Eqs. (9A.89) and (9A.91), we get:

tp 0.13542 ms (9A.92)
23199
and:
12692

vp 1 e 23199
1.17929 V (9A.93)
Thus, the output exhibits a peak overshoot of about 18%.

9A.38
Summary
For the RLC circuit, we define the undamped natural frequency:
1
0
LC
For the RLC circuit, we define the exponential damping coefficient:
1
parallel
2 RC
R
series
2L
The RLC circuit exhibits three different forms for the natural response:
Overdamped ( 0 )
The natural response is of the form:
f n K 1e s1t K 2 e s2t
Critically Damped ( 0 )
f n K 1 K 2 t e t
Underdamped ( 0 )
f n e t K1 cos d t K 2 sin d t , d 02 2
The complete response is the sum of the forced response and the natural
response.
References
Hill, 1984.

9A.39
Exercises
1.
iR iL iC
1
2 L v 20 F

Let L 1.25 H and determine vt if v 0 100 V and:
(a)
iC 0 20 A (b)
iL 0 20 A
2.
50 125 mH
8 F v
200 50
120 V 6V
Both switches close at t 0 after having been open for a very long time.
(a) Find vt .
(b) Determine the maximum and minimum values of vt .

9A.40
3.
The capacitor voltage in a parallel RLC circuit that is critically damped is given
by vt 1000e 500t t 0.01 V . If the energy stored in the capacitor is 2 mJ at
t 0 , find:
(a) R.
(b) The initial energy stored in the inductor.
4.
5 L v C
The voltage is given by vt e 7 t 20 cos 24t 5 sin 24t V for t 0 . Find:
(a) L and C .
(b) The initial energy stored in the circuit.
5.
1 t=0
10
1.5 V H 1 nF 50 k
101
What is the maximum voltage magnitude present across the switch after t 0 ?
(Note that it is much safer to solve this problem analytically than to do so
experimentally.)

9A.41
6.
20
3
F
10 2 mH
100 V t=0 40 vR
The switch closes at t 0 . Find vR t .
7.
t=0 2.5 mH
10 V v 10 nF 0.5
The switch has been closed for hours. It is opened at t 0 . Show that a 10 V
battery can create a high voltage by finding v at t 2.5 s .
8.
A 2.5 H inductor, a 4 resistor, and a 25 mF capacitor are in parallel.
An 80 V battery is then placed in series with the inductor.
(a) After a long time has passed, find the energy stored in the inductor and in
the capacitor.
(b) The battery voltage drops suddenly to 40 V at t 0 . Find the energy

stored in the inductor and in the capacitor 0.25 s later.

9A.42
9.
20 H 1.2 k
iC
10
20 V vC F 50 u (t ) V
3
Find vC t and iC t .
10.
25
iL
2.5 mH 25 F vC
3-5u (t ) A
Find:
(a) vC t
(b) iL t
11.
A 5 mH inductor, a 50 F capacitor, and a 25 resistor are in series with a
voltage source vs t . The source voltage is zero prior to t 0 . At t 0 it
jumps to 75 V, at t 1 ms it drops to zero, at t 2 ms it again jumps to 75 V,

and it continues in this periodic manner thereafter. Find the source current at:
(a) t0 (b) t 1 ms (c) t 2 ms

9A.43
William Thomson (Lord Kelvin) (1824-1907)
William Thomson was probably the first true electrical engineer. His
engineering was firmly founded on a solid bedrock of mathematics. He
invented, experimented, advanced the state-of-the-art, was entrepreneurial, was
a businessman, had a multi-disciplinary approach to problems, held office in
the professional body of his day (the Royal Society), published papers, gave
lectures to lay people, strived for an understanding of basic physical principles
and exploited that knowledge for the benefit of mankind.
William Thomson was born in Belfast, Ireland. His father was a professor of
engineering. When Thomson was 8 years old his father was appointed to the
chair of mathematics at the University of Glasgow. By age 10, William
Thomson was attending Glasgow University. He studied astronomy, chemistry
and natural philosophy (physics, heat, electricity and magnetism). Prizes in
Greek, logic (philosophy), mathematics, astronomy and physics marked his
progress. In 1840 he read Fouriers The Analytical Theory of Heat and wrote:
I had become filled with the utmost admiration for the splendour and
poetry of Fourier I took Fourier out of the University Library; and in a
fortnight I had mastered it - gone right through it.
At the time, lecturers at Glasgow University took a strong interest in the

approach of the French mathematicians towards physical science, such as
Lagrange, Laplace, Legendre, Fresnel and Fourier. In 1840 Thomson also read
Laplaces Mcanique Cleste and visited Paris.
In 1841 Thomson entered Cambridge and in the same year he published a

paper on Fourier's expansions of functions in trigonometrical series. A more
important paper On the uniform motion of heat and its connection with the
mathematical theory of electricity was published in 1842.
The examinations in Cambridge were fiercely competitive exercises in problem

solving against the clock. The best candidates were trained as for an athletics
contest. Thomson (like Maxwell later) came second. A day before he left
Cambridge, his coach gave him two copies of Greens Essay on the

9A.44
Application of Mathematical Analysis to the Theories of Electricity and
Magnetism.
After graduating, he moved to Paris on the advice of his father and because of
his interest in the French approach to mathematical physics. Thomson began
trying to bring together the ideas of Faraday, Coulomb and Poisson on
electrical theory. He began to try and unify the ideas of action-at-a-distance,
the properties of the ether and ideas behind an electrical fluid. He also
became aware of Carnots view of heat.
In 1846, at the age of twenty two, he returned to Glasgow on a wave of

testimonials from, among others, De Morgan, Cayley, Hamilton, Boole,
Sylvester, Stokes and Liouville, to take up the post of professor of natural
philosophy. In 1847-49 he collaborated with Stokes on hydrodynamic studies,
which Thomson applied to electrical and atomic theory. In electricity Thomson
provided the link between Faraday and Maxwell. He was able to mathematise
Faradays laws and to show the formal analogy between problems in heat and
electricity. Thus the work of Fourier on heat immediately gave rise to theorems
on electricity and the work of Green on potential theory immediately gave rise
to theorems on heat flow. Similarly, methods used to deal with linear and
rotational displacements in elastic solids could be applied to give results on
electricity and magnetism. The ideas developed by Thomson were later taken
up by Maxwell in his new theory of electromagnetism.
Thomsons other major contribution to fundamental physics was his

combination of the almost forgotten work of Carnot with the work of Joule on
the conservation of energy to lay the foundations of thermodynamics. The
thermodynamical studies of Thomson led him to propose an absolute
temperature scale in 1848 (The Kelvin absolute temperature scale, as it is now
known, was defined much later after conservation of energy was better
understood).

9A.45
The Age of the Earth

In the first decades of the nineteenth century geological evidence for great
changes in the past began to build up. Large areas of land had once been under
water, mountain ranges had been thrown up from lowlands and the evidence of
fossils showed the past existence of species with no living counterparts. Lyell,
in his Principles of Geology, sought to explain these changes by causes now
in operation. According to his theory, processes such as slow erosion by
wind and water; gradual deposition of sediment by rivers; and the cumulative
effect of earthquakes and volcanic action combined over very long periods of
time to produce the vast changes recorded in the Earths surface. Lyells so-
called uniformitarian theory demanded that the age of the Earth be measured
in terms of hundreds of millions and probably in terms of billions of years.
Lyell was able to account for the disappearance of species in the geological
record but not for the appearance of new species. A solution to this problem
was provided by Charles Darwin (and Wallace) with his theory of evolution by
natural selection. Darwins theory also required vast periods of time for
operation. For natural selection to operate, the age of the Earth had to be
measured in many hundreds of millions of years.
Such demands for vast amounts of time run counter to the laws of
thermodynamics. Every day the sun radiates immense amounts of energy. By
the law of conservation of energy there must be some source of this energy.
Thomson, as one of the founders of thermodynamics, was fascinated by this
problem. Chemical processes (such as the burning of coal) are totally
insufficient as a source of energy and Thomson was forced to conclude that
gravitational potential energy was turned into heat as the sun contracted. On
this assumption his calculations showed that the Sun (and therefore the Earth)
was around 100 million years old.

9A.46
However, Thomsons most compelling argument concerned the Earth rather

than the Sun. It is well known that the temperature of the Earth increases with
depth and
this implies a continual loss of heat from the interior, by conduction

outwards through or into the upper crust. Hence, since the upper crust does
not become hotter from year to year there must be aloss of heat from the
whole earth. It is possible that no cooling may result from this loss of heat
but only an exhaustion of potential energy which in this case could scarcely
be other than chemical.
Since there is no reasonable mechanism to keep a chemical reaction going at a

steady pace for millions of years, Thomson concluded that the earth is
merely a warm chemically inert body cooling. Thomson was led to believe
that the Earth was a solid body and that it had solidified at a more or less
uniform temperature. Taking the best available measurements of the
conductivity of the Earth and the rate of temperature change near the surface,
he arrived at an estimate of 100 million years as the age of the Earth
(confirming his calculations of the Suns age).
The problems posed to Darwins theory of evolution became serious as

Thomsons arguments sank in. In the fifth edition of The Origin of Species,
Darwin attempted to adjust to the new time scale by allowing greater scope for
evolution by processes other than natural selection. Darwin was forced to ask
for a suspension of judgment of his theory and in the final chapter he added
With respect to the lapse of time not having been sufficient since our planet
was consolidated for the assumed amount of organic change, and this
objection, as argued by [Thomson], is probably one of the gravest yet
advanced, I can only say, firstly that we do not know at what rate species
change as measured by years, and secondly, that many philosophers are not
yet willing to admit that we know enough of the constitution of the universe
and of the interior of our globe to speculate with safety on its past duration.
(Darwin, The Origin of Species, Sixth Edition, p.409)

9A.47
The chief weakness of Thomsons arguments was exposed by Huxley
this seems to be one of the many cases in which the admitted accuracy of
mathematical processes is allowed to throw a wholly inadmissible
appearance of authority over the results obtained by them. Mathematics A variant of the
may be compared to a mill of exquisite workmanship, which grinds you stuff adage:
of any degree of fineness; but nevertheless, what you get out depends on Garbage in equals
garbage out.
what you put in; and as the grandest mill in the world will not extract
wheat-flour from peascods, so pages of formulae will not get a definite
result out of loose data.
(Quarterly Journal of the Geological Society of London, Vol. 25, 1869)
However, Thomsons estimates were the best available and for the next thirty
years geology took its time from physics, and biology took its time from
geology. But Thomson and his followers began to adjust his first estimate
down until at the end of the nineteenth century the best physical estimates of
the age of the Earth and Sun were about 20 million years whilst the minimum
the geologists could allow was closer to Thomsons original 100 million years.
Then in 1904 Rutherford announced that the radioactive decay of radium was
accompanied by the release of immense amounts of energy and speculated that
this could replace the heat lost from the surface of the Earth.
The discovery of the radioactive elementsthus increases the possible limit

of the duration of life on this planet, and allows the time claimed by the
geologist and biologist for the process of evolution.
(Rutherford quoted in Burchfield, p.164)
A problem for the geologists was now replaced by a problem for the physicists.
The answer was provided by a theory which was just beginning to be gossiped
about. Einsteins theory of relativity extended the principle of conservation of
energy by taking matter as a form of energy. It is the conversion of matter to
heat which maintains the Earths internal temperature and supplies the energy
radiated by the sun. The ratios of lead isotopes in the Earth compared to
meteorites now leads geologists to give the Earth an age of about 4.55 billion
years.

9A.48
The Transatlantic Cable

The invention of the electric telegraph in the 1830s led to a network of
telegraph wires covering England, western Europe and the more settled parts of
the USA. The railroads, spawned by the dual inventions of steam and steel,
were also beginning to crisscross those same regions. It was vital for the
smooth and safe running of the railroads, as well as the running of empires, to
have speedy communication.
Attempts were made to provide underwater links between the various separate
systems. The first cable between Britain and France was laid in 1850. The
operators found the greatest difficulty in transmitting even a few words. After
12 hours a trawler accidentally caught and cut the cable. A second, more
heavily armoured cable was laid and it was a complete success. The short lines
worked, but the operators found that signals could not be transmitted along
submarine cables as fast as along land lines without becoming confused.
In spite of the record of the longer lines, the American Cyrus J. Fields proposed
a telegraph line linking Europe and America. Oceanographic surveys showed
that the bottom of the Atlantic was suitable for cable laying. The connection of
existing land telegraph lines had produced a telegraph line of the length of the
proposed cable through which signals had been passed extremely rapidly. The
British government offered a subsidy and money was rapidly raised.
Faraday had predicted signal retardation but he and others like Morse had in
mind a model of a submarine cable as a hosepipe which took longer to fill with
water (signal) as it got longer. The remedy was thus to use a thin wire (so that
less electricity was needed to charge it) and high voltages to push the signal
through. Faradays opinion was shared by the electrical adviser to the project,
Dr Whitehouse (a medical doctor).
Thomsons researches had given him a clearer mathematical picture of the

problem. The current in a telegraph wire in air is approximately governed by
the wave equation. A pulse on such a wire travels at a well defined speed with
no change of shape or magnitude with time. Signals can be sent as close
together as the transmitter can make them and the receiver distinguish them.

9A.49
In undersea cables of the type proposed, capacitive effects dominate and the
current is approximately governed by the diffusion (i.e. heat) equation. This
equation predicts that electric pulses will last for a time that is proportional to
the length of the cable squared. If two or more signals are transmitted within
this time, the signals will be jumbled at the receiver. In going from submarine
cables of 50 km length to cables of length 2400 km, retardation effects are
2500 times worse. Also, increasing the voltage makes this jumbling (called
intersymbol interference) worse. Finally, the diffusion equation shows that the
wire should have as large a diameter as possible (small resistance).
Whitehouse, whose professional reputation was now involved, denied these

conclusions. Even though Thomson was on the board of directors of Fields
company, he had no authority over the technical advisers. Moreover the
production of the cable was already underway on principles contrary to
Thomsons. Testing the cable, Thomson was astonished to find that some
sections conducted only half as well as others, even though the manufacturers
were supplying copper to the then highest standards of purity.
Realising that the success of the enterprise would depend on a fast, sensitive
detector, Thomson set about to invent one. The problem with an ordinary
galvanometer is the high inertia of the needle. Thomson came up with the
mirror galvanometer in which the pointer is replaced by a beam of light.
In a first attempt in 1857 the cable snapped after 540 km had been laid. In
1858, Europe and America were finally linked by cable. On 16 August it
carried a 99-word message of greeting from Queen Victoria to President
Buchanan. But that 99-word message took 16 hours to get through. In vain,
Whitehouse tried to get his receiver to work. Only Thomsons galvanometer
was sensitive enough to interpret the minute and blurred messages coming
through. Whitehouse ordered that a series of huge two thousand volt induction
coils be used to try to push the message through faster after four weeks of
this treatment the insulation finally failed; 2500 tons of cable and 350 000 of
capital lay useless on the ocean floor.

9A.50
In 1859 eighteen thousand kilometres of undersea cable had been laid in other
parts of the world, and only five thousand kilometres were operating. In 1861
civil war broke out in the United States. By 1864 Field had raised enough
capital for a second attempt. The cable was designed in accordance with
Thomsons theories. Strict quality control was exercised: the copper was so
pure that for the next 50 years telegraphists copper was the purest available.
Once again the British Government supported the project the importance of
quick communication in controlling an empire was evident to everybody. The
new cable was mechanically much stronger but also heavier. Only one ship
was large enough to handle it and that was Brunels Great Eastern. She was
fives time larger than any other existing ship.
This time there was a competitor. The Western Union Company had decided to
build a cable along the overland route across America, Alaska, the Bering
Straits, Siberia and Russia to reach Europe the long way round. The
commercial success of the cable would therefore depend on the rate at which
messages could be transmitted. Thomson had promised the company a rate of
8 or even 12 words a minute. Half a million pounds was being staked on the
correctness of the solution of a partial differential equation.
In 1865 the Great Eastern laid cable for nine days, but after 2000 km the cable
parted. After two weeks of unsuccessfully trying to recover the cable, the
expedition left a buoy to mark the spot and sailed for home. Since
communication had been perfect up until the final break, Thomson was
confident that the cable would do all that was required. The company decided
to build and lay a new cable and then go back and complete the old one.
Cable laying for the third attempt started on 12 July 1866 and the cable was
landed on the morning of the 27th. On the 28th the cable was open for business
and earned 1000. Western Union ordered all work on their project to be
stopped at a loss of $3 000 000.

9A.51
On 1 September after three weeks of effort the old cable was recovered and on
8 September a second perfect cable linked America and Europe. A wave of
knighthoods swept over the engineers and directors. The patents which
Thomson held made him a wealthy man.
For his work on the transatlantic cable Thomson was created Baron Kelvin of
Largs in 1892. The Kelvin is the river which runs through the grounds of
Glasgow University and Largs is the town on the Scottish coast where
Thomson built his house.
Other Achievements
There are many
Thomson worked on several problems associated with navigation sounding other factors
influencing local
machines, lighthouse lights, compasses and the prediction of tides. Tides are tides such as
channel width
primarily due to the gravitational effects of the Moon, Sun and Earth on the which produce
phenomena akin to
oceans but their theoretical investigation, even in the simplest case of a single resonance in the
tides. One example
ocean covering a rigid Earth to a uniform depth, is very hard. Even today, the of this is the narrow
study of only slightly more realistic models is only possible by numerical Bay of Fundy,
between Nova
computer modelling. Thomson recognised that the forces affecting the tides Scotia and New
Brunswick, where
change periodically. He then approximated the height of the tide by a the tide can be as
high as 21m. In
trigonometric polynomial a Fourier series with a finite number of terms. The contrast, the
Mediterranean Sea
coefficients of the polynomial required calculation of the Fourier series is almost tideless
coefficients by numerical integration a task that required not less than because it is a
broad body of water
twenty hours of calculation by skilled arithmeticians. To reduce this labour with a narrow
entrance.
Thomson designed and built a machine which would trace out the predicted
height of the tides for a year in a few minutes, given the Fourier series Michelson (of
Michelson-Morley
coefficients. fame) was to build a
better machine that
used up to 80
Thomson also built another machine, called the harmonic analyser, to perform Fourier series
coefficients. The
the task which seemed to the Astronomer Royal so complicated and difficult production of blips
at discontinuities by
that no machine could master it of computing the Fourier series coefficients this machine was
from the record of past heights. This was the first major victory in the struggle explained by Gibbs
in two letters to
to substitute brass for brain in calculation. Nature. These blips
are now referred to
as the Gibbs
phenomenon.

9A.52
Thomson introduced many teaching innovations to Glasgow University. He
introduced laboratory work into the degree courses, keeping this part of the
work distinct from the mathematical side. He encouraged the best students by
offering prizes. There were also prizes which Thomson gave to the student that
he considered most deserving.
Thomson worked in collaboration with Tait to produce the now famous text
Treatise on Natural Philosophy which they began working on in the early
1860s. Many volumes were intended but only two were ever written which
cover kinematics and dynamics. These became standard texts for many
generations of scientists.
In later life he developed a complete range of measurement instruments for

physics and electricity. He also established standards for all the quantities in
use in physics. In all he published over 300 major technical papers during the
53 years that he held the chair of Natural Philosophy at the University of
Glasgow.
During the first half of Thomson's career he seemed incapable of being wrong
while during the second half of his career he seemed incapable of being right.
This seems too extreme a view, but Thomson's refusal to accept atoms, his
opposition to Darwin's theories, his incorrect speculations as to the age of the
Earth and the Sun, and his opposition to Rutherford's ideas of radioactivity,
certainly put him on the losing side of many arguments later in his career.
William Thomson, Lord Kelvin, died in 1907 at the age of 83. He was buried
in Westminster Abbey in London where he lies today, adjacent to Isaac
Newton.
References
Burchfield, J.D.: Lord Kelvin and The Age of the Earth, Macmillan, 1975.
Krner, T.W.: Fourier Analysis, Cambridge University Press, 1988.
Encyclopedia Britannica, 2004.
Morrison, N.: Introduction to Fourier Analysis, John Wiley & Sons, Inc., 1994.
Thomson, S.P.: The Life of Lord Kelvin, London, 1976.
Kelvin & Tait: Treatise on Natural Philosophy, Appendix B.

10A.1
Lecture 10A Second-Order Frequency Response

Resonance. Parallel resonance. Quality factor. Second-order circuit relations.
Bandwidth. Series resonance. Other resonant forms. The second-order lowpass
frequency response. The second-order highpass frequency response. Standard
forms of second-order frequency responses.
Introduction
The use of both capacitors and inductors in a circuit gives rise to an important
phenomenon the exchange of electric and magnetic stored energy in a
sinusoidal fashion between the ideal elements without dissipation. The peak
energy stored in the elements reaches a maximum when a sinusoidal forcing
function drives the circuit into resonance at a frequency close to the natural
frequency of the circuit.
The resonance phenomenon occurs in many natural systems, and is the result
of systems being able to store energy in different ways. For example, a
pendulum, without friction, will oscillate forever once set in motion, and there
will be a continual exchange of kinetic energy (the velocity of the pendulum
mass) and potential energy (stored by virtue of the position of the pendulum in
a gravitational field) as the pendulum oscillates in a sinusoidal fashion.
In an electric circuit the resonant condition can be used to create a highly

selective circuit in the sense that a narrow band of frequencies of the forcing
function will cause the circuit response to be large whilst all other frequencies
result in a response which is much smaller. This property of second-order
circuits is not exhibited by passive first-order circuits (or cascades of them).
The width of the magnitude response, or bandwidth B, for a second-order

circuit will be seen to be related to a quantity called the quality factor, Q. The
quality factor, together with the undamped natural frequency, 0 , uniquely
determine the properties of many second-order systems. These parameters can
be determined for a parallel passive RLC circuit, an electronic op-amp circuit, a
mechanical system, a hydraulic system, etc. Thus, it pays to express second-
order frequency response quantities in terms of 0 and Q for the sake of
uniformity across the disciplines.

10A.2
Resonance
A system being driven by a sinusoidal forcing function will produce a
sinusoidal steady-state response at the frequency of the driving force. The
amplitude of the response may be larger than the forcing function when the
frequency of the driving force is near a natural frequency of oscillation of the
system.1 This dramatic increase in amplitude near a natural frequency is called
resonance, and we denote the frequency at which it occurs as r , which is
called the resonance frequency of the system.2
The phenomenon of resonance is a familiar one (at least qualitatively). For

example, a child using a swing realizes that if the pushes are properly timed,
the swing can move with a very large amplitude. The quartz crystal in a
computer or watch is made to mechanically vibrate at a resonance frequency
which is determined by its cut (its shape and size). A further illustration is
furnished by the shattering of a crystal wineglass when exposed to a musical
tone of the right pitch. The condition of resonance may or may not be
desirable, depending on the circumstances.
The reason for large-amplitude oscillations at the resonance frequency is that

energy is being transferred to the system under favourable conditions. In fact,
for a parallel RLC circuit, maximum power will be dissipated by the circuit
when the forcing functions voltage and current are in phase i.e. the circuit
appears to be purely resistive. This leads us to a precise definition for a
resonance frequency, r .
1
A natural frequency of oscillation only occurs in systems that are second-order or higher.
Even then, a large amplitude response will occur only under certain conditions.
2
This assumes that one natural frequency gives rise to one resonance frequency. Certain
topologies of circuit components can give rise to two resonance frequencies even though there
is only one natural frequency. Circuits and systems of high-order can have multiple natural
frequencies and consequently may have multiple resonance frequencies. Also, even though the
amplitude response is very large at a resonance frequency, it is not necessarily the maximum
amplitude response it depends on how we define response.

10A.3
In a two-terminal electrical network containing at least one inductor and one
capacitor, we define resonance to be the condition which exists when the input
impedance of the network is purely resistive:
Resonance defined
n th -order
circuit
Z ( j ) (n 2)
at resonance when Z is real
Figure 10A.1
The resonant condition can be achieved by adjusting any circuit parameter. We

normally devote our attention to the case for which is the variable, and since
we have denoted the resonance frequency as r , then resonance occurs if:
ImZ jr 0 (10A.1)
Thus, a two-terminal circuit is said it be in resonance when the sinusoidal

voltage and current at the circuit input terminals are in phase.
This definition also implies that at resonance:
ImY jr 0 (10A.2)
since Z R is real at resonance and therefore Y 1 R G is real also.
It should be noted that the resonance condition ImZ j r 0 may not be

satisfied at any real positive frequency r . Thus, resonance is a condition
which a circuit may achieve, but only if its topology and component values
allow it.

10A.4
Parallel Resonance
We shall apply the definition of resonance to the parallel RLC circuit:
IR IL IC
1
Is R jL V
jC
Figure 10A.2
The admittance offered to the ideal current source is:
1 1
Y j C (10A.3)
R L
Obviously, if the impedance is to be purely resistive at resonance, then so is the
admittance. Thus, resonance occurs when:
1
r C 0 (10A.4)
r L
Hence, the resonance frequency for this simple case is:
1
r 0 (10A.5)
LC
This resonance frequency is identical to the undamped natural frequency that
was defined whilst considering the step-response of the parallel RLC circuit.
Let us examine the magnitude of the response, the voltage V as indicated in

Figure 10A.2. For a constant-amplitude sinusoidal current source input, the
response is proportional to the input impedance Z .

10A.5
The admittance as a function of can be written as:
(10A.6)
1 1
Y jC
R jL
2 j RC 1 LC
C
j
The response is therefore given by:
1 j (10A.7)
V ZI
C 1 LC 2 j RC
Is
The response starts at zero, reaches a maximum value at the resonance

frequency, and then drops again to zero as becomes infinite:
Typical magnitude
and phase response
|V| of a second-order
bandpass circuit
R |I s|
R |I s|/ 2
1 pole roll-off
0
0 1 0 2
90
45
V
() 0 2
0 1
-45
-90
Figure 10A.3
The frequency response is referred to as a bandpass response, since it passes

only midband frequencies whilst attenuating low and high frequencies.

10A.6
To show that the maximum response is R I s , and that this occurs at resonance,
we could take the magnitude of Eq. (10A.7), differentiate and equate to zero to
obtain the frequencies of any relative maxima and minima, and then use these
values to obtain the magnitude of the response. However, there is a simpler
way.
The admittance contains a constant conductance and a susceptance which has a

minimum magnitude (of zero) at resonance. The minimum admittance
magnitude therefore occurs at resonance, and it is 1 R . Hence, the maximum
impedance magnitude is R , and it occurs at resonance.
At the resonance frequency, therefore, the voltage across the parallel circuit is
simply RI s , and we can see that the source current goes through the resistor.
However, there is also a current in L and C, since there is a voltage across

them:
(10A.8)
RI
I Lr
jr L
I Cr jr CRI
Since 1 r C r L at resonance, we find that:
I Cr I Lr (10A.9)
and therefore I Lr I Cr 0 . We thus have a circulating current around the LC
part of the circuit which is the cause of the never-ending exchange of energy
between the inductor and the capacitor at resonance.
Although the height of the response curve depends only upon the value of R,
the width of the curve or the steepness of the sides depends upon the other two
element values also. The width of the response curve is most easily expressed
when we introduce a very important parameter, the quality factor Q.

10A.7
Phasor Diagram of the Parallel RLC Circuit
It is instructive to illustrate resonance with the aid of phasor diagrams. The

figure below shows phasor diagrams, as well as illustrations of the sinusoidal
currents, for the parallel RLC circuit at three different frequencies:
i
IC
iR
iL
iC
IR V
t
I s = I R+ I L + I C
IL + IC
< r
IL
i
IC iR iL iC
IR V
t
I s = I R+ I L + I C
IL + IC = 0
= r
IL
IC
i
IL + IC iR
I s = I R+ I L + I C iC
iL
IR V
t
IL
> r
Figure 10A.4

10A.8
Quality Factor
We define quality factor, Q, as:
maxmium energy stored

Q 2 (10A.10)
total energy lost per period
The constant 2 is put into the definition in order to simplify the more useful
expressions for Q that occur in the study of second-order systems.
For the parallel RLC circuit, energy is only lost in the resistor. We can
therefore express Q in terms of the instantaneous energy associated with each
of the reactive elements and the average power dissipated in the resistor:
Q 2
wL t wC t max
(10A.11)
PRT0
We will apply this definition and determine the value of Q at the resonance
frequency r 0 , which is denoted by Q0 . We select the forcing function:
it I m cos0t (10A.12)
and obtain the corresponding voltage at resonance:
vt Rit RI m cos0t (10A.13)
The instantaneous energy stored in the capacitor is:
2 2
wC t 12 Cv cos2 0t
CR
2 Im
(10A.14)
2

10A.9
The instantaneous energy stored in the inductor is:
2
1 t
wL t 12 LiL2 12 L vdt
L 0
R 2 I m2 CR 2 I m2
sin 2
t sin 2
0t (10A.15)
20 L
2 0
2
The total instantaneous energy stored is therefore constant:
(10A.16)
CR 2 I m2
wL t wC t
2
and this constant value must also be the maximum value.
In order to find the energy lost in the resistor in one period, we take the average
power absorbed by the resistor:
(10A.17)
PR 12 RI m2
and multiply by one period, obtaining:
(10A.18)
RI m2
PRT0
2 f0
We thus find the quality factor at resonance:
(10A.19)
CR 2 I m2 2
Q0 2 2 2f 0CR 0CR
RI m 2 f 0
This equation holds only for the simple parallel RLC circuit.

