Chapter 5

AC electricity

While direct current (DC) refers to the flow of electrical charge carriers in a continuous direction,
alternating current (or AC ) refers to a periodic reversal of charge flow direction1 . As a mode of
transferring electrical power, AC is tremendously useful because it allows us to use transformers to
easily and efficiently step voltage up or down at will. If an electro-physical sensor senses a physical
quantity that oscillates, the electric signal it produces will oscillate (AC) as well. For both these
reasons, an instrument technician needs to be aware of how AC circuits work, and how to understand
them mathematically.

1 It is also acceptable to refer to electrical voltages and/or currents that vary periodically over time even if their

directions never alternate, as AC superimposed on DC.

351
352 CHAPTER 5. AC ELECTRICITY

5.1 RMS quantities

It is often useful to express the amplitude of an alternating quantity such as voltage or current in
terms that are equivalent to direct current (DC). Doing so provides an “apples-to-apples” comparison
between AC and DC quantities that makes comparative circuit analysis much easier.
The most popular standard of equivalence is based on work and power, and we call this the
root-mean-square value of an AC waveform, or RMS for short. For example, an AC voltage of 120
volts “RMS” means that this AC voltage is capable of delivering the exact same amount of power
(in Watts) at an electrical load as a 120 volt DC source powering the exact same load.
The problem is exactly how to calculate this “RMS” value if all we know about the AC waveform
is its peak value. If we compare a sine wave and a DC “wave” side by side, it is clear that the sine
wave must peak at a greater value than the constant DC level in order to be equivalent in terms of
doing the same work in the same amount of time:

peak
voltage constant
voltage

AC circuit DC circuit

+
R − R

At first, it might seem like the correct approach would be to use calculus to integrate (determine
the area enclosed by) the sine wave over one-half of a cycle: from 0 to π radians. This is close,
but not fully correct. The ability of an electrical voltage to dissipate power at a resistor is not
directly proportional to the magnitude of that voltage, but rather proportional to the square of the
magnitude of that voltage! In mathematical terms, resistive power dissipation is predicted by the
following equation:

V2
P =
R
If we double the amount of voltage applied to a resistor, the power increases four-fold. If we
triple the voltage, the power goes up by a factor of nine. If we are to calculate the “RMS” equivalent
value of a sine wave, we must take this nonlinearity into consideration.
5.1. RMS QUANTITIES 353

First let us begin with a mathematical equivalence between the DC and AC cases. For DC, the
amount of work done is equal to the constant power of that circuit multiplied by time. The unit of
measurement for power is the Watt, which is defined as 1 Joule of work per second. So, multiplying
the steady power rate in a DC circuit by the time we keep it powered will result in an answer of
joules (total energy dissipated by the resistor):
 2
V
Work = t
R
Showing this equivalence by dimensional analysis:
 
[Joules]
[Joules] = [s]
[s]
We cannot calculate the work done by the AC voltage source quite so easily, because the power
dissipation varies over time as the instantaneous voltage rises and falls. Work is still the product of
power and time, but we cannot simply multiply one by the other because the voltage in this case is
a function of time (V (t)). Instead, we must use integration to calculate the product of power and
time, and sum those work quantities into a total work value.
Since we know the voltage provided by the AC source takes the form of a sine wave (V (t) = sin t
if we assume a sine wave with a peak value of 1 volt), we may write the formula for instantaneous
AC power as follows:
2
(V (t)) sin2 t
Power = =
R R
To calculate the work done by this sinusoidal voltage on the resistor, we will integrate this
instantaneous power with respect to time, between the intervals of t = 0 and t = π (one half-cycle
of the sine wave):
Z π
sin2 t
Work = dt
0 R
In order to solve for the amount of DC voltage equivalent (from the perspective of resistive power
dissipation) to a one-volt AC sine wave, we will set the DC work and AC work equations equal to
each other, making sure the DC side of the equation has π for the amount of time (being the same
time interval as the AC side):
Z π
sin2 t
 2
V
π= dt
R 0 R
Our goal is to solve for V on the left-hand side of this equation.
354 CHAPTER 5. AC ELECTRICITY

First, we know that R is a constant value, and so we may move it out of the integrand:

1 π 2
 2
V
Z
π= sin t dt
R R 0
Multiplying both sides of the equation by R eliminates it completely. This should make intuitive
sense, as our RMS equivalent value for an AC voltage is defined strictly by the ability to produce
the same amount of power as the same value of DC voltage for any resistance value. Therefore the
actual value of resistance (R) should not matter, and it should come as no surprise that it cancels:
Z π
V 2π = sin2 t dt
0

Now, we may simplify the integrand by substituting the half-angle equivalence for the sin2 t
function
Z π
1 − cos 2t
V 2π = dt
0 2
Factoring one-half out of the integrand and moving it outside (because it’s a constant):

1 π
Z
V 2π = 1 − cos 2t dt
2 0
We may write this as the difference between two integrals, treating each term in the integrand
as its own integration problem:
Z π Z π 
2 1
V π= 1 dt − cos 2t dt
2 0 0

The second integral may be solved simply by using substitution, with u = 2t, du = 2 dt, and
dt = du
2 :
Z π Z π 
1 cos u
V 2π = 1 dt − du
2 0 0 2
Moving the one-half outside the second integrand:
Z π
1 π

1
Z
2
V π= 1 dt − cos u du
2 0 2 0
5.1. RMS QUANTITIES 355

Finally we are at a point where we may perform the integrations:

Z π π 
1 1
Z
2
V π= 1 dt − cos u du
2 0 2 0

 
2 1 π 1 π
V π= [t]0 − [sin 2t]0
2 2
 
2 1 1
V π= [π − 0] − [sin 2π − sin 0]
2 2
 
2 1 1
V π= [π − 0] − [0 − 0]
2 2

1
V 2π = (π − 0)
2

1
V 2π = π
2
We can see that π cancels out of both sides:

1
V2 =
2
Taking the square root of both sides, we arrive at our final answer for the equivalent DC voltage
value:

1
V =√
2
So, for a sinusoidal voltage with a peak value of 1 volt, the DC equivalent or “RMS” voltage
value would be √12 volts, or approximately 0.707 volts. In other words, a sinusoidal voltage of 1 volt
peak will produce just as much power dissipation at a resistor as a steady DC voltage of 0.7071 volts
applied to that same resistor. Therefore, this 1 volt peak sine wave may be properly called a 0.7071
volt RMS sine wave, or a 0.7071 volt “DC equivalent” sine wave.
This factor for sinusoidal voltages is quite useful in electrical power system calculations, where
the wave-shape of the voltage is nearly always sinusoidal (or very close). In your home, for example,
the voltage available at any wall receptacle is 120 volts RMS, which translates to 169.7 volts peak.
Electricians and electronics technicians often memorize the √12 conversion factor without realizing
it only applies to sinusoidal voltage and current waveforms. If we are dealing with a non-sinusoidal
wave-shape, the conversion factor between peak and RMS will be different! The mathematical
procedure for obtaining the conversion factor will be identical, though: integrate the wave-shape’s
function (squared) over an interval sufficiently long to capture the essence of the shape, and set that
equal to V 2 times that same interval span.
356 CHAPTER 5. AC ELECTRICITY

5.2 Resistance, Reactance, and Impedance

Resistance (R) is the dissipative opposition to an electric current, analogous to friction encountered
by a moving object. In any example of electrical resistance, the electrical energy is converted into
some other form of energy that cannot (or does not) return back to the circuit. Resistance may take
the form of an actual resistor, in which case the electrical energy is converted into heat. Resistance
may also take the form of an electric motor, an electric light, or an electrochemical cell where the
electrical energy is converted into mechanical work, photons, or enables an endothermic chemical
reaction, respectively.
Reactance (X) is the opposition to an electric current resulting from energy storage and release
between certain components and the rest of the circuit, analogous to inertia of a moving object.
Capacitors and inductors are classic examples of “reactive” electrical components, behaving either as
electrical loads or as electrical sources depending on whether the applied electrical signal is increasing
or decreasing in intensity at that instant in time. When a purely reactive component is subjected
to a sinusoidal signal, it will spend exactly half the time behaving as a load (absorbing energy from
the circuit) and half the time behaving as a source (returning energy to the circuit). Thus, a purely
reactive component neither contributes nor dissipates any net energy in the circuit, but merely
exchanges2 energy back and forth. Even though the fundamental mechanism of reactance (energy
storage and release) is different from the fundamental mechanism of resistance (energy conversion
and dissipation), reactance and resistance are both expressed in the same unit of measurement: the
ohm (Ω).
Impedance (Z) is the combined total opposition to an electric current, usually some combination
of electrical resistance (energy dissipation) and electrical reactance (energy storage and release). It
is also expressed in the unit of the ohm. In order to represent how much of a particular impedance is
due to resistance and how much is due to reactance, the value of an impedance may be expressed as a
complex number with a “real” part (representing resistance) and an “imaginary” part (representing
reactance). This concept will be explored in much more detail later in this chapter.

The amount of electrical reactance offered by a capacitor or an inductor depends on the frequency
of the applied signal. The faster the rate at which an AC signal oscillates back and forth, the more
a reactive component tends to react to that signal. The formulae for capacitive reactance (XC ) and
inductive reactance (XL ) are as follows:

1
XC = XL = 2πf L
2πf C
Just as conductance (G) is the reciprocal of resistance (1/R), a quantity called susceptance (B) is
the reciprocal of reactance (1/X). Susceptance is useful when analyzing parallel-connected reactive
components while reactance is useful for analyzing series-connected reactive components, in much
the same way that conductance and resistance are useful when analyzing parallel-connected and
series-connected resistors, respectively.

Impedance (Z) also has a reciprocal counterpart known as admittance (Y ).

2 Charles Proteus Steinmetz, in his book Theoretical Elements of Electrical Engineering, refers to the voltage and

current values of a reactive component being “wattless” in honor of the fact that they transfer zero net power to or
from the circuit (page 41). The voltage and current values of resistive components, by contrast, constitute real power
dissipated in the circuit.
5.3. SERIES AND PARALLEL CIRCUITS 357

5.3 Series and parallel circuits

Impedance in a series circuit is the orthogonal sum of resistance and reactance:
q
Z = R2 + (XL2 − XC2 )
Equivalent series and parallel circuits are circuits that have the exact same total impedance
as one another, one with series-connected resistance and reactance, and the other with parallel-
connected resistance and reactance. The resistance and reactance values of equivalent series and
parallel circuits may be expressed in terms of those circuits’ total impedance:

Rseries

Rparallel Xparallel

Xseries

If the total impedance of one circuit (either series or parallel) is known, the component values of
the equivalent circuit may be found by algebraically manipulating these equations and solving for
the desired R and X values:

Z 2 = Rseries Rparallel Z 2 = Xseries Xparallel

5.4 Transformers
The transformer is one of the most important components in all of AC circuitry. Principally used to
“step” between different values of AC voltage and current in power systems, transformers find uses
in many other types of circuits including electronic amplifiers (for impedance matching) and even
sensor circuits (sensing physical position).
358 CHAPTER 5. AC ELECTRICITY

5.4.1 Basic principles

Before exploring the operation of a transformer, it is useful to review the operation of a simple
inductor, which is nothing more than a coil of wire usually wrapped around a ferromagnetic core
material:

iron core
φ

I
AC
voltage
source
I
If we apply an alternating (AC) voltage to this coil, it will generate an alternating magnetic
field in the core. Just how much magnetic flux (φ) will develop in the core depends on how much
voltage we apply to the coil. The fundamental relationship between voltage and magnetic flux for
any conductive coil is given by Faraday’s Law of Electromagnetic Induction3 :
dφ
V =N
dt
Where,
V = Voltage applied to the coil or induced by the coil (volts)
N = Number of turns of wire
dφ
dt = Rate of change of magnetic flux (Webers per second)

3 At first it may seem strange to apply Faraday’s Law here, because this formula is typically used to describe the

amount of voltage produced by a coil of wire exposed to a changing magnetic field, not the amount of magnetic field
produced by an applied voltage. However, the two are closely related because the inductor must produce a voltage
drop in equilibrium with the applied voltage just like any other component, in accordance with Kirchhoff’s Voltage
Law. In a simple circuit such as this where the voltage source directly connects to the inductor (barring any resistive
losses in the connecting wires), the coil’s induced voltage drop must exactly equal the source’s applied voltage at all
points in time, and so Faraday’s Law works just as well to describe the source’s applied voltage as it does to describe
the coil’s induced voltage. This is the principle of self-induction.
5.4. TRANSFORMERS 359

If the applied voltage is sinusoidal (i.e. shaped like a sine wave), then the magnetic flux magnitude
will trace a cosine wave over time. We may demonstrate this mathematically by substituting sin ωt
(the sine of some frequency ω at any particular point in time t) for V in Faraday’s equation and
integrating:
dφ
V =N
dt
dφ
sin ωt = N
dt

sin ωt dt = N dφ
Z Z
sin ωt dt = N dφ
Z Z
sin ωt dt = N dφ

1
− cos ωt + φ0 = N φ
ω
1
φ=− cos ωt + φ0
Nω
Thus, the amount of magnetic flux (φ) in the core at any point in time t is proportional to the
cosine of the frequency-time function ωt plus any residual magnetism (φ0 ) the core happened to
start out with before any voltage was applied to the coil.

The amount of current drawn by this inductor depends on the reluctance of the core’s magnetic
“circuit” and the number of turns in the coil (N ). The less reluctance offered by the magnetic path,
the less current will be necessary to generate the requisite magnetic field to balance the applied
voltage. If we were to take two perfect inductors (i.e. lacking wire resistance) – one with a heavy
ferrous core and one with a light ferrous core (or even an air core) – and apply the same AC voltage
to them, they would both generate the exact same strength of alternating magnetic field, but the
inductor with the lesser core would draw more current from the source in doing so. In other words,
the latter inductor would exhibit less reactance (i.e. fewer ohms) to oppose current.
360 CHAPTER 5. AC ELECTRICITY

Things get interesting if we wrap a second coil of wire around the same core as the first. For the
sake of analysis we will label voltage polarities at one of the peaks of the AC source:

iron core
φ

I
AC I I
voltage
source R

I
At that moment in time when the top terminal of the source is positive and the bottom terminal
is negative, we see that the first coil drops the same voltage (due to self-induction), and that the
second coil drops the same voltage as well (due to mutual induction). The polarity of both coils’
voltages are identical because they are wrapped in the same direction around the core and they
both experience the same magnetic flux (φ). When we examine the directions of current through
each coil, however, we see they are opposite one another: the left-hand coil acts as a load (drawing
energy from the AC voltage source) while the right-hand coil acts as a source (providing energy to
the resistive load).
What we have created here is a true transformer : an electromagnetic component transferring
energy from electric form to magnetic form and back again to electric form. The AC voltage source
is able to energize the resistive load without direct conductive connection between the two, since the
magnetic flux serves as the energy “link” between the two circuits.

Transformers are typically drawn as a set of coils sharing a common core. The coil connected to
an electrical source is called the primary, while the coil connected to an electrical load is called the
secondary. If the core is ferromagnetic, it is shown as a set of parallel lines between the coils:

A simple transformer circuit

Primary Secondary
5.4. TRANSFORMERS 361

5.4.2 Loading effects

We may explore transformer behavior by observing the effects of powering one with a constant4 -
voltage AC source and varying the load resistance:

iron core Normal load

φ
R = Normal
φ VS = Normal
IS = Normal
IP φ = Normal
AC VP VS
IS IS
voltage VP = Normal
source R
IP = Normal
IP

iron core Light load

φ
R = Large
φ VS = Normal
IS = Small
IP φ = Normal
AC VP VS
IS IS
voltage VP = Normal
source R
IP = Small
IP

iron core Heavy load

φ
R = Small
φ VS = Normal
IS = Large
IP φ = Normal
AC VP VS
IS IS
voltage VP = Normal
source R
IP = Large
IP

Observe how voltage at both coils is unaffected by load, and similarly how the magnetic flux
remains unchanged under different load conditions. The secondary coil acts like a voltage source
to the resistive load, reflecting the nature of the primary coil’s source behavior. The magnetic
flux amplitude is unaffected by secondary loading in order to satisfy Kirchhoff’s Voltage Law and
Faraday’s Law at the primary coil: the coil’s voltage drop must be equal and opposite to the source’s
4 In this context, “constant” means an alternating voltage with a consistent peak value, not “constant” in the sense

that a DC source is constant at all points in time.

362 CHAPTER 5. AC ELECTRICITY

applied voltage, and so the magnetic flux must alternate at the same rates and reach the same peaks
so long as the primary source voltage does the same.

Continuing our exploration of transformer behavior, we will now power one with a constant5 -
current AC source and vary the load resistance:

iron core Normal load

φ
R = Normal
φ VS = Normal
IS = Normal
IP φ = Normal
AC VP VS
IS IS
current VP = Normal
source R
IP = Normal
IP

iron core Light load

φ
R = Small
φ VS = Small
IS = Normal
IP φ = Small
AC VP VS
IS IS
current VP = Small
source R
IP = Normal
IP

iron core Heavy load

φ
R = Large
φ VS = Large
IS = Normal
IP φ = Large
AC VP VS
IS IS
current VP = Large
source R
IP = Normal
IP

Observe how current now is the unaffected quantity, while voltage and magnetic flux are load-
dependent. The secondary coil now acts like a current source to the resistive load, reflecting the
nature of the primary coil’s source behavior. As load resistance varies, the secondary coil’s voltage
varies proportionately, which in turn demands a commensurate change in magnetic flux.
5 In this context, “constant” means an alternating voltage with a consistent peak value, not “constant” in the sense

that a DC source is constant at all points in time.

5.4. TRANSFORMERS 363

5.4.3 Step ratios

Transformers are principally used to step between different levels of voltage and current. This is
achieved by building the transformer with primary and secondary coils having different numbers of
turns. Since both coils share the same magnetic flux, the number of turns will be proportionate to
how much voltage is developed at each coil. We may prove this mathematically with Faraday’s Law,
using dφ
dt as the quantity shared between primary and secondary coils:

dφ dφ
V P = NP V S = NS
dt dt

VP dφ VS dφ
= =
NP dt NS dt

VP VS
=
NP NS

VP NP
=
VS NS
That is to say, the ratio of primary to secondary voltage is the same as the ratio of primary to
secondary turns. We may exploit this principle to build transformers delivering the same amount
of power to two different load resistances from the same power source, the only difference being the
number of turns in the secondary coil:

iron core Step-up transformer

φ
R = Large
φ VS = Large
IS = Small
IP φ = Normal
AC VP VS
IS IS
voltage VP = Normal
source R
IP = Normal
IP

iron core Step-down transformer

φ
R = Small
φ VS = Small
IS = Large
IP φ = Normal
AC VP VS
IS IS
voltage VP = Normal
source R
IP = Normal
IP
364 CHAPTER 5. AC ELECTRICITY

Whichever way a transformer steps voltage from primary to secondary, it must step current the
other way.

Here are some quantitative examples, assuming lossless transformers:

Step-up transformer
1:5 ratio
20 A 4A

120 VAC 600 VAC 150 Ω

PP = IP VP PS = IS VS
PP = (120 V) (20 A) PS = (600 V) (4 A)
PP = 2400 W PS = 2400 W

Step-down transformer
10:1 ratio
0.8 A 8A

480 VAC 48 VAC 6Ω

PP = IP VP PS = IS VS
PP = (480 V) (0.8 A) PS = (48 V) (8 A)
PP = 384 W PS = 384 W

Note how primary and secondary powers are always equal to each other for any given transformer
arrangement. Real transformers suffer some internal6 power loss, and as such will exhibit secondary
power levels slightly less than primary, but assuming equality provides an easy way to check our
voltage and current ratio calculations.
6 These power losses take the form of core losses due to magnetic hysteresis in the ferrous core material, and winding

losses due to electrical resistance in the wire coils. Core losses may be minimized by reducing magnetic flux density
(H), which requires a core with a larger cross-section to disperse the flux (φ) over a wider area. Winding losses may
be minimized by increasing wire gauge (i.e. thicker wire coils). In either case, these modifications make for a bulkier
and more expensive transformer.
5.5. PHASORS 365

5.5 Phasors
Phasors are to AC circuit quantities as polarity is to DC circuit quantities: a way to express the
“directions” of voltage and current waveforms. As such, it is difficult to analyze AC circuits in
depth without using this form of mathematical expression. Phasors are based on the concept of
complex numbers: combinations of “real” and “imaginary” quantities. The purpose of this section is
to explore how complex numbers relate to sinusoidal waveforms, and show some of the mathematical
symmetry and beauty of this approach.
Since waveforms are not limited to alternating current electrical circuits, phasors have
applications reaching far beyond the scope of this chapter.
366 CHAPTER 5. AC ELECTRICITY

5.5.1 Circles, sine waves, and cosine waves

Something every beginning trigonometry student learns (or should learn) is how sine and cosine
waves may be derived from a circle. First, sketch a circle, then sketch a set of radius vectors from
the circle’s center to the circle’s perimeter at regular angle intervals. Mark each point of intersection
between a vector and the circle’s perimeter with a dot and label each with the vector angle. Sketch
rectangular graphs to the right and below the circle, with regularly-spaced divisions. Label those
divisions with the same angles as the vectors and then sketch dashed “projection” lines from each
vector tip to the respective divisions on the rectangular graphs, marking each intersection with a
dot. Connect the dots with curves to reveal sinusoidal waveshapes.
The following illustration shows the dots and projection lines for the first five vectors (angles 0
through π2 radians7 ) only. As you can see, the circle’s vertical projection forms a sine wave, while
the circle’s horizontal projection forms a cosine wave:

π/2
3π/8
π/4

π/8
Sine wave

π 0

π/8 3π/8
3π/2 0 π/4 π/2 π 3π/2 2π
0
π/8
π/4
3π/8
π/2

Cosine wave
π
3π/2
2π

7 A full circle contains 360 degrees, which is equal to 2π radians. One “radian” is defined as the angle encompassing

an arc-segment of a circle’s circumference equal in length to its radius, hence the name “radian”. Since the
circumference of a circle is 2π times as long as its radius, there are 2π radians’ worth of rotation in a circle. Thus,
while the “degree” is an arbitrary unit of angle measurement, the “radian” is a more natural unit of measurement
because it is defined by the circle’s own radius.
5.5. PHASORS 367

The Swiss mathematician Leonhard Euler (1707-1783) developed a symbolic equivalence between
polar (circular) plots, sine waves, and cosine waves by plotting the circle on a complex plane where
the vertical axis is an “imaginary”8 number line and the horizontal axis is a “real” number line.
Euler’s Relation expresses the vertical (imaginary) and horizontal (real) projections of an imaginary
exponential function as a complex (real + imaginary) trigonometric function:

ejθ = cos θ + j sin θ

Where,
e = Euler’s number (approximately equal to 2.718281828)
θ = Angle of vector, in radians
cos θ = Horizontal projection of a unit
√ vector (along a real number line) at angle θ
j = Imaginary “operator” equal to −1, represented by i or j
j sin θ = Vertical projection of a unit vector (along an imaginary number line) at angle θ

To illustrate, we will apply Euler’s relation to a unit9 vector having an angular displacement of
3π
8 radians:

+j1
+j1

Sine wave
+j0.9239
θ = 3π/8
-1 +1 j0

-j1 π/8 3π/8

-j1 0 π/4 π/2 π 3π/2 2π
-1 0 +1
ej3π/8 = cos 3π/8 + j sin 3π/8
0
π/8

ej3π/8 = 0.3827 + j0.9239

π/4
3π/8
π/2

+0.3827
Cosine wave
π
3π/2
2π

8 The
√
definition of an imaginary number is the square root of a negative quantity. −1 is the simplest case, and
is symbolized by mathematicians as i and by electrical engineers as j.
9 The term “unit vector” simply refers to a vector with a length of 1 (“unity”).
368 CHAPTER 5. AC ELECTRICITY

This mathematical translation from circles to sinusoidal waves finds practical application in AC
electrical systems, because the rotating vector directly relates to the rotating magnetic field of an
AC generator while the sinusoidal function directly relates to the voltage generated by the field’s
motion. The principle of an AC generator is that a magnet rotates on a shaft past stationary coils
of wire. When these wire coils experience the changing magnetic field produced by the rotating
magnet, a sinusoidal voltage is induced:

V0

coil coil vcoil

S
vcoil θ
N magnet 0 π/2 π 3π/2 2π

-V0

vcoil = V0 cos θ

If we mark the shaft of this simple two-pole generator such that an angle of zero is defined as the
position where the stator coils develop maximum positive voltage (V0 ), the AC voltage waveform
will follow the cosine function as shown above. If we were to add another set of stationary (“stator”)
coils to the generator’s frame perpendicular to this set, that second set of coils would generate an
AC voltage following the sine function. As it is, there is no second set of stator coils, and so this
sine-function voltage is purely an imaginary quantity whereas the cosine-function voltage is the only
“real” electricity output by the generator.
The canonical form of Euler’s Relation assumes a circle with a radius of 1. In order to realistically
represent the output voltage of a generator, we may include a multiplying coefficient in Euler’s
Relation representing the function’s peak value:

Aejθ = A cos θ + jA sin θ

In the case of the generator shown above, the A coefficient is the peak10 voltage (V0 ) output by
the stator coils at the precise moment in time when the shaft angle is zero. In other words, the
exponential term (ejθ ) tells us which way the vector points while the coefficient (A) tells us how
long the vector is.

10 Although A truly should represent a waveform’s peak value, and θ should be expressed in units of radians to be

mathematically correct, it is more common in electrical engineering to express A in RMS (root-mean-square) units
and θ in degrees. For example, a 120 volt RMS sine wave voltage at a phase angle of 30 degrees will be written by an
engineer as 120ej30 even though the true phase angle of this voltage is π6 radians and the actual peak value is 169.7
volts.
5.5. PHASORS 369

In electrical studies it is commonplace to use various “shorthand” notations for vectors. An

instantaneous quantity (e.g. voltage or current at some moment in time) described simply in terms
of real and imaginary values is called rectangular form, for example 0.3827 + j0.9239 volts. An
instantaneous quantity described simply in terms of vector angle and length is called polar form, for
example 1 volt 6 67.5o . Euler’s Relation is the mathematical link connecting rectangular and polar
forms.

Polar Rectangular
Shorthand notation A6 θ x + jy
Full notation Aejθ A cos θ + jA sin θ

All of these vectors described by the imaginary exponential function ejθ are of a special kind –
vectors inhabiting a space defined by one real axis and one imaginary axis (the so-called complex
plane). In order to avoid confusing this special quantity with real vectors used to represent quantities
in actual space where every axis is real, we use the term phasor instead. From this point on in the
book, I will use the term “phasor” to describe complex quantities as represented in Euler’s Relation,
and “vector” to describe any other quantity possessing both a magnitude and a direction.
370 CHAPTER 5. AC ELECTRICITY

Euler’s Relation may also be used to describe the behavior of the rotating generator shaft and
its corresponding AC voltage output as a function of time. Instead of specifying a shaft position
(angle θ) we now specify a point in time (t) and an angular velocity (rotational speed) represented
by the lower-case Greek letter “omega” (ω). Since angular velocity is customarily given in units of
radians per second, and time is customarily specified in seconds, the product of those two will be an
angle in radians (ωt = θ). Substituting ωt for θ in Euler’s Relation gives us a function describing
the state of the generator’s output voltage in terms of time:

Aejωt = A cos ωt + jA sin ωt

Where,
A = Peak amplitude of generator voltage, in volts
e = Euler’s number (approximately equal to 2.718281828)
ω = Angular velocity, in radians per second
t = Time, in seconds
cos ωt = Horizontal projection of phasor
√ (along a real number line) at time t
j = Imaginary “operator” equal to −1, represented by i or j
j sin ωt = Vertical projection of phasor (along an imaginary number line) at time t

With just a little bit of imagination we may visualize both the exponential and complex sides
of Euler’s Relation in physical form. The exponential (ejωt ) is a phasor rotating counter-clockwise
about a central point over time. The complex (cos ωt + j sin ωt) is a pair of cosine and sine waves
oscillating along an axis of time. Both of these manifestations may be joined into one visual image
if you imagine a corkscrew where the centerline of the corkscrew is the time axis, and the evolution
of this function over time follows the curve of the corkscrew down its length. Viewed from one end
(looking along the centerline axis), the path appears to be a circle, just as a rotating phasor traces
a circle with its tip. Viewed perpendicular to that axis, the path appears to trace a sinusoid, just
as the sine or cosine function goes up and down as it progresses from left to right on a rectangular
graph. A compression-type coil spring serves just as well as a corkscrew for a visual aid: viewed
from the end the spring looks like a circle, but viewed from the side it looks like a sine wave.
To view a flip-book animation of this three-dimensional visualization, turn to Appendix A.3
beginning on page 2733.
5.5. PHASORS 371

5.5.2 Phasor expressions of phase shifts

As previously discussed, Euler’s Relation provides a mathematical tool to express AC circuit
quantities such as voltage, which change over time and therefore are more challenging to represent
than the static quantities found in DC circuits. At its core, Euler’s Relation expresses the
mathematical equivalence between an imaginary exponential function (e raised to some imaginary
power) and a complex number (the sum of a real number and an imaginary number):

ejθ = cos θ + j sin θ

If we wish to represent AC quantities having magnitudes other than 1, we may modify Euler’s
Relation to include a multiplying coefficient specifying the function’s peak value:

Aejθ = A(cos θ + j sin θ) = A cos θ + jA sin θ

Phasor notation proves extremely useful to compare or combine AC quantities at the same
frequency that are out-of-phase with each other. Consider the following example, showing two AC
voltage waveforms of equal magnitude (5 volts peak) that are a constant 60 degrees ( π3 radians) out
of step with each other:

B A

5 V ∠ 60o 5 V ∠ 0o

π/3
phase
π/2 shift
VB VA

VB
π/3
phase
shift
π 0
VA

3π/2 0 π/2 π 3π/2 2π

Recalling that the assumed direction of rotation for phasors is counter-clockwise, we see here
that phasor B leads phasor A (i.e. phasor B is further counter-clockwise than phasor A). This is
372 CHAPTER 5. AC ELECTRICITY

also evident from the rectangular graph11 , where we see sinusoid A lags behind sinusoid B by π3
radians. If we were to connect a dual-channel oscilloscope to these two voltage sources, we would
see the same phase shift on its display.

Suppose we wished to calculate the amount of voltage between points B and A, sensing the
difference in potential between these two AC voltage sources. Phasor notation makes this calculation
rather easy to do, since all it entails is calculating the distance between the two phasor tips. We will
designate this new measurement VBA representing the potential at point B with reference to point
A12 :

VBA

B A
VΩ

A COM

5 V ∠ 60o 5 V ∠ 0o

π/2
VB VA

VB VBA
π/3
phase
shift
π 0
VA

3π/2 0 π/2 π 3π/2 2π

Simply sketching a new phasor stretching between the tips of phasors B and A yields a graphical
solution for VBA . The length of phasor VBA represents the magnitude of that voltage, while its angle
from horizontal represents the phase shift of that voltage (compared to phasor A at 0o ).

11 The fact that this graph shows the vertical (imaginary) projections of both phasors rather than the horizontal

(real) projections is irrelevant to phase shift. Either way, the voltage waveform of source B will still lead the voltage
waveform of source A by 60o .
12 One way to think of this is to imagine an AC voltage-measuring instrument having red and black test leads just

like a regular voltmeter. To measure VBA you would connect the red test lead to the first point (B) and the black
test lead to the second point (A).
5.5. PHASORS 373

If we wished to plot the sine wave represented by this new phasor, all we would have to do is
move the VBA phasor until its tail rested at the center of the circular graph (being careful to preserve
its angle), then project its tip to the rectangular graph and plot the resulting wave through the new
phasor’s complete rotation:

VBA

B A
VΩ

A COM

5 V ∠ 60o 5 V ∠ 0o

π/2
VB VA VBA

VBA VB
120o
60o
π 0
VA

3π/2 0 π/2 π 3π/2 2π

We see clearly now how phasor VBA is phase-shifted 120 ahead of (leading) phasor A, and 60o
ahead of (leading) phasor B, possessing the same magnitude as both A and B. We may express each
of these voltages in polar form:

VA = 5 volts 6 0o

VB = 5 volts 6 60o

VBA = 5 volts 6 120o

We may also arrive at this conclusion mathematically by subtracting VB from VA to arrive at
VAB :

VAB = VA − VB

(5 volts 6 0o ) − (5 volts 6 60o ) = 5 volts 6 120o

374 CHAPTER 5. AC ELECTRICITY

When using phasors to describe differential voltages (i.e. voltages between two non-grounded
points), it is important to specify which point is the reference. Suppose, for example, we took this
same two-source system where each source has a peak voltage of 5 volts with reference to ground,
but we connected our voltage-measuring instrument backwards from before such that the black
(reference) lead touches point B and the red (measurement) lead touches point A (sensing VAB
instead of VBA ). Our new phasor VAB must then be sketched with its tail (reference) at point B
and its tip (measurement) at point A:

VAB

B A
VΩ

A COM

5 V ∠ 60o 5 V ∠ 0o

π/2
VB VA VAB

VB VAB
π/3
phase
shift
π 0
VA

VAB

3π/2 0 π/2 π 3π/2 2π

The dashed-line phasor is the same angle as the solid-line phasor reaching from point B to point
A, just moved so that its tail rests at the circle’s center in order to plot its vertical projection on
the rectangular graph to the right of the circle. As you can see, the new phasor VAB is 180o shifted
from the old phasor VBA . This makes sense, as a reversal of the test leads on our voltage-sensing
instrument must reverse the polarity of that voltage at every point in time.
Thus, the new phasor VAB actually lags behind both VA (by 60 degrees) and VB (by 120 degrees).
If the voltage-sensing instrument is nothing more than a voltmeter, there will be no difference in its
reading of VBA versus VAB : in both cases it will simply register 5 volts AC peak. However, if the
voltage-sensing instrument is an oscilloscope with its sweep triggered by one of the other voltage
signals, this difference in phase will be clearly evident.
5.5. PHASORS 375

It is important to remind ourselves that voltage and current phasors, like the AC waveforms
they represent, are actually in constant motion. Just as the generator shaft in an actual AC power
system spins, the phasor representing that generator’s voltage and the phasor representing that
generator’s current must also spin. The only way a phasor possessing a fixed angle makes any sense
in the context of an AC circuit is if that phasor’s angle represents a relative shift compared to some
other phasor at the same frequency13 . Here, we see Euler’s Relation written in fixed-angle and
rotating-angle forms:

Fixed-angle phasor Aejθ = A cos θ + jA sin θ

Rotating phasor Aejωt = A cos ωt + jA sin ωt

This concept can be very confusing, representing voltage and current phasors as possessing fixed
angles when in reality they are continuously spinning at the system’s frequency. In order to make
better sense of this idea, we will explore an analysis of the same two-source AC system using a
special instrument designed just for the purpose.

13 The necessity of a shared frequency is easily understood if one considers a case of two sine waves at different

frequencies: their respective phasors would spin at different speeds. Given two phasors spinning at different speeds,
the angle separating those two phasors would be constantly changing. It is only when two phasors spin around at
precisely the same speed that we can sensibly talk about there being a fixed angular displacement between them.
Fortunately this is the usual case in AC circuit analysis, where all voltages and currents share the same frequency.
376 CHAPTER 5. AC ELECTRICITY

Imagine building a small synchronous electric motor with an arrow-shaped magnetic rotor. We
will wind the stator coils and magnetize the rotor such that the arrow points toward the 0o mark
whenever the red lead is positive and the black lead is negative (connected to a DC test source).
When connected to an AC source, this “phasometer” will spin in sync14 with the AC generator
powering the circuit, the rotating arrow keeping pace with the generator’s shaft15 . The intent of
this arrow is to serve as a mechanical analogue of an electrical phasor, pointing in the same direction
as the phasor points for any sinusoidal voltage or current signal connected to the phasometer:

A mechanical "phasometer"
o
+90

coil coil
Arrow points to 0o when
180o 0o red lead is positive and + DC test
S N − source
black lead is negative

-90o

For a system frequency of 60 Hz as is standard for AC power systems in North America, the
rotational speed of our phasometer’s rotor will be 3600 revolutions per minute – too fast for the
rotating arrow to be anything but a blur of motion to an unaided human eye. Therefore, we will
need to add one more component to the phasometer to make it practical: a strobe light connected
to an AC voltage in the system to act as a synchronizing pulse. Just like using a strobe light (also
called a “stroboscope”) to make a moving machine part16 appear to “freeze” in time, this strobe
light will visually “freeze” the arrow so we will be able to read its position with our eyes.

14 An important detail is that our phasometer must always spin counter-clockwise in order to maintain proper phasor

convention. We can ensure this will happen by including a pair of shading coils (small copper rings wrapped around
one corner of each magnetic pole) in the stator structure. For a more detailed discussion of shading coils, refer to the
section on AC induction motors (10.4.1) starting on page 696.
15 This, of course, assumes the generator powering the system is also a two-pole machine like the phasometer. If the

generator has more poles, the shaft speed will not match the phasometer’s rotor speed even though the phasometer
will still faithfully represent the generator’s cosine wave rotation.
16 Automobile mechanics may be familiar with a tool called a timing light, consisting of a strobe light connected to

the engine in such a way that the light flashes every time the #1 cylinder spark plug fires. By viewing the marks
etched into the engine’s crankshaft with this strobe light, the mechanic is able to check the ignition timing of the
engine.
5.5. PHASORS 377

Returning to our two-source AC system, we may perform a “thought experiment” to see what
multiple phasometers would register if connected to various points in the system, using a single
strobe light connected to source A (pulsing briefly every time its sensed voltage reaches the positive
peak). Imagine this one strobe light is bright enough to clearly illuminate the faces of all phasometers
simultaneously, visually “freezing” their arrows at the exact same point in every cycle:

Phasometer
105o 90o 75o
120o 60o
135o 45o
o
150 30o

165o 15o Phasometer arrow appears

180o
0o "frozen" at the -60o mark
-165o
VAB -15o

-150o -30o

-135o -45o
-120o -60o
-105o -90o -75o

Phasometer B A
105o 90o 75o
120o 60o
135o 45o
o o
150 30

165o VB 15o

180o

5 V ∠ 60o 5 V ∠ 0o
0o

-165o -15o

-150o -30o
o o
-135 -45
-120o
-105o -90o -75o
-60o
Strobe
Phasometer arrow appears
light
"frozen" at the 60o mark
Phasometer
105o 90o 75o
120o 60o
135o 45o

150o 30o

Phasometer arrow appears 165o VA 15o

180o

"frozen" at the 0o mark 0o

-165o -15o

-150o -30o

-135o -45o
-120o -60o
-105o -90o -75o

The phasometer connected to source A must register 0o because the strobe light flashes whenever
source A hits its positive peak (at 0o on its cosine waveform), and the positive peak of a waveform
is the precise point at which the phasometer arrow is designed to orient itself toward the 0o mark.
The phasometer connected to source B will register 60o because that source is 60 degrees ahead of
(leading) source A, having passed its positive peak already and now headed downward toward zero
volts at the point in time when the strobe flashes. The upper phasometer registers −60o at that
same time because the voltage it senses at point A with respect to point B is lagging behind source
A by that much, the VAB cosine wave heading toward its positive peak.
If we were to substitute a constant illumination source for the strobe light, we would see all three
phasometers’ arrows spinning counter-clockwise, revealing the true dynamic nature of the voltage
phasors over time. Remember that these phasors are all continuously moving quantities, because
the voltages they represent are sinusoidal functions of time. Only when we use a strobe light keyed
to source A’s positive peak do we see fixed-angle readings of VA = 0o , VB = 60o , and VAB = −60o .
378 CHAPTER 5. AC ELECTRICITY

You may wonder what might happen if we keep the three phasometers connected to the same
points, but change the strobe light’s reference point. The answer to this question is that the new
reference source voltage will now be our zero-degree definition, with every phasor’s angle changing
to match this new reference. We may see this effect here, using the same two sources and three
phasometers, but moving the strobe light from source A to source B:

Phasometer
105o 90o 75o
120o 60o
135o 45o

150o 30o

165o VAB 15o Phasometer arrow appears

180o
0o "frozen" at the -120o mark
-15o
Strobe -165o

-150o -30o

light -135o

-120o -60o
-45o

-105o -90o -75o

Phasometer B A
105o 90o 75o
120o 60o
135o 45o

150o 30o

165o VB 15o

180o

5 V ∠ 0o 5 V ∠ -60o
0o

-165o -15o

-150o -30o

-135o -45o
-120o -60o
-105o -90o -75o

Phasometer arrow appears

"frozen" at the 0o mark
Phasometer
105o 90o 75o
120o 60o
135o 45o

150o 30o

Phasometer arrow appears 165o

180o
VA 15o

"frozen" at the -60o mark 0o

-165o -15o

o o
-150 -30

-135o -45o
-120o -60o
-105o -90o -75o

Having the strobe light keyed to source B makes it flash 60 degrees earlier than it did when it
was keyed to source A, which in turn “freezes” the arrows on the faces of all three phasometers −60o
from where they used to be when source A was the time-reference. The following table compares
the phasometer readings with both strobe light reference points:

Strobe reference VA VB VAB

o 0
Source A 0 60 −60o
Source B −60o 00 −120o

It is even possible to imagine connecting the strobe light to a special electronic timing circuit
pulsing the light at 60 Hz (the AC system’s frequency), synchronized to an atomic clock so as to
keep ultra-accurate time. Such an arrangement would permit phasor angle measurements based on
an absolute time reference (i.e. the atomic clock) rather than a relative time reference (i.e. one of
the AC voltage sources). So long as the two AC sources maintained the same frequency and phase
shift, the three phasometers would still be displaced 60o from each other, although it would only be
by blind luck that any of them would point toward 0o (i.e. that any one voltage would happen to
be precisely in-phase with the electronic timer circuit’s 60 Hz pulse).
5.5. PHASORS 379

Interestingly, this technique of measuring AC power system phasors against an absolute time
reference is a real practice and not just a textbook thought experiment. The electrical power industry
refers to this as a synchrophasor measurement. Special instruments called Phasor Measurement
Units or PMU s connect to various points within the power system to acquire real-time voltage
and current measurements, each PMU also connected to a GPS (Global Positioning System) radio
receiver to obtain an absolute time reference with sub-microsecond uncertainty. Each PMU uses
the ultra-precise time signal from the GPS receiver to synchronize a “standard” cosine wave at the
power system frequency such that this reference waveform is at its positive peak (0o ) at the top of
every second in time.
Synchrophasor technology makes it possible to perform simultaneous comparisons of phasor
angles throughout a power system, which is useful for such tasks as analyzing frequency stability
in a power grid or detecting “islanding” conditions where tripped circuit breakers segment a power
grid and allow distributed generators to begin drifting out of sync with each other.
380 CHAPTER 5. AC ELECTRICITY

5.5.3 Phasor expressions of impedance

The ultimate purpose of phasors is to simplify AC circuit analysis, so this is what we will explore
now. Consider the problem of defining electrical opposition to current in an AC circuit. In DC
(direct-current) circuits, resistance (R) is defined by Ohm’s Law as being the ratio between voltage
(V ) and current (I):
V
R=
I
There are some electrical components, though, which do not obey Ohm’s Law. Capacitors and
inductors are two outstanding examples. The fundamental reason why these two components do not
follow Ohm’s Law is because they do not dissipate energy like resistances do. Rather than dissipate
energy (in the form of heat and/or light), capacitors and inductors store and release energy from and
to the circuit in which they are connected. The contrast between resistors and these components
is remarkably similar to the contrast between friction and inertia in mechanical systems. Whether
pushing a flat-bottom box across a floor or pushing a heavy wheeled cart across a floor, work is
required to get the object moving. However, the flat-bottom box will immediately stop when you
stop pushing it due to the energy loss inherent to friction, while the wheeled cart will continue to
coast because it has kinetic energy stored in it. When a resistor disconnected from an electrical
source, both voltage and current immediately cease. Capacitors and inductors, however, may store
a “charge” of energy when disconnected from a source (capacitors retaining voltage and inductors
retaining current).