10A.10
Equivalent expressions for Q0 which are often useful may be obtained by
substitution:
R R C
Q0 0CR R
X C 0 X L0 L (10A.20)
It is apparent that Q0 is a dimensionless constant which is a function of all
three circuit elements in the parallel resonant circuit, and it turns out that it can
be evaluated from a knowledge of the natural response, as will be illustrated
later.
A useful interpretation of Q0 is obtained when we inspect the capacitor and

inductor currents at resonance:
I Cr I Lr j0CRI jQ0I (10A.21)
Each is Q0 times the source current in magnitude and they are 180 out of
phase. Thus if we apply 1 mA at the resonance frequency to a parallel resonant

circuit with a Q0 of 50, we find 1 mA in the resistor, and 50 mA in both the
inductor and capacitor. A parallel resonant circuit can therefore act as a current
amplifier (but not a power amplifier, since it is a passive network).

10A.11
Second-Order Circuit Relations

The two parameters and d were introduced in connection with the natural
response of a second-order circuit. These two parameters can be related to the

undamped natural frequency, 0 , and the quality factor at resonance, Q0 .
We have:
(10A.22)
1 1

2 RC 2Q0 0C C
and thus:
0

2Q0 (10A.23)
We also have:
(10A.24)
d 02 2
and thus:
2
1
d 0 1 (10A.25)
2
0
Q
When we analyze second-order circuits in the time-domain, we usually resort

to using , 0 and d (for an underdamped circuit). When we discuss
second-order circuits in the frequency-domain, we usually use 0 and Q0 .

10A.12
Bandwidth
The width of the response curve for the parallel RLC circuit can be defined
more carefully and related to Q0 .
|T|
R |I|
R |I| / 2
0.5
B
0
0 1 0 2
Figure 10A.5
The two half-power frequencies 1 and 2 are those frequencies at which the
magnitude of the voltage response is 1 2 times its maximum value (or -3 dB

down from the peak).
We select 1 as the lower half-power frequency and 2 as the upper half-
power frequency. These names arise from the fact that a voltage which is 1 2
times the resonance voltage is equivalent to a squared voltage which is one-half
the squared voltage (and therefore the power) at resonance.
The bandwidth of a resonant circuit is defined as the difference of these two

half-power frequencies:
B 2 1 rads-1 (10A.26)
We can also refer to bandwidth as:
B f 2 f1 Hz (10A.27)
The context of the analysis or design makes the units of bandwidth clear.

10A.13
The use of half-power frequencies in the definition of bandwidth is an arbitrary
but widely accepted criterion used by the engineering profession. The concept
of bandwidth is used in many other electrical systems and is a very important
parameter in the design of filters, amplifiers and electrical systems in general.
One should also be aware that bandwidth is only defined for systems with a
single peak response otherwise the definition is ambiguous.
We think of this bandwidth as the width of the response curve, even though
the curve actually extends from 0 to .
We can express the bandwidth B in terms of 0 and Q0 . The admittance of the

parallel RLC circuit is:
(10A.28)
1 1
Y j C
R L
1 1 0CR 0 R
j
R R 0 0 L
1 0
1 jQ0
R 0
The magnitude of the admittance at resonance is 1 R , and we seek frequencies
at which the magnitude reaches 2 R to achieve half-power. This must occur

when the imaginary part of the bracketed quantity has a magnitude of unity.
Thus:

Q0 1 0 1 and Q0 2 0 1 (10A.29)
0 1 0 2

10A.14
Solving, we have:

2
1 1
1 0 1
2Q0 2Q0

2
1 1
2 0 1
2Q0 2Q0
(10A.30)
Although individually complicated, their difference provides a very simple

formula for the bandwidth:
Bandwidth defined
for a parallel 0
resonant circuit B 2 1 rads1 (10A.31)
Q0
This equation tells us that Q0 and B are inversely related, as shown below:
The inverse
relationship between
B and Q0 |T|
|T| max
1.0 Low Q0 , large B
1/ 2
0.5
High Q0 , small B
0
0 0.5 1.0 1.5 2.0 0
Figure 10A.6

10A.15
Circuits possessing a higher Q0 have a narrower bandwidth they have greater

frequency selectivity or higher quality. Such circuits were used extensively in
receivers of the old analog broadcast systems, such as AM and FM radio and
TV, to tune into a station whilst rejecting all others.
If we multiply the two half-power frequencies together, we can show that:
(10A.32)
12 02
and therefore:
(10A.33)
0 12
That is, the resonance frequency is the geometric mean of the two half-power
frequencies.
For high-Q circuits ( Q0 5 ), we can show that:

2
1 1 0 B
1, 2 0 1 (10A.34)
2Q0 2Q0 2

and:
1 2
0 12 (10A.35)
2
That is, for high-Q circuits, each half-power frequency is located
approximately one-half bandwidth from the resonance frequency the
resonance frequency is approximately the arithmetic mean of the half-power
frequencies.

10A.16
Series Resonance
The series resonant circuit finds less use than the parallel circuit. Consider the
series RLC circuit below:
R I jL
1
V
jC
Figure 10A.7
We can derive the important equations for the series resonant circuit by using a
dual language on the parallel circuit. We find that resonance occurs when:
1
r 0 (10A.36)
LC
which is the same as for the parallel RLC circuit. However, the quality factor at
resonance for the series RLC circuit is different:
0 L X L0 X C 0 1 L
Q0 (10A.37)
R R R R C

10A.17
If the response is taken across the resistor so that it is proportional to the

current, then we achieve a similar response to that of the parallel resonant
circuit (a bandpass response). In this case the equations for the half-power
frequencies, the bandwidth, and the resonance frequency are the same as
before:

2
1 1
1, 2 0 1 (10A.38)
2Q0 2Q0

0
B 2 1 rads1 (10A.39)
Q0
0 12
(10A.40)
The series resonant circuit is characterized by a low impedance at resonance.

The series resonant circuit provides inductor and capacitor voltages which are
greater than the source voltage by the factor Q0 . The series circuit thus
provides voltage amplification at resonance.

10A.18
Other Resonant Forms

The parallel and series RLC circuits of the previous two sections represent
idealized resonant circuits they are useful approximations to real physical
circuits where the resistance of the wire making up the inductor and the losses
in the capacitors dielectric are small. The network shown below is a
reasonably accurate model for the parallel combination of a physical inductor,
capacitor and resistor. The resistor RL represents the ohmic losses, core losses,
and radiation losses of the physical coil. The resistor R represents the losses in
the dielectric within the physical capacitor as well as the resistance of the
physical resistor that is placed in parallel with the inductor and capacitor.
RL
1
R
jC
Y jL
Figure 10A.8
In this circuit, there is no way to combine elements and produce a simpler

model. We have to resort to first principles to determine its resonant condition.
The definition of resonance is unchanged, and we determine the resonance
frequency by setting the imaginary part of the admittance to zero:
ImY jr 0
1 1
Im jr C 0 (10A.41)
R RL j r
L

10A.19
Realizing the denominator of the inductor branch, we get:
1 R jr L
Im jr C L2 0
2 2
(10A.42)
R RL r L
Thus:
L
C (10A.43)
RL2 r2 L2
and:
2
1 RL
r (10A.44)
LC L
We note that r is less than 1 LC , but sufficiently small values of the ratio
RL L may result in a negligible difference between r and 1 LC .
The maximum magnitude of the input impedance is not R , and it does not
occur at r (or at 1 LC ). The proof is algebraically cumbersome, but
the theory is straightforward (set the derivative of the impedance magnitude to
zero to find relative maxima and minima, etc.).
It should also be pointed out that if R L L 1 LC then resonance will never

occur, since Eq. (10A.44) results in an imaginary quantity. Thus, we must be
careful in any analysis we undertake to check the conditions under which a
circuit may exhibit the phenomenon of resonance.

10A.20
Example
Consider the simple RLC circuit shown below:
L= 14 H v
i
1
v s (t ) = cos (4 t ) V R= 2 C= 18 F
By KVL, for the mesh on the left:
di
L v vs 0
dt
while, by KCL, at node v:
dv v
iC
dt R
Substituting the second expression into the first, we get:
d 2 v L dv
LC 2 v vs 0
dt R dt
from which:
d 2v 1 dv 1 1
2
v vs
dt RC dt LC LC
Thus we see that the undamped natural frequency is:
1 1
0 32 4 2 rads -1
LC 1 1

4 8

10A.21
However, the impedance seen by the source is:
R1 jC
Z jL
R 1 j C
R
jL
1 jRC
R1 jRC
jL
1 2 R 2C 2
R R 2 C
j
L
1 2 R 2C 2 1 2 R 2C 2
The imaginary part of Z vanishes when:
r R 2C
r L 0
1 r2 R 2 C 2
r L1 r2 R 2 C 2 R 2 C 0
R 2C L 1 1
r 0 and r 2 2
R 2 LC 2 LC R C
We can see that neither of these resonance frequencies is equal to the

undamped natural frequency, i.e. r 0 . The two resonance frequencies are:
1 1
r 0 and r 4 rads -1

1 1 2
2 2
1
4 8 8

10A.22
Since the frequency of the source is 4 rads -1 , the circuit is in resonance.

The circuit is shown below in the frequency-domain:
j V
I
1
1 0 V 2 - j2
=4
By nodal analysis:
1 V V V

j 2 j2
from which:
2 2
V 2 45
1 j 245
and:
1 V 1 2 1 j 1 j 1 j 1 j
I 10
j j j 1 j
So:
vt 2 cos4t 45
and:
it cos4t

10A.23
The energy stored in the inductor is:
1 2 1
wL t Li t cos 2 4t
2 8
and the energy stored in the capacitor is:
1 2 1
wC t Cv t cos 2 4t 45
2 8
Using the trigonometric identity:
1
cos 2 x 1 cos 2 x
2
the total stored energy is:
1
wL t wC t 1 cos 8t 1 1 cos8t 90
16 16
1
2 cos 8t sin 8t
16

1
16

2 2 cos8t 45
and this has a maximum value of:
wL t wC t max 1
16

2 2 J
The power dissipated by the resistor is:
2
1 V 12 1
PR W
2 R 22 2
and the energy lost in a period is:
2 1 2
PR T0 PR J
r 2 4 4
Thus the Q0 of the circuit is:
Q0 2
1 162 2 1 1
1.707
4 2

10A.24
The Second-Order Lowpass Frequency Response

Consider the series RLC circuit again, but this time the response is taken as the
voltage across the capacitor:
R jL
1
Vi Vo
jC
Figure 10A.9
In this case the frequency response is given by the voltage divider rule:
T j
Vo
Vi
1 jC

R jL 1 jC
1 LC

1 LC 2 jR L (10A.45)
Noting that 0 1 LC and Q0 0 L R for the series RLC circuit, this can be
written as:
02
T j 2
0 2 j 0 Q0 (10A.46)
This has the form of a second-order lowpass frequency response it passes low
frequencies but attenuates high frequencies.

10A.25
The magnitude
2
T j
response of a
0
(10A.47) lowpass second-
2
0
2 2
0 Q0
2 order frequency
response
and the phase is:

The phase response
Q of a lowpass
T j tan 1 2 0 20 (10A.48)
second-order
0
frequency response
The magnitude and phase functions are plotted below for Q0 1.25 :
Typical magnitude
and phase
peak = 0 1-(1/2 Q02) responses of a
lowpass second-
|T| order frequency
Q0
(V/V) response
1
-40 dB / decade
0
0 0
T
0
() All Q0
-90
-180 asymptote
for all Q0
-180
0 0
Figure 10A.10

10A.26
Peak Frequency
The peak of the frequency response does not correspond to the resonance
frequency 0 (nor does it have any relation with d , which arises in the
description of the time-domain natural response). To obtain the peak of the
frequency response, we find the relative maximum in the usual way.
To simplify the algebra, we let:
u 0
2 (10A.49)
Then the magnitude response can be written as:
1 1
T j
G u
(10A.50)
1 u 2
u Q 2
0
where:
G u 1 u u Q02
2
(10A.51)
2

u 1 Q 2 u 1 2
0
We want to find p so that T j p max , or equivalently, G u p min .
To find u that minimizes G u , we let:
d
du
G u 2u 1 Q02 2 0 (10A.52)
Solving this for u we get:
1
up 1 (10A.53)
2Q02

10A.27
Thus, the frequency at which the magnitude response reaches a peak is:
1
p 0 1 , Q0 1 2
2Q02
(10A.54)
p 0, Q0 1 2
Notice that the peak response always occurs before the resonance frequency for
the lowpass response, and we approach p 0 for high Q0 (say Q0 5 ).
We can also see that a relative peak will not occur in the magnitude response if
Q0 1 2 (for then p is an imaginary quantity!). In this case, the absolute
peak occurs at 0 Hz, or DC.
The special value of Q0 1 2 causes the relative and absolute peak to

coincide at DC, and the magnitude response in this special case is known as
maximally flat (all derivatives of the magnitude response at DC are zero).
At the peak frequency, the magnitude of the frequency response is:
T j p
Q0
, Q0 1 2
1 1 2Q0
2
T j p 1,
(10A.55)
Q0 1 2
For high Q0 (say Q0 5 ), the magnitude response is T j p Q0 .

10A.28
Bandwidth
The bandwidth of the lowpass RLC circuit is the difference between the two
half-power frequencies on each side of the peak frequency:
|T|
|T ( jp)|
|T ( jp)|
2 1
0
0 1 p 2
Figure 10A.11
That is, the usual definition of bandwidth applies:
(10A.56)
B 2 1 rads-1
By definition, the two half-power frequencies must satisfy:
T j1, 2
2
1 (10A.57)

T j p
2
2
However, as will be seen, the relative peak in the magnitude response only
occurs when Q0 1 2 0.7071 , and when this is the case the peak is only
greater than 2 when Q0 1 1 2 1.307 . Therefore, we need to consider

three separate cases to determine the bandwidth of the lowpass frequency
response.

10A.29
Case I Relative Peak > 2 ( Q0 1 1 2)
As shown before:
1 1
T j
2
2

G u u 1 Q02 2 u 1 (10A.58)
and for the case of a relative peak:
T j p
2 Q02
(10A.59)
1 1 2Q0
2
so we have:
T j1, 2
2

1 1 2Q Q 0
2 2
0

1 (10A.60)
T j p
2 2

u 1 Q 2 u 1
2
0 2
This can be rewritten as:

u 2 1 Q02 2 u 1 2 1 1 2Q0
2
Q 2
0 0
(10A.61)
which can be solved for u to get:
2
1 1 1
u1, 2 1 2
1 (10A.62)
2Q0 Q0 2Q0
The two half-power frequencies are therefore:
1 u10
(10A.63)
2 u 2 0

10A.30
and the bandwidth is thus:
B 2 1 u2 u1 0 (10A.64)
For high Q0 (say Q0 5 ), we can simplify this result in the following way.
Firstly:
(10A.65)
1
u1, 2 1
Q0
then, using the binomial series:
(10A.66)
1 x 1 1 x 1 x 2
1
2
2 8
we have:
B u2 u1 0
1 1

1
Q
1
Q 0
0 0
1 1

1 1 0
2 2
Q 0 Q 0
0 (10A.67)

Q0
The lowpass circuit in this case exhibits bandpass behaviour, and it is debatable
whether we should still call it a lowpass filter. However, since the circuit still
passes low frequencies down to DC (but at levels which are below half-power),
the circuit is still classified as a lowpass filter. Perhaps the best name would be
a lowpass filter with band enhancement.

10A.31
Case II Relative Peak < 2 (1 2 Q0 1 1 2)
From Eq. (10A.62), it can be shown that when Q0 1 1 2 1.307 , then u1
will be negative and the lower half-power frequency will cease to exist. In this
case, the bandwidth is:
(10A.68)
B 2 u 2 0
2
1 1 1
0 1 1
2Q02 Q0 2
0
Q
This is because the peak response falls below 2 , and therefore
T j p 2 1 . This case is illustrated below:
|T|
|T ( jp)|
|T ( jp)| 1
2
0
0 p 2
Figure 10A.12

10A.32
Case III No Relative Peak ( Q0 1 2)
If there is no relative peak in the magnitude response (for Q0 1 2 0.7071),
then the peak response occurs at DC with a magnitude of 1. We then have only
one half-power frequency.
In a manner similar to the derivation for Case I, we can show that the
bandwidth is given by:
2
1 1
B 0 1 1
2Q 2
1 (10A.69)
2Q02 0
This case is illustrated below:
|T|
1
1
2
0
0 B
Figure 10A.13

10A.33
Bode Plots
The magnitude and phase Bode plots for a range of values of Q0 are shown
below for a normalised resonance frequency of 0 1 :
Figure 10A.14
The asymptotic Bode plot decreases at the rate of -40 dB / decade, and this is
sometimes described as two-pole rolloff.

10A.34
The Second-Order Highpass Frequency Response

Consider the series RLC circuit again, but this time the response is taken as the
voltage across the inductor:
R
1
jC
Vi jL Vo
Figure 10A.15
The frequency response is given by the voltage divider rule:
T j
Vo
Vi
jL

R jL 1 jC
2
(10A.70)
1 LC 2 jR L
Noting that 0 1 LC and Q0 0 L R for the series RLC circuit, this can be
written as:
2
T j 2
0 2 j 0 Q0 (10A.71)
This has the form of a second-order highpass frequency response it attenuates

low frequencies but passes high frequencies.

10A.35
The magnitude
2
T j
response of a
(10A.72) highpass second-
2
0
2 2
0 Q0
2 order frequency
response
and the phase is:

The phase response
Q of a highpass
T j 180 tan 1 2 0 20 (10A.73)

second-order
0
frequency response
The magnitude and phase functions are plotted below for Q0 1.25 :
Typical magnitude
and phase
peak = 0 responses of a
1-(1/2 Q02 ) highpass second-
|T| order frequency
Q0 response
(V/V)
1
0
0 0
T 180
All Q0
()
90
0 asymptote
for all Q0
0
0 0
Figure 10A.16

10A.36
Peak Frequency
In a similar manner to the lowpass response, the peak of the frequency

response for the highpass response does not correspond to the resonance
frequency 0 . If we proceed in a similar manner to that shown for the lowpass
response, we find that the frequency at which the magnitude response reaches a
peak is:
0
p , Q0 1 2
1
1 2
2Q0 (10A.74)
p , Q0 1 2
Notice that the peak response always occurs after the resonance frequency for
the highpass response, and we approach p 0 for high Q0 (say Q0 5 ).
At the peak frequency, the magnitude of the frequency response is:
T j p
Q0
, Q0 1 2
1 1 2Q0
2
T j p 1,
(10A.75)
Q0 1 2
For high Q0 (say Q0 5 ), the magnitude response is T j p Q0 .
Bandwidth
If there is a relative peak in the magnitude response, then the two half-power
frequencies are the same as for the lowpass case (the circuit exhibits bandpass
behaviour) and the bandwidth is thus:
B 2 1 u2 u1 0 (10A.76)
If Q0 1 1 2 1.307 then the peak is below 2 and the bandwidth is
infinite.

10A.37
Standard Forms of Second-Order Frequency Responses

The table below shows the possibilities and names associated with the
second-order frequency response.
Frequency Response Magnitude Response Name
02
TLP Lowpass
02 2 j 0 Q0
j 0 Q0
TBP Bandpass
2 j 0 Q0
2
0
2
THP Highpass
02 2 j 0 Q0
02 2
TBS Bandstop
02 2 j 0 Q0
notch
02 2 j 0 Q0
TAP 2 Allpass
0 2 j 0 Q0
Table 10A.1 Standard Forms of Second-Order Frequency Responses

10A.38
Summary
Resonance is a phenomenon that only occurs in 2nd-order or higher circuits,
and even then, only under certain conditions. It occurs when the forcing
function drives the circuit near one of its natural frequencies of oscillation.
For the RLC circuit, we define the resonance frequency:
1
r 0
LC
For the RLC circuit, we define the quality factor at resonance:
R R C
Q0 0CR R parallel
X C 0 X L0 L
0 L X L0 X C 0 1 L
Q0 series
R R R R C
The RLC circuit can be used to create a lowpass, bandpass or highpass

filter.
The two half-power frequencies 1 and 2 are those frequencies at which
the magnitude of the response is 1 2 times its maximum value

(or -3 dB down from the peak).
The bandwidth of an RLC circuit is defined as the difference of the two

half-power frequencies (if they exist):
B 2 1 rads-1 or B f 2 f1 Hz
For a lowpass RLC circuit, if there is only one half-power frequency then
the bandwidth is equal to it (the formulae above apply with f1 1 0 ).
References
Hill, 1984.

10A.39
Exercises
1.
Find 0 and Q0 for a parallel resonant circuit in which:
(a) C 1 4 F , L = 4 H, R 20 k .
(b) 1000 s -1 , C 5 nF , L 1 72 H .
(c) 50 s -1 , d 600 rads-1
2.
Let vs t 100 cos0t V in the circuit shown below:
10 k
i1 i2 i3
vs 40 k 50 mH 1.25 F v( t )
(a) Find the equivalent parallel RLC circuit and then determine
0 , Q0 and vt .
(b) Find i1 t , i2 t and i3 t .
(c) Calculate the average power loss in the 10 k resistor and the maximum
energy stored in the inductor.

10A.40
3.
Find the resonance frequency of the circuit shown:
10
0.2 H 25
500 F
4.
400 H 20 k
1 0 mV 100 pF 80 k V0
Find:
(a) r (b) Q0 (c) B (d) 1 (e) 2 (f) V0 at r
5.
A parallel RLC circuit used in a radio frequency (RF) amplifier is intended to
have an impedance magnitude of 5 k at resonance, 0 107 rads-1 , and
3 k at a frequency 5 kHz below resonance. Specify R, L and C.

10A.41
6.
In the series RLC circuit shown below:
100 pF
i 50
100 cos ( t ) V 400 H
(a) At what value of is the amplitude of i a maximum?
(b) By how many rads-1 would have to be increased to reduce I by 5%?
7.
Find the effective values of 0 and Q0 for the network shown below:
0.1 H
0.1 F vC
20 0.04vC

10A.42
8.
Determine reasonably accurate values of r and Q0 for the resonant circuits
shown below:
250
0.1 0.15 1 k
2 mH
2 mH 20 F 20 F
(a) (b)
9.
A series RLC circuit has an impedance of 10 j 40 at 100 rads-1 . After it

is scaled in magnitude and frequency by the same factor (i.e. k m k f ), it is
found to have an impedance of 30 j180 at 50 rads-1 . Determine the

elements in the original network.

10A.43
James Clerk Maxwell (1831-1879)
Maxwell produced a most spectacular work of individual genius he unified

electricity and magnetism. Maxwell was able to summarize all observed
phenomena of electrodynamics in a handful of partial differential equations
known as Maxwells equations:

E

B
E
t
B 0
E
B J
t
From these he was able to predict that there should exist electromagnetic waves
which could be transmitted through free space at the speed of light. The
revolution in human affairs wrought by these equations and their experimental
verification by Heinrich Hertz in 1888 is well known: wireless
communication, control and measurement - so spectacularly demonstrated by
television and radio transmissions across the globe, to the moon, and even to
the edge of the solar system!
James Maxwell was born in Edinburgh, Scotland. His mother died when he
was 8, but his childhood was something of a model for a future scientist. He
was endowed with an exceptional memory, and had a fascination with
mechanical toys which he retained all his life. At 14 he presented a paper to the
Royal Society of Edinburgh on ovals. At 16 he attended the University of
Edinburgh where the library still holds records of the books he borrowed while
still an undergraduate they include works by Cauchy on differential
equations, Fourier on the theory of heat, Newton on optics, Poisson on
mechanics and Taylors scientific memoirs. In 1850 he moved to Trinity
College, Cambridge, where he graduated with a degree in mathematics in 1854.
Maxwell was edged out of first place in their final examinations by his
classmate Edward Routh, who was also an excellent mathematician.

10A.44
Maxwell stayed at Trinity where, in 1855, he formulated a theory of three
primary colour-perceptions for the human perception of colour. In 1855 and
1856 he read papers to the Cambridge Philosophical Society On Faradays
Lines of Force in which he showed how a few relatively simple mathematical
equations could express the behaviour of electric and magnetic fields.
In 1856 he became Professor of Natural Philosophy at Aberdeen, Scotland, and

started to study the rings of Saturn. In 1857 he showed that stability could be
achieved only if the rings consisted of numerous small solid particles, an
explanation now confirmed by the Voyager spacecraft.
In 1860 Maxwell moved to Kings College in London. In 1861 he created the

first colour photograph of a Scottish tartan ribbon and was elected to the
Royal Society. In 1862 he calculated that the speed of propagation of an
electromagnetic wave is approximately that of the speed of light:
We can scarcely avoid the conclusion that light consists in the transverse
undulations of the same medium which is the cause of electric and magnetic
phenomena.
Maxwells famous account, A Dynamical Theory of the Electromagnetic
All the mathematical
sciences are Field was read before a largely perplexed Royal Society in 1864. Here he
founded on relations
between physical brought forth, for the first time, the equations which comprise the basic laws of
laws and laws of
numbers, so that the electromagnetism.
aim of exact science
is to reduce the Maxwell also continued work he had begun at Aberdeen, on the kinetic theory
problems of nature
to the determination of gases (he had first considered the problem while studying the rings of
of quantities by
operations with Saturn). In 1866 he formulated, independently of Ludwig Boltzmann, the
numbers. James kinetic theory of gases, which showed that temperature and heat involved only
Clerk Maxwell
molecular motion.

10A.45
Maxwell was the first to publish an analysis of the effect of a capacitor in a

circuit containing inductance, resistance and a sinusoidal voltage source, and to
show the conditions for resonance. The way in which he came to solve this
problem makes an interesting story:
Maxwell was spending an evening with Sir William Grove who was then
engaged in experiments on vacuum tube discharges. He used an induction
coil for this purpose, and found the if he put a capacitor in series with the
primary coil he could get much larger sparks. He could not see why. Grove
knew that Maxwell was a splendid mathematician, and that he also had
mastered the science of electricity, especially the theoretical art of it, and so
he thought he would ask this young man [Maxwell was 37] for an
explanation. Maxwell, who had not had very much experience in
experimental electricity at that time, was at a loss. But he spent that night in
working over his problem, and the next morning he wrote a letter to Sir
William Grove explaining the whole theory of the capacitor in series
connection with a coil. It is wonderful what a genius can do in one night!
Maxwells letter, which began with the sentence, Since our conversation
yesterday on your experiment on magneto-electric induction, I have considered
it mathematically, and now send you the result, was dated March 27, 1868.
Preliminary to the mathematical treatment, Maxwell gave in this letter an
unusually clear exposition of the analogy existing between certain electrical
and mechanical effects. In the postscript, or appendix, he gave the
mathematical theory of the experiment. Using different, but equivalent,
symbols, he derived and solved the now familiar expression for the current i in
such a circuit:
di 1
L Ri idt V sin t
dt C
The solution for the current amplitude of the resulting sinusoid, in the steady-
state is:
V
I
2
1
R L
2

C
from which Maxwell pointed out that the current would be a maximum when:
1
L
C

10A.46
Following Maxwell, Heinrich Hertz later showed a thorough acquaintance with
electrical resonance and made good use of it in his experimental apparatus that
proved the existence of electromagnetic waves, as predicted by Maxwells
equations. In the first of his series of papers describing his experiment, On
Very Rapid Electric Oscillations, published in 1887, he devotes one section to
a discussion of Resonance Phenomena and published the first electrical
resonance curve:
The first electrical

resonance curves
published, by Hertz,
1887
When creating his standard for electrical resistance, Maxwell wanted to design
a governor to keep a coil spinning at a constant rate. He made the system stable
by using the idea of negative feedback. It was known for some time that the
governor was essentially a centrifugal pendulum, which sometimes exhibited
hunting about a set point that is, the governor would oscillate about an
equilibrium position until limited in amplitude by the throttle valve or the
travel allowed to the bobs. This problem was solved by Airy in 1840 by fitting
a damping disc to the governor. It was then possible to minimize speed
j
fluctuations by adjusting the controller gain. But as the gain was increased,
the governors would burst into oscillation again. In 1868, Maxwell published

s -plane
his paper On Governors in which he derived the equations of motion of
engines fitted with governors of various types, damped in several ways, and
explained in mathematical terms the source of the oscillation. He was also able
to set bounds on the parameters of the system that would ensure stable
operation. He posed the problem for more complicated control systems, but
thought that a general solution was insoluble. It was left to Routh some years
later to solve the general problem of linear system stability: It has recently
come to my attention that my good friend James Clerk Maxwell has had
difficulty with a rather trivial problem.

10A.47
In 1870 Maxwell published his textbook Theory of Heat. The following year he
returned to Cambridge to be the first Cavendish Professor of Physics he
designed the Cavendish laboratory and helped set it up.
The four partial differential equations describing electromagnetism, now

known as Maxwells equations, first appeared in fully developed form in his
Treatise on Electricity and Magnetism in 1873. The significance of the work
was not immediately grasped, mainly because an understanding of the atomic
nature of electromagnetism was not yet at hand.
The Cavendish laboratory was opened in 1874, and Maxwell spent the next 5
years editing Henry Cavendishs papers.
Maxwell died of abdominal cancer, in 1879, at the age of forty-eight. At his

death, Maxwells reputation was uncertain. He was recognised to have been an
exceptional scientist, but his theory of electromagnetism remained to be
convincingly demonstrated. In 1888 Hertz performed a series of experiments
which produced and measured electromagnetic waves and showed how they
behaved like light. Thereafter, Maxwells reputation continued to grow, and he
may be said to have prepared the way for twentieth-century physics.
References
Blanchard, J.: The History of Electrical Resonance, Bell System Technical
Journal, Vol. 20 (4), p. 415, 1941.

10B.1
Lecture 10B Second-Order Op-Amp Filters

Filter design parameters. The lowpass biquad circuit. The universal biquad
circuit. Approximating the ideal lowpass filter. The Butterworth lowpass filter.
Introduction
With the advent of op-amps and circuit miniaturization, engineers developed
what is known as a universal filter. Its frequency response takes the form of a
biquadratic equation, and so it is also known as a biquad. Depending on the
connections made and the point at which the output is taken, the universal filter
can deliver lowpass, highpass, bandpass, bandstop (notch) and allpass
responses. It is one of the most useful circuits to the electrical engineer and is
widely available.
Filter Design Parameters

Jargon fills a special need for the engineer. It is shorthand that permits the
expression of ideas quickly and compactly. Design a lowpass filter with an
0 of 10,000 and a Q0 of 5. We will explore how this can be carried out.
We begin with the RLC circuit shown below, which has the now familiar form
of a voltage-divider circuit.
R jL
1
Vi Vo
jC
Figure 10B.1

10B.2
The frequency response for this circuit is given by:
T j
Vo 1 LC

Vi 1 LC 2 jR L (10B.1)
This result can be put into a standard form by noting that 0 1 LC and
Q0 0 L R for the series RLC circuit. We can then write:
02
T j 2
0 2 j 0 Q0 (10B.2)
0 and Q0 The specification of the two parameters 0 and Q0 uniquely specify the
uniquely specify a
second-order standard form of the second order frequency response.
frequency response
The historical identification of Q0 with RLC circuits is no longer appropriate,
since we can identify many kinds of circuits with the parameter Q0 . We can
now make the association of 0 and Q0 with any second-order circuit, as
suggested by the figure below:
R jL
1 0
Vi Vo
jC Vi
Q0
Vo
any second-order circuit standard form
Figure 10B.2

10B.3
The Lowpass Biquad Circuit

The standard form of a lowpass second order frequency response, as in
Eq. (10B.2), does not recognise the availability of gain that is possible with
active circuits. Also, an active circuit may be inverting or non-inverting. A
more general form for T j is therefore:
H02 Standard form of a

T j 2 lowpass second-
0 2 j 0 Q0 (10B.3)
order frequency
response with gain
We seek a circuit that will implement this second-order frequency response, as

well as any other biquadratic frequency response. (A biquadratic, or biquad
frequency response is similar to the way a bilinear frequency response was
defined a biquadratic function is a ratio of second-order polynomials).
Normalising so that 0 1 , and anticipating an inverting realisation for the

frequency response, we have:
Standard form of a
H
T j
Vo normalised lowpass

1 2 j 1 Q0 Vi
second-order
frequency response
(10B.4) with gain
We can manipulate this equation so that it has a form that can be identified
with simple circuits we have seen before. We first rewrite Eq. (10B.4) as:
1 j j 1 Q0 Vo HVi (10B.5)
Dividing by j j 1 Q0 , it becomes:
1 H
1 o
V Vi
j j 1 Q
0 j j 1 Q0 (10B.6)

10B.4
We can manipulate further to form:
Second-order
1 H 1
Vo Vo Vi 1
frequency response

made from first-
order parts j j j 1 Q0 (10B.7)
The (1) term can be realised by an inverting circuit of gain 1. The factor
1 j 1 Q0 is realised by a lossy inverting integrator. Two operations are
indicated by the remaining factor. The circuit realisation must produce a sum
of voltages, and it must have a frequency response of the form 1 j .
The three circuits that provide for these three operations are shown below:
The three first-order
circuits that make a
second-order circuit 1
1
1
1
vo
1
vi 1 Q0
1/H
Figure 10B.3

10B.5
If we connect the three circuits together, including a feedback connection of the

output vo to the input, the result is a scaled version of the Tow-Thomas biquad
circuit:
The normalised
Tow-Thomas biquad
circuit
1
1 1
1/H
vi 1 Q0
1
vo
Figure 10B.4
There are many circuits that implement biquadratic frequency responses. The
Tow-Thomas circuit is one of them, the Kerwin-Huelsman-Newcomb (KHN)
circuit is another. For brevity, we will simply refer to the Tow-Thomas biquad
circuit as the biquad.
With the elements identified by Rs and Cs, we have:

The Tow-Thomas
biquad circuit
R5
C1 C2
R2
R1
vi R2 R4
R3
vo
Figure 10B.5

10B.6
Show that the frequency response is:
1 R1R3C1C2
The biquads
T j
1 R3 R5C1C2 2 j 1 R4C2
frequency response
(10B.8)
Comparing this with Eq. (10B.3), we have:
1
0
R3 R5C1C2 (10B.9a)
R42C2
Q0
The biquads design
equations
(10B.9b)
R3 R5C1
R5
H (10b.9c)
R1
The biquad can be

An important property of the biquad is that it can be orthogonally tuned. Using
orthogonally tuned the above equations, we can devise a tuning algorithm:
1. R3 can be adjusted to the specified value of 0 .
2. R4 can then be adjusted to give the specified value of Q0 without

The biquads tuning
algorithm changing 0 .
3. R1 can then be adjusted to give the specified value of H without

affecting either 0 or Q0 .
Other advantages of the circuit are:
the input impedance is purely resistive
there is effectively pre-amplification built-in to the topology via the

gain setting resistor R1 (the incoming signal amplitude is amplified and
then filtered, which eliminates more noise than filtering and then
amplifying).

10B.7
Example
We require a circuit that will provide an 0 1000 rads-1 , a Q0 0.866 and a
DC gain of H 2 . We set 0 1 and use the biquad circuit of Figure 10B.4

with the values of Q0 and H specified above.
We then perform frequency scaling to meet the specifications, by setting

k f 1000 . The biquad circuit then becomes:
1 mF 1 mF
1
0.5 0.866
vi 1
1
vo
Figure 10B.6
We then select km 10 000 to give convenient element values. A realistic

circuit that meets the specifications is then:
10 k
100 nF 100 nF
10 k
5 k
vi 10 k 10 k 8.66 k
vo
Figure 10B.7

10B.8
The Universal Biquad Circuit

By applying a feedforward scheme to the lowpass Tow-Thomas biquad circuit,
A universal filter can
implement any a universal filter can be implemented. A universal filter is one that can be
biquadratic
frequency response made either a lowpass, highpass, bandpass, notch or allpass filter by
appropriate selection of component values.
The normalised
Tow-Thomas 1
universal filter
1 1
1 Q0
1
vo
R1 R2 C3
vi
Figure 10B.8
In terms of the quantities in Figure 10B.8, show that:
Vo 1 R1 C3 2 j 1 R2

Vi 1 2 j 1 Q0 (10B.10)
If we choose C 3 1 and R1 R2 , then the first and third terms in the
numerator vanish, leaving only the 2 term. Writing this result we have:
The universal
biquad circuit can
2
implement a
T j
1 2 j 1 Q0
highpass second-
order frequency (10B.11)
response
which means we have now created a highpass filter.

10B.9
Starting with the universal biquad circuit, it is possible to realise a lowpass, The universal
biquad can
highpass, bandpass, bandstop or allpass filter by making simple changes such implement many
second-order
as the removal of a resistor. The normalised design values for the various frequency
responses
responses are given in the table below, where H is the passband gain.
TABLE 10B.1 Design Values for the Tow-Thomas Universal Filter

Design Values Table of design
Filter Type R1 R2 C3 values for a
universal filter
Lowpass 1H 0
Bandpass Q0 H 0
Highpass H
Notch 0 n H
2
H
Allpass 1H Q0 H H
An example of a commercially available universal filter is the UAF42 from

Texas Instruments. This uses the state-variable biquad topology rather than
the Tow-Thomas. It requires just four external resistors to set the filter
parameters, and makes available lowpass, highpass and bandpass outputs:
The UAF42
universal filter from
HP Out BP Out LP Out Texas Instruments
R F1 R F2
12 13 8 7 14 1
R1
50 k
R2 C1 C2
50 k
2
1 nF 1 nF
50 k
RG A1 A2 A3
v i
3
R4
50 k
RQ
UAF42
11
Figure 10B.9

10B.10
Approximating the Ideal Lowpass Filter

The ideal lowpass filter characteristic is the brick wall filter. Although we
cannot achieve the ideal, it provides a basis on which we can rate an
The features we approximation. We want T to be as constant as possible in the passband, so
want when
approximating the that different low frequency signals are passed through the filter without a
ideal lowpass filter
change in amplitude. In the stopband we require n-pole rolloff, where n is
large, in contrast to the n 2 rolloff for the biquad circuit. We want the
transition from passband to stopband to be as abrupt as possible.
This is summarised in the figure below:
Approximating the
ideal lowpass filter
|T| Small error
1 Brick wall
n-pole rolloff
0
0 1
Figure 10B.10

10B.11
The method we will use in the approach to this problem is illustrated below:
We achieve the
approximation to the
|T| ideal lowpass filter
by cascading
1.5 |T1 |
Brick wall
1.0 |T2 |
|T3 |
0.5
|T1 ||T2 ||T3 | product

0
0 0.5 1.0 1.5
Vi T1 T2 T3 Vo
Figure 10B.11
We will connect modules in cascade such that the overall frequency response is
of the form given in Figure 10B.10. For the example in Figure 10B.11, large
The cascaded
values of T1 are just overcome by the small values of T2 and T3 to achieve circuits have the
same 0 but
the approximation to the brick wall. The frequency responses have the same
different Q0
value of 0 , but different values of Q0 . Determining the required values of Q0
is a part of filter design.

10B.12
The Butterworth Lowpass Filter

The Butterworth1 response is the name given to the following magnitude
function:
Tn j
The Butterworth 1
magnitude response
1 0
defined 2n
(10B.12)
Normalising such that 0 1 gives:
The normalised
Tn j
Butterworth
1
1 2n
magnitude response
(10B.13)
From this equation we can observe some interesting properties of the

Butterworth response:
1. Tn j 0 1 for all n.
1. Tn j1 1 2 0.707 for all n.

Properties of the
normalised
Butterworth
magnitude response 2. For large , Tn j exhibits n-pole rolloff.
3. Tn j has all derivatives but one equal to zero near 0 . The response
is known as maximally flat.
1
Stephen Butterworth was a British engineer who described this type of response in connection
with electronic amplifiers in his paper On the Theory of Filter Amplifiers, Wireless Eng.,
vol. 7, 1930, pp. 536-541.

10B.13
These properties are shown below:
Butterworth lowpass
magnitude
1.0 responses
0.9
0.8
0.7
0.6
n =1
|Tn ( j )| 0.5
0.4
2
0.3
6 4
0.2
8
0.1 10
0
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0

Figure 10B.12
The Butterworth response is well-known and has been studied extensively.

Since all normalised Butterworth responses have 0 1 , we can draw up a
table of Q0 values for each value of n:
TABLE 10B.2 Q0 of Butterworth Poles
n
2 3 4 5 6 7 8 9 10
Table of Q0 for
0.707 1.000 0.541 0.618 0.518 0.555 0.510 0.532 0.506
1.307 1.618 0.707 0.802 0.601 0.653 0.561 Butterworth poles
1.932 2.247 0.900 1.000 0.707
2.563 2.879 1.101
3.196
Note: For n odd, there is an additional first-order factor with a cutoff frequency
at 0 1 .

10B.14
Example
We are required to implement a 5th-order Butterworth lowpass filter with an

0 1248 rads -1 .
From Table 10B.2, for n 5 the required values of Q0 are 0.618 and 1.618.
There is also a first-order term with a cutoff frequency at 0 .
The realisation to meet these specifications can be shown in block diagram

form:
cutoff at 0 =1248 0 =1248

Vi Vo
0 =1248 Q0 =0.618 Q0 =1.618
1st-order 2nd-order 2nd-order
Figure 10B.13
Each second-order block could be realised using the biquad circuit. The first-
order circuit can be realised with a simple buffered RC circuit. If we use the
circuit of Figure 10B.4 for the biquad, then we must frequency scale using
k f 1248 . We then perform magnitude scaling to achieve practical element
values.
The Butterworth response is not the only response that can be used to
approximate the ideal brickwall filter. Other common responses are the
Chebyshev response, inverse Chebyshev response and Cauer (or elliptic)
response. Each has their advantages and disadvantages.

10B.15
Summary
The universal biquad circuit is a ready-made module which provides a
variety of second-order frequency responses lowpass, highpass, bandpass,
notch or allpass.
The design of a biquad circuit is specified by just three parameters the

resonant frequency 0 , the resonant Q0 , and the passband gain H.
Butterworth filters are easily designed using a table of Q0 values and can
be implemented as a cascade of first-order and second-order circuits.
References
Huelsman, L. P.: Active and Passive Analog Filter Design, McGraw-Hill,
Singapore, 1993.
Sedra, A. & Smith, K.: Microelectronic Circuits, Saunders College Publishing,

Sydney, 1991.
Van Valkenburg, M. E.: Analog Filter Design, Holt-Saunders, Tokyo, 1982.

10B.16
Exercises
1.
R
Vo
2R R
Vi1
2R 1
Vi2
j C
(a) Show that:
1 1
Vo Vi1 Vi 2
jCR jCR
(b) Show that the use of this circuit in the lowpass biquad circuit permits us to
reduce the number of op-amps by one.
2.
(a) Design a lowpass second-order filter with 0 10 000 , Q0 5 and H 2
using 1 nF capacitors.
(b) Design a highpass second-order filter with 0 5 000 , Q0 8 and H 1

using 10 nF capacitors.

10B.17
3.
Derive the frequency response of the so-called Sallen-Key circuit:
2Q0
1 1
vi
vo
1
2Q0
This can be used instead of the biquad as a single op-amp lowpass second-
order circuit.
4.
Derive the frequency response of the so-called Friend circuit:
1
2Q0
2
4 Q0
1
vi
vo
1
2Q0
This can be used instead of the biquad as a single op-amp bandpass second-
order circuit.

11A.1
Lecture 11A Complex Frequency

Complex frequency. Damped sinusoidal forcing function. Generalized
impedance and admittance. Frequency response as a function of . Frequency
response as a function of .The complex-frequency plane. Visualization of the
frequency response from a pole-zero plot.
Introduction
We have analysed circuits with DC sources and sinusoidal sources DC can be
considered a special case of a sinusoidal source (one that has zero frequency).
We have also seen that sinusoids are made up of exponentials with imaginary
exponents, and that the exponential function plays a special and key role in the
determination of the natural response of circuits.
It is only natural, then, to try and extend our study of circuits to a more general
class of exponential functions specifically to complex exponential functions,
which then include DC, exponential, and sinusoidal functions as special cases.
The introduction of a complex exponential forcing function leads to a natural

definition of complex frequency. Analysis of circuits using the concept of
complex frequency leads to special insight into circuit behaviour. We will see
that complex frequency unifies concepts and results in many advantages from
a deeper insight into understanding circuit behaviour, to analytical techniques
associated with establishing the natural, forced and complete response of a
circuit, and eventually to intuition and circuit design.

11A.2
Complex Frequency
Consider an exponentially damped sinusoidal function, such as the voltage:
vt Vm et cost (11A.1)
(This is termed a damped function because is generally negative).
There are several special cases of this function.
When both and are zero, we have DC:
vt Vm cos V0 (11A.2)
When only is zero, we have a sinusoid:
vt Vm cost (11A.3)
If only is zero, we have an exponential voltage:
vt Vmet cos V0et (11A.4)
The form of this exponential voltage reminds us of the complex exponential

representation of a sinusoid:
V0e jt (11A.5)
The only difference is that the exponent is real in one case, and imaginary in
the other. Since t must be dimensionless, we define to be a frequency.
Perhaps we should generalise our complex exponential representation of a

sinusoid by including this new frequency, ? Thus, we define complex
frequency:
s j (11A.6)

11A.3
Now, suppose we declare a function of the form:
f t Kest (11A.7)
where K and s are complex constants (independent of time). We say that the
function is characterised by the complex frequency s . For example, a constant
voltage V0 may be declared in the required format as:
vt V0e0t (11A.8)
The complex frequency of a DC voltage or current is thus s 0 .
The exponential function:
vt V0et (11A.9)
is already in the required format, and we identify s j 0 .
A sinusoidal function cannot be expressed in the required format. Given:
vt Vm cost (11A.10)
we use Eulers identity to express it as:

vt 12 Vm e j t e j t (11A.11)
12 Vme j e jt 12 Vme j e jt
K1es1t K 2es2t
Therefore, we have the sum of two complex exponentials a pair of conjugate,

counter-rotating phasors. Therefore, two complex frequencies are present, one
for each term. The complex frequency of the first term is s s1 j and that
of the second term is s s 2 j . Thus, s1 s*2 and the two values of K are
also conjugate, K 1 12 Vm e j and K 2 K1* 12 Vm e j .

11A.4
Lastly, let us determine the complex frequency associated with the
exponentially damped sinusoidal function. We again use Eulers identity:
vt Vmet cost (11A.12)

12 Vmet e j t e j t
12 Vme j e j t 12 Vme j e j t
We therefore find once again that a conjugate complex pair of frequencies is

required to describe the exponentially damped sinusoid, s1 j and
s 2 s1* j . In general, neither nor is zero, and we see that the

exponentially varying sinusoidal waveform is the general case the constant,
exponential and sinusoidal waveforms are special cases.
As numerical illustrations, we should now recognize by inspection the complex

frequencies associated with these voltages:
(11A.13)
vt 100 s0
vt 5e 2t s 2 j 0
s1 j500
vt 2 sin500t
s 2 s1 j500
*
s1 3 j 6
vt 4e 3t sin6t 10
s 2 s1 3 j 6
*

11A.5
The reverse should also be true given a complex frequency or a pair of

conjugate complex frequencies, we should be able to identify the nature of the
function with which they are associated. Thus, the most special case s 0
defines a constant:
s0 f t Ke0t constant (11A.14)
where K must be real if the function is to be real.
A negative real value, such as s 5 j 0 , defines a decaying exponential:
s 5 j 0 f t Ke 5t (11A.15)
where again K must be real if the function is to be real.
A purely imaginary value of s , such as s j10 , can never be associated with a

real quantity. In order to construct a real function, it is necessary to consider
conjugate values of s , such as s1, 2 j10 , with which must be associated
conjugate values of K . Loosely speaking, however, we may identify either of

the complex frequencies s j10 or s j10 with a sinusoidal voltage at the
radian frequency 10 rad/s. The presence of the conjugate complex frequency is
understood. Thus:
s1 j10 (11A.16)
f t A cos10t
s 2 j10
In a similar manner, a general value for s , such as s 3 j 5 , can be

associated with a real quantity only if it is accompanied by its conjugate:
s1 3 j5 (11A.17)
f t Ae 3t
cos5t
s 2 3 j5

11A.6
In general, the complex frequency s describes an exponentially varying
sinusoid. The real part of s is associated with the exponential variation if it is
negative, the function decays to zero as t increases, if positive, the function
increases, and if it is zero, the sinusoidal amplitude is constant. The imaginary
part of s describes the sinusoidal variation it is specifically the radian
frequency.
The Damped Sinusoidal Forcing Function

The general exponentially varying sinusoidal voltage:
vt Vm et cost (11A.18)
is expressible in terms of the complex frequency s by making use of Eulers

identity:

vt Re Vmet e j t (11A.19)
Collecting factors:

vt Re Vme j e j t (11A.20)
we now substitute s j , and obtain:

vt Re Vme j est (11A.21)
We note the resemblance to the corresponding representation of the undamped

sinusoid:

vt Re Vme j e jt (11A.22)
The only difference is that we now have s where we previously had j . Thus,
our approach will be to develop a frequency-domain description of the
exponentially varying sinusoid in exactly the same way as we did for the
undamped sinusoid: omit the Re notation and suppress est .

11A.7
Thus, if we apply the forcing function:
vt Vmet cost (11A.23)
or vt Re Vme j est
then we expect the forced response, say a current in some branch of the circuit,
to be:
it I met cost (11A.24)

or it Re I me j est
where the complex frequency of the source and the response must be identical.
We recall that the application of a complex source results in a complex

response, and that the real part of the source resulted in the real part of the

response. Thus, given the real forcing function vt Re Vme j est we apply the
complex forcing function Vm e j est . The resultant forced response, I m e j est is

complex and it must have as its real part the desired time-domain forced

response it Re I m e j est .

11A.8
Example
i (t ) 2 3H
v( t ) 0.1 F
We shall apply the forcing function:
vt 60e2t cos4t 10
and we desire the forced response:
it I me 2t cos4t
We first express the forcing function as the real part of a complex function:
vt 60e 2t cos4t 10

Re 60e 2t e j 4t 10
Re60e j10
e 2 j 4 t

or:

vt Re Vest
where:
V 6010 and s 2 j 4
After dropping the Re, we are left with the complex forcing function Vest . In a
similar manner, we represent the unknown response by the complex quantity
Iest where I I m .

11A.9
From KVL we obtain:
idt vt
di 1
dt C
Ri L
2i 3 10 idt vt
di
dt
and we substitute the given complex forcing function and the assumed complex
forced response:
10 st
2Iest 3sIest Ie 6010est
s
The common factor est is suppressed:
10
2I 3sI I 6010
s
and:
6010
I
2 3s 10 s
The left side of this equation is a current. On the right, the numerator is a
voltage, and so the denominator must be interpreted as an impedance.
We now let s 2 j 4 and solve for the complex current I :
6010
I
2 3 2 j 4 10 2 j 4
After manipulating the complex numbers, we find:
I 5.37 106.6
Thus, the forced response is:
i t 5.37e 2t cos4t 106.6

11A.10
Generalized Impedance and Admittance

In order to apply Kirchhoffs Laws directly to a circuit with a complex forcing
function, we need to determine the relationship between the complex voltage
across an element and the complex current through it, i.e. the impedance or
admittance of an element.
We will consider the inductor, and then state the relationships for the other
elements, since the derivations are similar.
The defining time-domain equation for the inductor is:
di t
vt L (11A.25)
dt
d st
Vest L Ie (11A.26)
dt
Vest sLIest (11A.27)
Suppressing est , we find the desired voltage-current relationship:
V sLI (11A.28)
Thus, the impedance of the inductor is:

Generalized
impedance of an (11A.29)
Zs sL
inductor V
I

11A.11
and the admittance is:
Ys
I 1
(11A.30)
V sL
We still call V and I phasors. These complex quantities have a magnitude
and angle which, along with a specific complex frequency value, enable us to
characterize the exponentially varying sinusoidal waveform completely. The
phasor is still a frequency-domain description, but we have now expanded its
definition to include complex frequency.
The frequency-domain equivalent of an inductor is shown below:
i I
di
v=L L V=s LI sL
dt
Figure 11A.1
In the table below, we show how we can represent a resistor, inductor or

capacitor in the time-domain with its frequency-domain impedance:
Time-domain Frequency-domain Generalized

impedances of the
three passive
R R elements
L sL
C 1
sC

11A.12
Example
If we now reconsider the series RLC example in the frequency domain, the
source voltage:
vt 60e 2t cos4t 10
is transformed to the phasor voltage:
V 6010
A phasor current I is assumed, and the impedance of each element at the

complex frequency s 2 j 4 is determined and placed on a frequency-
domain circuit diagram:
I 2 -6+ j 12
60 10 10 = -1- j 2
-2+ j 4
s =-2+ j 4
The phasor current is now easily found by dividing the phasor voltage by the
sum of the three impedances:
6010 6010
I 5.37 106.6 A
2 6 j12 1 j 2 5 j10
The forced response is then:
i t 5.37e 2t cos4t 106.6
Thus, the previous result is obtained, but much more easily and rapidly.

11A.13
Frequency Response as a Function of

Suppose we have a circuit which is excited by a single source, Vs Vs 0 ,
and that the desired response is a current I . This phasor response is a complex
number, and it can be represented by two numbers a magnitude and a phase.
If the source varies as a function of only radian frequency , then we can plot
the magnitude response as a function of and the phase response as a
function of .
Example
Suppose we have a series RL circuit:
I R
Vs sL
The phasor voltage Vs Vs 0 is applied, and the response is the phasor

current I . The forced response is:
Vs
I
R sL
If Vs represents a sinusoidal forcing function, then s j and we have:
Vs
I
R jL
The magnitude of the response is:
Vs
I
R 2 2 L2

11A.14
and the phase of the response is:
L
I tan 1
R
These are the analytical expressions for the magnitude and phase angle of the
response as a function of we can now present the same information
graphically.
For the magnitude curve, we note that we are plotting the absolute value of
some quantity and so the entire curve lies above the axis. The response is
Vs R at zero frequency, and the response approaches zero as frequency
approaches infinity. We graph both positive and negative values of frequency:
|I |
Vs / R magnitude of
Vs
I ( j ) = R+ j L
0
- 2LR -R
L
R
L
2R
L
The second part of the response, the phase angle of I versus , is an inverse
tangent function:
90
angle of
Vs
I ( j ) = R+ j L
0
- 2LR -R
L
R
L
2R
L
-90

11A.15
Example
Suppose we have a parallel LC circuit:
1
Is V sL sC
The voltage response is easily obtained:
V Is
sL 1 sC I sL
sL 1 s C s LC 1
s 2
For a sinusoidal forcing function, then s j and the circuit presents an

impedance to the current source of:
V j L 1
Z j
I s LC 1
2
C 1 LC
2
By letting 0 1 LC , we can express the magnitude of the impedance in a
form which enables those frequencies to be identified at which the response is

zero or infinite:
1
Z
C 0 0
Such frequencies are called critical frequencies, and their early identification
simplifies the construction of the response curve. We note that the response has
zero amplitude when 0 . When this happens we say that the response has a
zero at 0 , and we also describe the frequency at which it occurs as a zero.
A response of infinite amplitude is noted at 0 and 0 . These
frequencies are called poles, and the response is said to have a pole at each of
these frequencies. Finally, we note that the response approaches zero as
, and thus is also a zero (it is customary to consider plus
infinity and minus infinity as being the same point).

11A.16
The locations of the critical frequencies should be marked on the axis, by
using small circles for the zeros and crosses for the poles. The drawing of the
graph is made easier by adding broken vertical lines as asymptotes at each pole
location. The completed graph of magnitude versus is shown below:
| Z|
2
0C
1
0C
- 2 0 -0 0 0 2 0
The corresponding phase response is, from an inspection of the original

equation:
90
- 2 0 -0 0 0 2 0
-90

11A.17
Frequency Response as a Function of

By analogy with the response due to a varying , we now consider a response
due to a varying . Suppose we have a circuit which is excited by a single
source, Vs Vs 0 , and that the desired response is a current I . This phasor
response is a complex number, and it can be represented by two numbers a
magnitude and a phase. If the source varies as a function of only , then we
can plot the magnitude response as a function of and the phase response as a
function of .
Example
Suppose we have a series RL circuit:
I R
Vs sL
The phasor voltage Vs Vs 0 is applied, and the response is the phasor

current I . The forced response is:
Vs
I
R sL
We now set 0 , s j 0 , thus restricting ourselves to time-domain
sources of the form v s Vs et . Thus:
Vs
I
R L
As varies, we can obtain a qualitative description of the circuits behaviour.