The relationships between voltage and current for capacitors (C) and inductors (L) are as follows:
dV dI
I=C V =L
dt dt
Expressed verbally, capacitors pass electric current proportional to how quickly the voltage across
them changes over time. Conversely, inductors produce a voltage drop proportional to how quickly
current through them changes over time. The symmetry here is beautiful: capacitors, which store
energy in an electric field that is proportional to the applied voltage, oppose changes in voltage.
Inductors, which store energy in a magnetic field that is proportional to applied current, oppose
changes in current. The manner in which a capacitor or an inductor reacts to changes imposed upon
it is a direct consequence of the Law of Energy Conservation: since energy can neither appear from
nothing nor simply vanish, an exchange of energy must take place in order to alter the amount of
energy stored within a capacitor or an inductor. The rate at which a capacitor’s voltage may change
is directly related to the rate at which electric charge (current) enters or exits the capacitor. The
rate at which an inductor’s current may change is directly related to the amount of electromotive
force (voltage) impressed across the inductor.
When either type of component is placed in an AC circuit and subjected to oscillating signals,
it will pass a finite amount of alternating current. Even though the mechanism of a capacitor’s or
inductor’s opposition to current (called reactance) is fundamentally different from that of a resistor
(called resistance), just like inertia differs in its fundamental nature from friction, it is still convenient
to express the amount of electrical opposition in a common unit of measurement: the ohm (Ω). To
do this, we will have to figure out a way to take the above equations and manipulate them to express
each component’s behavior as a ratio of VI .
5.5. PHASORS 381

Let’s start with capacitors. Suppose we impress a 1 volt peak AC voltage across a capacitor,
representing that voltage as the exponential ejωt where ω is the angular velocity (frequency) of the
signal and t is time:

AC voltage
source
C Capacitor
V = ejωt
(1 volt peak)

We will begin by writing the current/voltage relationship for a capacitor, along with the
imaginary exponential function for the impressed voltage:
dV
I=C V = ejωt
dt
Substituting ejωt for V in the capacitor formula, we see we must apply the calculus function of
differentiation to the time-based voltage function:
d jωt

I=C e
dt
Fortunately for us, differentiation is a very simple17 process with exponential functions:

I = jωCejωt
Remember that our goal here is to solve for the ratio of voltage over current for a capacitor.
So far all we have is a function for current (I) in terms of time (t). If we take this function for
current and divide that into our original function for voltage, however, we see that ratio simplify
quite nicely:

V ejωt
=
I jωCejωt
V 1 1
= = −j
I jωC ωC
Note18 how the exponential term completely drops out of the equation, leaving us with a clean
ratio strictly in terms of capacitance (C), angular velocity (ω), and of course j.

17 Recall from calculus that the derivative of the function ex with respect to x is simply ex . That is, the value of an

exponential function’s slope is equal to the value of the original exponential function! If the exponent contains any
constants multiplied by the independent variable, those constants become multiplying coefficients after differentiation.
Thus, the derivative of ekx with respect to x is simply kekx . Likewise, the derivative of ejωt with respect to t is
jωejωt .
18 Note also one of the interesting properties of the imaginary operator: 1 = −j. The proof of this is quite simple:
j
1 j j
j
= j2
= −1
= −j.
382 CHAPTER 5. AC ELECTRICITY

Next, will will apply this same analysis to inductors. Recall that voltage across an inductor and
current through an inductor are related as follows:
dI
V =L
dt
If we describe the AC current19 through an inductor using the familiar imaginary exponential
expression I = ejωt (representing a 1 amp peak AC current at frequency ω), we may substitute this
expression for current into the inductor’s characteristic equation to solve for the inductor’s voltage
as a function of time:
dI
V =L I = ejωt
dt
d jωt

V =L e
dt

V = jωLejωt
Now that we have the inductor’s voltage expressed as a time-based function, we may include the
original current function and calculate the ratio of V over I:

V jωLejωt
=
I ejωt
V
= jωL
I

In summary, we may express the impedance (voltage-to-current ratio) of capacitors and inductors
by the following equations:
1 1
ZL = jωL ZC = or − j
jωC ωC
Most students familiar with electronics from an algebraic perspective (rather than calculus) find
the expressions XL = 2πf L and XC = 2πf1 C easier to grasp. Just remember that angular velocity
(ω) is really “shorthand” notation for 2πf , so these familiar expressions may be alternatively written
1
as XL = ωL and XC = ωC .

19 Note that we begin this analysis with an exponential expression of the current waveform rather than the voltage

waveform as we did at the beginning of the capacitor analysis. It is possible to begin with voltage as a function of
time and use calculus to determine current through the inductor, but unfortunately that would necessitate integration
rather than differentiation. Differentiation is a simpler process, which is why this approach was chosen. If ejωt = L dI
dt
R R jωt
then ejωt dt = L dI. Integrating both sides of the equation yields ejωt dt = L dI. Solving for I yields ejωL plus
a constant of integration representing a DC component of current that may or may not be zero depending on where
the impressed voltage sinusoid begins in time. Solving for Z = V /I finally gives the result we’re looking for: jωL.
Ugly, no?
5.5. PHASORS 383

Furthermore, recall that reactance (X) is a scalar quantity, having magnitude but no direction.
Impedance (Z), on the other hand, possesses both magnitude and direction (phase), which is why
the imaginary operator j must appear in the impedance expressions to make them complete. The
impedance offered by pure inductors and capacitors alike are nothing more than their reactance
values (X) scaled along the imaginary (j) axis (phase-shifted 90o ).
These different representations of opposition to electrical current are shown here for components
exhibiting 50 ohms, resistances (R) and reactances (X) shown as scalar quantities near the
component symbols, and impedances (Z) as phasor quantities on the complex plane:

+imag +imag +imag

R = 50 Ω XC = 50 Ω XL = 50 Ω

ZL = 50 Ω ∠ 90o
ZR = 50 Ω ∠ 0o ZL = 0 + j50 Ω
ZR = 50 + j0 Ω
-real +real -real +real -real +real

ZC = 50 Ω ∠ -90o
ZC = 0 - j50 Ω

-imag -imag -imag

384 CHAPTER 5. AC ELECTRICITY

5.5.4 Phasor arithmetic

Another detail of phasor math that is both beautiful and practical is the famous expression of Euler’s
Relation, the one all math teachers love because it directly relates several fundamental constants
in one elegant equation (remember that i and j mean the same thing, just different notational
conventions for different disciplines):

eiπ = −1
This equation is actually a special case of Euler’s Relation, relating imaginary exponents of e to
sine and cosine functions:

eiθ = cos θ + i sin θ

What this equation says is really quite amazing: if we raise e to an imaginary exponent of some
angle (θ), it is equivalent to the real cosine of that same angle plus the imaginary sine of that
same angle. Thus, Euler’s Relation expresses an equivalence between exponential and trigonometric
functions. Specifically, the angle θ describes which way a phasor points on a complex plane, the real
and imaginary coordinates of a unit phasor’s20 tip being equal to the cosine and sine values for that
angle, respectively.
If we set the angle θ to a value equal to π, we see the general form of Euler’s relation transform
into eiπ = −1:

eiπ = cos π + i sin π

eiπ = −1 + i0

eiπ = −1

20 A “unit” phasor is one having a length of 1.

5.5. PHASORS 385

After seeing this, the natural question to ask is what happens when we set θ equal to other,
common angles such as 0, π2 , or 3π π
2 (also known as − 2 )? The following examples explore these
angles:

Angle (θ) Exponential Trigonometric Rectangular Polar

0 radians = 0o ei0 cos 0 + i sin 0 1 + i0 = 1 1 6 0o
π/2 radians = 90o eiπ/2 cos 90o + i sin 90o 0 + i1 = i 1 6 90o
π radians = 180o eiπ cos 180o + i sin 180o −1 + i0 = −1 1 6 180o
−π/2 radians = −90o ei−π/2 cos −90o + i sin −90o 0 − i1 = −i 1 6 −90o

We may show all the equivalences on the complex plane, as unit phasors:

+imaginary
iπ/2
e =i

eiπ = -1 ei0 = 1
-real +real

ei3π/2 = e-iπ/2 = -i
-imaginary
386 CHAPTER 5. AC ELECTRICITY

As we saw previously, the amount of opposition to electrical current offered by reactive

components (i.e. inductors and capacitors) – a quantity known as impedance – may be expressed as
functions of j and ω:

ZL = jωL

1
ZC = −j
ωC
Knowing that j is equal to ejπ/2 and that −j = e−jπ/2 , we may re-write the above expressions
for inductive and capacitive impedance as functions of an angle:

ZL = ωLejπ/2

1 −jπ/2
e
ZC =
ωC
Using polar notation as a “shorthand” for the exponential term, the impedances for inductors
and capacitors are seen to have fixed angles21 :
π
ZL = ωLejπ/2 = ωL6 radians = ωL6 90o
2
1 −jπ/2 1 π 1
ZC = e = 6 − radians = 6 − 90o
ωC ωC 2 ωC
Beginning electronics students will likely find the following expressions of inductive and capacitive
impedance more familiar, 2πf being synonymous with ω:

ZL = 2πf 6 90o

1
ZC = 6 − 90o
2πf C

21 The fact that these impedance phasor quantities have fixed angles in AC circuits where the voltage and current

phasors are in constant motion is not a contradiction. Since impedance represents the relationship between voltage
and current for a component (Z = V /I), this fixed angle represents a relative phase shift between voltage and current.
In other words, the fixed angle of an impedance phasor tells us the voltage and current waveforms will always remain
that much out of step with each other despite the fact that the voltage and current phasors themselves are continuously
rotating at the system frequency (ω).
5.5. PHASORS 387

When performing circuit analysis in AC systems we must have a way of managing phasor
quantities in the context of Ohm’s Law and Kirchhoff’s Laws. Kirchhoff’s Voltage Law and
Kirchhoff’s Current Law both require the addition and subtraction of quantities. Ohm’s Law requires
the multiplication and division of quantities.
The basic rules of phasor arithmetic are listed here, with phasors having magnitudes of A and
B, and angles of M and N , respectively:

AejM + BejN = (A cos M + B cos N ) + j(A sin M + B sin N )

AejM − BejN = (A cos M − B cos N ) + j(A sin M − B sin N )

AejM × BejN = ABej(M +N )

A h j(M −N ) i
AejM ÷ BejN = e
B
Addition and subtraction lend themselves readily to the rectangular form of phasor expression,
where the real (cosine) and imaginary (sine) terms simply add or subtract. Multiplication and
division lend themselves readily to the polar form of phasor expression, where magnitudes multiply
or divide and angles add or subtract.

To summarize:

• Voltages and currents in AC circuits may be mathematically represented as phasors, which are
imaginary exponential functions (e raised to imaginary powers)

• Phasors are typically written in either rectangular form (real + imaginary) or polar form
(magnitude @ angle)
• Ohm’s Law and Kirchhoff’s Laws still apply in AC circuits as long as all quantities are in
phasor notation

• Addition is best done in rectangular form: add the real parts, and add the imaginary parts

• Subtraction is best done in rectangular form: subtract the real parts, and subtract the imaginary
parts

• Multiplication is best done in polar form: multiply the magnitudes, and add the angles

• Division is best done in polar form: divide the magnitudes, and subtract the angles
388 CHAPTER 5. AC ELECTRICITY

5.5.5 Phasors and circuit measurements

A phasor is a special form of vector (a quantity possessing both magnitude and direction) lying in a
complex plane. Phasors relate circular motion to simple harmonic (sinusoidal) motion as shown in
the following diagram. In AC electrical circuits, this is the relationship between an electromechanical
generator’s shaft angle22 and the sinusoidal voltage it outputs:

π/2
3π/8
π/4

π/8
Sine wave

π 0

π/8 3π/8
3π/2 0 π/4 π/2 π 3π/2 2π
0
π/8
π/4
3π/8
π/2

Cosine wave
π
3π/2
2π

In any operating AC electrical circuit the phasors, just like the sinusoidal waveforms, never rest:
they are in continuous motion. When we speak of a phasor as having a fixed angle, what we really
mean is that the phasor is either leading or lagging with respect to some other “reference” phasor
in the system, not that the phasor itself is stationary. We explored this concept previously, where
we used an invented instrument called a “phasometer” to represent the direction of a measured
phasor in real time, and then used a strobe light connected to some point in the same system to
visually “freeze” the phasometer arrows so we could see which direction each arrow was pointed at
the moment in time when our chosen reference waveform reached its positive peak value (i.e. 0o on a
cosine wave). The fixed angle represented by each “strobed” phasometer arrow therefore represented
the amount of relative phase shift between each respective phasor and the reference phasor.

22 Assuming a two-pole generator, where each period of the sinusoidal waveform corresponds exactly to one revolution

of the generator shaft.

5.5. PHASORS 389

Phasor angles are to AC quantities what arithmetic signs are to DC quantities. If a phasometer
registers an angle of 180 degrees, it means the red lead is fully negative and the black lead is
fully positive at the moment in time when the strobe light flashes. If a DC voltmeter registers a
negative voltage, it means the red lead is negative and the black lead is positive at the time of the
measurement:

Digital multimeter
Voltmeter registers
a negative value
Phasometer arrow appears
"frozen" at the 180o mark Phasometer
90o

135o
120o
105o 75o
60o

45o
V A
150o 30o

165o 15o

180o
0o

Strobe -165o -15o V A

OFF
light -150 o

-135o -45o
-30 o

-120o -60o
-105o -90o -75o

AC DC + A COM

source source −

Reversing each instrument’s test lead connections to the circuit will reverse its indication: flipping
red and black leads on the phasometer will cause its indication to be 0o instead of 180o ; flipping red
and black leads on the DC voltmeter will cause it to register a positive voltage instead of a negative
voltage:

Digital multimeter
Voltmeter registers
a positive value
Phasometer arrow appears
"frozen" at the 0o mark Phasometer
90o

135o
120o
105o 75o
60o

45o
V A
150o 30o

o o
165 15

180o
0o

Strobe -165 o -15o V A

OFF
light -150o

-135o -45o
-30o

-120o -60o
-105o -90o -75o

AC DC + A COM

source source −

From these demonstrations we can see that the indication given by a measuring instrument
depends as much on how that instrument connects to the circuit as it does on the circuit quantity
itself. Likewise, a voltage or current value obtained in the course of analyzing a circuit depends as
much on how we label the assumed polarity (or direction) of that quantity on the diagram as it does
on the quantity itself.
390 CHAPTER 5. AC ELECTRICITY

To illustrate the importance of signs, assumed directions, and phasor angles in circuit analysis,
we will consider two multi-source circuits: one DC and one AC. In each case we will explore how
Kirchhoff’s Current Law relates to currents at a similar node in each circuit, relating the current
arrow directions to signs and phasor angles.

First, the DC circuit:

5 mA 5 mA 5 mA 5 mA

+ +
5V
− 5V
− 1 kΩ 1 kΩ

5 mA 5 mA
5 mA 5 mA

10 mA

With each 1 kΩ resistor powered by its own 5 volt DC source, the current through each of these
resistors will be 5 milliamps in accordance with Ohm’s Law (I = VR ). The direction of each current
is easily predicted by examining the polarity of each source (with current represented in conventional
flow notation exiting the positive terminal and entering the negative terminal of each source) and/or
by examining the polarity of the voltage drop across each resistor (with conventional-flow current
entering the positive terminal and exiting the negative terminal of each load). At the node below
the two resistors, these two currents join together to form a larger (10 milliamp) current headed
toward ground. Kirchhoff’s Current Law declares that the sum of all currents entering a node must
equal the sum of all currents exiting that node. Here we see this is true: 5 milliamps in plus 5
milliamps in equals 10 milliamps out.
The voltage polarities and current directions in this DC circuit are all clear and unambiguous
because the quantities are constant over time. Each power supply acts as an energy source and each
resistor acts as an energy load, all the time. Our application of Kirchhoff’s Current Law at the node
is so obvious it hardly requires explanation: the flow of charge carriers (current) in and out must be
in equilibrium.