When is large and negative, the current response is negative and relatively
small in amplitude. As increases, becoming a smaller negative number, the

11A.18
magnitude of the response increases, until we encounter R L , a pole of

the forced response, at which point the response is infinite. As continues to
increase, the next noteworthy point is 0 , where we have the case of a DC
source, v s Vs , and the forced response is obviously I Vs R . Positive values
of must all provide positive amplitude responses, and the response must
decrease as get larger. Finally, as , we have a zero-amplitude
response, and therefore a zero. The only critical frequencies are the pole at
R L and the zero at .
This information is plotted graphically below:
|I |
Vs
R
0
-R
L
The phase response, not shown, is either 0 or 180 .
In the example above, a finite critical frequency occurred at R L . That

is, when the circuit is excited at this frequency by the voltage:
vs Vs e Rt L (11A.31)
we get an infinite response. Why? The forcing function has a familiar form it
is the same form as the natural response of the circuit. We know that if we have
a source-free circuit, then the response due to an initial current would be:
i I 0e Rt L (11A.32)

11A.19
Thus, a zero-amplitude input produces a finite response. Since the circuit is
linear, we conclude that a non-zero-amplitude forcing function will produce an
infinite forced response.
This result is quite general:
When any circuit is excited at a frequency which is a pole of

the response, then an infinite response must result. (11A.33)
We also have:
The frequencies of the poles are directly related

to the natural response of the circuit. (11A.34)
Example
Suppose we have a series RC circuit, excited by an exponential current source

i s I s et :
i s I s e t

v C
The voltage across the source is:
1 1 RC
V ZI R I s RI s
C
We can now identify the poles and zeros of the response from the form of the
equation it indicates a pole at 0 and a zero at 1 RC .

11A.20
The response magnitude is plotted as a function of below:
|V |
RIs
0
- R1C
The reason for the pole at zero frequency may again be explained on physical
grounds. If I s 0 , then the current source is an open-circuit and the response
is the initial capacitor voltage. A non-zero I s must therefore produce an

infinite forced voltage response. In other words, if a constant current has been
applied to the circuit forever, then the capacitor voltage would be infinite.
Example
Suppose we have a series RLC circuit, excited by an exponential voltage

source:
i (t ) 6 1H
100 e t V 0.2 F
The current is easily found:
100
I 100
6 5 1 5

11A.21
The response curve is most easily found by first indicating the locations of all
poles and zeros on the axis, and placing vertical asymptotes at the poles. We
can then seek out relative minima and maxima, and make a sketch of the
response:
|I |
-5 -1 0
The two poles may again be used to construct the natural response:
in t A1e t A2e 5t
Compare this with the case of an overdamped series RLC circuit:
2
R R 1
s1, 2
2L 2L LC
3 3 2 5
1 and 5
Once again, at either of these two frequencies, a zero-amplitude forcing

function may be associated with a finite-amplitude response which can be
interpreted as the natural response of the circuit.

11A.22
The Complex-Frequency Plane

Now that we have considered the forced response of a circuit as varies (with
0 ) and as varies (with 0 ), we are prepared to develop a more
general graphical representation by graphing quantities as functions of s .
After we develop the concept of the complex-frequency plane, or s -plane, we

shall see how quickly the behaviour of a circuit can be obtained from a
graphical representation of its critical frequencies.
The response of a circuit to a complex frequency s is a complex function of s .

Thus, we need a method of representing both the magnitude and phase of the
response as s varies. Therefore, we represent the magnitude and phase
individually, using a three-dimensional model built over the s -plane, where the
height at any point is equal to the value of the quantity we are representing
either the magnitude or the phase.
The s -plane consists of a axis and a j axis, perpendicular to each other:
Im s
s= + j
j
0 Re s
complex plane
Figure 11A.2

11A.23
We are familiar with the type of time-domain function associated with a

particular value of the complex frequency s . It is now possible to associate the
functional form of a forcing function or forced response with the various
regions in the s -plane.
s -plane
Figure 11A.3
The origin, for example, must represent a DC quantity. Points lying on the
axis must represent exponential functions, decaying for 0 , increasing for
0 . Pure sinusoids are associated with points on the positive or negative
j axis. Points in the left half-plane (LHP) describe the frequencies of
exponentially decreasing (or damped) sinusoids. The RHP contains points
describing exponentially increasing sinusoids.
We may represent the magnitude of a function of s as a surface lying above (or

just touching) the s -plane.

11A.24
Example
Suppose we want to graph the magnitude of the admittance of the series

combination of a 1 H inductor and a 3 resistor:
Ys
1
s3
We identify the pole immediately as s 3 j 0 thus our 3D model of the

magnitude of Ys must have infinite height over this point. When s is
infinite, the magnitude of Ys must be zero. The model must have zero height
at points infinitely far away from the origin. A cut-away view of the magnitude
of Ys is shown below:
1
The graph of the
magnitude of Ys 0.8
over the s-plane

0.6
forms a surface with |Y|
poles and zeros
0.4
0.2
0
-6
10
-4 5
0
-2
-5
0 -10
j

We can completely specify Ys , apart from a constant gain factor, by drawing

a so-called pole-zero plot:
j
A pole-zero plot is a
shorthand way of
representing a -3

complex function
of s s -plane

11A.25
A pole-zero plot locates all the critical points in the s-plane that completely
specify the function Ys (to within an arbitrary constant), and it is a useful
analytic and design tool.
One cut of the surface has been fortuitously placed along the imaginary axis. If
we graph the height of the surface along this cut against , we get a picture of
the magnitude of the frequency response versus :
|Y| The frequency

1/3 magnitude of response is
obtained from the
1
Y( ) = 3+ j j-axis of a plot of a
function over the
entire s-plane
We can obtain a mental image of the surface that represents the magnitude of a
function over the s-plane quite quickly if we imagine a rubber sheet model. At
zeros, we secure the rubber sheet to the plane (tent pegs). At poles, we prop the
rubber sheet up with infinitely high and thin poles.

11A.26
Example
We are given a pole-zero plot of some impedance Zs .
j
-1+ j 5
-2

s -plane
-1- j 5
If we visualize a rubber-sheet model, tacked down at s 2 j 0 and propped

up at s 1 j5 and s 1 j5 , we should see a terrain with two mountains
and one valley. The portion of the model for the upper LHP is shown below:
0.8
0.6
|Z|
0.4
0.2
0
-10
10
8
-5 6
4
2
0 0
j

11A.27
We can build up the expression for Zs which leads to this pole-zero

configuration. The zero requires a factor of s 2 in the numerator, and the
two poles require the factors s 1 j 5 and s 1 j 5 in the denominator.
Except for a multiplying constant k, we now know the form of Zs :
s2
Zs k
s 1 j5s 1 j5
s2
k 2
s 2s 26
The plots Z versus and Z versus may be obtained exactly from
this expression, but the general form of the function is apparent from the pole-
zero configuration and the rubber-sheet analogy. Portions of these two curves
appear at the sides of the 3D model above.

11A.28
Visualization of the Frequency Response from a Pole-Zero Plot

The frequency response can be visualised in terms of the pole locations of the
response function. For example, for an underdamped second-order lowpass
system:
Standard form for a

02
lowpass second
Vo 2 Vi
s 2s 02
order response
function (11A.35)
the poles are located on a circle of radius 0 and at an angle with respect to the
negative real axis of cos 1 0 . These complex conjugate pole locations

are shown below:
Pole locations for an

underdamped
lowpass second- j
order response
function
p j 0
p* - j 0
Figure 11A.4
In terms of the poles shown in Figure 11A.4, the response function is:
02
Underdamped
lowpass second-
Vo
s p s p
order response

Vi
function using pole (11A.36)
factors

11A.29
When s j the two pole factors in this equation become:
j p m1 1 j p m2 2 (11A.37) Polar representation

and of the pole factors
In terms of these quantities, the magnitude response is:
Magnitude function
H j
Vo 1 written using the
polar representation
of the pole factors
Vi m1m2 (11A.38)
and the phase response is:

Phase function
H j 1 2
written using the
(11A.39) polar representation
of the pole factors
Vectors representing Eq. (11A.37) are shown below:
Determining the
magnitude and
j j j
phase response
m1 j 2
from the s-plane
m1
p p j 0 p
1
m1 1 1
j 1

m2 m2 m2
2 2 2
p* p* p*
|H| 0 H
1 0 2
2 0
1
-90
-180
0 1 0 2
Figure 11A.5

11A.30
Figure 11A.5 shows three different frequencies one below 0 , one at 0 , and
one above 0 . From this construction we can see that the short length of m1
near the frequency 0 is the reason why the magnitude function reaches a peak
near 0 . These plots are useful in visualising the frequency response of the
circuit.

11A.31
Summary
Complex frequency is defined as s j .
A whole class of forcing functions and responses (exponentially damped

sinusoids) can be expressed using f t Kest , where K is a complex
constant. If s is complex, then a real function must be composed of a
conjugate pair: f t Ke st K * e s t .
*
A complex forcing function for a voltage can be expressed as Vest , where

V Vm is a phasor with the complex frequency s.
The generalised impedances of the three passive elements are:
1
ZR R , Z L sL , ZC
sC
Critical frequencies of a forced response occur at zeros and poles. A zero is

a frequency at which the response is zero. A pole is a frequency at which
the response is infinite.
The poles of a circuit determine the form of its natural response.
The frequency response of a circuit can be evaluated at each point in the

s-plane. The response along the j axis represents the sinusoidal steady-
state frequency response.
A pole-zero plot encapsulates all of the information about a circuits

response, apart from a multiplicative constant. We can also use a pole-zero
plot to visualize the frequency response.
References
Hill, 1984.

11A.32
Exercises
1.
Find the complex frequencies associated with the natural response of a source-
free series RLC circuit in which L 10 H , C 250 F , and R :
(a) 500 (b) 320
2.
(a) If i1 t 2e 100t cos1000t 35 mA and
i2 t 4e 100t sin 1000t 20 mA , find the phasor representing the sum

of these two currents.
(b) Find the phasor corresponding to it 2e 5t 3 cos 50t 2 sin 50t mA .
3.
1 1H i (t )
vs 1F 1H
Find i t at t 0.5 s if:
(a) v s 10 V
(b) v s 10 cos2t V
(c) v s 10e 3t V
(d) v s 10e 3t cos2t V

11A.33
4.
vs 2H 20 is
If v s 20e 4t cos 3t V , while i s 3e 3t cos 4t A , find the energy stored in the
inductor at t 0 .
5.
0.1 F
is v 2 k 1 k 0.2 F
Find all the critical frequencies of the ratio V I s , and graph the magnitude of
the ratio as a function of (use a spreadsheet).
6.
One of the critical frequencies of the impedance of a series RLC circuit occurs
at s 4 j10 s -1 . If the minimum impedance magnitude is 100 ,
determine:
(a) R (b) L (c) C

11A.34
7.
The pole-zero plot of an input impedance displays a zero at s 50 and poles
at s 20 j 30 .
(a) Determine the magnitude and angle of each vector from a critical frequency
to the point s j 20 .
(b) Calculate the (complex) ratio of Z in j 20 to Z in j 40 .
8.
In 1945, Dr Hendrik W. Bode published Network Analysis and Feedback
Amplifier Design, codifying in one classic book the filter and feedback-
amplifier theory upon which much of the electronics industry still relies.
One of the many results his seminal work revealed was the constant-resistance
network:
R1 R2
L1 C2
Zin
The figure shows only one of many forms of this circuit.
Show that, as long as you let R1 R2 and scale the components such that the
time constant L1 R1 equals the time constant R2 C 2 , then the impedance of the
whole circuit remains constant at all frequencies, and is equal to R1 .
The circuit occasionally sees application in digital systems as a terminating

network, where C 2 represents the unavoidable capacitance of a logic gate
input.

11B.1
Lecture 11B Specialty Amplifiers

Differential and common-mode signals. Difference amplifiers. Instrumentation
amplifiers. Programmable gain amplifiers. Isolation amplifiers.
Introduction
There are several popular types of specialty amplifiers, or amplifiers that are
based in some way on op-amp techniques. In an overall application sense, they
are not generally used as universally as op-amps. Examples of specialty
amplifiers include difference amplifiers, instrumentation amplifiers,
programmable gain amplifiers (PGAs) and isolation amplifiers. These will be
looked at briefly because they are used in the important areas of data
acquisition and distribution systems and embedded systems.
In addition, there are many other types of amplifiers such as audio and video
amplifiers, cable drivers, high-speed variable gain amplifiers and various
communication-related amplifiers.

11B.2
Differential and Common-Mode Signals

A difference amplifier is one that responds to the difference between the two
signals applied at its input and ideally rejects signals that are common to the
two inputs. The representation of signals in terms of their differential and
common-mode components is given below:
vi1 = vicm vid /2
vid /2
vid = vi2 vi1
vicm = 12 (vi1 + vi2)
vicm vid /2
vi2 = vicm + vid /2
Figure 11B.1 Common-Mode and Differential Signals
Although the ideal difference amplifier will amplify only the differential input
signal vid and reject completely the common-mode signal vicm , practical
circuits will have an output voltage vo given by:
vo Ad vid Acm vicm (11B.1)
where Ad denotes the amplifier differential gain and Acm denotes its common-
mode gain (ideally zero). One measure of a differential amplifiers
performance is the degree of its rejection of common-mode signals in
preference to differential signals. This is usually quantified by a measure
known as the common-mode rejection ratio (CMRR), defined as:
Ad
CMRR 20 log10 dB
Acm (11B.2)

11B.3
Difference Amplifiers
A simple difference amplifier can be constructed with four resistors and an
op-amp, as shown below:
Difference amplifier
R2
R1
v i1
R3
v i2
vo
R4
Figure 11B.2 A Difference Amplifier
It is often used in applications where a simple differential to single-ended

conversion is required. Because of its popularity, this circuit will be examined
in more detail, in order to understand its fundamental limitations before
discussing instrumentation amplifiers.
We can use superposition to determine the output:
Analyzing a
R2 R2 using superposition
R1 R1
v i1
vo1 R3 vo2
v i2
R3 R4 R4
(a) (b)
Figure 11B.3 Analyzing a Difference Amplifier

11B.4
We have:
R2
vo1 vi1
R1 (11B.3)
and, since we normally select R4 R3 R2 R1 , we also have:
R R4 R
vo 2 1 2 vi 2 2 vi 2
R1 R3 R4 R1 (11B.4)
Thus:
The differential gain

of a difference
amplifier vo
R2
vi 2 vi1 R2 vid Ad vid
R1 R1 (11B.5)
Thus, for exact resistor ratios R4 R3 R2 R1 , the common-mode gain is zero.
Any mismatch, however, will lead to finite CMRR.
The differential resistance of the amplifier is calculated as shown below:

Determining the
input resistance of a
difference amplifier R2
ii R1
v id R1
ii
R id R2 Virtual short circuit
Figure 11B.4 Determining the Input Resistance of a Difference Amp
Thus, the input resistance is:

of a difference
amplifier Rid 2 R1 (11B.6)

11B.5
Difference Amplifier Deficiencies
If the amplifier is required to have a large differential gain R2 R1 , then R1 will

be relatively small and the input resistance will be correspondingly low, a
drawback of this circuit.
Another drawback is that it is not easy to vary the differential gain, since we
need to maintain the resistors in the ratio R4 R3 R2 R1 .
It is also extremely sensitive to source impedance imbalance (the two inputs

see different input impedances, so a balanced source is not necessarily a good
thing).
A major difference
A net matching tolerance of 0.1% in the resistor ratios yields a worst case DC amplifier deficiency
is due to inevitable
CMRR of 66 dB, which is quite low (compared to the op-amp which typically unequal resistor
ratios
has a CMRR of 100 dB).
Difference Amplifier ICs
To overcome the matching tolerance on the resistors, it is usual to seek out an

integrated circuit difference amplifier which includes on-chip laser trimmed
precision thin film resistor networks. An example is the INA146 from Texas
Instruments:
The Texas
Instruments INA146,
R G1 R G2 a difference
amplifier with laser-
V+ trimmed resistors
7 5 a very flexible IC!
R1 R2
2
100 k 10 k
v IN
R5
10 k 6 vo
(1%) A2
A1
R3 R4
3
100 k 10 k
v+ IN
INA146
4 1 8
V- vo 1
Figure 11B.5 The Texas Instruments INA146

11B.6
Instrumentation Amplifiers
The most popular among all of the specialty amplifiers is the instrumentation
amplifier (or simply in-amp). The in-amp is widely used in many industrial and
measurement applications where DC precision and gain accuracy must be
maintained within a noisy environment, and where large common-mode signals
(usually at the AC power line frequency) are present.
The internal configuration of an in-amp is shown below:

A 3 op-amp
instrumentation
amplifier (in-amp)
v i1
A1
R4
R2 R3
RG A3
R2 R3 vo
R4
A2
v i2
Figure 11B.6 A 3 Op-Amp Instrumentation Amplifier
It has a pair of differential input terminals, and a single-ended output that

works with respect to a reference or common terminal. The input impedances
are balanced and high in value, typically greater than 1 G . The in-amp uses
an internal feedback resistor network, plus one gain set resistance, RG , which
is external to the package.

11B.7
An analysis of the in-amp is shown below:

Analysis of the
in-amp to determine
the differential gain
v i1
vo1
0V A1
R4
v i1 0
v id /R G R 2 v id /R G
R3
(v i2 v i1) 2R2
RG v id (1+ RG ) A3
= v id
R 2 v id /R G R3 v o = R 4 (1+ 2R 2 ) v id
v i2 0 R4 R3 RG
0V A2
vo2
v i2
Figure 11B.7 Analysis of the In-Amp
Thus, the in-amp gain can be set by the user by selecting the external RG .
In-amp gain can also be preset via an internal RG by pin selection (again
isolated from the signal inputs). Typical in-amp gains range from 1 to 1000.
Furthermore, common-mode signals are only amplified in the first stage by a

factor of 1 regardless of gain. When the two input terminals are connected
together to a common-mode input voltage vicm , it is easy to see that an equal
The in-amp has a
voltage appears at the negative input terminals of A1 and A2 , causing the very high CMRR
current through RG to be zero. Thus there will be no current in the R2
resistors, and the voltages at the output terminals of A1 and A2 will be equal to
the input (i.e. vicm ). Thus the first stage does not amplify vicm , it simply
propagates vicm to its two output terminals where they are subtracted to
produce a zero common-mode output by A3 .
Another important feature of the in-amp is that the internal resistance network
and RG are isolated from the signal input terminals.

11B.8
In-Amp Application
An example of a commercially available in-amp is the AD620. An application

of the in-amp is shown below, where it is used in a data acquisition system to
amplify the output from a pressure transducer bridge (which is normally
balanced) from a single 5 V supply:
Typical application
of an in-amp (the
AD620 from Analog +5 V
Devices) as an
amplifier for a
pressure transducer 3 7
3 k 3 k 20 k
8
REF
G=100
AD620B 6 IN
499
3 k 3 k 1 5 DIGITAL
10 k ADC DATA
2 4 OUTPUT
AD705 AGND
20 k
1.7 mA 1.3 mA 0.1 mA 0.6 mA

MAX MAX
Figure 11B.8 A Pressure Monitor Circuit
The in-amp is ideally suited for this application because the bridge output is
fundamentally balanced, and the in-amp presents it with a truly balanced high
impedance load.
Full scale output voltages from a typical bridge circuit can range from
approximately 10 mV to several hundred mV. Typical in-amp gains in the
order of 100 to 1000 are therefore ideally suited for amplifying these small
voltages to levels compatible with popular analog-to-digital (ADC) input
voltage ranges (usually 1 V to 10 V full scale).
In addition, the in-amps high CMRR at power line frequencies allows

common-mode noise to be rejected, when the bridge must be located remotely
from the in-amp.

11B.9
Programmable Gain Amplifiers

Most data acquisition systems with wide dynamic range need some method of
adjusting the input signal level to the analog-to-digital converter (ADC). A PGA is used to
increase the
Typical ADC full scale input voltage ranges lie between 1 V and 10 V. To dynamic range of a
system
achieve the rated precision of the converter, the maximum input signal should
be fairly near its full scale voltage.
Transducers however, have a very wide range of output voltages. High gain is
needed for a small sensor voltage, but with a large output, a high gain will
cause the amplifier or ADC to saturate. So, some type of predictably
controllable gain device is needed.
Such a device has a gain that is controlled by a digital input. This device is
known as a programmable gain amplifier, or PGA. Typical PGAs may be
configured either for selectable decade gains such as 10, 100, 1000, etc., or
they might also be configured for binary gains such as 1, 2, 4, 8, etc.
A PGA is usually located between a sensor and its ADC, as shown below:
A PGAs location in
a system
gain
control
sensor
digital
PGA ADC output
Figure 11B.9 Use of a Programmable Gain Amplifier
Thus, a digital processor can combine PGA gain information with the digital
output of the ADC to increase its resolution. Some ADCs have on-chip PGAs.

11B.10
PGA Design Issues
How to switch the gain
Effects of the switch on-resistance ( RON )
Gain accuracy
Gain linearity
Bandwidth
DC offset
Gain and offset drift over temperature
Settling time after switching gain
PGA Example
An example of a PGA with programmable gains of 1, 10, 100 and 1000 is
shown below:
A PGA with a
decade scale that is
not influenced by 100 +15 V
switch resistance v in
AD8610 v out
20 pF
-15 V
1 12 +5 V
1 k
2 3 G=1
8 13 +15 V 10 k
0-5 V GAIN 7 6 G=10
CONTROL
9 4 -15 V 1 k
10 11 G= 100
16 5 100
15 14 G 1000
ADG1412 11
Figure 11B.10 A PGA with Decade Gains

11B.11
Isolation Amplifiers
There are many applications where is it desirable, or even essential, for a
sensor to have no direct (galvanic) electrical connection with the system to Electrical isolation is
necessary in a wide
which it is supplying data. This might be in order to avoid the possibility of variety of
dangerous voltages or currents from one half of the system doing damage in applications
the other. Such a system is said to be isolated, and the arrangement that passes
a signal without galvanic connections is known as an isolation barrier.
Examples include the need to prevent the ignition of explosive gases by sparks
at sensors and the protection from electric shock of patients whose ECG or
EEG is being monitored. In the ECG case, protection may be required in both
directions: the patient must be protected from accidental electric shock, but if
the patients heart should stop, the ECG machine must be protected from the
very high voltages (> 7.5 kV) applied to the patient by the defibrillator which
will be used to attempt to restart it.
A summary of applications for isolation amplifiers (both analog and digital) is

given below:
Sensor is at a high voltage relative to other circuitry (or may become so
under fault conditions)
Sensor may not carry dangerous voltages, irrespective of faults in other
circuitry (e.g. patient monitoring and intrinsically safe equipment for
use with explosive gases)
To break ground loops (two grounds in the same system at different
voltages, causing current between them)
The most common isolation amplifiers use transformers, which exploit

magnetic fields, and another common type uses small high voltage capacitors,
exploiting electric fields. Optoisolators, which consist of an LED and a
photocell, provide isolation by light. Different isolators have differing
performance: some are sufficiently linear to pass high accuracy analog signals
across an isolation barrier. With others, the signal may need to be converted to
digital form before transmission for accuracy to be maintained.

11B.12
Example
An example of a 3-port isolation transformer is the Analog Devices AD210. It

allows the user to select gains from 1 to 100, using external resistors, with the
input section op-amp. It uses transformers to achieve 2500 V RMS (3500 V
peak) isolation.
A typical application using the AD210 is shown below:

Motor control
current sensing
using an analog High voltage
3-port isolator AC Input = 230 V
providing power, FB
INPUT OUTPUT
T1
input and output +15 V
-IN DEMOD
isolation +IN
MOD
FILTER
vo
0.1 RG AD620
REF COM i COM o
-15 V
+Viss T2 POWER T3 +Voss

INPUT OUTPUT
POWER POWER
M -V iss SUPPLY SUPPLY
-V oss
POWER
OCILLATOR
AD210
PWR COM PWR
+15 V
Figure 11B.11 An Analog 3-Port Isolator
The AD210 is used with an AD620 in-amp in a high-side current-sensing

system for motor control. Current is sensed on the high (230 V) side of the
motor (as opposed to using a resistor on the ground side so-called low-side
sensing) so that any fault current from the motor to the chassis (earth) is also
detected.
The input of the AD210, being isolated, can be directly connected to a 230 V
power line without protection being necessary. The input sections isolated
15 V powers the AD620, which senses the voltage drop in a small value
current sensing resistor. The AD210 input stage op-amp is simply connected as
a unity-gain follower. The 230 V RMS common-mode voltage is ignored by
this isolated system.

11B.13
Example
An example of a digital isolator is an LED / phototransistor optocoupler:
A digital isolator
using a LED and
425 10 k phototransistor
VDD1 VDD2 (5 V)
(5 V)
CMOS IIN IOUT v OUT
GATE
v IN
HIGH VOLTAGE GND 2
GND 1 ISOLATION BARRIER
Figure 11B.12 A Digital Optocoupler
A current of approximately 10 mA drives an LED transmitter, with light

output received by a phototransistor. The light produced by the LED saturates
the phototransistor (drives the collector-emitter voltage low). Input / output
isolation of 50007000 V RMS is common. Although fine for digital signals,
most optocouplers are too nonlinear for most analog applications. Also, since
the phototransistor is being saturated, response times can be around 10-20 s
for slower (cheap) devices, limiting high speed applications.
The availability of low cost digital isolators solves most system isolation
problems in data acquisition systems as shown below:
Practical application
of digital isolation in
ISOLATION BARRIER a data acquisition
system
DIGITAL MICRO-
SENSOR ADC
ISOLATORS CONTROLLER
ground reference A ground reference B
Figure 11B.13 Digital isolation in a Data Acquisition System
In this system, digitizing the signal first using an ADC with serial output, then
using digital isolation, eliminates the problem of analog isolation amplifiers
(which are expensive).

11B.14
Summary
The two signals that appear at the inputs of difference amplifiers can be
split into a common-mode signal and a differential signal. This aids in
calculating the output of a difference amplifier, and gives rise to a measure
of difference amplifier performance, known as the common-mode rejection
ratio, or CMRR. It is usually expressed in dB.
Difference amplifiers are used to amplify the difference between two

signals, and to reject the common-mode. However, they are very sensitive
to the values (i.e. tolerance, drift due to temperature, etc.) of the resistors
used in the circuit. Therefore, only integrated circuit packages with laser-
trimmed resistors in close proximity are used in practice.
Instrumentation amplifiers are an arrangement of op-amps that overcomes

some of the deficiencies of the standard difference amplifier. They have a
very large balanced input impedance, and can provide large differential
gain whilst rejecting the common-mode signal. They are readily available
in integrated circuit packages, and are commonly used in bridge circuits.
Programmable gain amplifiers are used in data acquisition systems where

the transducer exhibits a very large range of output voltages. They can be
controlled digitally.
Isolation amplifiers provide electrical isolation between sensors and

electronic circuitry, and are used in environments where a direct electrical
connection would be hazardous, such as in mining and biomedical
applications.
References
Jung, W: Op-Amp Applications, Analog Devices, 2002.

11B.15
Exercises
1.
Consider the instrumentation amplifier implementation shown below:
v i1 R3 R5
vo1
A1
10 k 10 k
R1
RG 100 k
A3 vo
1 k R2
100 k
R4 R6
A2
v i2 vo2 10 k 10 k
(a) Determine expressions for the output voltages of A1 and A2 in terms of

the input voltages vi1 and vi 2 .
(b) Find the overall differential voltage gain Ad of the in-amp.
(c) The following voltages are applied to the in-amp: vi1 230 mV ,
vi 2 235 mV . Determine the final output voltage.
(d) What value of RG must be used to change the gain of the in-amp to
1000?

11B.16
2.
Consider the two-op-amp instrumentation amplifier shown below:
v i1
A1 vo
R4
40 k
RG R3
16 k 10 k
R2
10 k
A2
v i2
R1
40 k
Determine the value of the overall differential voltage gain Ad of the in-amp.

12A.1
Lecture 12A Transfer Functions

Transfer functions. Forced response. Frequency response. Natural response.
Complete response.
Introduction
In any particular circuit, the phasor ratio of the desired forced response to the
forcing function, written in terms of the complex frequency s, is called the
transfer function. Thus, a transfer function is an input-output description of the
behaviour of a circuit, and it does not include any information concerning the
internal structure of the circuit and its behaviour (we have already seen that the
RLC circuit can be replaced by a biquad circuit utilising op-amps both have
the same input-output behaviour within certain bounds).
The transfer function is intimately related to the pole-zero plot it completely

characterises a circuit. We can thus dispense with circuit schematics, and start
to think in terms of cascaded and interconnected blocks that are described by
transfer functions so-called block diagrams.
The transfer function will also be seen to hold information about the form of
the circuits natural response. Thus, given a transfer function of a circuit, we
can write down an expression for the natural response by inspection. If we are
given a forcing function and the initial conditions, we can then determine the
complete response.
We shall also see that circuits are not special we can model any system
described by linear differential equations (e.g. mechanical, hydraulic,
electrical, thermal, fluid) with block diagrams.

12A.2
Transfer Functions
Let us analyse the following simple circuit:
vs ( t ) C v( t )
Figure 12A.1
We can find the forced response of this circuit by working in the frequency-
domain (if the source has a functional form that we can transform to the
frequency-domain, such as an exponentially damped sinusoid). Then we have:
1 sC
V Vs
R 1 sC
1 RC
Vs
s 1 RC (12A.1)
A transfer function is the ratio of the desired forced response to the forcing
function, using phasor notation and the complex frequency s. It is usually
designated H s . With this notation, we can see that the transfer function is a
function of the complex variable s . Thus, in this case, we have:
V
Hs
1 RC
Vs s 1 RC (12A.2)
Several important conclusions can be drawn from the form of the transfer
function.

12A.3
Characteristic Equation
Setting the denominator of the transfer function to zero, we get the

characteristic equation of the circuit, s 1 RC 0 . This corresponds exactly
to that obtained from the original source-free homogeneous differential
equation for the circuit, dv dt v RC 0 . Thus we see that the denominator
of the transfer function seems to come directly from the source-free circuit
itself, and is thus dependent only on the topology and types of passive circuit
components. We will see later that the form of the natural (i.e. source-free)
response is directly related to the characteristic equation.
Pole-Zero Plot
Roots of the characteristic equation give us the poles of the transfer function. In
this case there is one pole at s 1 RC j 0 . Zeros are obtained by finding
those frequencies for which the transfer function is zero. In this case, there is a
zero at s . When plotting poles and zeros, we normally dont show poles
and zeros at infinity, and so the pole-zero plot corresponding to this particular
transfer function is:
-1

RC
s -plane
Figure 12A.2
This pole-zero plot conveys exactly the same information as the transfer
function, apart from the multiplicative factor 1 RC . For example, from the plot
K
above we know the transfer function has the form . In many instances,
s 1 RC
just the form of the transfer function tells us a lot about the circuits behaviour,
so engineers make a lot of use of pole-zero plots.

12A.4
Transfer Function Form
The form of a transfer function is such that it can always be written as a ratio of
two polynomials. The numerator polynomial can be factored into m zero
terms, the denominator can be factored into n pole terms.
Hs K
s z1 s z 2 s z m
s p1 s p 2 s p n (12A.3)
The number of pole terms determines the order of the circuit. For our simple
example RC circuit, we have a first-order circuit, since there is one pole. There
are no zeros (except for the implicit one at s ).
Some of the poles in the transfer function may occur as complex conjugate
pairs. For example, if the first two pole terms are complex conjugates, then the
transfer function can be written as:
Hs K
s z1 s z 2 s z m
s p1 s p1* s p n (12A.4)
K 2
s z1 s z 2 s z m
s p1 p1* s p1p1* s pn
If we let p1 x jy then p1* x jy , then p1 p1* 2 x and p1p1* x 2 y 2 .
Thus, if we define the real numbers a 2 x and b x 2 y 2 , then the transfer

function can be written as:
Hs K
s z1 s z 2 s z m
s2 as bs pn (12A.5)
This is an extremely important observation it tells us that a transfer function

can be written with real coefficients as a product of first-order and second-
order factors! Thus, if we become familiar with the properties of first-order and
second-order circuits, we can handle circuits of any order!

12A.5
Relationship to Differential Equation
Returning to the original circuit, we can do KVL around the loop and write the
describing differential equation in the form:
dv v v
s (12A.6)
dt RC RC
or, using the D operator:
1 (12A.7)
v
D v s
RC RC
Now, putting aside all mathematical formality, we substitute s for D, and

transform into the frequency-domain:
1 (12A.8)
1
s V Vs
RC RC
We now get:
(12A.9)
V
Hs
1 RC
Vs s 1 RC
Thus, we can directly convert a differential equation into the frequency-domain

by setting all derivatives to s.
A more mathematically formal way to transform from the differential equation

(the time-domain) to the transfer function (the frequency-domain) will come
later with a study of the Laplace Transform.

12A.6
Circuit Abstraction
The transfer function is a relationship between the input and output of a circuit
only, expressed in the frequency-domain using the notion of complex
frequency and phasors. It tells us nothing about internal voltages, currents,
consumed power, topology or even components. Thus, we could represent the
RC circuit as a simple black box that converts one phasor into another:
vs ( t ) C v( t )
Vs 1/RC V
s 1/RC
Figure 12A.3
The representation of a circuit in terms of its input signal, transfer function and
output signal is known as a block diagram. The block diagram is a concept that
is used often across all disciplines of engineering, and can be applied to any
linear system, not just electrical circuits. It is a way to characterize a system
without resorting to writing differential equations instead we represent a
system by its transfer function and work with input and output phasors. The
only drawback to this approach is that we must work with algebraic equations
involving complex numbers but most of us would prefer this to solving
differential equations!

12A.7
Forced Response
For now, we can only consider cases where the forced response can be
expressed as an exponentially damped sinusoid (in the most general case). This
includes DC, exponential and sinusoidal waveforms as special cases.
To determine the forced response, the transfer function is evaluated at the

complex frequency of the forcing function. Then, to determine the phasor
representation of the desired response, we simply multiply the phasor
representation of the forcing function by the transfer function (which is now
just a complex number).
For example, suppose the forcing function is v s t V0 , a DC voltage. Then we
know that this forcing function has a complex frequency s 0 , so evaluation

of the transfer function at s 0 gives:
Hs s0
1 RC
1
s 1 RC s0 (12A.10)
Thus, the desired forced response voltage phasor is:
V f Hs Vs 1 V0 V0 (12A.11)
Converting back to the time-domain, we find that the forced response is V0 ,

which we know is true from other circuit analysis methods.

12A.8
If we had a sinusoidal forcing function, v s t Vm cost , then we
evaluate the transfer function at s j :
Hs s j
1 RC
s 1 RC s j
1 RC

j 1 RC
1
, 0 1 RC (12A.12)
1 j 0
Thus, the desired forced response voltage phasor is:
V f Hs Vs
1
Vm
1 j 0
Vm
(12A.13)
1 j 0
In more advanced circuit analysis, when the Laplace Transform is introduced,

we will be able to handle almost any type of forcing function, rather than just
exponentially damped sinusoids (or its special cases), by replacing the forcing
function phasor (which is a constant) by a forcing function which is a
mathematical expression containing s.

12A.9
Example
0.2 H 0.01 F
is (t) v1(t) 1 3 v2(t)
We transform to the frequency-domain:
100
s
0.2 s
Is V1 1 3 V2
and, using the current-divider rule, we write down the transfer functions:
3 0.2s 100 s s 2 15s 500

H1 s
V1

Is 4 0.2s 100 s s 2 20s 500
H 2 s 3
V2 1 15s
2
Is 4 0.2s 100 s s 20s 500
Note that the denominators are the same!

12A.10
Suppose we want to find V2 for I s 0.20 A and s 0 . Then:
15 0
V2 H 2 0I s 0.20 0
0 20 0 500
2
Thus, the output voltage for a DC current source is zero, as can be seen by
inspection of the circuit the capacitor acts as an open circuit.
Suppose now that s . Then:
15
V2 H 2 I s
15 s
0.20 0.20 0
1 20 s 500 s 2
1 20 500 2
Thus, the output voltage for an infinitely high frequency source (say a
sinusoidal source) is zero, as can be seen by inspection of the circuit the
inductor acts as an open circuit.
Suppose now that s 10 . Then:
15 10
V2 H 2 10I s 0.20 0.075
10 20 10 500
2
Thus v2 t 0.075e 10t . Note that this is a forced response due to the forcing
function v2 t 0.2e 10t , and not the natural response.
Now apply a sinusoidal forcing function, with s j10 . Then:
V2 H 2 j10I s
15 j10
0.20
j10 20 j10 500
2
j3
0.20
8 j4
0.0670863.4
Thus v2 t 67.08 cos10t 63.4 mV .
Lastly, let s 10 j 20 . Then since we are exciting the system at the same
frequency as a pole, the response must be infinite.

12A.11
Frequency Response
Frequency response, by definition, is the sinusoidal steady-state response, i.e.
the forced response to a sinusoid. As shown above, we can derive the
frequency response directly from the transfer function. All we have to do is
evaluate the transfer function at an arbitrary sinusoidal frequency s j :
Hs s j H j frequency response
(12A.14)
For example, for the simple RC circuit, we have:
Hs s j
1
, 0 1 RC
1 j 0 (12A.15)
Thus, it is a simple matter to derive expressions for the frequency response

from the transfer function. We can now determine the magnitude response and
phase response for the simple RC circuit:
H j
1
1 0
2
(12A.16)
H j tan 1 0

12A.12
Example
For the circuit seen previously, in the frequency-domain:
100
0.2 s s
Is V1 1 3 V2
we can use the transfer function to establish the frequency response quite
quickly. For example, we know:
H 2 s
15s
s 20s 500
2
Then the frequency response is given by:
15 j
H 2 j
j
20 j 500
2
j15

500 2 j 20
The magnitude and phase responses are, respectively:
15
H 2 j
500 2 2
400 2
20
H 2 j 90 tan 1 2
500

12A.13
Bode plots of these responses are shown below:
The circuit, with the output taken as v2 t , is thus a bandpass filter.

12A.14
Natural Response
The natural response is, by definition, of a form independent of the forcing
function. Thus, we should be able to find the natural response by setting the
forcing function to zero. In the example RC circuit, we work in the frequency-
domain and set Vs 0 . Thus, we have:
1 RC
V Vs
s 1 RC
1 RC
0
s 1 RC (12A.17)
At first glance, it appears as though the response phasor must be zero. This is
true for most frequencies, but what happens when the forcing function happens
to be at a complex frequency s 1 RC j 0 ?
We know that if we excite the circuit at s 1 RC j 0 with a finite voltage,

we get an infinite voltage (the definition of a pole). Mathematically, we have:
Vs
V (12A.18)
0
So we conclude that if we excite the circuit at the same frequency with a zero
voltage, that we get a finite response. Mathematically, we have:
0
V
0
V0 0
finite 0 0 (12A.19)
Thus, the natural response must occur at frequencies corresponding to the poles
of the transfer function.

12A.15
For the example RC circuit, the pole of the transfer function occurs and infinite
voltage results when the operating frequency is s 1 RC j 0 . A finite
voltage at this frequency thus represents the natural response:
Vn A at s 1 RC j 0 (12A.20)
Transforming this natural response to the time-domain:
vn t Aet RC (12A.21)
For an arbitrary transfer function:
Hs K
s z1 s z 2 s z m
s p1 s p 2 s p n (12A.22)
the poles of Hs occur at s p1 , p 2 , p n , and if the input forcing function is

operating at any of these frequencies, an infinite forced response will result.
Thus, a finite response at each of these frequencies is possible for the natural
response. We thus have the form of the natural response by inspection of the
poles of the transfer function. The most general expression is:
f n t A1e p1t
B1ebt B2tebt B2t 2ebt
C1e1t cosd 1t D2e1t sind 1t
E1 E2t e2t cosd 2t F1 F2t e2t sind 2t (12A.23)
or, in words:
natural response reponse due to real and distinct poles

response due to real and repeated poles (12A.24)
reponse due to complex and distinct poles
response due to complex and repeated poles
The coefficients of the natural response must be evaluated using a knowledge

of the forcing function and the initial conditions.

12A.16
Example
Consider the circuit seen previously in the frequency-domain:
100
0.2 s s
Is V1 1 3 V2
The transfer functions were derived as:
3 0.2s 100 s s 2 15s 500

H1 s
V1

Is 4 0.2s 100 s s 2 20s 500
H 2 s 3
V2 1 15s
2
Is 4 0.2s 100 s s 20s 500
Consideration of either transfer function will determine the form of the natural
response. First, we find the poles by setting the denominator to zero, and thus
obtain the roots by solving the characteristic equation:
s 2 20s 500 0
s 102 400 0
s 10 j 20
Thus, we have two complex and distinct poles. The form of the natural
response for either v1 t or v2 t is thus:
vn t Ae10t cos20t Be10t sin20t
To evaluate A and B, we need to know the initial conditions and the forcing
function.

12A.17
If the natural response is desired for a circuit that contains no forcing function,
then we can insert any source we like into the circuit, evaluate the transfer
function, and then determine the natural response by inspection of the poles.
If the circuit already contains a source, then we are allowed to set it to zero,
apply a forcing function in a more convenient location, and then determine the
poles from the resulting transfer function.
These methods will work since the poles of the transfer function are a
characteristic of the circuit only, and not of the forcing function we will
obtain the same poles for any of the many source locations that are possible
so long as we dont derive the degenerate transfer function H s 0 (e.g. a
voltage forcing function that is naively applied in parallel with a short-circuit).
Example
Consider the source-free circuit:
3F
v1(t) 1 2F
1
2 v2(t)
We seek expressions for v1 t and v2 t for t 0 , given the initial conditions,

v1 0 v2 0 11 V .
Firstly, note that this is not a single time constant (STC) circuit, since we
cannot reduce the circuit further. We thus need to investigate this circuit by
examination of a transfer function.
Let us install a current source I s in parallel with the 1 resistor, and find the
transfer function Hs V1 I s , which also happens to be the input impedance

seen by the current source.

12A.18
We have:
V1
Is 3s 2I s
2
1 2s 6s 3s 2 6s 13s 2
1 s 2
Hs
V1
2 3
I s s 2s 16
Thus, v1 t must be of the form:
v1 t Ae2t Bet 6
The solution is completed by using the given initial conditions to establish the
values of A and B. Since v1 0 is given as 11, then:
11 A B
The necessary additional equation is obtained by differentiating v1 and

applying KCL at the bottom node:
dv1 i 11 22
C 2 A 1 6 B
dt t 0 C 2
Thus, A 8 and B 3 , and the desired solution is:
v1 t 8e 2t 3e t 6
The natural frequencies comprising v2 are the same as those of v1 , and a

similar procedure to evaluate the arbitrary constants leads to:
v2 t 12e 2t e t 6

12A.19
Complete Response
As we know, the complete response is the sum of the forced response plus
natural response:
complete forced
natural (12A.25)
response response response
The transfer function can be used to give us both the forced response and
natural response, and therefore the complete response!
Example
We wish to find the complete response it of the following circuit:
1
3 F
t=0 2
i(t )
vC1
vs ( t ) vC2 1F

1
2
=e -tcos(2t)
The switch is in an open position prior to t 0 , and thus all currents and
voltages to the right of the switch are assumed to be zero. At t 0 the switch is
closed, and the current through the switch is to be found. This response is
composed of both a forced response and a natural response:
it i f t in t
Each may be found through a knowledge of the transfer function, Hs I Vs ,

which is also the input admittance of the circuit to the right of the switch.
We have:
Vs ss 2Vs
I
2 3 s 1 s 2 2ss 2 3s 2 s

12A.20
After combining and factoring:
ss 2
Hs
I

Vs 2s 1s 3
In order to find the forced response, the frequency-domain voltage source

Vs 10 at s 1 j 2 may be multiplied by the transfer function, evaluated
at s 1 j 2 :
I f Hs Vs
1 j 21 j 2 1 5
2 j 22 j 2 j81 j
and thus:
5 2
If 45
16
Transforming to the time-domain, we have:
i f t e cos2t 45
5 2 t
16
The form of the natural response can be written by inspection of the poles of
the transfer function:
in t Aet Be3t
The complete response is therefore:
it e cos2t 45 Aet Be3t

5 2 t
16
The solution is completed by using the given initial conditions to establish the
values of A and B. Since the voltage across both capacitors is initially zero, the
initial source voltage of 1 V must appear across the 2 resistor. Thus:
i0
1 5 2 1
A B
2 16 2

12A.21
Again, it is necessary to differentiate and then to obtain an initial condition for
di dt . From the expression for the complete response, we first find:
di 5 2 2 1
A 3B
dt t 0 16 2 2
15
A 3B
16
The initial value of this rate of change is obtained by analysing the circuit.
However, those rates of change which are most easily found are the derivatives
of the capacitor voltages, since i C dv dt , and the initial values of the
capacitor currents should not be too difficult to find. KVL around the circuit
gives:
2i vs vC1 vC 2
Dividing by 2 and taking the derivative:
di 1 dvs 1 dvC1 1 dvC 2

dt 2 dt 2 dt 2 dt
The first term on the right-hand side is obtained by differentiation of the source
function and evaluation at t 0 , the result is 1 2 As-1 .The second term is
numerically equal to 3 2 of the initial current through the 1 3 F capacitor, or
3 4 As-1 . Similarly, the last term is 1 4 As-1 . Thus:
di 1 3 1 3

dt t 0 2 4 4 2
We may now use our two equations in A and B to determine the unknown
coefficients of the natural response:
A 0 and B 3 16
The complete response is therefore:
it 5 16 2e t cos2t 45 3 16 e 3t

12A.22
Summary
A transfer function is the ratio of the desired forced response to the forcing
function, using phasor notation and the complex frequency s. It is usually
designated Hs .
A transfer function is always the ratio of two polynomials in s. It can be

written as:
Hs K
s z1 s z 2 s z m
s p1 s p 2 s p n
where z i is termed a zero and p j is termed a pole.
A pole-zero plot provides an alternative description of a transfer function

(to within an arbitrary constant K) as it specifies all zeros and poles.
The transfer function can be derived from the differential equation

describing a circuit by replacing the D operator with s.
The transfer function can be derived by working directly in the frequency-

domain with the concept of generalized impedance ( R , sL and 1 sC ).
The transfer function can be used to determine the forced response phasor
by evaluating it at the complex frequency of the forcing function and
multiplying the resultant complex number by the input phasor.
The transfer function can be used to determine the frequency response of a

circuit by letting s j :
Hs s j H j frequency response
The transfer function can be used to determine the natural response of a

circuit, since the poles of a circuit determine the form of its natural
response.
The transfer function is a complete description of a circuit from it we can

derive the forced response, the natural response, the complete response, and
the frequency response!

12A.23
References
Hill, 1984.

12A.24
Exercises
1.
The capacitors in the circuit shown below carry charge at t 0 .
250 k
i1 i2
1 F v 50 k 8 F
Find the frequencies present in the specified response for t 0 :
(a) i1n t (b) i2 n t (c) v n t
2.
Find vt for all values of time in the circuit shown below:
12
i s(t) t=0 v( t ) 3 2H 6H
=e -tcos(2t)
3.
Find it for all values of time in the circuit shown below:
t=0 10 0.8 H
12 V 2 0.1 H

12A.25
4.
2.5i 2
5 2
1
2 H 0.1 F v2
i2
Determine:
(a) The transfer function V2 I s1 , if I s1 is in parallel with the inductor with

its arrow directed upward.
(b) The transfer function V2 I s 2 , if I s 2 is in parallel with the capacitor with

its arrow directed upward.
(c) The transfer function V2 Vs1 , if Vs1 is in series with the inductor with its
positive reference on top.
(d) Specify the form of the natural response v2 n t .

12A.26
5.
1 0.5 H 1
i (t )
2 u(t ) V vs 0.8 F 0.5 H
(a) Write the transfer function for I Vs and find the three natural
frequencies (possibly complex) associated with the response i t .
Hint: There is a pole at s 2 .
(b) If i t is represented as a forced response plus a natural response,

i f t in t , find i f t by working in the frequency-domain.
(c) Write down the form of the natural response, in t .
(d) Determine the complete response i t .
6.
There is no initial energy stored in the following circuit:
L R2
i2(t )
vs R1 C
(a) Transform the circuit into the s-domain and formulate mesh-current
equations.
(b) Solve the mesh equations to establish the transfer function I 2 Vs .
(c) Find i2 t if vs t 12u t V , R1 1 k , R2 2 k , L 4 H and

C 500 nF .

12A.27
Pierre Simon de Laplace (1749-1827)
The application of mathematics to problems in physics became a primary task

in the century after Newton. Foremost among a host of brilliant mathematical
thinkers was the Frenchman Laplace. He was a powerful and influential figure,
contributing to the areas of celestial mechanics, cosmology and probability.
Laplace was the son of a farmer of moderate means, and while at the local
military school, his uncle (a priest) recognised his exceptional mathematical
talent. At sixteen, he began to study at the University of Caen. Two years later
he travelled to Paris, where he gained the attention of the great mathematician
and philosopher Jean Le Rond dAlembert by sending him a paper on the
principles of mechanics. His genius was immediately recognised, and Laplace
became a professor of mathematics.
He began producing a steady stream of remarkable mathematical papers. Not

only did he make major contributions to difference equations and differential
equations but he examined applications to mathematical astronomy and to the
theory of probability, two major topics which he would work on throughout his
life. His work on mathematical astronomy before his election to the Acadmie
des Sciences included work on the inclination of planetary orbits, a study of
how planets were perturbed by their moons, and in a paper read to the
Academy on 27 November 1771 he made a study of the motions of the planets
which would be the first step towards his later masterpiece on the stability of
the solar system.
In 1773, before the Academy of Sciences, Laplace proposed a model of the

solar system which showed how perturbations in a planets orbit would not
change its distance from the sun. For the next decade, Laplace contributed a
stream of papers on planetary motion, clearing up discrepancies in the orbits
of Jupiter and Saturn, he showed how the moon accelerates as a function of the
Earths orbit, he introduced a new calculus for discovering the motion of
celestial bodies, and even a new means of computing planetary orbits which
led to astronomical tables of improved accuracy.

12A.28
The 1780s were the period in which Laplace produced the depth of results
which have made him one of the most important and influential scientists that
the world has seen. Laplace let it be known widely that he considered himself
the best mathematician in France. The effect on his colleagues would have
been only mildly eased by the fact that Laplace was right!
In 1784 Laplace was appointed as examiner at the Royal Artillery Corps, and
in this role in 1785, he examined and passed the 16 year old Napoleon
Bonaparte.
In 1785, he introduced a field equation in spherical harmonics, now known as

Laplaces equation, which is found to be applicable to a great deal of
phenomena, including gravitation, the propagation of sound, light, heat, water,
electricity and magnetism.
Laplace presented his famous nebular hypothesis in 1796 in Exposition du

systeme du monde, which viewed the solar system as originating from the
"Your Highness, I contracting and cooling of a large, flattened, and slowly rotating cloud of
have no need of this
hypothesis. "
incandescent gas. The Exposition consisted of five books: the first was on the
- Laplace, to apparent motions of the celestial bodies, the motion of the sea, and also
Napoleon on why
his works on atmospheric refraction; the second was on the actual motion of the celestial
celestial mechanics
make no mention of bodies; the third was on force and momentum; the fourth was on the theory of
God.
universal gravitation and included an account of the motion of the sea and the
shape of the Earth; the final book gave an historical account of astronomy and
included his famous nebular hypothesis which even predicted black holes.
Laplace stated his philosophy of science in the Exposition:
If man were restricted to collecting facts the sciences were only a sterile
nomenclature and he would never have known the great laws of nature. It is
in comparing the phenomena with each other, in seeking to grasp their
relationships, that he is led to discover these laws...
Exposition du systeme du monde was written as a non-mathematical

introduction to Laplace's most important work. Laplace had already discovered
the invariability of planetary mean motions. In 1786 he had proved that the
eccentricities and inclinations of planetary orbits to each other always remain
small, constant, and self-correcting. These and many of his earlier results

12A.29
formed the basis for his great work the Trait du Mcanique Cleste published
in 5 volumes, the first two in 1799.
The first volume of the Mcanique Cleste is divided into two books, the first
on general laws of equilibrium and motion of solids and also fluids, while the
second book is on the law of universal gravitation and the motions of the
centres of gravity of the bodies in the solar system. The main mathematical
approach was the setting up of differential equations and solving them to
describe the resulting motions. The second volume deals with mechanics
applied to a study of the planets. In it Laplace included a study of the shape of
the Earth which included a discussion of data obtained from several different
expeditions, and Laplace applied his theory of errors to the results.
In 1812 he published the influential study of probability, Thorie analytique

des probabilits. The work consists of two books. The first book studies
generating functions and also approximations to various expressions occurring
in probability theory. The second book contains Laplace's definition of
probability, Bayes's rule (named by Poincar many years later), and remarks on
mathematical expectation. The book continues with methods of finding
probabilities of compound events when the probabilities of their simple
components are known, then a discussion of the method of least squares, and
inverse probability. Applications to mortality, life expectancy, length of
marriages and probability in legal matters are given.
After the publication of the fourth volume of the Mcanique Cleste, Laplace
continued to apply his ideas of physics to other problems such as capillary
action (1806-07), double refraction (1809), the velocity of sound (1816), the
theory of heat, in particular the shape and rotation of the cooling Earth
(1817-1820), and elastic fluids (1821).
Many original documents concerning his life have been lost, and gaps in his
biography have been filled by myth. Some papers were lost in a fire that
destroyed the chateau of a descendant, and others went up in flames when
Allied forces bombarded Caen during WWII.
Laplace died on 5 March, 1827 at his home outside Paris.

12B.1
Lecture 12B Sensor Signal Conditioning

Bridge circuits. Strain, force, pressure and flow measurements. High
impedance sensors. Temperature sensors.
Introduction
A sensor is a device that receives a signal or stimulus and responds with an
electrical signal. Sensors and their associated circuits are used to measure
Sensors translate a
various physical properties such as temperature, force, pressure, flow, position, physical quantity to
an electrical quantity
light intensity, etc. These properties act as the stimulus to the sensor, and the
sensor output is conditioned and processed to provide the corresponding
measurement of the physical property.
Sensors which measure different properties may have the same type of
electrical output. For example, a Resistance Temperature Detector (RTD) is a
variable resistance, as is a resistive strain gauge. Both RTDs and strain gauges
are often placed in bridge circuits, and the conditioning circuits are therefore
quite similar. Therefore bridges and their conditioning circuits will be looked at
in detail.
Active sensors require an external source of excitation. Examples include RTDs

and strain gauges.
Passive or self-generating sensors do not require external power. Examples

include thermocouples and photodiodes.
The full-scale outputs of most sensors are relatively small voltages, currents, or
resistance changes, and therefore their outputs must be properly conditioned
before further analog or digital processing can occur. Amplification, level
translation, galvanic isolation, impedance transformation, linearization and
filtering are fundamental signal-conditioning functions that may be required.

12B.2
Sensors
There are many types of sensors we will briefly look at those which lend
themselves to measurement systems, data acquisition systems and process
control systems.
Some typical sensors and their output formats are shown in the table below:
Some typical
sensors and their Property Sensor Active / Output
output formats
Passive
Temperature Thermocouple Passive Voltage

Silicon Active Voltage / Current
RTD Active Resistance
Thermistor Active Resistance
Force / Pressure Strain Gauge Active Resistance
Piezoelectric Passive Voltage
Acceleration Accelerometer Active Capacitance
Position Linear Variable Active AC Voltage
Differential Transformer
(LVDT)
Light Intensity Photodiode Passive Current
Table 12B.1 Typical Sensors

12B.3
Process Control Systems
A common application of sensors is within process control systems. One

example would be control of a physical property, such as temperature. A
sample block diagram of how this might be implemented is given below:
A typical process
control system,
REMOTE CONTROL ROOM showing where the
SIGNAL 4 TO 20mA 4 TO 20mA SIGNAL
sensor fits into the
CONDITIONING TRANSMITTER RECEIVER CONDITIONING overall system
SENSOR ADC
PC or MICRO-
PROCESS
SCADA CONTROLLER
ACTUATOR DAC
SIGNAL 4 TO 20mA 4 TO 20mA SIGNAL

CONDITIONING RECEIVER TRANSMITTER CONDITIONING
Figure 12B.1 A Typical Process Control System
In this system, an output from a temperature sensor is conditioned, transmitted

over some distance, received, and then digitized by an analog-to-digital
converter (ADC). The microcontroller or host computer determines if the
temperature is above or below the desired value, and outputs a digital value to
the digital-to-analog converter (DAC). The DAC output is conditioned and
drives the remotely located actuator, in this case a heater. The interface
between the control room and the remote process is via an industry standard
4-20 mA current loop.