With AC, however, we know things will not be so simple. AC circuit quantities are not constant
over time as they are in DC circuits. Reactive (energy-storing) components such as inductors and
capacitors play alternating roles as sources and loads, complicating their analysis. The good news
in all of this is that phasor representations of voltage and current allow us to apply all the same
fundamental principles we are accustomed to using for DC circuits (e.g. Ohm’s Law, Kirchhoff’s
Voltage Law, Kirchhoff’s Current Law, network theorems, etc.). In order to use these principles to
calculate AC quantities, though, we must be careful in how we label them in the circuit.
5.5. PHASORS 391

Here, we will modify the circuit to include two AC power sources (phase-shifted by 60o ), and
replace one of the 1 kΩ resistors with a capacitor exhibiting 1 kΩ of reactance at the system frequency.
As in the DC circuit, each of the two loads sees the full 5 volts of its respective source. Unlike the
DC circuit, we must represent each of the voltage and impedance quantities in complex (phasor)
form in order to apply Ohm’s Law to calculate load currents.
One more thing we will do to this circuit before beginning any phasor arithmetic is to label each
voltage and each current as we did in the DC circuit. Placing + and − symbols at the terminals
of AC voltage sources, and drawing conventional-flow notation arrows showing alternating current
through loads at first may seem absurd, since we know this is an AC circuit and so by definition these
quantities lack fixed polarities and directions. However, these symbols will become very important
to us because they serve to define what 0o means for each of the respective phasors. In other words,
these polarities and arrows merely show us which way the voltages and currents are oriented when
each of those phasors is at its positive peak – i.e. when each phasor angle comes around to its own
0o mark. As such, the polarities and arrows we draw do not necessarily represent simultaneous
conditions. We will rely on the calculated phasor angles to tell us how far these quantities will
actually be phase-shifted from each other at any given point in time:

5 V ∠ 60o 5 V ∠ 0o 1 kΩ ∠ 0o 1 kΩ ∠ -90o

First, calculating current through the resistor23 , recalling that the impedance of a resistor has a
0 degree phase angle (i.e. no phase shift between voltage and current):
VR 5 V6 0o
IR = = = 5 mA6 0o
ZR 1 kΩ6 0o

Next, calculating current through the capacitor24 , recalling that the impedance for a capacitor
has a −90 degree phase angle because voltage across a capacitor lags 90 degrees behind current
through a capacitor:
VC 5 V6 60o
IC = = = 5 mA6 150o
ZC 1 kΩ6 − 90o

23 When dividing two phasors in polar form, the arithmetic is as follows: divide the numerator’s magnitude by the

denominator’s magnitude, then subtract the denominator’s angle from the numerator’s angle. The result in this case
is 5 milliamps (5 volts divided by 1000 ohms) at an angle of 0 degrees (0 minus 0).
24 The same arithmetic applies to this quotient as well: the current’s magnitude is 5 volts divided by 1000 ohms,

while the current’s phase angle is 60 degrees minus a negative 90 degrees (150 degrees).
392 CHAPTER 5. AC ELECTRICITY

From the arrows sketched at the node we can see the total current headed toward ground must be
equal to the sum of the two currents headed into the node, just as with the DC circuit. The difference
here is that we must perform the addition using phasor quantities instead of scalar quantities in order
to account for phase shift between the two currents:

Itotal = IR + IC

Itotal = 5 mA6 0o + 5 mA6 150o

Representing these current phasors in rectangular mode so we may sum their real and imaginary
parts:

Itotal = (5 + j0 mA) + (−4.33 + j2.5 mA)

Itotal = 0.67 + j2.5 mA

Converting the rectangular form into polar form:

Itotal = 2.59 mA6 75o

This result is highly non-intuitive. When we look at the circuit and see two 5 milliamp currents
entering a node, we naturally expect a sum total of 10 milliamps to exit that node. However, that
will only be true if those 5 mA quantities are simultaneous: i.e. if the two currents are 5 mA in
magnitude at the same point in time. Since we happen to know these two currents are phase-shifted
from each other by 150 degrees (nearly opposed to each other), they never reach their full strength
at the same point in time, and so their sum is considerably less than 10 milliamps:

1 kΩ ∠ 0o 1 kΩ ∠ -90o

5 mA ∠ 0o 5 mA ∠ 150o

2.59 mA ∠ 75o
5.5. PHASORS 393

We may make more sense of this result by adding current-sensing phasometers to the circuit,
their red and black test leads connected in such a way as to match the arrows’ directions (current
entering the red lead and exiting the black lead) as though these were DC ammeters being connected
into a DC circuit:

5 mA ∠ 0o 5 mA ∠ 150o

5 V ∠ 60o 5 V ∠ 0o 1 kΩ ∠ 0o 1 kΩ ∠ -90o

105o 90o 75o 105o 90o 75o

120o 60o 120o 60o
135o 45o 135o 45o

IR IC
o o o
150 30 150 30o

165o 15o 165o 15o

o o
180 180
0o 0o

-165o -15o -165o -15o

-150o -30o -150o -30o

Strobe -135o

-120o
-105o -75o
-60o
-45o -135o

-120o
-105o -75o
-60o
-45o

-90o -90o
light

105o 90o 75o

120o 60o
135o 45o

150o 30o

165o

180o
Itotal 15o

0o

-165o -15o

o o
-150 -30

-135o -45o
-120o -60o
-105o -90o -75o

2.59 mA ∠ 75o
394 CHAPTER 5. AC ELECTRICITY

A very common and practical example of using “DC notation” to represent AC voltages and
currents is in three-phase circuits where each power supply consists of three AC voltage sources
phase-shifted 120 degrees from each other. Here, we see a “wye” connected generator supplying
power to a “wye” connected load. Each of the three generator stator windings outputs 277 volts,
with the entirety of that voltage dropped across each of the load resistances:

2.77 A ∠ 0o

2.77 A ∠ 240o

277 V ∠ 0o 100 Ω ∠ 0o 100 Ω ∠ 0o

277 V ∠ 240o
(277 V ∠ -120o)

277 V ∠ 120o 100 Ω ∠ 0o

2.77 A ∠ 120o

Note the directions of the three currents (red arrows) in relation to the node at the center of
the generator’s “wye” winding configuration, and also in relation to the node at the center of the
resistive load’s “wye” configuration. At first this seems to be a direct violation of Kirchhoff’s Current
Law: how is it possible to have three currents all exiting a node with none entering or to have three
currents all entering a node with none exiting? Indeed, this would be impossible if the currents were
simultaneously moving in those directions, but what we must remember is that each of the current
arrows simply shows which way each current will be moving when its respective phasor comes around
to the 0o mark. In other words, the arrows simply define what zero degrees means for each current.
Similarly, the voltage polarities would suggest to anyone familiar with Kirchhoff’s Voltage Law in
DC circuits that the voltage existing between any two of the power conductors between generator
and load should be 0 volts, reading the series-opposed sum of two 277 volt sources. However, once
again we must remind ourselves that the + and − symbols do not actually represent polarities at
the same point in time, but rather serve to define the orientation of each voltage when its phasor
happens to point toward 0o .
5.5. PHASORS 395

If we were to connect three current-sensing phasometers to measure the phase angle of each line
current in this system (keying the strobe light to the voltage of the 0o generator winding), we would
see the true phase relationships relative to that time reference. The directions of the current arrows
and the orientation of the + and − polarity marks serve as references for how we shall connect the
phasometers to the circuit (+ on red, − on black; current entering red, exiting black):

105o 90o 75o

120o 60o
135o 45o

150o 30o

165o 15o

180o
0o

-165o -15o

o o
-150 -30

Strobe -135o

-120o -60o
-45o

light -105o -90o -75o

2.77 A ∠ 0o

2.77 A ∠ 240o

105o 90o 75o

120o 60o
135o 45o

150o 30o

165o 15o

180o
0o

277 V ∠ 0 o
100 Ω ∠ 0o 100 Ω ∠ 0o
277 V ∠ 240o -165o -15o

(277 V ∠ -120o)
o o
-150 -30

-135o -45o
-120o -60o
-105o -90o -75o

277 V ∠ 120o 100 Ω ∠ 0o

2.77 A ∠ 120o

105o 90o 75o

120o 60o
135o 45o
o
150 30o

165o 15o
o
180
0o

-165o -15o

-150o -30o

-135o -45o
-120o -60o
-105o -90o -75o
396 CHAPTER 5. AC ELECTRICITY

We may use the three arrows at the load’s center node to set up a Kirchhoff’s Current Law
equation (i.e. the sum of all currents at a node must equal zero), and confirm that the out-of-phase
currents all “entering” that node do indeed amount to a sum of zero:

Inode = 2.77 A6 0o + 2.77 A6 120o + 2.77 A6 240o

Converting polar expressions into rectangular so we may see how they add together:

Inode = (2.77 + j0 A) + (−1.385 + j2.399 A) + (−1.385 − j2.399 A)

Combining all the real terms and combining all the imaginary terms:

Inode = (2.77 − 1.385 − 1.385) + j(0 + 2.399 − 2.399) A

Inode = 0 + j0 A
Looking closer at these results, we may determine which way the three currents were actually
flowing at the time of the strobe’s flash (when the upper-right generator winding is at its peak
positive voltage). The first phasor has a real value of +2.77 amps at that instant in time, which
represents 2.77 amps of actual current flowing in the actual direction of our sketched current arrow.
The second and third phasors each have real values of −1.385 amps at that same instant in time,
representing 1.385 amps each flowing against the direction of our sketched arrows (i.e. leaving that
node). Thus, we can see that the “snapshot” view of currents at the time of the strobe’s flash makes
complete sense: one current of 2.77 amps entering the node and two currents of 1.385 amps each
exiting the node. Viewed at any instant in time, the principles of DC circuits hold true. It is only
when we sketch polarity marks and draw arrows representing voltages and currents reaching their
positive peak values at different times that the annotations seem to violate basic principles of circuit
analysis.

5.6 The s variable

A powerful mathematical concept useful for analyzing practically any physical system – electrical
circuits included – is something called the s variable. The s variable is closely related to Euler’s
Relation and phasor expressions of waveforms, which is why a discussion of it is included here.
5.6. THE S VARIABLE 397

5.6.1 Meaning of the s variable

As we saw previously, Euler’s Relation allows us to express rotating phasors as imaginary exponents
of e. For example, Aejθ represents a phasor of length A at an angle of θ radians. Aejωt represents
a phasor of length A rotating at a velocity of ω radians per second at a particular instant in time
t. This happens to be an incredibly useful mathematical “trick” for representing sinusoidal waves
in physical systems. For example, if we wished to mathematically express a sinusoidal AC voltage
as a function of time with a peak voltage value of 10 volts and a frequency of 60 hertz (377 radians
per second, since ω = 2πf ), we could do so like this:

V (t) = 10ej377t
Exponential functions aren’t just useful for expressing sinusoidal waves, however. They also work
well for expressing rates of growth and decay, as is the case with RC and L/R time-delay circuits
where exponential functions describe the charging and discharging of capacitors and inductors. Here,
the exponent is a real number rather than an imaginary number: the expression e−t/τ approaching
zero as time (t) increases. The Greek letter “tau” (τ ) represents the time constant of the circuit,
which for capacitive circuits is the product of R and C, and for inductive circuits is the quotient of
L and R. For example, if we wished to mathematically express the decaying voltage across a 33 µF
capacitor initially charged to 10 volts as it dissipates its stored energy through a 27 kΩ resistor (the
circuit having a time constant of 0.891 seconds, since τ = RC), we could do so like this:

V (t) = 10e−(t/0.891)
The sign of the exponential term here is very important: in this example we see it is a negative
number. This tells us the function decays (approaches zero) over time, since larger positive values of
t result in larger negative values of t/τ (recall from algebra that a negative exponent is the equivalent
of reciprocating the expression, so that e−x = e1x ). If the exponent were a real positive number, it
would represent some quantity growing exponentially over time. If the exponent were zero, it would
represent a constant quantity. We expect a discharging resistor-capacitor circuit to exhibit decaying
voltage and current values, and so the negative exponent sign shown here makes sense.

If imaginary exponents of e represent phasors, and real exponents of e represent growth or decay,
then a complex exponent of e (having both real and imaginary parts) must represent a phasor that
grows or decays in magnitude over time. Engineers use the lower-case Greek letter “omega” (ω) along
with the imaginary operator j to represent the imaginary portion, and the lower-case Greek letter
“sigma”25 (σ) to represent the real portion. For example, if we wished to mathematically express
a sine wave AC voltage with a frequency of 60 hertz (377 radians per second) and an amplitude
beginning at 10 volts but decaying with a time constant (τ ) of 25 milliseconds (σ = 1/τ = 40 time
constants per second), we could do so like this:

V (t) = 10e−40t+j377t
We may factor time from the exponential terms in this expression, since t appears both in the
real and imaginary parts:

V (t) = 10e(−40+j377)t
25 σ is equal to the reciprocal of the signal’s time constant τ . In other words, σ = 1/τ .
398 CHAPTER 5. AC ELECTRICITY

With t factored out, the remaining terms −40 + j377 completely describe the sinusoidal wave’s
characteristics. The wave’s decay rate is described by the real term (σ = −40 time constants per
second), while the wave’s phase is described by the imaginary term (jω = 377 radians per second).
Engineers use a single variable s to represent the complex quantity σ + jω, such that any growing
or decaying sinusoid may be expressed very succinctly as follows:

Aest = Ae(σ+jω)t = Aeσt ejωt

Where,
A = Initial amplitude of the sinusoid (e.g. volts, amps) at time t = 0 (arbitrary units)
s = Complex growth/decay rate and frequency (sec−1 )
σ = τ1 = Real growth/decay rate (time constants per second, or sec−1 )
jω = Imaginary frequency (radians per second, or sec−1 )
t = Time (seconds)

Separating the expression Aeσt ejωt into three parts – A, eσt , and ejωt – we get a complete
description of a rotating phasor:

A = Initial amplitude of the phasor (t = 0)

eσt = How much the phasor’s magnitude has grown (σ > 0) or decayed (σ < 0) at time t
ejωt = Unit phasor (length = 1) at time t
5.6. THE S VARIABLE 399

If we set ω at some constant value and experiment with different values of σ, we can see the
effect σ has on the shape of the wave over time:

Growing sinusoid Steady sinusoid A

Decaying sinusoid
time time time
A

Ae(σ + jω)t Ae(σ + jω)t Ae(σ + jω)t

t=0 where σ > 0 t=0 where σ = 0 t=0 where σ < 0

If we set σ at zero and experiment with different values26 of ω, we can see the effect ω has on
the shape of the wave over time:

Zero frequency Low frequency High frequency

time time time
A A A

Ae(σ + jω)t Ae(σ + jω)t Ae(σ + jω)t

t=0 where ω = 0 t=0 where ω > 0 t=0 where ω >> 0

As we will soon see, characterizing a sinusoidal response using the complex variable s allows us
to mathematically describe a great many things. Not only may we describe voltage waveforms
using s as shown in these simple examples, but we may also describe the response of entire
physical systems including electrical circuits, machines, feedback control systems, and even chemical
reactions. In fact, it is possible to map the essential characteristics of any linear system in terms of
how exponentially growing, decaying, or steady sinusoidal waves affect it, and that mapping takes
the form of mathematical functions of s.
When engineers or technicians speak of a resonant system, they mean a circuit containing
inductive and capacitive elements tending to sustain oscillations of a particular frequency (ω).
A lossless resonant system (e.g. a superconducting tank circuit, a frictionless pendulum) may
be expressed by setting the real portion of s equal to zero (σ = 0 ; no growth or decay) and
26 One value of ω not shown in this three-panel graphic comparison is a negative frequency. This is actually not as

profound as it may seem at first. All a negative value of ω will do is ensure that the phasor will rotate in the opposite
direction (clockwise, instead of counter-clockwise as phasor rotation is conventionally defined). The real portion of
the sinusoid will be identical, tracing the same cosine-wave plot over time. Only the imaginary portion of the sinusoid
will be different, as j sin −θ = −j sin θ.
400 CHAPTER 5. AC ELECTRICITY

letting the imaginary portion represent the resonant frequency (jω = j2πf ). Real-life resonant
systems inevitably dissipate some energy, and so a real resonant system’s expression will have both
an imaginary portion to describe resonant frequency and a negative real portion to describe the
oscillation’s rate of decay over time.
Systems exhibiting a positive σ value are especially interesting because they represent instability:
unrestrained oscillatory growth over time. A feedback control loop with excessive gain programmed
into the loop controller is a simple example of a system where σ > 1. This situation, of course, is
highly undesirable for any control system where the goal is to maintain the process variable at a
steady setpoint.

5.6.2 Impedance expressed using the s variable

Previously, we saw how the impedance of inductors and capacitors could be calculated using jω
to represent the frequency of the applied signal. Doing so greatly simplified the mathematics by
eliminating the need to manipulate trigonometric functions such as sine and cosine. Here, we will
discover that s works just as nicely for the same task, with the added benefit of showing how
inductors and capacitors react to exponentially growing or decaying signals.
First, let’s begin with capacitors. We know that voltage across a capacitor and current “through”
a capacitor are related as follows:
dV
I=C
dt
Next, we substitute an expression27 for voltage in terms of s and then use calculus to differentiate
it with respect to time:
d st

I=C e
dt

I = sCest
V
The ratio of I (the definition of impedance) will then be:

V est
ZC = =
I sCest
1
ZC =
sC
Instead of the common scalar expression for capacitive impedance (ZC = 2πf1 C ) which only tells
us the magnitude of the impedance (in ohms) but not the phase shift, we have a complex expression
1
for capacitive impedance (ZC = sC ) describing magnitude, phase shift, and its reaction to the
growth or decay of the signal.

27 The expression used here to represent voltage is simply est . I could have used a more complete expression such as

Aest (where A is the initial amplitude of the signal), but as it so happens this amplitude is irrelevant because there
will be an “A” term in both the numerator and denominator of the impedance quotient. Therefore, A cancels out
and is of no consequence.
5.6. THE S VARIABLE 401

Likewise, we may do the same for inductors. Recall that voltage across an inductor and current
through an inductor are related as follows:
dI
V =L
dt
Substituting an expression for current in terms of s and using calculus to differentiate it with
respect to time:
d st

V =L e
dt

V = sLest
V
The ratio of I (the definition of impedance) will then be:

V sLest
ZL = = st
I e

ZL = sL
As with capacitors, we now have a complex expression for inductive impedance describing
magnitude, phase shift, and its reaction to signal growth or decay (ZL = sL) instead of merely
having a scalar expression for inductive impedance (ZL = 2πf L).

Resistors directly oppose current by dropping voltage, with no regard to rates of change.
Therefore, there are no derivatives in the relationship between voltage across a resistor and current
through a resistor:

V = IR
If we substitute e for current into this formula, we will see that voltage must equal Rest . Solving
st

for the ratio of voltage over current to define impedance:

V Rest
ZR = = st
I e

ZR = R
Not surprisingly, all traces of s cancel out for a pure resistor: its impedance is exactly equal to
its DC resistance.

In summary:

Inductive impedance (ZL ) Capacitive impedance (ZC ) Resistive impedance (ZR )

sL 1/sC R
402 CHAPTER 5. AC ELECTRICITY

Now let’s explore these definitions of impedance using real numerical values. First, let’s consider
a 22 µF capacitor exposed to a steady AC signal with a frequency of 500 Hz. Since the signal in
this case is steady (neither growing nor decaying in magnitude), the value of σ will be equal to zero.
ω is equal to 2πf , and so a frequency of 500 Hz is equal to 3141.6 radians per second. Calculating
impedance is as simple as substituting these values for s and computing 1/sC:
1 1
ZC = =
sC (σ + jω)C
1
ZC =
(0 + j3141.6 sec−1 )(22 × 10−6 F)
1
ZC =
j0.0691
−j
ZC =
0.0691

ZC = 0 − j14.469 Ω (rectangular notation)

ZC = 14.469 Ω 6 − 90o (polar notation)

Thus, the impedance of this capacitor will be 14.469 ohms at a phase angle of −90o . The purely
imaginary nature of this impedance (its orthogonal phase shift between voltage and current) tells us
there is no net power dissipated by the capacitor. Rather, the capacitor spends its time alternately
absorbing and releasing energy to and from the circuit.
5.6. THE S VARIABLE 403

Next, we will consider the case of a 150 mH inductor exposed to an exponentially rising DC
signal with a time constant (τ ) of 5 seconds. 5 seconds per time constant (τ ) is equal to 0.2 time
constants per second (σ). Since the signal in this case is DC and not AC, the value of ω will be
equal to zero. Calculating impedance, once again, is as simple as substituting these values for s and
computing sL:

ZL = sL = (σ + jω)L

ZL = (0.2 + j0 sec−1 )(150 × 10−3 H)

ZL = 0.03 + j0 Ω (rectangular notation)

ZL = 0.03 Ω 6 0o (polar notation)

Thus, the impedance of this inductor will be 0.03 ohms at a phase angle of 0o . The purely real
nature of this impedance (i.e. no phase shift between voltage and current) tells us energy will be
continually absorbed by the inductor, and for this reason it will be seen by the rest of the circuit as
though it were a resistor dissipating energy for however long the signal continues to exponentially
grow.

A phase shift of 0 degrees for a reactive component such as an inductor may come as a surprise
to students accustomed to thinking of inductive impedances always having 90 degree phase shifts!
However, the application of the complex variable s to impedance mathematically demonstrates
we can indeed have conditions of no phase shift given just the right circumstances. This makes
conceptual sense as well if we consider how inductors store energy: if the current through an inductor
increases exponentially over time, never reversing direction, it means the inductor’s magnetic field
will always be growing and therefore absorbing more energy from the rest of the circuit.
404 CHAPTER 5. AC ELECTRICITY

We see something even more interesting happen when we subject a reactive component to a
decaying DC signal. Take for example a 33,000 µF capacitor exposed to a decaying DC signal
with a time constant of 65 milliseconds. 65 milliseconds per time constant (τ ) is equal to 15.38
time constants per second (σ). Once again ω will be zero because this is a non-oscillating signal.
Calculating capacitive impedance:
1 1
ZC = =
sC (σ + jω)C
1
ZC =
(−15.38 + j0 sec−1 )(33000 × 10−6 F)
1
ZC =
−0.508

ZC = −1.970 + j0 Ω (rectangular notation)

ZC = 1.970 Ω 6 180o (polar notation)

A negative real impedance figure represents a phase shift of 180o between voltage and current.
Once again, this may surprise students of electronics who are accustomed to thinking of capacitive
impedances always having phase shifts of −90 degrees. What a 180 degree phase shift means is
the direction of current with respect to voltage polarity has the capacitor functioning as an energy
source rather than as a load. If we consider what happens to a capacitor when it discharges, the
180 degree phase shift makes sense: current flowing in this direction depletes the capacitor’s plates
of stored charge, which means the electric field within the capacitor weakens over time as it releases
that energy to the rest of the circuit.
5.7. TRANSFER FUNCTION ANALYSIS 405

5.7 Transfer function analysis

An extremely important topic in engineering is that of transfer functions. Simply defined, a transfer
function is the ratio of output to input for any physical system, usually with both the output and
input being mathematical functions of s. In other words, we express both the output of a system
and the corresponding input of that system in terms of exponentially growing/decaying sinusoidal
waves and then solve for the ratio of those two expressions.
Unfortunately, the teaching of transfer functions and their relation to real-world phenomena is
often obfuscated by a heavy emphasis on mathematics. The intent of this section is to introduce
this concept in a way that is very “gentle” and continually referenced to real-world applications. If
I can do anything to help lift the veil of mystery surrounding concepts such as transfer functions,
the s variable, and pole-zero plots, then technicians as well as engineers will be able to appreciate
the power of this analytical technique and be able to exchange more ideas in the same “language”.