12B.4
Programmable Logic Controllers
Many industrial processes are controlled at the remote end, with the
microcontroller taking the form of a Programmable Logic Controller, (PLC).
A PLC is a digital computer used for automation of processes, such as control
of machinery on factory assembly lines, chemical processes, etc. Unlike
general-purpose computers, the PLC is designed for multiple input and output
arrangements, extended temperature ranges, immunity to electrical noise, and
resistance to vibration and impact. They are generally programmed with
proprietary software from the PLC vendor.
A Programmable
Logic Controller
(PLC)
Figure 12B.2 A Programmable Logic Controller (PLC)

12B.5
PLCs communicate with a Supervisory Control and Data Acquisition

(SCADA) system using industrial Ethernet or a vendor proprietary protocol.
The control room has a human interface to the SCADA system.
A process control
scheme that uses
REMOTE to other PLCs
CONTROL ROOM
PLCs and a SCADA
system
SIGNAL
CONDITIONING
SENSOR ADC
MICRO- HUMAN
PROCESS SCADA
CONTROLLER INTERFACE
ACTUATOR DAC
SIGNAL
CONDITIONING
to other PLCs
PLC
Figure 12B.3 A Process Control Systems with PLCs and SCADA
Smart Transducers
By including the microcontroller (with integrated ADC and DAC), the sensor,
and the actuator into one device, a smart transducer can be implemented with
self-contained calibration and linearization features, among others.
A smart transducer
Smart Transducer
Digital
PHYSICAL MICRO- NETWORK Communication
PROCESS
TRANSDUCER CONTROLLER INTERFACE System
Figure 12B.4 A Smart Transducer

12B.6
Programmable Automation Controllers
Modern industrial automation is starting to use a new device known as a

Programmable Automation Controller (PAC). PACs are used to interface with
simple sensors and actuators, just like a PLC, but they also have advanced
control features, network connectivity, device interoperability and enterprise
data integration capabilities. PACs are multifunctional, handling the digital,
analog, and serial signal types common in all types of industrial applications.
The same hardware can be used for data collection, remote monitoring, process
control and discrete and hybrid manufacturing. PACs also use standard IT
components and protocols, with Ethernet and TCP/IP being very common.
A PAC in a modern
industrial application
(Courtesy Opto 22)
Figure 12B.5 A PAC in a Modern Industrial Application
In the example shown above, the PAC is operating in multiple domains to

monitor and manage a production line, a chemical process, a test bench, and
shipping activities. To do so, the PAC must simultaneously manage analog
values such as temperatures and pressures; digital on/off states for valves,
switches, and indicators; and serial data from inventory tracking and test
equipment. At the same time, the PAC is exchanging data with an OLE for
Process Control (OPC) server, an operator interface, and a Structured Query
Language (SQL) database. Simultaneously handling these tasks without the
need for additional processors or middleware is a hallmark of a PAC.

12B.7
Bridge Circuits
Resistive elements are some of the most common sensors. They are
inexpensive, and relatively easy to interface with signal-conditioning circuits.
Resistive elements can be made sensitive to temperature, strain (by pressure or
by flex) and light. Using these basic elements, many complex physical
phenomena can be measured.
Sensor element resistance can range from less than 100 to several hundred
Resistive elements
k , depending on the sensor design and the physical environment to be form the basis for
many types of
measured. The table below shows the wide range of sensor resistances used in physical
measurements
bridge circuits.
Sensor Resistance Range

Strain Gauges 120 , 350 , 3500
Weigh-Scale Load Cells 350 - 3500
Pressure Sensors 350 - 3500
Relative Humidity 100 k - 10 M
Resistance Temperature Devices (RTDs) 100 , 1000
Thermistors 100 - 10 M
Table 12B.2 Sensor Resistance Ranges
Resistive sensors produce relatively small percentage changes in resistance in

The change in a
response to a change in a physical variable. For example, RTDs and strain resistive element is
gauges present a significant measurement challenge because the typical change small and therefore
accurate
in resistance over the entire operating range may be less than 1% of the measurement is
essential
nominal resistance value. Accurately measuring small resistance changes is
therefore critical when applying resistive sensors.

12B.8
A resistance bridge, or Wheatstone bridge, shown in the figure below, is used

to measure small resistance changes accurately.
The Wheatstone
bridge is the basis of
many types of
sensor VB
The Wheatstone bridge:
R4 R3
R1 R2
vo VB
1
R R4 R2 R3
vo
At balance,
R1 R2
R1 R2 vo if
R4 R3
Figure 12B.6 The Wheatstone Bridge
It consists of four resistors connected to form a quadrilateral, a source of

excitation voltage VB (or, alternately, a current) connected across one of the
diagonals, and a voltage detector connected across the other diagonal. The
detector measures the difference between the outputs of the two voltage
dividers connected across VB . The general formula for the output vo is:
R1 R2
vo VB
1
R R4 R2 R3
(12B.1)
For sensor applications, the deviation of one or more of the resistors in a bridge
from an initial value is measured as an indication of the change in the measured
variable. In this case, the output voltage change is an indication of the
resistance change. Since very small resistance changes are common, the output
voltage change may be as small as tens of millivolts, even with an excitation
voltage of VB 10 V (a typical value).

12B.9
In many bridge applications, there may not just be a single variable element,
but two, or even four elements, all of which may vary. The figure below shows
those bridges most commonly suited for sensor applications:
Bridges that are
commonly suited for
VB VB VB VB sensor applications
R+ R R- R
R R R+ R R R R- R
vo vo vo vo
R R+ R R R+ R R R+ R
R- R R+ R
VB R VB R VB R R
vo : R R VB
4 R+ 2 R+ 2 R R
2 2
(A) Single-Element (B) Two-Element (C) Two-Element (D) All-Element

Varying Varying (1) Varying (2) Varying
Figure 12B.7 Bridges for Sensors
Note that since the bridge output is always directly proportional to VB , the
measurement accuracy can be no better than that of the accuracy of the
excitation voltage.
In each case, the value of the fixed bridge resistor R is chosen to be equal to
the nominal value of the variable resistor(s). The deviation of the variable
resistor(s) about the nominal value is assumed to be proportional to the
quantity being measured, such as strain (in the case of a strain gauge), or
temperature (in the case of an RTD).
The single-element varying bridge (A) is most suited to temperature sensing

using RTDs or thermistors. All the resistances are nominally equal, but one of The single-element
varying bridge is
them (the sensor) is variable by an amount R . As the equation indicates, the most suited to
relationship between the bridge output and R is not linear. Also, in practice, temperature sensing
most sensors themselves will exhibit a certain specified amount of nonlinearity,
which must be taken into account. Software calibration is used to remove the
linearity error in digital systems.

12B.10
The two-element varying bridge (C) requires two identical elements that vary
in opposite directions. For example, this could correspond to two identical
strain gauges: one mounted on top of a flexing surface, and one on the bottom.
This configuration is linear, and the terms R R and R R can be viewed
as two sections of a linear potentiometer.
The all-element varying bridge (D) produces the most signal for a given
resistance change, and is inherently linear. It is also an industry-standard
configuration for loads cells constructed from four identical strain gauges.
Load cells bridge
elements are strain
gauges arranged in
a particular
orientation
Figure 12B.8 Load Cells

12B.11
Bridge Design Issues
Selecting the configuration (1, 2, 4-element varying)
Selection of voltage or current excitation
Stability of excitation or ratiometric operation
Bridge sensitivity: Full-Scale Output / Excitation

1 mV/V to 10 mV/V typical
Full-scale bridge outputs

10 mV to 100 mV typical
Precision, low-noise amplification / conditioning techniques required
Linearization techniques may be required
Remote sensors present challenges
Regardless of the absolute level, the stability of the excitation voltage or

current directly affects the accuracy of the bridge output. Therefore stable
references and / or ratiometric drive techniques are required to maintain
highest accuracy.
Ratiometric refers to the use of the bridge drive voltage of a voltage-driven

bridge as the reference input to the ADC that digitizes the amplified bridge
output voltage.

12B.12
Amplifying and Linearizing Bridge Outputs
The output of a single-element varying bridge may be amplified with an

instrumentation amplifier (in-amp):
The output of a
single-element
varying bridge is
normally amplified VB
with an in-amp
VREF = VB
optional ratiometric output
R R +VS
RG
VB R
IN AMP
vo =
4 R+ R
GAIN
REF
2
R R+ R -VS
Figure 12B.9 Bridge Output Using an In-Amp
The in-amp provides a large and accurate gain that is set with a single resistor,
RG . The in-amp also provides dual, high-impedance loading to the bridge
nodes it does not unbalance or load the bridge. Using modern in-amps with
gain ranging from 10-1000, excellent common-mode rejection and gain
accuracy can be achieved with this circuit.
However, due to the intrinsic characteristics of the bridge, the output is still
nonlinear. In a system where the output of the in-amp is digitized using an
ADC and fed into a microcontroller, this nonlinearity can be corrected in
software.
The bridge in this example is voltage driven, by the voltage VB . This voltage
can optionally be used for an ADC reference voltage, in which case it is an
additional output of the circuit, VREF .

12B.13
An analog circuit for linearizing a single-element bridge is shown below:

An analog circuit for
linearizing a single-
element varying
VB bridge
VB R R2
R R vo = 1+
2 R R1
vo
R+ R R R2
R1
Figure 12B.10 Linearizing a One-Element Varying Bridge
The top node of the bridge is excited by the voltage VB . The bottom of the
bridge is driven in complementary fashion by the left op-amp, which maintains
a constant current of VB R in the varying resistance element, R R , which is
the mechanism for linearity improvement. Also, the bridge left-side centre
node is ground-referenced by the op-amp, making this configuration
suppress common-mode voltages.
The output signal is taken from the right-hand leg of the bridge, and is
amplified by a second op-amp connected as a noninverting gain stage.
The circuit requires two op-amps operating on dual supplies. In addition,

paired resistors R1 and R2 must be ratio matched and stable for overall
accurate and stable gain. The circuit is practical if a dual precision op-amp
(with high gain, low offset / noise and high stability) is used.

12B.14
A circuit for linearizing a two-element bridge is shown below:

An analog circuit for
linearizing a two-
element varying
bridge
R+ R R +VS
IB RG IN AMP
vo = IB R
GAIN
REF 2
R R+ R -VS
+VS
IB
R SENSE
-VS
VREF
Figure 12B.11 Linearizing a Two-Element Varying Bridge
This circuit uses an op-amp, a sense resistor, and a voltage reference, set up in
a feedback loop containing the sensing bridge. The net effect of the loop is to
maintain a constant current through the bridge of I B VREF RSENSE . The
current through each leg of the bridge remains constant ( I B 2 ) as the

resistance changes. Therefore the output is a linear function of R . An in-amp
provides the additional gain. If ratiometric operation of an ADC is desired, the
VREF voltage can be used to drive the ADC.

12B.15
Driving Remote Bridges
Wiring resistance and noise pickup are the biggest problems associated with
remotely located bridges. The figure below shows a 350 strain gauge which
is connected to the rest of the bridge circuit by 30 m of twisted pair copper
wire. The temperature coefficient of the copper wire is 0.385 %/C. The figure
shows nominal resistor values at 25 C .
Wiring resistance is
a big problem for
+10 V remotely located
bridges
350 350
R LEAD
10.5 (10.904 )
vo
0 23.45 mV Strain Gauge
350 (5.44 mV 28.84 mV)
R LEAD 350 353.5 FS
R COMP 10.5 (10.904 )
21
Offset error over temperature = +23% FS

Gain error over temperature = -0.21% FS
Figure 12B.12 A Bridge with Wiring Resistance
The full-scale variation of the strain gauge resistance above its nominal 350
value is +1% ( 3.5 ), corresponding to a full-scale strain gauge resistance of
353.5 which causes a bridge output voltage of +23.45 mV. Notice that the
addition of the 21 resistor, RCOMP , is used to compensate for the wiring
resistance and balances the bridge when the strain gauge is 350 .
Assume that the cable temperature increase 10 C above a nominal room

temperature of 25 C . The values in parentheses in the diagram indicate the
values at 35 C . With no strain, the additional lead resistance produces an
offset of +5.44 mV in the bridge output. Full-scale strain produces a bridge
output of +28.84 mV (a change of +23.4 mV from no strain). Thus, the
increase in temperature produces an offset voltage error of +5.44 mV, or +23%
full-scale, and a gain error of -0.05 mV (23.4 mV 23.45 mV), or -0.21% full-
scale.

12B.16
The effects of wiring resistance on the bridge output can be minimized by the
3-wire connection shown below:
Wiring resistance for
a remotely located
bridge can be +10 V
overcome using a 3-
wire connection
350 350
R LEAD
10.5 (10.904 )
vo
0 24.15 mV I=0 Strain Gauge
(0 24.13 mV)
R LEAD 350 353.5 FS
350
10.5 (10.904 )
Offset error over temperature = 0% FS

Gain error over temperature = -0.08% FS
Figure 12B.13 A 3-Wire Connection for a Bridge
We assume that the bridge output voltage is measured by a high impedance

device, therefore there is no current in the sense lead. The sense lead measures
the voltage output of a divider: the top half is the bridge resistor plus the lead
resistance, and the bottom half is strain gauge resistance plus the lead
resistance. The nominal sense voltage is therefore independent of the lead
resistance. When the strain gauge resistance increases to full-scale ( 353.5 ),
the bridge output increases to 24.15 mV.
Increasing the temperature to 35 C increases the lead resistance in each half

of the divider. The full-scale bridge output voltage decreases to +24.13 mV,
but there is no offset error. The gain error due to the temperature increase of
10 C is therefore only -0.02 mV, or -0.08% full-scale.
The three-wire method works well for remotely located resistive elements
which make up one leg of a single-element varying bridge. However, four-
element varying bridges are generally housed in a complete assembly, as in the
case of a load cell. When these bridges are remotely located from the
conditioning electronics, special techniques must be used to maintain accuracy.

12B.17
Most four-element varying bridges (such as load cells) are six-lead assemblies:
two leads for the bridge output, two leads for the bridge excitation, and two
sense leads. To take full advantage of the additional accuracy that these two
extra leads allow, a method called Kelvin or 4-wire sensing is employed, as
shown below:
A 6-lead bridge uses

Kelvin sensing to
VB overcome lead
resistance in both
R LEAD the sensing
+FORCE terminals and the
excitation terminals
6-lead +SENSE R LEAD

bridge
R LEAD
vo
R LEAD
-SENSE R LEAD
-FORCE R LEAD
Figure 12B.14 A 6-Lead Bridge Using Kelvin Sensing
In this setup the drive voltage VB is not applied directly to the bridge, but goes
instead to the input of the upper precision op-amp, which is connected in a
feedback loop around the bridge (+) terminal. Although there may be a
substantial voltage drop in the +FORCE lead resistance of the remote cable, the
op-amp will automatically correct for it, since it has a feedback path through
the +SENSE lead. The net effect is that the upper node of the remote bridge is
maintained at a precise level of VB . A similar situation occurs with the bottom
precision op-amp, which drives the bridge (-) terminal to ground level. Again,
the voltage drop in the FORCE lead is relatively immaterial, because of the
sensing at the SENSE terminal.

12B.18
Integrated Bridge Transducers
A very powerful combination of bridge circuit techniques is shown below:

A bridge transducer
ADC can be used to
connect directly to a
+5V +5V/+3V
6-lead bridge +FORCE R LEAD
+SENSE R LEAD AVDD DVDD

6-lead
bridge +VREF
R LEAD
+A IN
AD7730
R LEAD
-A IN
-SENSE R LEAD
-VREF
GND
-FORCE R LEAD
Figure 12B.15 An Integrated Bridge Transducer
This is an example of a basic DC operated bridge, utilising ratiometric

conversion with a high performance ADC, combined with Kelvin sensing to
minimize errors due to wiring resistance.
The Analog Devices AD7730 Bridge Transducer ADC can be driven from a
single supply voltage of 5 V, which in this case is used to excite the bridge.
Both the analog input and the reference input to the ADC are high impedance
and fully differential. By using the +SENSE and SENSE outputs from the
bridge as the differential reference voltage to the ADC, there is no loss in
measurement accuracy as the actual bridge excitation voltage varies.
The AD7730 is one of a family of sigma-delta ADCs with high resolution

(24 bits) and internal programmable gain amplifiers (PGAs) and is marketed as
a Bridge Transducer ADC. The chip has a self-calibration feature which
allow offset and gain errors due to the ADC to be minimized. A system
calibration feature allows offset and gain errors to be reduced to a few
microvolts.

12B.19
Strain, Force, Pressure and Flow Measurements

The most popular electrical elements used in force measurements include the
resistance strain gauge, the semiconductor strain gauge, and piezoelectric
A strain gauge can
transducers. The strain gauge measures force indirectly by measuring the be used to measure
strain, force,
deflection it produces in a calibrated carrier. Pressure can be converted into a pressure and flow
force using an appropriate transducer, and strain gauge techniques can then be
used to measure pressure. Flow rates can be measured using differential
pressure measurements, which also make use of strain gauge technology.
Measurement Sensor
Strain Strain gauge, Piezoelectric transducer
Force Load cell
Pressure Diaphragm to force to strain gauge
Flow Differential pressure techniques
Table 12B.3 Sensors used for Typical Measurements
The resistance-based strain gauge uses a resistive element which changes in

length, hence resistance, as the force applied to the base on which it is mounted
causes stretching or compression. It is the most well known transducer for
converting force into an electrical variable.

12B.20
A bonded strain gauge consists of a thin wire or conducting film arranged in a

coplanar pattern and cemented to a base or carrier. The basic form of this type
of gauge is shown below:
A bonded strain
gauge
FORCE
Small surface area

Low leakage
High isolation
FORCE
Figure 12B.16 A Bonded Strain Gauge
The strain gauge is normally mounted so that as much as possible of the length
of the conductor is aligned in the direction of the stress that is being measured.
Lead wires are attached to the base and brought out for interconnection.
Semiconductor strain gauges have a greater sensitivity and higher-level output

than wire strain gauges. They can also be produced to have either positive or
negative changes when strained. However, they are temperature sensitive, are
difficult to compensate, and the change in resistance is nonlinear.

12B.21
High Impedance Sensors

Many popular sensors have output impedances greater than several megohms,
and thus the associated signal-conditioning circuitry must be carefully designed
to meet the challenges of low bias current, low noise, and high gain. A few
examples of high impedance sensors are:
High impedances
Photodiode preamplifiers sensors
Piezoelectric sensors
Humidity monitors
pH monitors
Chemical sensors
Smoke detectors
Very high gain is usually required to convert the output signal of these sensors
into a usable voltage. For example, a photodiode application typically needs to
detect outputs down to 30 pA of current, and even a gain of 106 will only yield
30 mV. To accurately measure photodiode currents in this range, the bias
current of the op-amp should be no more than a few picoamps. A high
performance JFET-input op-amp is normally used to achieve this specification.
Special circuit layout techniques are required for the signal conditioning
require special
circuitry. For example, circuit layouts on a printed circuit board (PCB) interfacing circuits
typically need very short connections to minimise leakage and parasitic
elements. Inputs tend to be guarded with ground tracks to isolate sensitive
amplifier inputs from voltages appearing across the PCB.

12B.22
Temperature Sensors
Temperature measurement is critical in many electronic devices, especially
Temperature is an expensive laptop computers and other portable devices their densely packed
extremely important
physical property to circuitry dissipates considerable power in the form of heat. Knowledge of
measure
system temperature can also be used to control battery charging, as well as to
prevent damage to expensive microprocessors.
Accurate temperature measurements are required in many other measurement

systems, for example within process control and instrumentation applications.
Some popular types of temperature sensors and their characteristics are
indicated in the table below:
Popular types of
temperature sensor Sensor Range Accuracy Excitation Feature
and their
characteristics Thermocouple -184C to High accuracy Needs cold Low-
+2300C and junction voltage
repeatability compensation
RTD -200C to Fair linearity Requires Low cost
+850C excitation
Thermistor 0C to Poor linearity Requires High
+100C excitation sensitivity
Semiconductor -55C to Linearity: 1C Requires 10 mV/K,
+150C Accuracy: 1C excitation 20 mV/K
or 1A/K
typical
output
Table 12B.4 Popular Types of Temperature Sensors
In most cases, because of low-level and/or nonlinear outputs, the sensor output
must be properly conditioned and amplified before further processing can
occur. Sensor outputs may be digitized directly by high resolution ADCs
linearization and calibration can then be performed in software, reducing cost
and complexity.

12B.23
Resistance Temperature Devices (RTDs) are accurate, but require excitation

current and are generally used within bridge circuits.
Thermistors have the most sensitivity, but are also the most nonlinear. They are
popular in portable applications for measurement of battery and other critical
system temperatures.
Modern semiconductor temperature sensors offer both high accuracy and

linearity over about a -55C to +150C operating range. Internal amplifiers can
scale output to convenient values, such as 10 mV/C.

12B.24
Summary
A sensor is a device that receives a signal or stimulus and responds with an
electrical signal. The full-scale outputs of most sensors are relatively small
voltages, currents, or resistance changes, and therefore their outputs must
be properly conditioned before further analog or digital processing can
occur.
Amplification, level translation, galvanic isolation, impedance

transformation, linearization and filtering are fundamental signal-
conditioning functions that may be required with sensors.
A resistance bridge, or Wheatstone bridge, is used to measure small

resistance changes accurately. There are a variety of different bridge
circuits, and a variety of amplifying and linearizing techniques to suit each
type.
There are a variety of methods for interfacing to remote bridges. Many

integrated bridge transducers are available as one-chip solutions to bridge
driving and measurement.
There are many types of sensors their use in a certain application requires
an understanding of their physical construction and operation, as well as the
required performance and cost demanded by the overall system.
References
Jung, W: Op-Amp Applications, Analog Devices, 2002.

12B.25
Exercises
1.
For temperature measurements only one active transducer is used and so it is
not possible to have a linear output if it is placed in a bridge.
(a) Show that the output from a single-element varying bridge is given by:
VB R
vo
4 R R
2
(b) Since the active transducer resistance change can be rather large (up to
100% or more for RTDs), the nonlinearity of the bridge output
characteristic (the formula above) can become quite significant. It is
therefore desired to linearize the output of a temperature transducer using
the following circuit:
R1 R 2+ R
VB
vo
R1
R2
Derive an equation for the output voltage.

13A.1
Lecture 13A System Modelling

Differential equations of physical systems. Linear approximations of physical
systems. The transfer function. Block diagrams. Feedback.
Introduction
In order to understand, analyse and design complex systems, we must obtain
quantitative mathematical models of these systems. Since most systems are
dynamic in nature, the descriptive equations are usually differential equations.
If the system stays within bounds, then the equations are usually treated as
linear differential equations, and the method of transfer functions can be used
to simplify the analysis.
In practice, the complexity of systems and the ignorance of all the relevant
factors necessitate the introduction of assumptions concerning the system
operation. Therefore, we find it useful to consider the physical system,
delineate some necessary assumptions, and linearize the system. Then, by
using the physical laws describing the linear equivalent system, we can obtain
a set of linear differential equations. Finally, utilizing mathematical tools, such
as the transfer function, we obtain a solution describing the operation of the
system.
In summary, we:
1. Define the system and its components.
2. List the necessary assumptions and formulate the mathematical model.
3. Write the differential equations describing the model.
4. Solve the equations for the desired output variables.
5. Examine the solutions and the assumptions.
6. Reanalyse or design.

13A.2
Differential Equations of Physical Systems

The differential equations describing the dynamic performance of a physical
system are obtained by utilizing the physical laws this approach applies
equally well to electrical, mechanical, fluid and thermodynamic systems.
For mechanical systems, Newtons laws are applicable.
Example
Consider the simple spring-mass-damper mechanical system shown below:
K
f
Friction Mass y
M
r (t )
Force
This is described by Newtons second law of motion (this system could
represent, for example, a cars shock absorber). We therefore obtain:
d2y dy
M 2
f Ky r
dt dt
where K is the spring constant of the ideal spring and f is the friction constant.

13A.3
Example
Consider the electrical RLC circuit below:
r (t ) R L C v( t )
This is described by Kirchhoffs current law. We therefore obtain:
dv v 1
dt R L
C vdt r
In order to reveal the close similarity between the differential equations for the
mechanical and electrical systems, we can rewrite the mechanical equation in
terms of velocity:
dy
v
dt
Then we have:
dv
M fv K vdt r
dt
The equivalence is immediately obvious where velocity vt and voltage vt

are equivalent variables, usually called analogous variables, and the systems
are analogous systems.
The concept of analogous systems is a very useful and powerful technique for
system modelling. Analogous systems with similar solutions exist for
electrical, mechanical, thermal and fluid systems. The existence of analogous
systems and solutions allows us to extend the solution of one system to all
analogous systems with the same describing differential equation.

13A.4
Linear Approximations of Physical Systems

Many physical systems are linear within some range of variables. However, all
systems ultimately become nonlinear as the variables are increased without
limit. For example, the spring-mass-damper system is linear so long as the
mass is subjected to small deflections yt . However, if yt were continually
increased, eventually the spring would be overextended and break. Therefore,
the question of linearity and the range of applicability must be considered for
each system.
A necessary condition for a system to be linear can be determined in terms of a

forcing function xt and a response yt . A system is linear if and only if:
ax1 t bx2 t ay1 t by2 t (13A.1)
That is, linear systems obey the principle of superposition, [excitation by

x1 t x 2 t results in y1 t y 2 t ] and they also satisfy the homogeneity
property [excitation by ax1 t results in ay1 t ].
It may come as a surprise that a system obeying the relation y mx b is not

linear, since it does not satisfy the homogeneity property. However, the system
may be considered linear about an operating point x 0 , y 0 for small changes
x and y . When x x0 x and y y 0 y , we have:
y mx b
(13A.2)
y0 y mx0 mx b
and, since y 0 mx0 b , then y mx , which is linear.
In general, we can often linearize nonlinear elements by assuming small-signal

conditions. This approach is the normal approach used to obtain linear
equivalent circuits for electronic circuits and transistors.

13A.5
Consider a general element of a system which can be described by a

relationship between the excitation variable xt and the response yt :
y
y=f ( x)
tangent
y0
operating point
0 x0 x
Figure 13A.1
The normal operating point is designated by x0 , y 0 . A Taylor series
expansion about the operating point gives:
y f x0 x f x0 x
df
(13A.3)
dx x x0
The tangent to the curve at the operating point is a good approximation to the
curve for small x . Thus, for a small region about the operating point, a
reasonable first-order approximation of the element is:
y f x0 x
df
y0 mx (13A.4)
dx x x0
which can be written as the approximate linear equation:
y mx (13A.5)
This linear approximation is only accurate for a range of small signals which
depends on the actual nonlinear elements characteristic.

13A.6
Example
Consider the pendulum oscillator:
Length L

Mass M
The magnitude of the torque on the mass is:
T MgL sin
where g is the gravity constant. The equilibrium position for the mass is
0 0 . The nonlinear relation between T and is shown graphically below:

2 2
The first derivative evaluated at equilibrium provides the linear approximation

which is:
T MgL
d
sin 0
d 0
MgLcos 0 0
MgL
This approximation is reasonably accurate for 4 4 . For example,

the response of the linear model for the swing through 30 is within 2% of
the actual nonlinear pendulum response.

13A.7
The Transfer Function

We have already seen that the transfer function, which represents the
input/output relationship of a system in terms of complex frequency, has a
simple relationship to the differential equation describing the system.
Differentiation in the time-domain turns into multiplication by s in the
frequency-domain:
(13A.6)
d
s
dt
Similarly, integration in the time-domain turns into division by s in the time-
domain.
(13A.7)
t 1
0 d
s
Example
For the simple spring-mass-damper mechanical system, the describing

differential equation was found to be:
d2y dy
M 2 f Ky r
dt dt
Transforming to the frequency-domain, we get:
Ms 2 Y fsY KY R
Thus, the transfer function is:
Y 1

R Ms fs K
2
Alternatively, the transfer function of a system can be obtained by analysis

performed entirely in the frequency-domain (as we do for the transfer function
of circuits).

13A.8
Block Diagrams
We can represent a transfer function graphically, with a block diagram. This
shows the relationship between the forced response and the forcing function
an input/output relationship. For example, the simple RC circuit can be
represented by:
vs ( t ) C v( t )
Vs 1/RC V
s 1/RC
Figure 13A.2
Example
The block diagram for the simple spring-mass-damper mechanical system is:
R 1 Y
2
M s + f s +K

13A.9
With this representation, we understand that we have to multiply the input

phasor by the transfer function in the box to obtain the output phasor. In
general, we use the following notation, where G s is the transfer function,
X is the input, and Y is the output:
A block represents
multiplication with a
X Y transfer function
G ( s)
Figure 13A.3
Blocks can be connected in cascade, but only if the outputs are buffered, i.e.
the connection does not cause the transfer function of each individual block to
be different from the unloaded or open condition:
Cascading blocks
implies multiplying
the transfer
X Y= G1X Z= G1G2 X functions
G 1( s ) G 2( s )
Figure 13A.4
Obviously, op-amp circuits are ideal for cascading.
Most systems have several blocks interconnected by various forward and

backward paths. Signals (e.g. voltages, currents) in block diagrams can not
only be transformed by a transfer function, they can also be added and
subtracted.
Addition and
subtraction of
X Z= X+ Y X Z= X - Y signals in a block
diagram
Y Y
Figure 13A.5

13A.10
Example
Consider the ideal integrator op-amp circuit:
R
vi
vo
We know that the gain can be expressed as:
Vo Z 1 sC
2
Vi Z1 R
Therefore, the op-amp integrator can be represented by:
Vi -1 Vo
s RC
Thus, an integrator circuit has a 1 s term in its transfer function. This makes
intuitive sense, since multiplication by s represents differentiation, and so
division by s must be representative of integration.

13A.11
Example
The op-amp circuit:
RF
R2
v2
CF
R1
v1
vo
can be represented by:
V2 1 I2
R2
V1 1 I1 -1/CF Vo
R1 s + 1/R FCF

13A.12
Example
The bilinear op-amp circuit:
K p
1 z
1 K
Vi
Vo
1
can be represented by:
Vi s +z Vo
-K s +p
It should be obvious that we can now establish a cookbook approach to

circuit design if we have a transfer function that we want to implement, we
simply find the appropriate op-amp circuit in the cookbook.

13A.13
Example
The DC motor is a power actuator device that delivers energy to a load as

shown below:
A schematic diagram of the DC motor is shown below:
Armature
Ra
La
Rf
ia
vf Lf
Inertia= J
if Friction= f
Field
Load

13A.14
For a field-controlled DC motor, the armature current is assumed to be constant

and the torque developed by the motor is assumed to be linearly related to the
field current:
Tm K mi f
The motor torque is delivered to the load, which can also be subjected to
external disturbances (e.g. wind-gust forces for a tracking antenna):
TL Tm Td
The load torque for a rotating system with inertia and friction comes from the
rotational form of Newtons second law:
d d
TL J f ,
dt dt
The blocks are derived from the differential equations governing the various
parts of the system The block diagram of a field-controlled DC motor is
therefore:
disturbance
torque
Td
field field armature load
voltage current torque torque speed position
Vf 1 If Tm TL 1 1
Km
Rf + s L f f +sJ s
motor windings motor friction and
action
inertia
If there is no disturbance torque, then the model of the field-controlled DC

motor is:
Vf Km
s ( R f + s L f ) (f + s J )

13A.15
Feedback
Perhaps the most important block diagram is that of a feedback connection,
shown below:
Standard form for
the feedback
connection
R E C
G ( s)
H ( s)
Figure 13A.6
We have the following definitions:
G s = forward path transfer function
H s = feedback path transfer function

r t = reference, input, or desired output
ct = controlled variable, or output
et = actuating error signal

r t ct = system error
C
= closed-loop transfer function
R
G s H s = loop gain

13A.16
To find the closed-loop transfer function, we solve the following two equations
which are self-evident from the block diagram:
C G s E
E R H s C (13A.8)
Then the output C is given by:
C G s R G s H s C
C1 G s H s G s R (13A.9)
and therefore:
Transfer function for
the standard
C G s
negative feedback

connection
R 1 G s Hs (13A.10)
Note that for negative feedback we get 1 G s H s and for positive feedback
we get 1 G s H s .
Negative and
positive feedback
negative feedback positive feedback
1+GH 1- GH
Figure 13A.7

13A.17
Example
Consider a non-inverting amplifier where the op-amp is modelled as having

infinite input impedance, zero output impedance, and a large but finite gain A:
vi
A vo
R2
R1
The voltage fed back to the inverting terminal is negative feedback. Let the
proportion of the voltage fed back be given by:
R1
vo vo
R1 R2
Then a block diagram that models the non-inverting amplifier is:
Vi Vo
A
The transfer function can be simplified to:
Vi A Vo
1+A
For large values of A such that A 1 , the transfer function reduces to:
1 R2
1
R1

13A.18
Summary
We model linear time-invariant systems by making necessary simplifying
assumptions before applying the basic physical laws. The result is a linear
differential equation describing the system.
A linear differential equation can be turned into a transfer function by

replacing the derivative operator with the complex frequency s.
Alternatively, the transfer function of a system can be obtained by
performing analysis directly in the frequency-domain.
A transfer function can be represented graphically in the form of a block

diagram. Block diagrams can be connected together in cascade, and signals
can be added and subtracted.
Negative feedback in a system is a very important concept. For a forward

path transfer function G and a feedback path transfer function H, the
closed-loop transfer function is:
Ts
G
1 GH
References
Dorf, R.: Modern Control Systems, 5th Ed., Addison-Wesley, 1989.

13A.19
Exercises
1.
RF
R1
R5
C1
R3 C5
v in1 R0
-1 R2 R4
A1
A2
v out1 A3 v out3
v7 R7 R6
v8 R8
A4 v out4
(a) Construct a block diagram of the circuit.
(b) Show that the transfer function T1 s Vout3 Vin1 is:
R3 R0 R2 R4 C1C5
T1 s
1 1 R3 1
s 2 s
1 1
R C R5 5
C 2 4 F 1 5
R R R C C R1 5 1 5
R C C
(c) By comparing with the standard form of a second-order all-pole

transfer function:
K102
T1 s 2
s 2s 02
write expressions for K1 , and 0 in terms of Rs and Cs.

13A.20
(d) For the special case of when R1 R5 , C1 C5 and RF R4 , and defining
d 02 2 , show that the poles of the transfer function are located
at p1, 2 j d where:
1 R 1
d 3
R1C1 R2 R4C1
(e) For the special case of when R1 R5 , C1 C5 and RF R4 , show that
the transfer function T2 s Vout1 Vin1 can be put in the form:
K 2 s
T2 s
s 2s 02
2
(f) For the special case of when R1 R5 , C1 C5 , RF R4 , R6 R8 and
2 R0 R8
, show that the transfer function T3 s Vout4 Vin1 can be put
R1 R7
in the form:
K 3 s 2 02 2 2
T3 s
s 2 2s 02
(g) Draw pole-zero plots for each of the three transfer functions.
(h) Perform a simulation of the circuit and determine each transfer functions
frequency response. Hence, classify each of the transfer functions in
terms of their frequency response (e.g. lowpass, highpass, notch, etc.).

13B.1
Lecture 13B Revision

Categorize material as know, familiar, or aware.
Identify important problems and applications.
Lecture 1A - Basic Laws and Op-Amp Amplifiers

Know
Current, voltage, power and circuits.
Circuit elements and types of circuits.
Ohms Law.
Kirchhoffs Current Law (KCL).
Kirchhoffs Voltage Law (KVL).
Resistance and source combination.
Voltage and current dividers.
Amplifiers.
The ideal op-amp.
Negative feedback.
The noninverting amplifier.
The inverting amplifier.
Know the passive and active sign conventions and have confidence in applying
KCL and KVL to any circuit. Know and use Ohms Law. Know circuit
analysis short-cuts with series / parallel combinations, voltage and current
dividers. Know what parameters are used for an ideal op-amp, the concept and
application of negative feedback and the concept of the virtual short-circuit.
You should know the gain formulas for both noninverting and inverting
amplifiers, and be able to derive them if necessary.
Lecture 1B Nodal and Mesh Analysis

Know
Nodal analysis.
Mesh analysis.
Know that nodal analysis applies KCL and finds nodal voltages, whilst mesh
analysis applies KVL and finds mesh current. Know how to perform, by hand,
analysis on a three-node or three-mesh circuit composed of resistors, ideal
independent and ideal dependent sources.

13B.2
Lecture 2A - Circuit Analysis Techniques

Know
Source transformations.
Linearity and superposition.
Thvenins and Nortons Theorems.
You should be able to convert between a real voltage source and a real current
source (a special application of Thvenins and Nortons Theorems). Know
when linearity applies (and does not apply), and be able to use superposition
appropriately (e.g. DC and AC sources, common-mode and differential mode).
Know how to apply Thvenins and Nortons Theorems to any linear circuit,
even if it contains dependent sources.
Lecture 2B - Linear Op-Amp Applications

Know
Summing amplifier.
Difference amplifier.
Integrator.
Differentiator.
Know the basic functionality of the summing and difference amplifier,

integrator and differentiator. You should be able to derive output equations
from first principles if needed, using the basic circuit laws and the concept of
the virtual short-circuit.

13B.3
Lecture 3A - Reactive Components

Know
Capacitor and inductor construction and v-i relationships.
Stored energy relationships for inductors and capacitors.
Series/parallel connection of inductors and capacitors.
Circuit analysis with inductors and capacitors.
Familiar
Physical characteristics of inductors; parasitic effects.
Physical characteristics of capacitors; parasitic effects.
Aware
Duality.
Know the v-i relationships and the energy stored for C and L, and how they can
be combined in series and parallel. Know how to write nodal or mesh equations
for any circuit. Know how to perform DC analysis for any circuit.
Familiar with the fact that real components exhibit other characteristics, and we
initially study idealized forms of components. The most important nonideality
of the passive components is the resistance of the inductor windings.
Aware of the concept of duality be able to recognise circuits that are duals.
Lecture 3B - Diodes and Basic Diode Circuits

Know
Ideal diode.
Familiar
Diode models.
Rectifier circuits.
Limiting and clamping circuits.
LEDs.
Know the operation of an ideal diode.
Familiar with the fact that real signal diodes exhibit a 0.7 V voltage drop when
conducting, the principles of rectification and why we rectify, application of
diodes to limiting and clamping voltage signals, light emitting diodes have a
forward voltage drop different to 0.7 V (depends on the colour).

13B.4
Lecture 4A - Source-free RC and RL Circuits

Know
Differential operators.
Properties of differential operators.
The characteristic equation.
The simple RC circuit.
Properties of the exponential response.
Single time constant RC circuits.
The simple RL circuit.
Single time constant RL circuits.
Know how to write a differential equation using the D operator, write a

characteristic equation by inspection of a differential equation, analyse any
first-order RC or RL circuit to obtain the natural response, expressed in terms
of the appropriate time constant.
Lecture 4B - Nonlinear Op-Amp Applications

Familiar
Super diode.
Precision full-wave rectifier.
Logarithmic amplifier.
Op-amp comparators.
Familiar with nonlinear op-amp circuits such as the super diode, precision
rectification, logarithmic amplifiers and the op-amp operated in open-loop as
a comparator.

13B.5
Lecture 5A First-Order Step Response

Know
The unit-step forcing function.
The driven RC circuit.
Forced and natural response.
RC circuits.
Analysis procedure for single time constant RC circuits.
RL circuits.
Analysis procedure for single time constant RL circuits.
Know the definition of a unit-step as applied to a voltage or current source, that

a driven circuit has a forced response and a natural response, and that these
correspond to the particular solution and complementary solution to the
circuits describing differential equation, how to analyse any first-order circuit
with a step (DC) forcing function.
Lecture 5B - Op-Amp Imperfections

Know
DC imperfections (offset voltage, bias and offset currents).
Finite open-loop gain.
Familiar
Finite bandwidth.
Output voltage saturation.
Output current limits.
Slew rate.
Full-power BW.
Know how to calculate the effect of DC imperfections on an op-amp circuits

output, and derive gain expressions if the op-amp has finite gain.
Familiar with other limitations of the op-amp, as observed in the laboratory.

13B.6
Lecture 6A - The Phasor Concept

Know
Sinusoidal signals.
Sinusoidal steady-state response.
Complex forcing function.
Phasors.
Phasor relationships for R, L and C.
Impedance.
Admittance.
Know the general expression for a sinusoid, the concepts of amplitude and
phase, that a sinusoidal input yields a sinusoidal output in the steady-state
(sinusoid in = sinusoid out), complex forcing functions can be created from
real sinusoidal forcing functions, the concept of a phasor (visualize a rotating
phasor, and how to obtain the real part), phasor relationships for R, L and C and
the concepts of impedance and admittance.
Lecture 6B - Circuit Simulation

Familiar
DC analysis.
AC analysis.
Transients.
Familiar with the schematic capture process, simulation in the time-domain

(transient analysis) and frequency-domain (AC analysis).

13B.7
Lecture 7A - The Sinusoidal Steady-State Response

Know
Nodal analysis.
Mesh analysis.
Superposition.
Source transformations.
Thvenins Theorem.
Nortons theorem.
Phasor diagrams.
Power in AC circuits.
Know how to apply nodal and mesh analysis to circuits in the frequency-
domain, apply superposition, source transformations, Thvenins and Nortons
Theorems and draw phasor diagrams.
Lecture 7B - Amplifier Models

Familiar
Circuit models for amplifiers.
Cascaded amplifiers.
Efficiency.
I/O impedances.
Ideal amplifiers.
Frequency response.
Linear distortion.
Transfer characteristic.
Nonlinear distortion.
Familiar with the circuit models used for real amplifiers, cascading amplifiers,
efficiency, input and output impedances, ideal amplifier characteristics,
frequency response including the concepts of bandwidth and half-power, linear
amplitude and phase distortion, nonlinear distortion.

13B.8
Lecture 8A - Frequency Response

Know
Decibels.
Cascading two-ports.
Logarithmic frequency scales.
Bode plots.
First-order lowpass filters.
First-order highpass filters.
Aware
Fourier analysis, cascade filters.
Know the definition of the decibel and its application to amplifier circuits, how
to cascade circuits, logarithmic scales, Bode plots, how to derive the equations
for the frequency response for first-order lowpass and highpass filters, and
sketch their responses.
Lecture 8B - First-Order Op-Amp Filters

Know
Bilinear frequency response.
First-order lowpass filters.
First-order highpass filters.
Know the form of the bilinear frequency response, and be able to derive a
suitable op-amp implementation of it, how to implement lowpass and highpass
filters using cascaded op-amp circuits.

13B.9
Lecture 9A Second-Order Step Response

Know
Solution of the homogeneous linear differential equation.
Source-free parallel RLC circuit.
Overdamped parallel RLC circuit.
Critical damping.
Underdamped parallel RLC circuit.
Source-free series RLC circuit.
Complete response of the RLC circuit.
Know the solution of the second-order homogeneous linear differential

equation from inspection of the roots of the characteristic equation, the natural
response of the parallel RLC circuit for the overdamped, critically damped and
underdamped cases, the natural response of the series RLC circuit and the
complete response of any RLC circuit.
Lecture 9B - Waveform Generation

Familiar
Open-loop comparator.
Comparator with hysteresis.
Astable multivibrator.
Waveform generator.
Familiar with the open-loop comparator, hysteresis, the astable multivibrator

and waveform generation from laboratory experience.

13B.10
Lecture 10A Second-Order Frequency Response

Familiar
Lowpass, bandpass and highpass responses.
Resonance.
Bandwidth.
Quality factor.
Familiar with the concept of lowpass, bandpass and highpass frequency

responses for second-order systems. The concept of resonance, bandwidth and
quality factor for a second-order system and the relationships between them.
The concept of frequency selectivity and its application to filtering signals in
the frequency-domain.
Lecture 10B Second-Order Op-Amp Filters

Familiar
Filter design parameters.
The lowpass biquad circuit.
The universal biquad circuit.
Aware
Approximating the ideal lowpass filter.
The Butterworth lowpass filter.
Familiar with op-amp circuits that implement second-order systems, such as

the universal biquad filter, and their advantages and disadvantages. Aware of
the concept of cascade filter design, the ideal filter and various practical
response types such as the Butterworth response.

13B.11
Lecture 11A - Complex Frequency

Familiar
Complex frequency.
Damped sinusoidal forcing function.
Generalized impedance and admittance.
Frequency response.
The complex-frequency plane.
Visualization of the frequency response from a pole-zero plot.
Familiar with the concept of complex frequency, the damped sinusoidal forcing
function, generalized impedance and admittance, frequency response for both
and , and visualization of the frequency response from a pole-zero plot.
Lecture 11B - Specialty Amplifiers

Familiar
Differential and common-mode signals.
Difference amplifiers.
Aware
Instrumentation amplifiers.
Programmable gain amplifiers.
Isolation amplifiers.
Familiar with the concept of the differential and common-mode signals, and
why we split signals up in this way. Familiar with the difference amplifier and
its limitations.
Aware of the existence of other types of amplifier, which are based on the op-
amp, such as the instrumentation amplifier which is suited to specific
applications, such as those found in data acquisition and distribution systems.

13B.12
Lecture 12A - Transfer Functions

Familiar
Transfer functions.
Forced response.
Frequency response.
Natural response.
Complete response.
Familiar with the concept of the transfer function, how to find the forced
response and natural response from it.
Lecture 12B - Sensor Signal Conditioning

Familiar
Bridge circuits.
Aware
Strain, force, pressure and flow measurements.
High impedance sensors.
Temperature sensors.
Familiar with bridge circuits and their linearization.
Aware of a few sensors and their signal-conditioning circuits, with an emphasis

on bridge circuits which are found in process control systems and data
acquisition systems.

13B.13
Lecture 13A System Modelling

Familiar
Differential equations of physical systems.
Linear approximations of physical systems.
The transfer function.
Block diagrams.
Feedback
Familiar with the concept of system modelling and linear approximations,

derivation of the transfer function from the describing differential equation,
block diagrams, modelling of electromechanical systems and the principles of
feedback.

M.1
Matrices - Quick Reference Guide
Definitions
Symbol Description
aij Element of a matrix. i is the row, j is the column.
a11 a12 a13 A is the representation of the matrix with elements
A a21 a 22
a 23 aij aij .
a31 a32 a33
x1 x is a column vector with elements x j .
x x2
x3
0 0 0 Null matrix, every element is zero.
0 0 0 0
0 0 0
1 0 0 Identity matrix, diagonal elements are one.
I 0 1 0
0 0 1
0 0 Scalar matrix.
I 0 0
0 0
1 0 0 Diagonal matrix, aij 0 i j .
0 2 0
0 0 3
Multiplication
Multiplication Description
Z kY Multiplication by a scalar: zij kyij
z Ax n
Multiplication by a vector: z j aik xk
k 1
Z AB n
Matrix multiplication: zij aik bkj .
k 1
AB BA In general, matrix multiplication is not commutative.

M.2
Operations
Terminology Description
a11 a 21 a31 Transpose of A (interchange rows and columns):
A a12 a22
t
a32 aijt a ji .
a13 a23 a33
a11 a12 a13 Determinant of A.
A det A a 21 a22 a 23 If A 0 , then A is singular.
a31 a32 a33 If A 0 , then A is non-singular.
a11 a1 j a1n Minor of aij . Delete the row and column containing
. . . . . the element aij and obtain a new determinant.
aij ai1 aij ain
. . . . .
a n1 a nj a nn
Aij 1 aij
i j
Cofactor of aij .
A11 A21 A31 Adjoint matrix of A. Replace every element aij by its
adj A A12 A22 A32 cofactor in A , and then transpose the resulting
A13 A23 A33 matrix.
A 1
adj A Reciprocal of A: A 1 A AA 1 I .
A Only exists if A is square and non-singular.
Formula is only used for 3x3 matrices or smaller.
Linear Equations
Terminology Description
a11 x1 a12 x2 a13 x3 b1 Set of linear equations written explicitly.
a21 x1 a22 x2 a23 x3 b2
a31 x1 a32 x2 a33 x3 b3
a11 a12 a13 x1 b1 Set of linear equations written using matrix elements.
a a 22 a23 x2 b2
21
a31 a32 a33 x3 b3
Ax b Set of linear equations written using matrix notation.
x A 1b Solution to set of linear equations.
Eigenvalues
Equations Description
Ax x are the eigenvalues. x are the column
eigenvectors.
A I 0 Finding eigenvalues.

T.1
Textbook Answers
1A - Basic Laws and Op-Amp Amplifiers

Hambley
P1.9
i t 2 2t A
P1.20
(a) 30 W absorbed
(b) 30 W absorbed
(c) 60 W supplied
P1.29
40.32 hours, 1.008 kWh, $0.3373 / kWh
P1.36
ia 2 A , ic 1 A , id 4 A . Elements A and B are in series.
P1.46
(a) A, B, C+D (b) v a vb (c) vb 12 V , v d 5 V
P1.60
5J
P2.8
(a) 18 (b) 10
P2.19
R1 R2 R3
Is
To turn a load off, the corresponding switch needs to be closed. The short-
circuit created by the switch will cause 0 V across the load, and it will carry
all the current hence no voltage across, or current through, the load.

T.2
P2.34
i1 1.5 A , i2 0.5 A
P4 A 30 W delivering
P2 A 15 W absorbing
P5 11.25 W absorbing
P15 3.75 W absorbing

P2.35
5 5
i1 A , i2 A
2 6
P2.38
10
v V
3
P2.43
Rg 25 m
P2.45
76
12.6 V 5V 50
P14.10
AV 8
P14.17
For an op-amp, the fact that the differential input voltage and the input
current are forced to zero is called the summing-point constraint. It does not
apply when positive feedback is present.
P14.21
R R2 v A RB vB RA
vo 1
R1 RA RB

T.3
P14.24
(a) vo R f iin
(b) Since vo is independent of RL the output behaves as a perfect voltage

source, and the output impedance is zero.
(c) The input voltage is zero because of the summing-point constraint, and
the input impedance is zero.
(d) This is an ideal transresistance amplifier.
P14.27
R22
(a) G
R1 R L
(b) G
The noninverting amplifier has the larger power gain.
P14.32
R
io 1 1 iin
R2
Rin 0
The output impedance is infinite.

T.4
HKD
2.19
(a) 1.1 s
(b) See figures below
0.14 80
0.12
60
0.1
40
0.08
q (C)
i (A)
0.06 20
0.04
0
0.02
-20
0
-0.02 -40
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
Time (s) Time (s)
2.26
(a) 6 pW
(b) 10 mW
(c) -20 W
2.36
The 4 A fuse is the optimum choice among the values available. 10 A is too
large, and the 1 A fuse only allows operation at minimum power.
2.47
Start with VS I R1 I 2 R2 I 2 .
3.2
(a) 4 nodes
(b) 7 elements
(c) 6 branches
3.8
(a) 1 A
(b) 3A. There is a typographical error in the 1st printing. The source should
be labelled 9 A.
(c) 0 A
3.23
(a) v1 2 V , v 2 2 V , v3 13 V , v 4 5 3 V , v5 5 3 V
i1 10 13 A , i2 13 A , i3 10 A , i4 9 23 A , i5 13 A
(b) p1 20 23 W , p 2 23 W , p3 103 W , p 4 16 19 W , p5 95 W

T.5
3.40
-1.29 A
3.46
i3 245
452 A , p x 1 452
283
W
3.57
697.0 mV
6.1
(a) -5 V
(b) -200 V
(c) 200 sin 5t V
6.6
1k0
1k8
5V
9V
Note the inverted 9 V.

6.12
27.0 V
6.14
(a) No current can enter either input pin of an ideal op-amp.
(b) There can be no voltage difference between the input pins of an ideal
op-amp.
6.29
1 1 R3 1
vout R4 R6 vin
R1 R4 R2 R3 R1

T.6
1B - Nodal and Mesh Analysis

Hambley
P2.48
2 3 6
v1 14 V , v2 11 V , i1 2 A
7 7 7
P2.57
200 270
v1 V , v2 V
37 37
P2.66
v2 0.5 V , P 6 W
P2.67
2
i1 A
7

T.7
HKD
4.13
8.833 V, 10.65 W
4.16
-2.268 A
4.22
500 mW
4.25
-1.636 A
4.34
(a) 318.4 A
(b) 0 W
4.38
3.257 W
4.44
i1 874.3 mA , 3.141 W
4.50
22.5 W
4.54
(a) 0 V
(b) 0 V
(c) No significant difference - nodal analysis is slightly easier in this case.

T.8
2A - Circuit Analysis Techniques

Hambley
P2.80
10
3
20 10
V 2A
3 3
P2.81
Rt 50
P2.91
Rt 0 , Pmax 80 W
P2.94
iv 2 A , ic 2 A , i iv ic 4 A
P2.95
10
is A
3

T.9
HKD
5.4
(a) 7.0 A
(b) 1.1%
(c) 111.1 mA
5.11
339 mA
5.18
5.249 W
5.21
-4.160 V
5.28
(a) I sc 1.364 mA , RTh 1.04 k
(b) 289 W
5.40
I N 116.3 mA , RTh 999.8
5.41
(a) 5.587 nW
(b) 1.282 nW
(c) 578.5 pW
5.45
Ri Ro AR f
VTh vin
R1Ro Ri Ro R1 R f Ri R f R1Ri AR1Ri
Ro Ri R f R1 R f R1 Ri
RTh
R1 Ro Ri Ro R1 R f Ri R f R1 Ri AR1 Ri

T.10
2B - Linear Op-Amp Applications

Hambley
P14.37
One possible solution is:
200 k *
R s2 100 k *
v2 100 k *
R s1
*1%-tolerance
resistors
v1
P14.42
2 M
200 k
v1
vo
v2
150 k
56.2 k
P14.70
100 k
10 k
v1
vo
v2
10 k
100 k

T.11
P14.74
vo ( t) (V)
0 2 4 6 8 10 t (ms)
-0.5
-1
20 pulses are required to produce vo 10 V .
P14.75
v o ( t ) (V)
10
t (ms)
0 1 2 3 4
-5

T.12
3A - Reactive Components
Hambley
P3.13
vC(t ) (V)
10
0 1 2 3 4 5 t (ms)
p ( t ) (W)
0.1
4 5
0 1 2 3 t (ms)
-0.1
-0.2
w (t ) (J)
100
0 1 2 3 4 5 t (ms)
P3.22
p 0 0 W , p 0.5ms 93.02 mW into the capacitor
P3.31
C 398 nF
P3.32
W1 500 J , C2 500 pF
V2 2 kV , W2 1 mJ
The additional energy is supplied by the force needed to pull the plates
apart.