A simple example of a transfer function is the gain of an electronic amplifier. As all students of
electronics learn, “gain” is the ratio of output signal to input signal for a circuit. Beginning students
learn to represent circuit gains as scalar values (e.g. “The amplifier has a voltage gain of 24”), first
as plain ratios and later as decibel figures (e.g. “The amplifier has a voltage gain of 27.6 dB”).
One limitation of this approach is that it oversimplifies the situation when the gain of the circuit in
question varies with the frequency and/or the growth/decay rate of the signal, which is quite often
the case. If we take the engineering approach of expressing output and input signals as functions of
s, we obtain a more complete picture of that circuit’s behavior over a wide range of conditions.
Another simple example of a transfer function is what we have just seen in this book: the
impedance of a reactive electrical component such as a capacitor or an inductor. Here, the ratio
in question is between voltage and current. If we consider current through the component to be
the “input” signal and voltage across the component to be the “output” signal – both expressed in
terms of s – then impedance Z(s) = VI(s) (s)
is the transfer function for that component. This raises
an important point about transfer functions: what we define as the “input” and the “output” of the
system is quite arbitrary, so long as there is an actual relationship between the two signals.

If we write generalized output/input transfer functions of s for an AC circuit, we may

mathematically analyze that transfer function to gain insight into the behavior and characteristics
of that circuit. Features of transfer functions of interest to us include:

• Zeros: any value(s) of s resulting in a zero value for the transfer function (i.e. zero gain)

• Poles: any value(s) of s resulting in an infinite value for the transfer function (i.e. maximum
gain)

An AC circuit’s zeros tell us where the circuit is unresponsive to input stimuli. An AC circuit’s
poles tell us where the circuit is able to generate an output signal with no input stimulus (i.e. its
natural or un-driven mode(s) of response).

In order to clearly understand the concept of transfer functions, practical examples are very
helpful. Here we will explore some very simple AC circuits in order to grasp what transfer functions
are and how they benefit system analysis.
406 CHAPTER 5. AC ELECTRICITY

5.7.1 Example: LR low-pass filter circuit

First, let’s begin with a simple low-pass filter circuit comprised of an inductor and a resistor
connected in series:

L Z(s) = sL

Vin

R Z(s) = R Vout

The impedance of each component as a function of s is shown in the diagram: the inductor’s
impedance is sL while the resistor’s impedance is simply R. It should be clear to any student of
electronics that these two components will function as a voltage divider, with the output voltage
being some fraction of the input voltage. Knowing this, we may write a transfer function for this
circuit based on the voltage divider formula, which tells us the ratio of output voltage to input
voltage is the same as the ratio of output impedance to total impedance:

Vout (s) R R
Transfer function = = =
Vin (s) R + sL R + (σ + jω)L
This transfer function allows us to calculate the “gain” of the system for any given value of s,
which brings us to the next step of our analysis. At this point we will ask ourselves three questions28 :

1. How does this system respond when s = 0?

2. What value(s) of s make the transfer function approach a value of zero?

3. What value(s) of s make the transfer function approach a value of infinity?

The first of these questions refers to a condition where we apply a steady DC signal to the input
of the system. If s = 0 then both σ and ω must each be equal to zero. A value of zero for σ means
the signal is neither growing nor decaying over time, but remains at some unchanging value. A value
of zero for ω means the signal is not oscillating. These two conditions can only refer to a steady DC
signal applied to the circuit. Substituting zero for s we get:
R
R + 0L
R
=1
R
28 What we are really doing here is applying a problem-solving technique I like to call limiting cases. This is where we

simplify the analysis of some system by considering scenarios where the mathematical quantities are easy to compute.
5.7. TRANSFER FUNCTION ANALYSIS 407

Therefore the transfer function of this circuit is unity (1) under DC conditions. This is precisely
what we would expect given an inductor connected in series with a resistor, with output voltage
taken across the resistor. If there is no change in the applied signal, then the inductor’s magnetic
field will be unchanging as well, which means it will drop zero voltage (assuming a pure inductor
with no wire resistance) leaving the entire input voltage dropped across the resistor.
The second question refers to a condition where the output signal of this circuit is zero. Any
values of s resulting in zero output from the system are called the zeros of the transfer function.
Examining the transfer function for this particular low-pass LR filter circuit, we see that this can
only be true if s becomes infinitely large, because s is located in the denominator of the fraction:
R
=0
R ± ∞L
This is consistent with the behavior of a low-pass filter: as frequency (ω) increases, the filter’s
output signal diminishes. The transfer function doesn’t just tell us how this circuit will respond to
change in frequency, however – it also tells us how the circuit will respond to growing or decaying
signals too. Here, we see that infinitely large σ values also result in zero output: the inductor, which
tends to oppose any current exhibiting a high rate of change, doesn’t allow much voltage to develop
across the resistor if the input signal is growing or decaying very rapidly.

The third question refers to a condition where either the transfer function’s numerator approaches
infinity or its denominator approaches zero. Any values of s having this result are called the poles
of the transfer function. Since the numerator in this particular case is a constant (R), only a
denominator value of zero could cause the transfer function to reach infinity:
R
= ∞ only if R + sL = 0
R + sL
If the necessary condition for a “pole” is that R + sL = 0, then we may solve for s as follows:

R + sL = 0

sL = −R

R
s=−
L
Thus, this transfer function for this simple low-pass filter circuit has one pole located at
s = −R/L. Since both R and L are real numbers (not imaginary) with positive values, then
the value of s for the pole must be a real number with a negative value. In other words, the solution
for s at this pole is all σ and no ω: this refers to an exponentially decaying DC signal.
408 CHAPTER 5. AC ELECTRICITY

It is important at this point to consider what this “pole” condition means in real life. The notion
that a circuit is able to produce an output signal with zero input signal may sound absurd, but it
makes sense if the circuit in question has the ability to store and release energy. In this particular
circuit, the inductor is the energy-storing component, and it is able to produce a voltage drop across
the resistor with zero input voltage in its “discharging” mode.
An illustration helps make this clear. If the “pole” condition is such that Vin (s) = 0, we may
show this by short-circuiting the input of our filter circuit to ensure a zero-input condition:

Acting as an
VL electrical source

shorting
wire
Vin = 0
Acting as an
VR electrical load Vout

Assuming the inductor has been “charged” with energy previous to the short-circuiting of the
input, an output voltage will surely develop across the resistor as the inductor discharges. In other
words, the inductor behaves as an electrical source while the resistor behaves as an electrical load :
connected in series they must of course share the same current, but their respective voltages are equal
in magnitude and opposing in polarity in accordance with Kirchhoff’s Voltage Law. Furthermore,
the value of s in this “pole” condition tells us exactly how rapidly the output signal will decay: it
will do so at a rate σ = −R/L. Recall that the growth/decay term of the s variable the reciprocal of
the system’s time constant (σ = 1/τ ). Therefore, a σ value of R/L is equivalent to a time constant
of L/R, which as all beginning students of electronics learn is how we calculate the time constant
for a simple inductor-resistor circuit.
5.7. TRANSFER FUNCTION ANALYSIS 409

Transfer functions are easier to understand when graphically plotted as three-dimensional

surfaces: the real and imaginary portions of the s variable occupying the horizontal axes, and
the magnitude of the transfer function fraction displayed as height. Here is a pole-zero plot of this
low-pass filter circuit’s transfer function, with a resistor value of R = 5 Ω and an inductor value of
L = 10 H:

R 5
R+sL
= 5+s10

−σ +jω

−jω +σ

This surface plot makes the meaning of the term “pole” quite obvious: the shape of the function
looks just like a rubber mat stretched up at one point by a physical pole. Here, the “pole” rises to
an infinite29 height at a value of s where σ = −0.5 time constants per second and ω = 0 radians
per second. The surface is seen to decrease in height at all edges of the plot, as σ and ω increase in
value.
The “zero” of this transfer function is not as obvious as the pole, since the function’s value does
not equal zero unless and until s becomes infinite, which of course cannot be plotted on any finite
domain. Suffice to say that the zero of this transfer function lies in all horizontal directions at an
infinite distance away from the plot’s origin (center), explaining why the surface slopes down to zero
everywhere with increasing distance from the pole.

29 Of course, the mathematical plotting software cannot show a pole of truly infinite height, and so the pole has

been truncated. This is why it appears to have a “flat” top.

410 CHAPTER 5. AC ELECTRICITY

One of the valuable insights provided by a three-dimensional pole-zero plot is the system’s
response to an input signal of constant magnitude and varying frequency. This is commonly referred
to as the frequency response of the system, its graphical representation called a Bode plot. We
may trace the Bode plot for this system by revealing a cross-sectional slice of the three-dimensional
surface along the plane where σ = 0 (i.e. showing how the system responds to sinusoidal waves of
varying frequency that don’t grow or decay over time):

R 5
R+sL
= 5+s10

−σ +jω

−jω +σ

Only one-half of the pole-zero surface has been plotted here, in order to better reveal the cross-
section along the jω axis. The bold, red curve traces the edge of the transfer function surface as
it begins at zero frequency (DC) to increasingly positive values of jω. The red trace is therefore
the Bode plot for this low-pass filter, starting at a maximum value of 1 (Vout = Vin for a DC input
signal) and approaching zero as frequency increases.
5.7. TRANSFER FUNCTION ANALYSIS 411

As insightful as three-dimensional pole-zero plots are, they are laborious to plot by hand, and
even with the aid of a computer may require significant30 time to set up. For this reason, pole-zero
plots have traditionally been drawn in a two-dimensional rather than three-dimensional format, from
a “bird’s eye” view looking down at the s plane. Since this view hides any features of height, poles
and zeros are instead located on the s plane by × and ◦ symbols, respectively. An example of a
traditional pole-zero plot for our low-pass filter appears here:

+jω

-σ +σ

Pole at s = (-0.5 + j0)

-jω

Admittedly, this type of pole-zero plot is a lot less interesting to look at than a three-dimensional
surface plotted by computer, but nevertheless contains useful information about the system. The
single pole lying on the real (σ) axis tells us the system will not self-oscillate (i.e. ω = 0 at the
pole), and that it is inherently stable: when subjected to a pulse, its natural tendency is to decay
to a stable value over time (i.e. σ < 0).

It should be noted that transfer functions and pole-zero plots apply to much more than just
filter circuits. In fact, any physical system having the same “low-pass” characteristic as this filter
circuit is describable by the same transfer function and the same pole-zero plots. Electric circuits
just happen to be convenient applications because their individual component characteristics are so
easy to represent as functions of s. However, if we are able to characterize the components of a
different physical system in the same terms31 , the same mathematical tools apply.
30 My first pole-zero plot using the ePiX C++ mathematical visualization library took several hours to get it just

right. Subsequent plots went a lot faster, of course, but they still require substantial amounts of time to adjust for a
useful and aesthetically pleasing appearance.
31 A powerful mathematical technique known as a Laplace Transform does this very thing: translate any differential

equation describing a physical system into functions of s, which may then be analyzed in terms of transfer functions
and pole-zero plots.
412 CHAPTER 5. AC ELECTRICITY

5.7.2 Example: RC high-pass filter circuit

For our next exploratory example we will consider another simple filter circuit, this time comprised
of a capacitor and a resistor, with the output signal taken across the resistor. As before, we may
derive a transfer function by expressing Vout /Vin as a ratio of the resistor’s impedance to the total
series resistor-capacitor impedance (treating this as a voltage divider circuit):

C Z(s) = 1 / sC

Vin

R Z(s) = R Vout

Vout (s) R
Transfer function = = 1
Vin (s) R + sC
After writing this initial transfer function based on component impedances, we will algebraically
manipulate it to eliminate compound fractions. This will aid our analysis of the circuit’s DC response,
zeros, and poles:
R
1
R + sC
R
sRC 1
sC + sC

R
1+sRC
sC

sRC
1 + sRC
This transfer function allows us to calculate the “gain” of the system for any given value of s,
which brings us to the next step of our analysis. Once again we will ask ourselves three questions
about the transfer function:

1. How does this system respond when s = 0?

2. What value(s) of s make the transfer function approach a value of zero?

3. What value(s) of s make the transfer function approach a value of infinity?

5.7. TRANSFER FUNCTION ANALYSIS 413

In answer to the first question, we see that the transfer function is equal to zero when s = 0:
0RC
1 + 0RC
0 0
= =0
1+0 1
Of course, a value of 0 for s means exposure to a steady DC signal: one that neither grows nor
decays over time, nor oscillates. Therefore, this resistor-capacitor circuit will output zero voltage
when exposed to a purely DC input signal. This makes conceptual sense when we examine the circuit
itself: a DC input signal voltage means the capacitor will not experience any change in voltage over
time, which means it will not pass any current along to the resistor. With no current through
the resistor, there will be no output voltage. Thus, the capacitor “blocks” the DC input signal,
preventing it from reaching the output. This behavior is exactly what we would expect from such
a circuit, which any student of electronics should immediately recognize as being a simple high-pass
filter: DC is a condition of zero frequency, which should be completely blocked by any filter circuit
with a high-pass characteristic.

The answer to our first question is also the answer to the second question: “what value of s
makes the transfer function equal to zero?” Here we see that it is only at a value of s = 0 that the
entire transfer function’s value will be zero. Any other values for s – even infinite – yield non-zero
results. In contrast to the last circuit (the resistor-inductor low-pass filter) this circuit exhibits a
singular “zero” point in its transfer function: one specific location on the pole-zero plot where the
function’s value diminishes to nothing.

When we consider the third question (“What value(s) of s make the transfer function approach
a value of infinity?”) we proceed the same as before: by finding value(s) of s which will make the
denominator of the transfer function fraction equal to zero. If we set the denominator portion equal
to zero and solve for s, we will obtain the pole for the circuit:

1 + sRC = 0

sRC = −1

1
s=−
RC
We know that both R and C are real numbers, not imaginary. This tells us that s will likewise
be a real number at the pole. That is to say, s will be comprised of all σ and no ω. The fact that
the value of σ is negative tells us the pole represents a condition of exponential decay, just the same
as in the case of the resistor-inductor low-pass filter. As before, this means the circuit will produce
an output voltage signal with no32 input voltage signal when the rate of signal decay is σ = −1/RC.
Recall that the rate of decay in the s variable (σ) is nothing more than the reciprocal of the
system’s time constant (τ ). Thus, a rate of decay equal to 1/RC equates to a time constant τ = RC,
32 As before, this counter-intuitive condition is possible only because the capacitor in this circuit has the ability to

store energy. If the capacitor is charged by some previous input signal event and then allowed to discharge through
the resistor, it becomes possible for this circuit to develop an output voltage even with short-circuited input terminals.
414 CHAPTER 5. AC ELECTRICITY

which as all electronics students know is how we calculate the time constant for any simple resistor-
capacitor circuit.

Using a computer to plot a three-dimensional representation of this transfer function, we clearly

see both the pole and the zero as singularities. Here I have assumed a 10 Ω resistor and a 0.2 F
capacitor to place the pole at the same location as with the low-pass filter circuit s = −0.5 + j0, for
an equitable comparison:

sRC s(10)(0.2)
1+sRC
= 1+s(10)(0.2)

−σ +jω

−jω +σ

Here we see both the pole at s = −0.5 + j0 and the zero at s = 0 + j0 quite clearly: the pole
is a singular point of infinite height while the zero is a singular point of zero height. The three-
dimensional surface of the transfer function looks like a rubber sheet that has been stretched to an
infinite height at the pole and stretched to ground level at the zero.
5.7. TRANSFER FUNCTION ANALYSIS 415

As in the last example, we may re-plot the transfer function in a way that shows a cross-sectional
view at σ = 0 in order to reveal the frequency response of this high-pass filter circuit:

sRC s(10)(0.2)
1+sRC
= 1+s(10)(0.2)

−σ
+jω

−jω +σ

Once again the bold, red curve traces the edge of the transfer function surface as it begins at zero
frequency (DC) to increasingly positive values of jω. The red trace is therefore the Bode plot for this
high-pass filter, starting at a minimum value of 0 (Vout = 0 for a DC input signal) and approaching
unity (1) as frequency increases. Naturally, this is the type of response we would expect to see
exhibited by a high-pass filter circuit.
416 CHAPTER 5. AC ELECTRICITY

A more traditional two-dimensional pole-zero plot for this circuit locates the pole with a “×”
symbol and the zero with a “◦” symbol:

+jω

Zero at s = (0 + j0)

-σ +σ

Pole at s = (-0.5 + j0)

-jω
5.7. TRANSFER FUNCTION ANALYSIS 417

5.7.3 Example: LC “tank” circuit

Next, we will explore the transfer function for a tank circuit, comprised of a capacitor and an
inductor. We will assume the use of pure reactances here with no electrical resistance or other
energy losses of any kind, just to analyze an ideal case. The output voltage in this particular circuit
will be taken across the inductor:

C Z(s) = 1 / sC

Vin

L Z(s) = sL Vout

Writing the transfer function for this tank circuit is (once again) a matter of expressing the ratio
between the output component’s impedance versus the total circuit impedance:

Vout (s) sL
Transfer function = = 1
Vin (s) sL + sC
Algebraically manipulating this function to eliminate compound fractions:
sL
1
sL + sC
sL
sLsC 1
sC + sC

sL
s2 LC 1
sC + sC

sL
s2 LC+1
sC

sLsC
s2 LC + 1

s2 LC
s2 LC + 1
Note how this transfer function contains s2 terms rather than s terms. This makes it a second-
order function, which will yield very different results on the pole-zero plot than what we saw with
either the resistor-inductor or resistor-capacitor filter circuits.
418 CHAPTER 5. AC ELECTRICITY

Asking ourselves the same three questions again:

1. How does this system respond when s = 0?

2. What value(s) of s make the transfer function approach a value of zero?

3. What value(s) of s make the transfer function approach a value of infinity?

In answer to the first question, we see that the transfer function is equal to zero when s = 0:

s2 LC
s2 LC + 1
0 0
= =0
0+1 1
As with the RC low-pass filter, its response at DC also happens to be a “zero” for the transfer
function. With a DC input signal, the output signal of this circuit will be zero volts.

In order to find poles for this transfer function, we must solve for values of s that will make the
denominator term of the transfer function equal to zero:

s2 LC + 1 = 0

s2 LC = −1

1
s2 = −
LC
r
1
s= −
LC
r
1
s = ±j
LC
Given the fact that both L and C are real, positive numbers, and therefore solving for s requires
we take the square root of a negative real number, we see that the value of s mustq
be imaginary. We
1
also see here that there are two poles in this transfer function: one at s = 0 + j LC and another
q
1
at s = 0 − j LC .
5.7. TRANSFER FUNCTION ANALYSIS 419

Using a computer to plot a three-dimensional representation of this transfer function, we clearly

see a single zero at s = 0 and two poles symmetrically positioned along the jω axis. Here I have
assumed a 0.2 F capacitor and a 5 H inductor for component values:

s2 LC
s2 LC+1

−σ +jω

−jω +σ

The two33 poles located on the jω axis (one at s = 0 + j1 and the other at s = 0 − j1) tell us
the circuit is able to generate an oscillating output signal (ω = 1 radian per second frequency) at
constant magnitude (σ = 0) with no input signal. This is only possible because we have assumed a
perfect capacitor and a perfect inductor with no energy losses whatsoever. If we charge up either
or both of these components and then immediately short-circuit the input of the circuit to ensure
Vin = 0, it will oscillate at its resonant frequency forever.

33 The two solutions for ω (one at +1 radian per second and the other at −1 radian per second) merely indicate the

circuit is able to oscillate “forward” as well as “backward”. In other words, it is able to oscillate sinusoidally where
the positive peak occurs at time t = 0 (+1 rad/sec) as well as oscillate sinusoidally where the negative peak occurs
at time t = 0 (-1 rad/sec). We will find that solutions for s in general are symmetrical about the real axis, meaning
if there is any solution for s requiring an imaginary number value, there will be two of them: one with a positive
imaginary value and the other with a negative imaginary value.
420 CHAPTER 5. AC ELECTRICITY
q q
1 1
Earlier we noted that the poles in this circuit were s = 0 + j LC and s = 0 − j LC . In other
q
1
words, its resonant frequency is ω = LC . Recalling that the definition for ω is radians of phasor
rotation per second, and that there are 2π radians in one complete revolution (cycle), we can derive
the familiar resonant frequency formula for a simple LC circuit:
r
1
ω=
LC

. . . substituting 2πf for ω . . .

r
1
2πf =
LC
1
f= √
2π LC
Taking a cross-section of this surface plot at σ = 0 to obtain a frequency response (Bode plot) of
the LC tank circuit, we see the output of this circuit begin at zero when the frequency (ω) is zero,
then the output peaks at the resonant frequency (ω = 1 rad/sec), then the output approaches unity
(1) as frequency increases past resonance:

s2 LC
s2 LC+1

−σ +jω

−jω +σ
5.7. TRANSFER FUNCTION ANALYSIS 421

A more traditional two-dimensional pole-zero plot for this circuit locates shows the zero and the
two poles using “◦” and “×” symbols:

+jω

Pole at s = (0 + j1)
Zero at s = (0 + j0)
-σ +σ

Pole at s = (0 - j1)

-jω
422 CHAPTER 5. AC ELECTRICITY

5.7.4 Example: RLC band-pass filter circuit

For our next example circuit, we will add a resistor in series with the inductor and capacitor to
explore its effects on the transfer function. Taking our output voltage across the resistor, we should
expect to see band-pass filtering behavior from this circuit, with maximum voltage developing across
the resistor at one frequency where the inductor’s and capacitor’s impedances cancel:

C Z(s) = 1 / sC

L Z(s) = sL
Vin

R Z(s) = R Vout

As usual, the transfer function for this circuit is the ratio between the output component’s
impedance (R) and the total series impedance, functioning as a voltage divider:

Vout (s) R
Transfer function = = 1
Vin (s) R + sL + sC

Algebraically manipulating this function to eliminate compound fractions:

R
1
R + sL + sC

R
sRC s2 LC 1
sC + sC + sC

R
sRC+s2 LC+1
sC

sRC
sRC + s2 LC + 1
As with the pure tank circuit analyzed previously, we can see that this LC circuit exhibits a
second-order transfer function because it contains an s2 term. Running through our three questions
again:

1. How does this system respond when s = 0?

2. What value(s) of s make the transfer function approach a value of zero?

3. What value(s) of s make the transfer function approach a value of infinity?

5.7. TRANSFER FUNCTION ANALYSIS 423

The answers to the first two questions are one and the same: the numerator of the transfer
function will be zero when s = 0, this being the single zero of the function. Recalling that a
condition of s = 0 represents a DC input signal (no growth or decay over time, and no oscillation),
this makes perfect sense: the presence of the DC-blocking series capacitor in this circuit ensures the
output voltage under steady-state conditions must be zero.
In answering the third question to identify any poles for this circuit, we encounter a more
complicated mathematical problem than seen with previous example circuits. The denominator
of the transfer function’s fraction is a second-degree polynomial in the variable s. As you may
recall from your study of algebra, any solution resulting in a polynomial having an over-all value
of zero is called a root of that polynomial expression. Since this particular expression is found in
the denominator of the transfer function where we know zero values mark poles of the system, and
solutions for s resulting are roots of the polynomial, then roots of the expression sRC + s2 LC + 1
must mark the locations of the poles on the s plane.

A very useful algebraic tool for finding roots of a second-degree polynomial expression is the
quadratic formula:
√
−b ± b2 − 4ac
x=
2a
Where,
ax2 + bx + c is a polynomial expression
x is the independent variable of that polynomial expression
a is the coefficient of the second-degree (x2 ) term
b is the coefficient of the first-degree (x) term
c is the coefficient of the zero-degree (constant) term

Reviewing the denominator of our transfer function again, we see that s is the independent
variable, and so LC must be the “a” coefficient, RC must be the “b” coefficient, and 1 must be the
“c” coefficient. Substituting these variables into the quadratic formula will give us a formula for
computing the poles of this resistor-inductor-capacitor circuit:
p
−RC ± (RC)2 − 4LC
s=
2LC
Perhaps the most interesting part of this formula is what lies beneath the radicand (square-root)
symbol: (RC)2 − 4LC. This portion of the quadratic formula is called the discriminant, and its
value determines both the number of roots as well as their real or imaginary34 character. If the
discriminant is equal to zero, there will be a single real root for our polynomial and therefore only
one pole for our circuit. If the discriminant is greater than zero (i.e. a positive value), then there
will be two real roots and therefore two poles lying on the σ axis (i.e. no imaginary jω parts). If
the discriminant is less than zero (i.e. a negative value), then there will be two complex roots for
our polynomial and therefore two complex poles having both real and imaginary parts.
Let us consider for a moment what the sign of the discriminant means in practical terms. A pole
that is purely real means a value for s that is all σ and no ω: representing a condition of growth
34 The only way to obtain a purely imaginary root for this polynomial is for the “b” coefficient to be equal to zero.

For our example circuit, it means either R or C would have to be zero, which is impossible if both of those components
are present and functioning. Thus, our RLC filter circuit will have either real poles or complex poles.
424 CHAPTER 5. AC ELECTRICITY

or decay but with no oscillation. This is similar to what we saw with the LR or RC low/high pass
filter circuits, where the circuit in a state of discharge could generate an output signal even with its
input terminals shorted to ensure no input signal.
If we have a positive value for the discriminant which yields two real poles, it means two different
possible values for σ (rate of growth/decay) that could occur with no signal input to the circuit.
This behavior is only possible with two energy-storing components in the circuit: a kind of double
time-constant where different portions of the circuit discharge at different rates. The lack of any
imaginary part within s means the circuit still will not self-oscillate.
If we have a negative value for the discriminant which yields two complex poles, it means two
different values for s both having real and imaginary parts. Since the real part (σ) represents
growth/decay while the imaginary part (ω) represents oscillation, complex poles tell us the circuit
will be able to self-oscillate but not at a constant magnitude as with an ideal (lossless) tank circuit.
In fact, intuition should tell us these complex poles must have negative real values representing
decaying oscillations, because it would violate the Law of Energy Conservation for our circuit to
self-oscillate with increasing magnitude.

Looking at the discriminant (RC)2 − 4LC we see that it is possible to push the circuit into any
one of these three modes of operation merely by adjusting the value of R and leaving both L and
C unchanged. If we wish to calculate the critical value of R necessary to produce a single real pole
for any given values of L and C, we may set the discriminant equal to zero and algebraically solve
for R as follows:

(RC)2 − 4LC = 0

(RC)2 = 4LC

√
RC = 4LC

√ √
RC = 2 L C

√ √
2 L C
R=
C
√ √
2 L C
R= √ √
C C
√ r
2 L L
R= √ =2
C C
This critical value of R resulting in one real pole is the minimum amount of resistance necessary
to prevent self-oscillation. If the circuit is operating at this point, it is said to be critically damped.
Larger values of R will result in multiple real poles, where the circuit is said to be over-damped.
Smaller values of R will permit some self-oscillation to occur, and the circuit is said to be under-
damped.
5.7. TRANSFER FUNCTION ANALYSIS 425

Sometimes electrical engineers intentionally install resistors into circuits containing both
inductance and capacitance for the express purpose of damping oscillations. In such cases, the
resistor is called an anti-resonance resistor, because its purpose is to combat resonant oscillations
that would otherwise occur as the inductive and capacitive elements of the circuit exchange energy
back and forth with each other. If the engineer’s intent is to install just enough resistance into the
circuit to prevent oscillations without creating unnecessary time delays, then the best value of the
resistor will be that which causes critical damping.
Recall that the subject of transfer functions, poles, and zeros applies to any linear system, not
just AC circuits. Mechanical systems, feedback control loops, and many other physical systems may
be characterized in the same way using the same mathematical tools. This particular subject of
damping is extremely important in applications where oscillations are detrimental. Consider the
design of an automobile’s suspension system, where the complementary energy-storing phenomena
of spring tension and vehicle mass give rise to oscillations following impact with a disturbance in
the road surface. It is the job of the shock absorber to act as the “resistor” in this system and
dissipate energy in order to minimize oscillations following a bump in the road. An under-sized
shock absorber won’t do a good enough job dissipating the energy of the disturbance, and so the
vehicle’s suspension will exhibit complex poles (i.e. there will be some lingering oscillations following
a bump). An over-sized shock absorber will be too “stiff” and allow too much of the bump’s energy
to pass through to the vehicle frame and passengers. A perfectly-sized shock absorber, however, will
“critically damp” the system to completely prevent oscillation while presenting the smoothest ride
possible.
Likewise, oscillations following a disturbance are undesirable in a feedback control system where
the goal is to maintain a process variable as close to setpoint as possible. An under-damped feedback
loop will tend to oscillate excessively following a disturbance. An over-damped feedback loop won’t
oscillate, but it will take an excessive amount of time to converge back on setpoint which is also
undesirable because this means more time spent off setpoint. A critically-damped feedback loop
is the best-case compromise, where oscillations are eliminated and convergence time is held to a
minimum.

In order to fully illustrate the characteristics of this circuit’s transfer function, we will do so for
three different resistor values: one where R yields critical damping (one real pole), one where R
makes the circuit over-damped (two real poles), and one where R makes the circuit under-damped
(two complex poles). We will use three-dimensional plotting to show the transfer function response
in each case. To be consistent with our former tank circuit example, we will assume the same
capacitor value of 0.2 Farads and the same inductor value of 5 Henrys. The resistor value will be
modified in each case to create a different damping condition.
426 CHAPTER 5. AC ELECTRICITY

First, the critically-damped example, with a resistor value of 10 ohms:

sRC
sRC+s2 LC+1

L = 5 ; C = 0.2
R = 10
Critically damped

−σ +jω

−jω +σ

As expected, a single zero appears at s = 0, and a single35 pole at s = −1 + j0. Thus, this circuit
has a decay rate of -1 time constants per second (τ = 1 second) as seen when we use the quadratic
formula to solve for s:
p
−RC ± (RC)2 − 4LC
s=
2LC
p
−(10)(0.2) ± [(10)(0.2)]2 − (4)(5)(0.2)
s=
(2)(5)(0.2)
√
−2 ± 0
s= = (−1 + j0) sec−1
2
Interestingly, only L and R determine the decay rate (σ, the real part of s) in the critically
damped condition. This is clear to see if we set the discriminant to zero in the quadratic formula
and look for variables to cancel:
√
−RC ± 0 R
s= =−
2LC 2L

35 Or, one might argue there are two repeated poles, one at s = −1 + j0 and another at s = −1 − j0.
5.7. TRANSFER FUNCTION ANALYSIS 427

Next, we will plot the same transfer function with a larger resistor value (15 ohms) to ensure
over-damping:

sRC
sRC+s2 LC+1

L = 5 ; C = 0.2
R = 15
Over-damped

−σ +jω

−jω +σ

We clearly see two poles36 centered along the σ axis in this plot, representing the two real roots
of the transfer function’s denominator. Again, we will use the quadratic formula to solve for these
two values of s:
p
−RC ± (RC)2 − 4LC
s=
2LC
p
−(15)(0.2) ± [(15)(0.2)]2 − (4)(5)(0.2)
s=
(2)(5)(0.2)
p
−3 ± (3)2 − 4
s=
2
√
−3 + 5
s= = (−0.382 + j0) sec−1
2
√
−3 − 5
s= = (−2.618 + j0) sec−1
2
These real poles represent two different decay rates (time constants) for the over-damped circuit:
a fast decay rate of σ = −2.618 sec−1 and a slow decay rate of σ = −0.382 sec−1 , the slower of these
two decay rates dominating the circuit’s transient response over long periods of time.

36 The center of the pole farthest from the plot’s origin actually lies outside the plotted area, which is why that pole

appears to be vertically sliced. This plot’s domain was limited to the same values (±2) as previous plots for the sake
of visual continuity, the compromise here being an incomplete mapping of one pole.
428 CHAPTER 5. AC ELECTRICITY

Next, we will plot the same transfer function with a smaller resistor value (5 ohms) to ensure
under-damping:

sRC
sRC+s2 LC+1

L=5
C = 0.2
R=5
Under-damped
−σ +jω

−jω +σ

We clearly see two poles once again, but neither of them are located on an axis. These represent
two complex values for s describing the circuit’s behavior with zero input. The imaginary (jω) part
of s tells us the circuit has the ability to self-oscillate. The negative, real (σ) part of s tells us these
oscillations decrease in magnitude over time. Using the quadratic formula to solve for these two
poles:
p
−RC ± (RC)2 − 4LC
s=
2LC
p
−(5)(0.2) ± [(5)(0.2)]2 − (4)(5)(0.2)
s=
(2)(5)(0.2)
p
−1 ± (1)2 − 4
s=
2
√
−1 + −3
s= = (−0.5 + j0.866) sec−1
2
√
−1 − −3
s= = (−0.5 − j0.866) sec−1
2
The calculated ω value of 0.866 radians per second is slower than the 1 radian per second resonant
frequency calculated for the pure tank circuit having the same L and C values, revealing that the
damping resistor skews the “center” frequency of this RLC band-pass filter. The calculated σ value
of −0.5 time constants per second (equivalent to a time constant of τ = 2 seconds) describes the
5.7. TRANSFER FUNCTION ANALYSIS 429

rate at which the sinusoidal oscillations decay in magnitude. Here as well we see the under-damped
decay rate (σ = −0.5 sec−1 ) is slower than the critically damped decay rate (σ = −1 sec−1 ).

If we compare two-dimensional pole-zero plots for each of the three resistor values in this RLC
circuit, we may contrast the over-damped, critically-damped, and under-damped responses:

+jω

Pole at s = (-0.382 + j0)

Zero at s = (0 + j0)
-σ +σ Over-damped (R = 15 Ω)

Pole at s = (-2.618 + j0)

-jω

+jω

Pole at s = (-1 + j0)

Zero at s = (0 + j0)
-σ +σ Critically damped (R = 10 Ω)

-jω

+jω

Pole at s = (-0.5 + j0.866)

Zero at s = (0 + j0)
-σ +σ Under-damped (R = 5 Ω)

Pole at s = (-0.5 - j0.866)

-jω
430 CHAPTER 5. AC ELECTRICITY

If we take cross-sectional plots of the transfer function at σ = 0 to show the frequency response
of this RLC band-pass filter, we see the response become “sharper” (more selective) as the resistor’s
value decreases and the poles move closer to the jω axis. Electronic technicians relate this to
the quality factor or Q of the band-pass filter circuit, the circuit exhibiting a higher “quality” of
band-pass selection as the ratio of reactance to resistance increases:

R = 15 R = 10 R=8
Over-damped Critically damped Under-damped

−σ +jω −σ +jω −σ +jω

−jω +σ −jω +σ −jω +σ

R=7 R=5 R=4

Under-damped Under-damped Under-damped

−σ +jω −σ +jω −σ +jω

−jω +σ −jω +σ −jω +σ

R=3 R=2 R=1

Under-damped Under-damped Under-damped

−σ +jω −σ +jω −σ +jω

−jω +σ −jω +σ −jω +σ

Although each and every pole in a pole-zero plot is has the same (infinite) height, poles grow
narrower when moved farther away from each other, and wider when closely spaced. As the resistance
in this circuit decreases and the poles move farther away from each other and closer to the jω axis,
their widths narrow and the frequency response curve’s peak becomes narrower and steeper.
5.7. TRANSFER FUNCTION ANALYSIS 431

5.7.5 Summary of transfer function analysis

Here is a summary of some of the major concepts important to transfer function analysis:

• The s variable is an expression of growing/decaying sinusoidal waves, comprised of a real part

and an imaginary part (s = σ + jω). The real part of s (σ) is the growth/decay rate, telling us
how rapidly the signal grows or decays over time, with positive values of σ representing growth
and negative values of σ representing decay. This growth/decay rate is the reciprocal of time
constant (σ = 1/τ ), and is measured in reciprocal units of time (time constants per second,
or sec−1 ). The imaginary part of s (jω) represents the frequency of the sinusoidal quantity,
measured in radians per second (sec−1 ).

• An important assumption we make when analyzing any system’s transfer function(s) is that
the system is linear (i.e. its output and input magnitudes will be proportional to each other
for all conditions) and time-invariant (i.e. the essential characteristics of the system do not
change with the passage of time). If we wish to analyze a non-linear system using these tools,
we must limit ourselves to ranges of operation where the system’s response is approximately
linear, and then accept small errors between the results of our analysis and the system’s real-life
response.

• For any linear time-invariant system (an “LTI” system), s is descriptive throughout the system.
In other words, for a certain value of s describing the input to this system, that same value of
s will also describe the output of that system.

• A transfer function is an expression of a systems’ gain, measured as a ratio of output over

input. In engineering, transfer functions are typically mathematical functions of s (i.e. s is
the independent variable in the formula). When expressed in this way, the transfer function
for a system tells us how much gain the system will have for any given value of s.

• Transfer functions are useful for analyzing the behavior of electric circuits, but they are
not limited to this application. Any linear system, whether it be electrical, mechanical,
chemical, or otherwise, may be characterized by transfer functions and analyzed using the
same mathematical techniques. Thus, transfer functions and the s variable are general tools,
not limited to electric circuit analysis.

• A zero is any value of s that results in the transfer function having a value of zero (i.e. zero
gain, or no output for any magnitude of input). This tells us where the system will be least
responsive. On a three-dimensional pole-zero plot, each zero appears as a low point where
the surface touches the s plane. On a traditional two-dimensional pole-zero plot, each zero
is marked with a circle symbol (◦). We may solve for the zero(s) of a system by solving for
value(s) of s that will make the numerator of the transfer function equal to zero, since the
numerator of the transfer function represents the output term of the system.

• A pole is any value of s that results in the transfer function having an infinite value (i.e.
maximum gain, yielding an output without any input). This tells us what the system is capable
432 CHAPTER 5. AC ELECTRICITY

of doing when it is not being “driven” by any input stimulus. Poles are typically associated
with energy-storing elements in a passive system, because the only way an unpowered system
could possibly generate an output with zero input is if there are energy-storing elements within
that system discharging themselves to the output. On a three-dimensional pole-zero plot, each
pole appears as a vertical spike on the surface reaching to infinity. On a traditional two-
dimensional pole-zero plot, each pole is marked with a (×) symbol. We may solve for the
pole(s) of a system by solving for value(s) of s that will make the denominator of the transfer
function equal to zero, since the denominator of the transfer function represents the input term
of the system.

• Second-order systems are capable of self-oscillation. This is revealed by poles having

imaginary values. These oscillations may be completely undamped (i.e. s is entirely imaginary,
with σ = 0), in which case the system is able to oscillate forever on its own. If energy-dissipating
elements are present in a second-order system, the oscillations will be damped (i.e. decay in
magnitude over time).

• An under-damped system exhibits complex poles, with s having both imaginary (jω)
frequency values and real (σ) decay values. This means the system can self-oscillate, but
only with decreasing magnitude over time.

• A critically damped system is one having just enough dissipative behavior to completely
prevent self-oscillation, exhibiting a single pole having only a real (σ) value and no imaginary
(jω) value.

• An over-damped system is one having excessive dissipation, exhibiting multiple real poles.
Each of these real poles represents a different decay rate (σ) or time constant (τ = 1/σ) in the
system. When these decay rates differ in value substantially from one another, the slowest one
will dominate the behavior of the system over long periods of time.
5.8. POLYPHASE AC POWER 433

5.8 Polyphase AC power

“Polyphase” means “many phases,” describing a form of AC electrical system where multiple
sinusoidal voltages exist that are not in step with each other. The most common form of polyphase
AC power in industry is three-phase, but all polyphase systems share similar traits. A good way
to understand three-phase AC systems is to begin with an understanding of simpler, single-phase
systems.

A simple alternator (AC generator) is nothing more than a magnetized rotor spinning between
a pair of electromagnetic poles, the stationary wire coils (“stator windings”) developing AC voltage
as the spinning rotor’s magnet passes by:

Mechanical
diagram
Schematic
iron diagram

60 V
N S 120 V
60 V

Note that the stator is comprised of two windings connected in series, so that their respective AC
voltages directly add. If each winding of this machine develops 60 volts, the series pair will develop
120 volts. This machine is properly called a single-phase alternator, because all its stator winding
voltages are in-phase with each other.
434 CHAPTER 5. AC ELECTRICITY

A much more common alternator design uses three sets of stator poles, each one with its own
winding pair, to generate three AC voltages phase-shifted from one another by 120o :

Mechanical Schematic
diagram diagram
A 60 V B
A
60 V
60 V
B C

60 V Neutral
N S 208 V
60 V
C B

60 V
A

C
Neutral

Note that each pair of stator winding voltages directly add, because they are in phase with
each other. In the example shown, each individual stator winding develops 60 volts, with each
series pair (each “phase” of the alternator) developing 120 volts. However, the voltage appearing
between different stator winding pairs is neither the simple sum (120 + 120) nor the simple difference
(120 − 120) of each phase voltage. Rather, the phase-to-phase voltage is the trigonometric sum of
two phasor quantities, spaced 120o apart. In the example shown, 1206 0o + 1206 120o = 207.856 60o ,
which is approximately 208 volts. This machine is properly called a three-phase alternator. More
specifically, this alternator is one with a wye-connected stator winding set, because the geometric
configuration of the stator windings resembles that of the letter “Y”.
5.8. POLYPHASE AC POWER 435

The following oscilloscope screenshot shows the output of a three-phase alternator, each channel
of the oscilloscope connected across one phase of the alternator (e.g. Channel 1 across “A” and
Neutral, Channel 2 across “B” and Neutral, and Channel 3 across “C” and Neutral):

In this oscillograph image we can clearly see a 120 phase shift between successive phase voltage
waveforms. From this image we may also discern the phase sequence or phase rotation of the system:
the sequential order in which the three phases reach their respective peak values. Phase seqeunce is
determined by the direction of the alternator shaft’s rotation as well as the orientation of the stator
phase windings. Surveying the three sine waves from left to right (the forward direction of time) on
the oscillograph, we see channel 1 (Yellow) reaches its positive peak 120 degrees before channel 2
(Blue), which reaches its positive peak 120 degrees before channel 3 (Magenta). A common method
of describing phase rotation is to list the phase letter labels in their order of sequence over time, in
this case the triad ABC. It should be noted that a phase sequence of ABC is synonymous with BCA
and also with CAB. This is easy to see if we write the letters in sequence for several rotations of the
alternator, seeing that all three triads may be found within the longer sequence: ABCABCABCABC.
436 CHAPTER 5. AC ELECTRICITY

This trigonometric relationship between voltages in this “wye” connected alternator is clearly
shown in the following phasor diagram. Solid phasors express the phase voltage for each of the three
winding pairs in a three-phase alternator, while dashed lines express the line voltage between any
two of the three output terminals on the alternator. The direction of each phasors expresses its
phase shift in time, as the sine wave voltages produced by each of the three phase windings will by
shifted apart from each other by 120 degrees. Here, each phase winding in the alternator happens
to produce 120 VAC, resulting in a line voltage equal to the trigonometric sum of the 120 VAC
phasors, 208 VAC:

Voltage phasor diagram for a

wye-connected alternator
C

V
CA
=2
VC

08
=1

V
20
V
VBC = 208 V

o
0
12

VA = 120 V
120o A
12
0
o
V
20
=1

V
08
=2
B
V

V AB
B

A simple way to calculate the side lengths of these non-right triangles is to use the Law of Sines,
which simply states the ratio between the length of a triangle’s side and the sine of its opposite angle
is constant, for any triangle:

sin a sin b sin c

= =
A B C
Applying this to the solution of the voltage between phases A and B (120 volts each, individually):

sin 120o sin 30o sin 30o

= =
VAB VA VB

sin 120o sin 30o

=
VAB 120 volts

sin 120o
 
VAB = (120 volts)
sin 30o

VAB = 207.84 volts ≈ 208 volts

5.8. POLYPHASE AC POWER 437

120o
√
The ratio sin
sin 30o is equal to the square-root of three ( 3), and this factor frequently appears in
three-phase electrical system calculations.

Some three-phase alternators have their phase windings connected differently, in a “delta”
configuration rather than a “wye” configuration:

Mechanical Schematic
diagram diagram

A 60 V 60 V

B C
60 V 60 V

N S

C B
60 V 60 V
A 120 V

The phasor diagram for voltage in a delta-connected alternator is simpler, because phase voltage
is exactly equal to line voltage, with each phase coil directly connected to a pair of line terminals:

Voltage phasor diagram for a

delta-connected alternator

VCA = 120 V
C A
VA = 120 V
V

V
V BC

VC

20

20
=1

=1
=1

=1
20

B
V
V
20

AB
V
V

B
438 CHAPTER 5. AC ELECTRICITY

The following photograph shows the terminal connections on the rear side of a three-phase
alternator intended to provide electrical power on a heavy-duty truck or boat. Note the three power
terminals at the bottom of the photograph (with yellow-colored wire connectors attached), where
the three-phase AC power is output. Also note the two copper “slip rings” I am pointing to with
my finger, where DC power is conducted through stationary carbon “brushes,” through the copper
rings on the shaft, to a winding on the spinning rotor to control its magnetization. The shaft of this
alternator, normally coupled to the crankshaft of the vehicle’s engine by a V-belt, is on the far side
of the alternator hidden from view of the camera:

Larger alternator units, such as those found in power plants, do not differ substantially in design
from this small unit. Sets of stator windings around the circumference of the machine connected
in either a wye or a delta configuration generate three-phase AC power, while a magnetized rotor
spins at the center of the machine providing the changing magnetic field necessary to induce voltage
in those stator windings. Permanent-magnet rotors are seldom used because they offer no way to
control or regulate the machine’s output voltage during operation. Instead, the rotor has a single
winding of its own, energized by DC supplied externally by a voltage regulator circuit. If more AC
voltage is desired from the alternator, the regulator circuit sends more direct current to the rotor in
order to strengthen its magnetic field. If less AC voltage is desired, the regulator sends less current
to the rotor’s winding in order to weaken its magnetic field.
5.8. POLYPHASE AC POWER 439

One of the advantages granted by three-phase power is a more constant delivery of electrical
energy to the load over time. With three sets of stator windings at work, the combined effect is not
unlike a triple bicycle with the three riders’ legs staggered by 120o of rotation, or of a multi-cylinder
automobile engine with the pistons staggered apart from each other: at any given time, one of the
phases will be at or near its peak. Single-phase AC systems, by contrast, pulsate to a much greater
degree.
In a single-phase AC circuit, energy transfer actually stops completely twice per cycle, when the
current waveform passes through zero. This never happens in a polyphase system, because there
are always other phases at non-zero current values when any one phase is at its zero-crossing point,
owing to the fact that the phases in a polyphase power system are shifted from one another. This
fact results in more efficient transfer of energy in AC power systems: a three-phase power system
can actually transfer the same amount of power as a comparable single-phase power system using
less metal in the power line conductors, despite the fact that a greater number of conductors is
necessary (3 versus 2).
Another advantage of three-phase AC power is in applications where the AC is to be rectified
into DC. The rectified output of a three-phase alternator is “smoother” than the rectified output
of a single-phase alternator, with less ripple voltage to interfere with on-board electronic devices
such as radios, because the phase-shifted currents overlap each other. This is why all automotive
alternators are three-phase rather than single-phase machines.
440 CHAPTER 5. AC ELECTRICITY

A comparison of single-phase rectification versus three-phase rectification shows this clearly:

Single-phase alternator Single-phase rectifier

Rectified DC

Single-phase AC

Three-phase alternator

Three-phase rectifier

A
B

Rectified DC

Three-phase AC (ABC phase sequence)

Not only is the ripple of the rectified DC voltage less in a three-phase system than in a single-
phase system, but the frequency of that ripple is three times as great, making it easier to filter37
out of the DC power.
37 Low-pass filter circuits are typically used to “smooth” the ripple from the output of a rectifier. The greater the

frequency of this ripple voltage, the easier it is to filter from the DC (which has a frequency of zero). All other
factors being equal, a low-pass filter attenuates higher-frequency components to a greater extent than lower-frequency
components.
5.8. POLYPHASE AC POWER 441

5.8.1 Delta and Wye configurations

Two basic forms of three-phase sources and loads appearing in industrial power systems are the delta
(∆) and the wye (or simply Y ) configurations. A “wye-connected” device has its three elements
joined at one common point in the middle as such:

Wye-connected Wye-connected
motor or generator resistive load

By contrast, a “delta-connected” device has its three elements joined as sides to a triangle:

Delta-connected Delta-connected
motor or generator resistive load

Each configuration has its own unique advantages and disadvantages in the larger context of a
three-phase electrical power system. Either source type may connect to either load type (e.g. delta
to wye, delta to delta, wye to delta, wye to wye) so long as the voltage and current ratings of all
components are compatible.
442 CHAPTER 5. AC ELECTRICITY

The voltage appearing across the terminals of each element in a polyphase device is called the
phase voltage, and the current through each element in a polyphase device is called the phase current:

Iphase

Iphase
Vphase
Vphase

Voltage appearing between any two of the connecting conductors (power lines) is called the line
voltage of the polyphase system, and current through any of the connecting conductors (power lines)
is called the line current:

Iline

Vline

Vline

Iline

Line and phase quantities relate to each other differently between delta devices
√ and wye devices.
Line voltage for a balanced38 wye device exceeds phase voltage by a factor of 3, while line current
for a balanced delta device exceeds phase current by the same factor:

System type Voltage

√ Current
Wye (Y) Vline = 3 × Vphase Iline √
= Iphase
Delta (∆) Vline = Vphase Iline = 3 × Iphase

38 Here, the term “balanced” refers to a condition where all phase voltages and currents are symmetrically equal.

Unbalanced conditions can and do exist in real polyphase power systems, but the degree of imbalance is usually quite
small except in cases of component faults.
5.8. POLYPHASE AC POWER 443

While it may be tempting to simply memorize these mathematical relationships, it is far better
to understand why they are the way they are. In a wye-connected device, it should be clear upon
inspection that line current must be equal to phase current, because each power line is in series with
each (respective) phase element, and we know from our understanding of basic electrical circuits
that series-connected elements necessarily share the same current. Likewise, it should be clear that
line voltage must be equal to phase voltage in a delta-connected device, because each power line pair
connects in parallel fashion to each (respective) phase element, and we know that parallel-connected
elements necessarily share the same voltage:

Iline = Iphase Vline = Vphase

Iline
Iphase

Vline
Vphase

Phase and line voltages are unequal in wye-connected devices, as are phase and line currents in
delta-connected devices. In each of these cases, though, we may see once again by visual inspection
that these line and phase quantities cannot be equal because the line quantities are the result of
two joining phase quantities. In a wye network, line voltage is the series (phasor) sum of two phase
voltages. In a delta network, line current is the parallel (phasor) sum of two currents summing
at a node. If we know the system in question is balanced, however, we may be assured that√the
multiplying factor between
√ these line and phase quantities will be the square-root of three √ ( 3),
therefore line voltage is 3 times greater than wye phase voltage, and line current is 3 times
greater than delta phase current:

Vline = 3 Vphase Iline = 3 Iphase

Iline
Iphase

Vphase Vline
Iphase
Vphase
444 CHAPTER 5. AC ELECTRICITY

As an example of phase and line voltages in a wye-connected system, we see a great many three-
phase industrial power circuits in the United States of the “480/277” volt configuration. These are
a wye-connected devices exhibiting having phase voltages of 277 volts (each) and a balanced line
voltage of 480 volts.
In the wye-connected system we see how the two phase voltages add in series (indirectly) to form
the larger line voltage. In the delta-connected system we see how the larger line current splits up in
parallel branches (indirectly) to form two smaller phase currents. The key to understanding these
mathematical relationships is to recognize where the rules of series and parallel connections dictate
the quantities be identical or different, and then all we√need to remember is that if the two are
different, the line quantity will be greater by a factor of 3.
5.8. POLYPHASE AC POWER 445

5.8.2 Power in three-phase circuits

Suppose we have a 480/277 wye-connected alternator supplying electrical power to a delta-connected
load consisting of three 200 ohm resistive heating elements:

480/277 V source

200 Ω

200 Ω 200 Ω

To begin our calculation of all electrical quantities in this circuit, we will apply Ohm’s Law to
the calculation of phase current at the load, since we already know phase voltage (480 volts) and
phase resistance (200 ohms) there:
480 V
Iphase(load) = = 2.4 A
200 Ω
Now that we know phase current at the delta-connected load, we may calculate line current for
the whole system by multiplying by the square-root of three:
√
Iline = ( 3)(2.4 A) = 4.157 A
This line current must be the same as the phase current in the wye-connected alternator, since
line and phase currents are equal in wye-connected devices by virtue of their series connection. We
already know the phase voltage of the alternator (277 volts) because that was given to us, but we
could just as well calculate it from the line voltage of 480 volts as such:
√
Vline = ( 3)(Vphase(source) )

Vline
Vphase(source) = √
3
480 V
Vphase(source) = √ = 277.1 V ≈ 277 V
3
Tabulating all results for voltage and current in our balanced system with a line voltage of 480
V and a line current of 4.157 A:

Quantity Source Load

Vphase 277 V 480 V
Iphase 4.157 A 2.4 A
446 CHAPTER 5. AC ELECTRICITY

Power for each of the three source or three load elements in this balanced system is simply the
product of phase voltage and phase current (P = IV ) because voltage and current are in-phase at
each of the individual resistors. Expanding our table to include the power for each phase element:

Quantity Source Load

Vphase 277 V 480 V
Iphase 4.157 A 2.4 A
Pphase 1152 W 1152 W

Total generated power at the alternator (as well as total dissipated power at the resistive heater)
is the simple sum of all three phase elements: 3456 watts in each case. No “square-root-of-three”
factor is required in this calculation, because power (work over time) is not a phasor quantity39 .
The Law of Energy Conservation demands that all power be accounted for, and thus three resistors
dissipating 1152 watts each must be together dissipating 3456 watts total:

480/277 V source Ptotal = 3456 W

Ptotal = 3456 W
1152 W

1152 W 1152 W

1152 W 1152 W
1152 W

In the interest of convenience, though, it is helpful to have a formula to calculate power in a

balanced three-phase system knowing just the line voltage and current rather than phase voltages
and currents. You will note how the simple product of line voltage (480 V) and line current (4.157
A) does not yield total power (3456 W) in this system. We may develop a proper formula for
calculating total power from line voltage and current by beginning with the formula described in
calculating total power from the power of each resistor at the delta-connected load:

Ptotal = (3)(Pphase )

Ptotal = (3)(Iphase )(Vphase )

39 You may recall from basic physics that while force and displacement are both vector quantities (having direction as

well as magnitude), work and energy are not. Since power is nothing more than the rate of work over time, and neither
work nor time are vector quantities, power is not a vector quantity either. This is closely analogous to voltage, current,
and power in polyphase electrical networks, where both voltage and current are phasor quantities (having phase angle
“direction” as well as magnitude) but where power merely has magnitude. We call such “directionless” quantities
scalar. Scalar arithmetic is simple, with quantities adding and subtracting directly rather than trigonometrically.
5.8. POLYPHASE AC POWER 447

We may substitute Iline for Iphase and Vline for Vphase in this equation if we properly relate them
for the delta connection. While Vline = Vphase in a delta configuration, Iphase = I√ line
3
:
 
Iline
Ptotal = (3) √ (Vline )
3
√
We may consolidate
√ √ the two constants in this formula√ (3 and 3) by re-writing the number 3 as
the product 3 3, then canceling one of these with the 3 in the denominator:
√ √
 
Iline
Ptotal = ( 3 3) √ (Vline )
3
√
Ptotal = ( 3)(Iline )(Vline )
As a test, we may check to see that this new formula accurately calculates the total power of our
balanced three-phase system:
√
Ptotal = ( 3)(4.157 A)(480 V)

Ptotal = 3456 W
448 CHAPTER 5. AC ELECTRICITY

5.8.3 Grounded three-phase circuits

So far all the three-phase configurations shown have been ungrounded : that is, none of the terminals
or conductors have any direct connection to earth ground. While it is possible (and practical in
many cases) to use polyphase power without an explicit earth ground connection, it is not always
the safest. Establishing a firm connection to earth limits the voltage which may develop between any
non-grounded (“hot”) conductor and earth ground. This is especially important in power systems
with overhead lines, where lightning strikes may dramatically elevate common-mode voltages in the
system.
In wye-connected systems, the natural point to ground is the center of the “Y”, like this:

Wye-connected
power source
A
B

4-wire, 3-phase
N power wiring

C
The three “hot” terminals of the source are typically labeled “A”, “B”, and “C”, while the
grounded point is referred to as the “Neutral” (N). The√voltage measured between any two of the
“hot” terminals (A to B, B to C, or A to C) will be 3 times more than the voltage measured
between any “hot” terminal and the neutral (A to N, B to N, or C to N). Common voltages for
4-wire wye-connected systems include 208/120 and 480/277.
5.8. POLYPHASE AC POWER 449

The existence of dual voltage levels in a center-grounded wye system enables the use of loads
with different voltage ratings. For example, in a 208/120 wye system, a three-phase motor with
windings rated for 208 volts would draw power from the three “hot” conductors directly, while light
bulbs rated for 120 volts would connect between any “hot” conductor and the neutral:

Wye-connected
power source (208V/120V)
A
B

C
Fuse Fuses

Switch 3-pole switch

208 volt
120 volt 120 volt 120 volt motor
lamp lamp lamp

A good practice in such systems is to equally spread the 120 volt loads among the three phases,
so that (ideally) the phase loading on the source will be nicely balanced when all loads are operating.
If the loading is perfectly balanced, in fact, the neutral conductor will carry no current at the point
where it connects to the center of the wye.
450 CHAPTER 5. AC ELECTRICITY

In delta-connected systems, there is no “natural” point to connect to earth ground as there is

in wye-connected systems. The most common grounding configuration for a delta-connected source
is shown here, sometimes called the “high-leg” connection. Here, one of the source’s phase coils is
center-tapped to provide a connection point for earth ground:

Delta-connected
power source
A

B
This configuration yields three different voltages available to power loads. If the phase voltage
of the delta-connected source is 240 volts, the three available voltages are:

• 240 volts (between A-B, B-C, and A-C)

• 120 volts (between B-N and C-N)

• 208 volts (between A-N)40

A disadvantage of this configuration is that the lower-voltage loads cannot be balanced among
all three phase coils of the source as they can in wye-connected systems. Any single-phase loads
(those connected between any one “hot” conductor and the neutral conductor) inevitably place
more burden on the B-C phase coil than the other two phase coils. However, this imbalance is often
negligible in industrial settings where three-phase loads are predominant and single-phase loads are
few (and low-wattage).

40 A colorful term for this odd voltage is bastard voltage.

5.8. POLYPHASE AC POWER 451

5.8.4 Symmetrical components

Balanced three-phase networks are relatively easy to calculate quantities in, as each line and phase
variable is symmetrical. For instance, if we happen to know one of the phase voltages in a balanced
Wye-connected power system component is 2400 volts, then we may safely conclude the other two
phase voltages are√2400 volts as well. The voltage between √ any two lines in this same system is
guaranteed to be 3 larger than any phase voltage: 2400 3 ≈ 4160 volts.
Calculations become much more complex, however, in unbalanced three-phase √ networks. One of
the assumptions we must discard in an unbalanced
√ network is the simple factor of 3 relating phase
and line quantities: while Vline = Vphase 3 in a balanced Wye-connected system, it is not necessarily
true in an unbalanced Wye-connected system. Compare these phasor diagrams of a balanced
versus unbalanced 2400/4160 volt Wye systems, the unbalanced system being representative of
a three-phase generator with a fault in one of the phase windings causing that phase voltage to be
substantially less than it ought to be:

Balanced Wye-connected system Unbalanced Wye-connected system

(phase B voltage 1/10 of other phases)
C C

V V
VC

VC
AC AC
=4 V BC = =4
=2

=2
16 16
0V 0V
40

40
0V

0V
VBC = 4160 V

2530
o

o
0

0
12

12
VA = 2400 V VA = 2400 V
120o A 120o A
V

12
0o
12
0

B
VAB = 2530 V
o
0V
40

V
=2

40

1 60
=2

=4
B
V

V AB
V

In fact, the existence of faults in three-phase power systems is the primary reason for considering
unbalanced systems, since the vast majority of three-phase electrical components are expressly
designed to be balanced. If power system engineers and technicians are to analyze faults, they
must have some means of quantifying unbalanced system conditions.

A breakthrough in mathematical analysis for unbalanced three-phase electrical circuits came in

1913 from a man named Charles Legeyt Fortescue, who presented his discovery in a paper entitled,
“Method of Symmetrical Co-ordinates Applied to the Solution of Polyphase Networks” after doing
research on AC induction motors operating under unbalanced conditions. His discovery condenses to
the fact that any set of phasors describing conditions of voltage or current in a three-phase network,
no matter how unbalanced and asymmetrical they may be, are mathematically equivalent to the
sum of three unique phasor sets with different rotations. Thus, it is possible to mathematically
decompose an unbalanced phasor triad into a multiple sets of balanced phasor triads, each of those
452 CHAPTER 5. AC ELECTRICITY

balanced sets being√relatively easy to analyze on its own because the simple rules of symmetrical
networks (e.g. the 3 factor between phase and line quantities) still apply.
Fortescue’s breakthrough is reminiscent of Jean Baptiste Joseph Fourier’s discovery roughly 100
years prior that any periodic waveform, no matter its shape, is mathematically equivalent to a
summation of pure sinusoidal waveforms of harmonic frequencies (called a Fourier Series). In both
cases, we see there is a mathematical equivalence between one entity that is ugly and asymmetrical,
and a set of pure and symmetrical entities that are easier to deal with on mathematical terms.

In Fortescue’s model, which is widely known under the title of symmetrical components, any set
of three phasors describing voltage or current in a three-phase system is equivalent to the summation
of three different three-phasor sets:

• One set of three phasors rotating in the normal A-B-C direction of the power system, called
the positive sequence. By convention, this sequence is designated by the number 1.

• One set of three phasors rotating in the reverse direction (A-C-B)41 , called the negative
sequence. By convention, this sequence is designated by the number 2.

• One set of three phasors all pointed in the same direction, having no sequence at all, called
the zero sequence. By convention, this sequence is designated by the number 0.

Positive sequence Negative sequence Zero sequence

(A-B-C rotation) (A-C-B rotation) (in-phase rotation)
C1 B2
o

o
0

0
12

12

B0
120o A1 120o A2 A0
C0
12

12
0

0
o

B1 C2

Fortescue’s discovery was that we may synthesize any possible three-phasor set simply by
superimposing these positive, negative, and/or zero sequence phasor sets at the appropriate
magnitudes and phase shifts. Each positive sequence, negative sequence, and zero sequence set
is perfectly balanced, although they must usually differ in magnitude from each other in order for
the summation of all three to equal the real-world unbalanced set.

41 If you are having difficulty seeing the A-B-C or A-C-B rotations of the positive and negative sequences, you may

be incorrectly visualizing them. Remember that the phasors (arrows) themselves are rotating about the center point,
and you (the observer) are stationary. If you imagine yourself standing where the tip of each “A” phasor now points,
then imagine all the phasor arrows rotating counter-clockwise, you will see each phasor tip pass by your vantage point
in the correct order.
5.8. POLYPHASE AC POWER 453

Another way of stating this is to say that the actual voltages and currents in a three-phase power
system, no matter how unbalanced they may be from each other, are really equivalent to multiple sets
of voltages and currents existing in the circuit simultaneously, each with its own rotational sequence
and perfectly balanced magnitude. Our hypothetical three-phase generator with the faulted phase
winding, therefore, is really equivalent to three healthy generators with their windings wired together
to be series-aiding: one generator spinning in the normal direction, the next generator spinning the
wrong direction, and the last generator a single-phase unit with three in-phase windings. The
combined sum of these three generators’ outputs would create the exact same patterns of voltages
and currents as the one faulted-winding generator:

Positive sequence Negative sequence Zero sequence

generator generator generator

3-phase generator 3-phase generator 1-phase generator

spinning forward spinning backward (no phase sequence)
(rotation = ABC) (rotation = CBA)

Any unbalanced phase pattern

imaginable may be produced by
combining the proper positive,
negative, and zero sequence
phasor sets from these three
generators!
454 CHAPTER 5. AC ELECTRICITY

Let us explore this in more detail by means of a concrete example. Taking the faulted-winding
generator whose phasor voltage diagram was shown earlier (phase voltages VA and VC at 2400 volts
each, but phase voltage VB at only 240 volts), our task is to see how to decompose this unbalanced
set of voltage phasors into three balanced sets of phasors (positive-, negative-, and zero-sequence):

Voltage phasor diagram for

faulted generator
C

VA = 2400 V ∠ 0o
VB = 240 V ∠ 240o
VC = 2400 V ∠ 120o

Somehow, we must determine the specific pattern of balanced positive-, negative-, and zero-
sequence phasor sets that will be equivalent to the unbalanced phasor set describing our faulted
generator. Once we determine these symmetrical components, we will then sum them together
to demonstrate how the respective phasors do indeed add up to the original (unbalanced) phasor
diagram.
5.8. POLYPHASE AC POWER 455

Thankfully, symmetrical component analysis provides us with a set of equations for calculating
the “A” phasor of each sequence42 . Please note that each variable in each of the following equations
is a phasor quantity, possessing both a magnitude and a phase angle. Also note that these equations
apply just as well to calculations of current as they do to voltage:
1
Va1 = (Va + Vb 6 + 120o + Vc 6 + 240o ) Positive-sequence phasor A
3
1
Va2 = (Va + Vb 6 + 240o + Vc 6 + 120o ) Negative-sequence phasor A
3
1
Va0 = (Va + Vb + Vc ) Zero-sequence phasor A
3
The “+120o ” and “+240o ” references deserve further explanation43 . What we are saying here
is that the calculations involve shifting the angle of the unbalanced system’s phasor by either +120
degrees or +240 degrees when calculating the phasors for the positive and negative sequences.
Performing the calculations:

Given phasors in unbalanced system:

Va = 24006 0o
Vb = 2406 240o
Vc = 24006 120o

1
Va1 = [24006 0o + 2406 (240o + 120o ) + 24006 (120o + 240o )] = 16806 0o
3
1
Va2 = [24006 0o + 2406 (240o + 240o ) + 24006 (120o + 120o )] = 7206 − 60o
3
1
Va0 =[24006 0o + 2406 240o + 24006 120o ] = 7206 60o
3
Recalling that our faulted generator is mathematically equivalent to three healthy generators with
their respective phase windings connected in series, these solutions tell us that the equivalence is
one generator spinning the correct direction at 1680 volts per phase, connected to another generator
spinning the wrong direction as well as phase-shifted −60 degrees at 720 volts per phase, connected
to a single-phase generator phase-shifted +60 degrees at 720 volts per phase. These three healthy
generators, each one producing a symmetrical (balanced) set of voltages, will together behave the
same as our one 2400 volt phase generator with the faulted “B” winding.
42 Itis good to remember that each of the symmetrical components is perfectly balanced (i.e. the “b” and “c” phasors
each have exactly the same magnitude as the “a” phasor in each sequential set), and as such each of the phasors for
each symmetrical set will have exactly the same magnitude. It is common to denote the calculated phasors simply
as V1 , V2 , and V0 rather than Va1 , Va2 , and Va0 , the “a” phasor implied as the representative of each symmetrical
component.
43 A “shorthand” notation commonly seen in symmetrical component analysis is the use of a unit phasor called

a, equal to 16 120o . Multiplying any phasor quantity by a shifts that phasor’s phase angle by +120 degrees while
leaving its magnitude unaffected. Multiplying any phasor quantity by a2 shifts that phasor’s phase angle by +240
degrees while leaving its magnitude unaffected. An example of this “a” notation is seen in the following formula for
calculating the positive sequence voltage phasor: Va1 = 31 (Va + aVb + a2 Vc )
456 CHAPTER 5. AC ELECTRICITY

Graphically, the decomposition of the original unbalanced phasor diagram into symmetrical
components is as follows:

Voltage phasor diagram for

faulted generator
C

VA = 2400 V ∠ 0o
VB = 240 V ∠ 240o

24
00
VC = 2400 V ∠ 120o

V
2400 V
A

B 240 V

Positive sequence Negative sequence Zero sequence

phasor diagram phasor diagram phasor diagram
C1
VA1 = 1680 V ∠ 0o
VB1 = 1680 V ∠ 240o
16

B2 B0 A0
VC1 = 1680 V ∠ 120o
80

C0
0V

0V
VA2 = 720 V ∠ -60 o
V

72

1680 V

72
720 V
VB2 = 720 V ∠ 60o
A1 C2
72

VC2 = 720 V ∠ 180o

0V

VA0 = 720 V ∠ 60o

V

A2
80

VB0 = 720 V ∠ 60o

16

VC0 = 720 V ∠ 60o

B1
5.8. POLYPHASE AC POWER 457

The graphical proof that these three symmetrical component phasor sets do indeed add up to be
equivalent to the original (unbalanced) phasor diagram of the faulted generator is as follows:

Summation of symmetrical phasor sets (+, -, and 0 sequences)

C0
0V

VA = VA1 + VA2 + VA0 = 1680∠0o + 720∠-60o + 720∠60o

72

C2
VB = VB1 + VB2 + VB0 = 1680∠240o + 720∠60o + 720∠60o
720 V
VC = VC1 + VC2 + VC0 = 1680∠120o + 720∠180o + 720∠60o
16

C1
80
V

1680 V A0

72
0V
B0 A1
0V

0V
A2
72

72
80

B2
0V

B1
16
72

Phasor diagram for (original) faulted generator

VA = 2400 V ∠ 0o
VB = 240 V ∠ 240o
24
00

VC = 2400 V ∠ 120o
V

2400 V
A

B 240 V
458 CHAPTER 5. AC ELECTRICITY

In a perfectly balanced three-phase power system, the only sequence in existence is the positive
sequence (1). Any and all measurements of voltage and current, therefore, represent the positive
sequence of the system. since the magnitudes of the negative- and zero-sequence components are
zero. If an imbalance occurs for any reason – from unbalanced single-phase loads on a three-phase
power system to actual faults (opens, shorts) – negative-sequence and/or zero-sequence components
arise.
While negative-sequence quantities cannot be measured directly by any voltmeter or ammeter,
zero-sequence components can. In Y-connected systems, zero-sequence current manifests itself
as current through the neutral conductor, while zero-sequence voltage manifests itself as voltage
between the center of an ungrounded Wye and ground. Zero-sequence current manifests in Delta-
connected systems as current circulating in the phase elements of the source or load.
The following illustration shows a medium-voltage (4160 volt) electric motor circuit equipped
with a protective relay to halt the motor in the event its electrical or thermal parameters indicate
the potential for damage:

"Zero sequence"
CT CT

(Bearing)
CT
4160 VAC RTD
3-phase Motor
CT

RTD

(Stator)
PT PT

Line voltage sensing Line current sensing Ground fault sensing Temperature sensing

Protective relay
Trip contacts

Control power in Analog signal output RS-485 comm

Trip coil on circuit breaker

125 VDC
CR

125 VDC

4-20 mA Modbus

Note the “zero sequence” current transformer (CT) encircling all three lines conducting power
to the motor. This large CT magnetically senses the instantaneous sum of currents into and out
of the motor, which should be zero under normal operating conditions. If, however, a ground fault
develops within the motor, the instantaneous sum of currents will become non-zero, causing the
“zero sequence CT” to register some current. If this zero-sequence current becomes large enough,
5.8. POLYPHASE AC POWER 459

the protective relay will command the contactor feeding power to the motor to trip (open) in order
to protect the motor against further damage.
This exact same principle is used in household “GFCI” (Ground Fault Current Interruptor)
receptacles and circuit breakers required by the National Electrical Code to be used in “wet” areas
of a residence such as bathrooms: a small current transformer senses the net sum of current through
the “hot” and “neutral” conductors. If this net current measurement is anything but zero, the
interruptor contacts trip, ceasing the flow of power to the load and thereby protecting the person
using that load from injury due to ground-fault current passing through their body:

Ground-Fault Current Interruptor (GFCI)

Interrupting
Hair dryer, toaster, etc.
contacts
Hot CT
120 VAC
120 VAC load
Neutral

Trip
circuit

Any current leaking from the load to earth ground

is sensed as an imbalance of current by the CT
460 CHAPTER 5. AC ELECTRICITY

5.9 Phasor analysis of transformer circuits

Three-phase AC is the dominant mode of electric power generation and distribution in modern
societies. Analyzing voltages and currents in these systems can be quite complex, but phasor
representation is a tool useful for simplifying that analysis by representing voltage and current
quantities in graphical form. This section of the book illustrates the use of phasors as an analytical
tool through examples. This section begins with single-phase phasor analysis and progresses to
three-phase phasor analysis.

Extensive use of an imaginary instrument I call the “phasometer” helps relate phasor arrow
direction to real circuit connections. For a detailed introduction to this instrument, see section 5.5.2
beginning on page 376. Suffice it to say, a “phasometer” is nothing more than a small synchronous
AC motor with an arrow-shaped rotor, which when coupled with a strobe light set to flash whenever
some reference voltage reaches its positive peak, graphically shows the relative phase of the measured
quantity with respect to that reference voltage. In other words, a phasometer/strobe combination
shows a “snapshot” view of the waveform’s position at that point in time when the reference wave it
at its positive peak. This is precisely how phasors are drawn: arrows with angles displaced from 0o
by an amount proportional to the phase shift between that quantity and some reference waveform.
5.9. PHASOR ANALYSIS OF TRANSFORMER CIRCUITS 461

Here is what a triplet of phasometers would indicate if connected to a three-phase power bus,
with phase A used as the reference and given an ABC phase sequence (also called “phase rotation”):

A B C
VA θ = 0o
105o 90o 75o
120o 60o

135o 45o

150o 30o

165o 15o
o
180
0o

-165o -15o

-150o -30o

-135o -45o
o o
-120 -60
-105o -90o -75o

VB θ = -120o
90o

135o
120o
105o 75o
60o

45o
Three-phase power bus
150 o
30o
(ABC phase sequence)
165o 15o

180o
0o

-165o -15o

Strobe light -150o

-135o -45o
-30o

-120o -60o
-105o -90o -75o

VC θ = +120o
105o 90o 75o
120o 60o

135o 45o

150o 30o

165o 15o

180o
0o

-165o -15o

-150o -30o

-135o -45o

-120o -60o
-105o -90o -75o

In an ABC phase sequence, phase B’s voltage lags behind phase A’s voltage by 120o , while phase
C’s voltage leads ahead of phase A’s voltage by 120o (or, one could say phase C lags a full 240o
behind phase A).
462 CHAPTER 5. AC ELECTRICITY

5.9.1 Phasors in single-phase transformer circuits

Phasors are to AC circuit quantities as polarity is to DC circuit quantities, and for this reason they
are particularly useful in analyzing transformer circuits because transformers by their very nature
allow polarities to be mixed and matched.
In the “Basic principles” discussion of the “Transformers” section of this book (subsection 5.4.1
beginning on page 358), we used standard DC polarity marks (+ and −) to denote the polarity
of voltage at an instant in time where the sinusoidal waves for each winding reached their peaks.
Although it is often important to show the polarity of transformer windings relative to each other,
the use of DC notation for this purpose can be misleading. An alternative symbology places a
dot, or square, or some other distinguishing mark at one end of each transformer winding to show
common polarity at any instant in time. The marked terminal of a transformer winding is called
the “polarity” terminal, while the unmarked terminal is called the “nonpolarity” terminal:

Transformer polarity symbols

Dot notation Square notation "X" notation Number notation

H1 X1

H2 X2

The marks should be interpreted in terms of voltage polarity, not current. To illustrate using a
“test circuit44 ” feeding a momentary pulse of DC to a transformer:

Pushbutton momentarily pressed

Source Load Source Load

If the battery polarity were reversed, the “polarity” (dot) terminal of each winding would be
negative with respect to the “nonpolarity” terminal at the moment in time when the switch is
pressed. That is to say, the “polarity” marks merely show which terminals will share the same
instantaneous voltage polarity at any given point in time – not which terminal is always positive.

44 The battery-and-switch test circuit shown here is not just hypothetical, but may actually be used to test the

polarity of an unmarked transformer. Simply connect a DC voltmeter to the secondary winding while pressing and
releasing the pushbutton switch: the voltmeter’s polarity indicated while the button is pressed will indicate the relative
phasing of the two windings. Note that the voltmeter’s polarity will reverse when the pushbutton switch is released
and the magnetic field collapses in the transformer coil, so be sure to pay attention to the voltmeter’s indication only
during the instant the switch closes!
5.9. PHASOR ANALYSIS OF TRANSFORMER CIRCUITS 463

Consider these examples, where the polarity of a transformer’s secondary winding is changed
to illustrate the effect on AC phase shift from primary to secondary. In the left-hand example,
the “polarity” terminal of the transformer’s secondary winding exhibits a phase angle of 0o (with
respect to ground) at the same point in time when the primary winding’s voltage is at 0o : the two
winding voltages are in-phase with each other. In the right-hand example, the secondary winding
polarity has been reversed, the effect being a 180o phase shift in voltage. A pair of phasometers
in each example, both synchronized to the positive peak of the primary voltage, shows the phase
relationships clearly:

Voltage θ = 0o Voltage θ = 0o Voltage θ = 0o Voltage θ = 180o

105o 90o 75o 105o 90o 75o 105o 90o 75o 105o 90o 75o
120o 60o 120o 60o 120o 60o 120o 60o
135o 45o 135o 45o 135o 45o 135o 45o

150o 30o 150o 30o 150o 30o 150o 30o

165o 15o Strobe 165o 15o 165o 15o Strobe 165o 15o

180o 180o 180o 180o

0o

light 0o
0o

light 0o

-165 o -15o -165 o -15o -165 o -15o -165 o -15o

-150o -30o -150o -30o -150o -30o -150o -30o

-135o -45o -135o -45o -135o -45o -135o -45o

-120o -60o -120o -60o -120o -60o -120o -60o
-105o -90o -75o -105o -90o -75o -105o -90o -75o -105o -90o -75o

"Polarity" terminal "Polarity" terminal

"Polarity" terminal "Nonpolarity" terminal

"Nonpolarity" terminal "Polarity" terminal

"Nonpolarity" terminal "Nonpolarity" terminal

464 CHAPTER 5. AC ELECTRICITY

A useful way to analyze phase shifts in transformer circuits is to sketch separate phasor diagrams
for the primary and secondary windings, and use the polarity marks on each winding to reference each
phasor’s position on the diagram. We begin this analysis by first sketching a phasor for any voltage
of which the phase angle is given to us. Since transformer windings share common phase shift by
virtue of those windings sharing the same magnetic field, the phasor representing the other winding’s
voltage must possess the same angle. The only distinction is the length of each phasor (representing
primary and secondary voltage magnitudes) and its origin (starting point on the diagram).
Applying this analysis to the in-phase transformer circuit example. We will regard the secondary
winding’s voltage as a phasor that is half the length of the primary winding’s voltage given the 2:1
turns ratio, but oriented at the same angle:

2:1

120 VAC

Primary phasor diagram Secondary phasor diagram

120 V ∠ 0o 60 V ∠ 0o

The result is a pair of phasors both at 0o angles. Note the placement of dots on each phasor
to mark the relationship of the phasor to each winding polarity. The center of the diagram where
the real and imaginary axes intersect is the point representing zero potential (ground). The length
of each phasor tells us the magnitude of each waveform, in this case 120 volts and 60 volts AC,
respectively. The “nonpolarity” end of each phasor line must be located at the origin because the
nonpolarity terminal of each transformer winding is connected to ground.
5.9. PHASOR ANALYSIS OF TRANSFORMER CIRCUITS 465

Now consider the same analysis applied to the out-of-phase transformer circuit:

2:1

120 VAC

Primary phasor diagram Secondary phasor diagram

120 V ∠ 0o 60 V ∠ 180o

The result of this transformer connection is a pair of horizontal phasors, angled 180o apart
from each other. Note how the secondary phasor still has its polarity mark on the right-hand end,
but now that end of the phasor is anchored at the diagram’s origin because now the “polarity”
(dot) terminal of the transformer’s secondary winding is connected to ground. Following our rule
of keeping primary and secondary phasors at equal angles, the only difference made by grounding
the opposite terminal of the secondary winding is the reference point of the secondary phasor. The
secondary phasor must remain locked at the same angle as the primary phasor, but its position on
the diagram is determined by where the transformer winding connects to ground.
466 CHAPTER 5. AC ELECTRICITY

The following analyses illustrate the relationships between phase angles, transformer polarity,
circuit connections, and phasor diagrams.

Here we have an “autotransformer45 ” connected to boost the primary voltage. The VP and VS
phasors representing voltage across the primary and secondary windings must have common angles
(i.e. both horizontal phasors), but the secondary phasor is “stacked” on the end of the primary
phasor to create a longer phasor representing the voltage Vout with reference to ground. We can tell
the phasors will be stacked polarity-end-to-nonpolarity-end because that is how the two transformer
windings are electrically connected to each other:

"Boost" autotransformer connection

2:1 Vout

Vin VP VS

Primary phasor diagram Secondary phasor diagram

VP = 120 V ∠ 0o
VS = 60 V ∠ 0o
VP = 120 V ∠ 0o Vout = 180 V ∠ 0o

The result of this “boost’ configuration is an output voltage that is the direct sum of the primary
and secondary winding voltages: 180 volts with a 2:1 transformer ratio and 120 volt source. If the
transformer ratio were 4:1 instead of 2:1, the output would be 150 volts (120 volts + 120 4 volts).

45 An autotransformer is any transformer configuration where the primary and secondary windings are connected

rather than galvanically isolated from each other as is typical.

5.9. PHASOR ANALYSIS OF TRANSFORMER CIRCUITS 467

Here we have another autotransformer, this time connected to buck the primary voltage. The
VP and VS phasors still share common angles, but now they are “stacked” differently: polarity-end-
to-polarity-end. This is challenging to see on the second phasor diagram because the VP and VS
overlap each other, sharing a common “polarity” (dot) end point:

"Buck" autotransformer connection