T.13
P3.43
vL (t ) (V)
60
0 0.1 0.2 0.3 0.4 t (s)
-60
p ( t ) (W)
180
0.2 0.4
0 0.1 0.3 t (s)
-180
w ( t ) (J)
9
0 0.1 0.2 0.3 0.4 t (s)
P3.57
v 10 V , i 20 A , leave

T.14
HKD
7.5
(a) 91.64 pF
(b) 3.321 nF
(c) 3.32 pF
(d) 13.28 pF
7.11
(a)
150
100
v (V)
50
0
0 0.2 0.4 0.6 0.8 1 1.2 1.4
Time (s)
(b) 48.48 V, 72.72 V, 121.2 V

7.17
(a)
v L (mV)
150
2 2.5
-1.5 -1 -0.5 0 0.5 1 1.5 3 t (ms)
-150
(b) 0 V, 0 V, 0V
7.25
(a) -10.65 mA
(b) -10.65 mA
(c) -8.473 mA

T.15
7.29
(a) 18.64 V
(b) 16.38 V
(c) 25.79 V
(d) 16.38 V
7.35
3.62 F
7.37
1.5R
2C 0.5L
2V C
R R
7.36
6.639 H
7.49
vC t 9.2 V , vL t 2.4 sin 103 t V
7.50
v1 v2
50
t

0.02 103 v4 40e 20 d 20 10 3 e 20t
0
0.8v1 0.2v2 v4 0
v2 v1 v2 40e 20 dv
106 2 0
50 100 dt
7.64
1
6H 3
1
4
-0.2 t
10e A 7H 1
1F
5
8H 2F

T.16
3B - Diodes and Basic Diode Circuits

Hambley
P10.14
n 1.680
P10.16
vx 2.199 V , ix 0.8013 A
P10.26
Rs
Vs Ro load
variable
source
voltage
P10.36
(a) I 3.704 mA , V 0 V (b) I 0 A , V 10 V
(c) I 0 A , V 0 V (d) I 5 mA , V 5 V
P10.37
(a) I 0 A , V 10 V
(b) I 6 mA , V 6 V
(c) I 33.64 mA , V 30 V
P10.54
C 833 F
ideal load
10cos ( t) C R L = 90
P10.70
vo( t )
15
10
20 t (ms)
-5 10

T.17
4A - Source-free RC and RL Circuits

HKD
8.7

(a) i L 0 4 mA
(b) i L 0 4 mA
(c) 703.5 mA
(d) 125.1 mA
(e) 668.0 A
8.10
(a) iW t 300 100e t 210 A
9
(b) 200 A, 247.8 A

8.11
(a)
10
6
f(t)
0
0 0.5 1 1.5 2 2.5
Time (s)
(b)
1
10
0
10
log10 f(t)
-1
10
-2
10
0 0.5 1 1.5 2 2.5
Time (s)
(c) s -1
(d) 52.68 ms, 111.6 ms, 1.151 s

T.18
8.17
(a) vt 9e t 2210 V
3
(b) 327.8 mW, 44.36 mW

(c) No significant change.
8.22
(a) 169 16 ms
60 5.91110 3 t
(b) vt e V
83
(c) 212.3 mW
8.38
(a) 3 11
23 V
(b) 3 11
23 V
(c) 3.2 ms
(d) 1.362 V
8.40
(a) 54 nJ
(b) 9.384 nJ
(c) 8.557 pJ

T.19
5A - First-Order Step Response

HKD
8.46
There is a typographical error in the 1st printing. The source should be
labelled 9u t 0.001 V .
(a) 0 A
(b) 0 A
(c) 0 A
(d) 0 A
(e) 2.993 mA
8.48

(a) i L t 33 15
11

33 13 e 75000t u t 52 u -t mA
(b) 17.99 mA, 26.30 mA, 32.95 mA
8.52
(a) i L t 9
80
1 e 1600 t
ut 9
40 u -t A
(b)
iL (A)
9/40
9/80
t (ms)
-1 0 1 2 3
8.54

i1 t 52 1 e t 2610 u t A
9

8.58
(a) 85.93 mA
8.62

(a) vC t 53 1 e 15010 t u t V
3

(b) 0 W, 3.433 W, 3.6 W

T.20
5B - Op-Amp Imperfections
Hambley
P14.52
| A ( f )|
(dB)
106
100 open loop
80
slope = -20 dB / decade
60
A0 CL= 100
40
A0 CL= 10
20
f t = 106 Hz
f (Hz)
1 5 10 100 103 104 105 106
P14.59
SRmin 6 V / s
P14.61
(a) f FP 15.92 kHz
(b) 2.5 V (output current limited)
(c) 10 V (output voltage limited)
(d) 1.592 V (slew rate limited)

T.21
6A - The Phasor Concept

HKD
10.2
(a) 5 cos300t 90 , 1.95 cost 182 , 10.13 cos50t 164.6
(b) 66 sin 9t 80 , 4.15 sin 10t 90 , 17.14 sin 100t 50
10.4
(a) 235
(b) 155
(c) 195
(d) 275
(e) 126
10.10
There is a typographical error in the 1st printing. The source should be equal
to 4.53 cos3000t 30 V .
(a) 359.1 mA, 26.61 mA, 332.5 mA
(b) 409.7 cos3000t 35.77 mV
10.15
1.233 cos 2 6t 20.56 W
10.18
(a) 32 245
(b) 2.060 4.706
(c) 9.487108.4
(d) 2.037 99.56
(e) 376.539.80
10.24
1.129 cos35t 128.3 A
10.26
(a) 11 90
(b) 110
(c) 11 90
(d) 3 245

T.22
10.34
(a) An inductor.
(b) 20 mH
10.40
(a) 23.25 12.92
(b) 12..09 13.82
(c) 11.141.577
10.42
(a) 15.0033.13 V
(b) 15.00 73.13 V
(c) 9 20 V
(d) 20.1243.43 V

T.23
7A - The Sinusoidal Steady-State Response

HKD
10.47
There is a typographical error in the 1st printing. The current source should
be equal to 2 cos100t A .
(a)
-j 50
11
V1 V2
3
3 62A j1 2 5 2 0 A
100
-j 47
(b) v1 t 2.56 cos100t 139.2 V , v 2 t 4.35 cos100t 138.3 V

10.60
i1 t 27.12 cos100t 132.5 A
i2 t 6.141 cos100t 33.98 A
i3 t 29.24 cos100t 132.5 A
i4 t 1.818 cos100t 0.6300 A
10.61
Left-hand source contributions: 5.58 91.8 V , 1.29 75.9 V
Right-hand source contributions: 1.29 75.9 V , 9.08 115 V
10.70
a
2- j 4
A 2.5 j1.25
5

T.24
10.71
(a)
I x = I R + IL
j5 IL
V , IR
-45 -40 -35 -30 -25 -20 -15 -10 -5 0 5 10
-j 5
-j 10
I s = I C+ I x -j 15
-j 20
-j 25
IC
(b) There is a typographical error in the 1st printing. The last phasor current
should read I x .
I s leads I R by 83, I s leads I C by -7, I s leads I x by 146
10.74
(a) I 1 3 30 A , I 2 2.05830.96 A , I 3 2.400 53.13 A
(b)
j2
I2
2 Vs 4 6
0 I1 + I2
I1
-j 2 I3
Is = I1 + I2 + I3
I s 6.265 22.14 A

T.25
7B - Amplifier Models
Hambley
P11.9
vo 30 cos(100t ) , G 300 W/W
P11.16
The amplifier must have a large current gain, Ai 100 A/A , Ri RL 1000 .
P11.40
Avoc 100 , Aisc 500 , Rmoc 100 k
P11.62
A transresistance amplifier, with Ri 10 , Ro 200 , Rmoc 1000 .

T.26
8A - Frequency Response
Hambley
P6.8
vout t 10 3.5 cos5000t 15 2.5 cos15000t 135
P6.25
vout t 4.472 cos500t 26.57 3.535 cos1000t 45 2.236 cos2000t 63.43
P6.40
(a) H f 0.3162
(b) H f 3.162
P6.46
1
(a) H f
1 j f f B 2
(b) f 3dB 0.6436 f B
P6.59
1
H f
R
, fB 1 MHz
1 j f f B 2L
20
0
-20
-40
-60
10 kHz 1 MHz 100 MHz
-45
-90
10 kHz 1 MHz 100 MHz

T.27
P6.69
j f f B 1
H f , fB 39.79 Hz
1 j f f B 2RC
20
0
-20
-40
-60
100 mHz 10 Hz 1 kHz
90
45
100 mHz 10 Hz 1 kHz

P6.93
(a) lowpass: R2 and C 2 , highpass: R1 and C1
(b) C1 1.592 F , C 2 159.2 pF
jR1C1
(c) H f
1 jR2 C 2 1 jR1 C1 C 2
Magnitude response - bandpass circuit
10
-10
|H(f)| dB
-20
-30
-40
-50
0 1 2 3 4 5 6
10 10 10 10 10 10 10
f (Hz)

T.28
9A - Second-Order Step Response

HKD
9.4
(a) 10 , 0 6.003 , s1 18.00 , s 2 2.002
(b) 3.131 1012 , 0 3.162 1010 , s1 6.261 1012 ,

s1 1.597 10 8
(c) a 1.666 , b 99
9.6
There is a typographical error in the 1st printing. The exponential damping
coefficient should be 10 4 s -1 .
(a) 1 H 1 Vs/A ,
(b) s1 50.13 , s 2 1.995 10 4
(c) vn K1e s1t K 2e s2t
(d) The proof relies on the fact (which is difficult to show) that with the
current defined up (which is strange) in the inductor of Fig. 9.1 then
1K K
it 0 1 2 .
L s1 s2
9.17
it 312.1e 15.16t 2.136e 183.3t mA , vt 4.731e 15.16t 0.3914e 183.3t V
9.20

(a) wC t 3.870 e 19398t 2e 10t e 19.98t mJ
(b) 460.5 s
9.23
(a) 0 1562500 s -1
(b) You can never achieve infinite accuracy with real physical components.
(c) 180.3 V, 185.4 V, 630 ns
9.28
(a) 282.2 H
(b) 1.850 mJ, 4.174 mJ

T.29
9.32
There is a typographical error in the 1st printing.
The resistor should be 15 .
(a) 100
3
s -1 , d 314.5 s -1 , 0 316.2 s -1
(b) it 2.862e
100 t
3
sin 314.5t
(c) 2.437, 4.659 ms
9.36
v L t e 4000t 5.769 cos 2000t 588.5 sin 2000t
9.46
(a) 35
6
s -1
(b) 0 1 6 s -1
(c) 0.5 A
(d) 0 As -1
(e) 0.4589 A
9.49
i1 t 99.23e 6.154t sin 44.30t A
9.50
(a) 50000 s -1 , 0 1000 5 s -1
(b) 3 mV, 3 mA, 3 mV, 3 mV, 3 mA, 3 mV, 2 mA, 2 mV
9.56
(a) vC t 5.127e 2.042t 0.1273e 97.96t V
(b) 4.976 V, 1.506 V

T.30
10A - Second-order Frequency Response

Hambley
P6.72
VL =10 90
f 0 1.125 MHz
Qs 10
B 112.5 kHz VR =1 0
f H 1.181 MHz
f L 1.069 MHz
VC =10 -90
P6.75
L 79.57 H
C 318.3 pF
VC 20
P6.79
f 0 1.592 MHz
QP 10
B 159.2 kHz

T.31
12B - Sensor Signal Conditioning

Hambley
P9.16
To avoid ground loops, we must not have grounds at both ends of the 5 m
cable. Because the sensor is grounded, we need to use a differential
amplifier. To reduce interference from magnetic fields, we should use a
twisted pair or coaxial cable. To reduce interference from electric fields we
should choose a shielded cable and connect the shield to ground at the
sensor. A schematic diagram of the sensor, cable and amplifier is:
shielded twisted
R sensor pair
vsensor
differential
amplifier

A.1
Answers
1A.1
(a) To the left (b) 300e 100t mA (c) 4.055 ms (d) 0.3820 A/mm2
q i
(mC) (mA)
3 300
1.8963
110.4
t (ms) t (ms)
(e) 0 10 20 30 40 0 10 20 30 40
(f) 1.860 1016
1A.2
(a) 10 C (b) 10 C (c) 12.71 mA
1A.3
(a) 975 C (b) 383 C
1A.4
(a) 31.1 kC (b) 48 W (c) 373 kJ (d) 24.9 W
1A.5
8.0 C
1A.6
(a) 120.8 V (b) 8.453 kW (c) 754.0 W/mm2
1A.7
(a) 10 V (b) 5 A (c) 50 W
1A.8
(a) -3 A (b) 3 V (c) 15 W

A.2
1A.9
-30 W
1A.10
Pvx 68.48 W , P50 V 168.9 W , P0.2 vx 13.70 W , P10 114.1 W
1A.11
(a) -20 V, -20 mA, 50 mA (b) 26 23 V , 26 23 mA , 3 13 mA
1A.12
(a) 6 23 V , 3 13 A , 66 23 W (b) 20 V, 10 A, 200 W
1A.13
(a) 8 (b) 3.7
1A.14
(a) 2.5 A (b) 4 V
1A.15
R1 R3 R2 R3
v3 is i1 is
R1 R2 R3 R1 R2 R3
1A.16
5.5 V, 3.975 A
1A.17
(a) 30 W (b) -2 A
1A.18
(a) 0 V, 1 mA, 1 mA, -10 V, -10 mA, -11 mA
(b) -10 V/V (c) -10 A/A (d) 100 W/W

A.3
1A.19
R2
(c) (b) R1 (c) same results, thanks to the VSC
R1
1A.20
R2 R
(a) 1 (b) 2 (c) different result for (b), due to the positive feedback
R1 R1 R3
1A.21
R1 can be any non-zero value, R2 200
1A.22
(a) R (b) 0
1A.23
R R1
(a) R1 1 1 R3 (b) 0 (c) 1
R2 R2
1A.24
R Ri
(a) 1 2 Ri (b) Ri (c)
R1 R1

A.4
1B.1
(a) -33 (b) 17, -34, -11
1B.2
2A
1B.3
25.64 W
1B.4
(a) 6 A (b) 3 A
1B.5
-3.5 mA
1B.6
20 mA, -80 mW

A.5
2A.1
80 W
2A.2
4A
2A.3
(a) 150 V (b) 110 V
2A.4
(a) 15 A, 2 (b) 2 (c) 112.5 W
2A.5
381 mW
2A.6
8 9

A.6
2A.7
R2 I S
(a) i A
rm R2
(b) It is indeterminant i A A . For this special case you can show that
KVL is violated the circuit becomes:
- rm IS iA
Thus, the branch containing i A , which has 0 V across it, appears in

parallel with a Thvenin equivalent ideal voltage source equal to
VTh rm I S , which is impossible.
Note that there is no Norton equivalent circuit!
(c) In a practical setting, it is impossible to create ideal circuit elements, so the

real circuits behaviour would be perfectly explainable and measureable if
modelled correctly (inclusion of finite wire resistance, linear range of
power supplies, etc).
A large but finite current i A would result either briefly (before a fuse
blows or a protection device trips), or continuously (limited by the power
supplys output current and voltage capability).

A.7
2B.1
25 nF
10 k
vi
vo
2B.2
100 k
1 M
100 k
v1 1 M
vo
100 k
v2
2B.3
vo v1 v 2 2v3
2B.4
10 k
10 k
1 k
vi vi 10 mV when the source is attached
2B.5
vi
io
R

A.8
3A.1
(a) 20.6 ms (b) 177.7 ms
3A.2
2 t 0.001 A , 0.01 / t 0.001 V
3A.3
(a) 9.6 V, 192 mW, 1.152 mJ (b) 16 V, 0 W, 3.20 mJ
3A.4
(a) 2 nF (b) 2.4 nF
3A.5
12 12
(a) F (b) F (c) 9 F
7 11
3A.6
(a) 60 cos10t V (b) 5 2 sin 10t A
3A.7
2 10 4 v 25 vdt i s i L 0
dv t
(a) 2 10 4
dt 0
5000i 5000 idt v s vC 0

di t
(b) 0.04
dt 0
3A.8
Use v1 , v 2 , vC left to right.
v1 vC 3v1
v1 vC v 2 dv
10 v1 v 2 dt 0.5 0.8e 100t
t
2 10 4 C 0
0 20 50 dt
v 2 vC
10 v 2 v1 dt 0.5 0.8e 100t
t
0
0 50

A.9
4A.1
100 , 2 F
4A.2
(b) 50 mC (b) 38.9 mC
4A.3
76.10 V
4A.4
3et 0.003 mA
4A.5
t 122.6 s (over 2 minutes)
4A.6
6.75e t 40 A
4A.7
153.7 ms
4A.8

(a) 9.6 A (b) 2.4 A (c) 9.6e 2t A (d) 2.4e 4t A (e) 19.2 e 4t e 2t V
4A.9
(a) 28.95 ms (b) 144.3 ms
4A.10
(a) 800 mA (b) 280 mA
4A.11
10e 80t V
4A.12
60 4e 250t 40e 200t mA

A.10
5A.1

20 1 u t 0.2n V
n
n 0
5A.2

50 1 e 10t u t V , 20 25e 10t u t V
5A.3

20 1 e 10 t 3 u t V
6

5A.4
(a)
100u - t 40 60e 6250t u t V
(b) 100 V
5A.5
76.28 k , 62.13 F
5A.6
10 30e 40000t V
5A.7
22.31 mA, 9.812 mA
5A.8
(a)
0.4 1 e 1250 t A
(b)
10 23 e 100t e 1250t A
(c)
0.2 2 cos1250t 45 e 1250t A
5A.9
40 200e 15000 t
ut V
5A.10
(a) 0 W (b) 200 W (c) 131.7 W (d) 0 W

A.11
6A.1
(a) 12.5 ms, 80 Hz, 502.7 rad/s
(b) 27.77 cos160t 43.92 V
(c) 66.08
6A.2
(a) 8.00 and 38.68 (b) 11.17 ms (c) 89.54 Hz (d) 562.6 rad/s
(e) 8.00 and 128.7 (f) 0.8983
6A.3
412.3 cos500t 116.0 V
6A.4
80 cos2000t 36.87 mA
6A.5
2 cos5t 45 1.342 cos10t 63.44 A
6A.6
(a) 18.83133.5 (b) 5.584 56.87
(c) 0.07248 j 0.04702 (d) 0.05243 j 0.1838
6A.7
15.98 cos1000t 71.23 4.598 cos500t 38.22 V
6A.8
(a) 95.79 cost 94.01 mA (b) 25.14 111.6 mA
6A.9
(a) 143.5 W (b) -135.0 W

A.12
6A.10
(a) 39.99 W (b) 9.512 W (c) -9.512 W
6A.11
(a) 26.00 47.38 mA (b) 0.769237.38 A
6A.12
(a) 384.2 and 65.92 Hz (b) 203.8 and 124.3 Hz
6A.13
1.25 mH
6A.14
125 nF
6A.15
1 , 1 H

A.13
6B.1

A.14
7A.1
(a) 7.809 128.7 V (b) 4 6
7A.2

(a) 6.468 cos 10 4 t 44.04 V
(b) 3.234 cos 10 4 t 44.04 A
7A.3
(a) 11.09 cos10 4 t 33.69 A (b) 3.288 cos10 4 t 170.5 A
7A.4

(a) 4 2 cos 106 t 45 V (b)
2 cos 106 t 135 V
7A.5
6 j17
7A.6
1.581 18.43 A
7A.7
65.05 and 60.72 , or 112.8 and 13.00
7A.8
p R1 4.8 mW , p L 110.9 mW , pC 110.9 mW and p R2 19.2 mW
7A.9
(a) P20 10 kW , P10 5 kW (b) P20 3.125 kW , P10 6.25 kW
7A.10
6.622 mA

A.15
7A.11
300
(a) 106.1 (b) 61.24 V RMS
2 2
7A.12
(a) 4.471 A RMS (b) 0.9150 lagging
7A.13
(a) 46.86 kW (b) 33.32 kvar (c) 57.535.41 kVA
(d) 57.5 kVA (e) 9235.41
7A.14
There are two possible solutions:
S V 98.04 j 35.57 VA S V 118.0 j124.4 VA

or
S I 81.96 j124.4 VA S I 298.0 j 35.57 VA

A.16
8B.1
| T(j ) |
(dB)
0 dB 10 3 10 4 10 5 10 6 10 7 10 8
rad/s (log scale)
-20
-20 dB/decade
-40
-60
-80 -40 dB/decade

-100
8B.2
100 k
1 F
10 k 10 nF
8B.3
10 k 10 k 10 k 10 k
vi vo
100 nF 20 nF 100 nF 20 nF

A.17
9A.1
(a) 200e 2t 100e 8t V, t 0 (b) 100e 2t 200e 8t V, t 0
9A.2
(a) 40e 500t 55e 2000t V, t 0 (b) 15.00 V and 17.00 V
9A.3
(a) 25 (b) 1.28 mJ
9A.4
(a) 14
125 H, 1
70 F (b) 3.630 J
9A.5
12.89 kV
9A.6

200 e 5000t e 15000t V, t 0
9A.7
9.992 kV
9A.8
(a) 500 J in L, 80 J in C (b) 335.4 J in L, 62.21 J in C
9A.9
20 250e 100t e 150t V , 1
24 3e 150 t
2e 100t A
9A.10
(a) 50 25 3 16e 2000t e 8000t V, t 0
(b)
2 5 3 4e 2000t e 8000t A, t 0
9A.11
(a) 0 A (b) 1.748 A (c) -1.073 A

A.18
10A.1
(a) 1000 rads-1 , 5 (b) 120 krads-1 , 60 (c) 602.1 rads-1 , 6.021
10A.2
(a) 4 krads-1 , 40, 80 cos4000t V
(b) 2 cos4000t mA , 400 sin 4000t mA , 400 sin 4000t mA
(c) 20 mW, 4 mJ
10A.3
115.5 rads-1
10A.4
(a) 4.999 Mrads -1 (b) 49.99 (c) 100.0 krads -1
(d) 4.949 Mrads -1 (e) 5.049 Mrads -1 (f) 40 88.85 mV
10A.5
5 k , 2.360 H , 4.237 nF
10A.6
(a) 5 Mrads-1 (b) 20.59 krads-1
10A.7
4.472 krads-1 , 22.36
10A.8
(a) 5 krads-1 , 40.01 (b) 5 krads-1 , 20.03
10A.9
10 , 514.3 mH, 875 F

A.19
10B.1
(a) If we analyse the circuit:
R
vo
R1 R2
v i1
R5
v i2 C
then the output voltage phasor is found to be:
2 R2 R5 Vi1 R1Vi 2
Vo
jCR1 R2 R5 R1 R2 R2 R5 R1 R5
Note that we must have R2 R1 || R5 for the circuit to be a pure
integrator. If R1 R5 2 R then R2 R and we have:
2Vi1 2Vi 2 1 1
Vo Vi1 Vi 2
jCR1 jCR5 jCR jCR

A.20
(b) The circuit is a noninverting integrator, so that it accomplishes the same

objective as cascading an inverting integrator and an inverter. A
normalized version of the biquad with the noninverting integrator is
shown below:
R5 Q0
R 1
1
vo
2 2
H 1+ H
vi
R1 R2
1
Note that we must have R2 R1 || R5 for the noninverting integrator to be
a pure integrator, and if H 1 then R1 R5 2 and R2 1 .

A.21
10B.2
(a) One possible solution is:
100 k
1 nF 1 nF
100 k
50 k
vi 100 k 100 k 500 k
vo
(b) One possible solution is:
20 k
10 nF 10 nF
20 k
20 k 160 k
20 k
vo
10 nF
vi
10B.3
1
T j
1 j 1 Q0
2
10B.4
j 2Q0
T j
1 2 j 1 Q0

A.22
11A.1
(a) -10, -40 (b) 16 j12
11A.2
(a) 5.754 58.50 mA (b) 7.211 33.69 mA
11A.3
(a) 10.00 A (b) -1.995 A (c) -97.01 mA (d) 53.90 mA
11A.4
2.508 J
11A.5
Zeros: s 3333, ; poles s 2500, 10 000
V()

Is ()
5000
-20 -10 0 10 20 (kHz)
11A.6
(a) 100 (b) 12.5 H (c) 689.7 F
11A.7
(a) 53.85 21.80 from zero at s 50 ,
53.85 68.20 from pole at s 20 j 30 ,
22.36 26.57 from pole at s 20 j 30
(b) 1.137 42.13

A.23
11B.1
R R
(a) vo1 1 1 vi1 1 vi 2
RG RG
R R
vo 2 1 2 vi 2 2 vi1
RG RG
(b) Ad 201
(c) 1.005 V
(d) 200.2
11B.2
Ad 10

A.24
12A.1
(a), (b) and (c) -2 and -5
12A.2
15 8 2e t cos2t 45 9 8 e 3t u t V
12A.3
6 5 4 5 e 25 t

2e 10t u t A
12A.4
(a) 0 (b) 10 s 17.5 (c) 0 (d) Ae 17.5t V
12A.5
(a)
I Vs 5 s 2 s 2 2s 5 , s 2, 1 j 2
(b) i f t 1 A
(c) in t Ae 2t Be t cos2t Ce t sin 2t
(d)
i t 1 e 2t e t sin 2t u t A
12A.6
R1 sL R1 I V
1
1
(a) s
R1 R1 R2 I 2 0
sC
R1
s
(b)
I2

R1 R2 L
Vs R1 R2 1 R1
s 2 s
R1 R2 L R1 R2 C R1 R2 LC
(c)
i2 t e 1000 3t e 500t u t

A.25
12B.1
VB
(b) vo R
R1 R2

A.26
13A.1
(a)
1
RF
Vin1 -1 -1/C1 Vout1 R 3 1 -1/C5 Vout3

R0 s + 1/R 1C1 -R R4 s + 1/R 5C5
2
1
R7
-1 Vout4
R8 - R6
(b) Considering just the output Vout3 , this can be reduced to:
Vin1 -1 - R 3 / R 2 R 4 C1 C 5 Vout3
R0 (s + 1/R 1C1 ) (s + 1/R 5C5 )
1
RF
Further reduction gives:
Vin1 R 3 /R 0 R 2 R 4 C1C5 Vout3

(s + 1/R 1C1 ) (s + 1/R 5C5 ) + R 3 / R 2 R 4 R F C1C5
Expansion of the denominator results in the given transfer function.
(c)
R1 R3 R5 RF
K1
R0 R1 R3 R5 R2 R4 RF
1 1 1

2 R1C1 R5C5
R3 1
0
R2 R4 RF C1C5 R1 R5C1C5

A.27
(d)
1 1 1 1 2 1

2 R1C1 R1C1 2 R1C1 R1C1
R3 1 1 R3 1
d 2 2
2 2 2 2
R2 R4 C1 R1 C1 R1 C1 R2 R4 C1
(e) From the block diagram, we can see that:
Vout 3 R3 1

Vout1 R2 R4C5 s 1 R5C5
Then:
s 1 R5C5
Vout1 Vout3 Vout3 R0C1

Vin1 Vin1 Vout1 1 1 R3 1
s 2 s
R1C1 R5C5 R2 R4 RF C1C5 R1 R5C1C5
For the special case of R1 R5 , C1 C5 and RF R4 , this reduces to:
Vout1

1 s K s
Vin1 R0C1 s 2s 0 s 2s 02
2 2 2 2
(f) The output voltage Vout 4 is given by:
R6 R
Vout 4 Vin1 6 Vout1
R8 R7
Then substituting for Vout1 gives:
R6 R 1 s
Vout 4 Vin1 6 Vin1
R8 R7 R0C1 s 2s 02
2
The transfer function is then:
s 2 2s 02 s
T3 s
R6 R8
2 2
s 2s 0 R0 R7C1 s 2s 0
2 2
R8

A.28
Substituting the special conditions on the values gives:
R8 R8
s 2 2 s 02
R0 R7C1 R0 R7C1 s 2 02 2 2
T3 s 2
s 2 2s 02 s 2s 02
(g) The pole-zero plots for T1 s , T2 s and T3 s are respectively:
j j j
j 0 j 0 j 0
j d j d j d
j 02 2 2
-j 02 2 2
-j d -j d -j d
-j 0 -j 0 -j 0
(h) Lowpass, lowpass, notch.

Electronic and Circuits

Uploaded by

Copyright:

Available Formats

Electronic and Circuits

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Electronic and Circuits

Uploaded by

Copyright:

Available Formats

48520

Electronics and Circuits 2013

LECTURE 1B NODAL AND MESH ANALYSIS

LECTURE 2A CIRCUIT ANALYSIS TECHNIQUES

Electronics and Circuits 2013

LECTURE 3A REACTIVE COMPONENTS

LECTURE 4A SOURCE-FREE RC AND RL CIRCUITS

Electronics and Circuits 2013

LECTURE 6A THE PHASOR CONCEPT

Electronics and Circuits 2013

LECTURE 7A SINUSOIDAL STEADY-STATE RESPONSE

Electronics and Circuits 2013

LECTURE 9A SECOND-ORDER STEP RESPONSE

Electronics and Circuits 2013

LECTURE 10B SECOND-ORDER OP-AMP FILTERS

LECTURE 11A COMPLEX FREQUENCY

Electronics and Circuits 2013

LECTURE 12A TRANSFER FUNCTIONS

LECTURE 12B SENSOR SIGNAL CONDITIONING

Electronics and Circuits 2013

LECTURE 13B REVISION

MATRICES - QUICK REFERENCE GUIDE

Electronics and Circuits 2013

Lecture 1A Basic Laws & Op-Amp Amplifiers

Electronics and Circuits 2013

The unit of current is the ampere (A) and is equivalent to Cs -1 . In a circuit

A current which is constant is termed a direct current, or simply DC. Examples

Electronics and Circuits 2013

Note the voltages across the circuit elements below:

In both (a) and (b), terminal B is 5 V positive with respect to terminal A.

In both (c) and (d), terminal A is 5 V positive with respect to terminal B.

Electronics and Circuits 2013

Circuit Elements and Types of Circuits

Circuit elements can be categorised as either active or passive.

Active Circuit Elements

Active circuit elements can deliver a non-zero average power indefinitely.

Passive Circuit Elements

Passive circuit elements cannot deliver a non-zero average power indefinitely.

There are three types of passive circuit element. They are:

The interconnection of two or more circuit elements forms an electrical

Electronics and Circuits 2013

The Independent Voltage Source

An independent voltage source is characterised by a terminal voltage which is

Electronics and Circuits 2013

Electronics and Circuits 2013

An independent current source establishes a current which is independent of Independent current

In other words, an ideal current source is a device that, when connected to

Electronics and Circuits 2013

The Resistor and Ohms Law

Consider a uniform cylinder of conducting material, to which a voltage has

We refer to a construction in which Ohms Law is obeyed as a resistor.

Electronics and Circuits 2013

The ideal resistor relationship is a straight line through the origin:

Even though resistance is defined as R v i , it should be noted that R is a

The circuit symbol

Electronics and Circuits 2013

Consider the circuit shown below.

Note that i2 i1 , as expected.

Consider the circuit shown below.

Ohms Law yields: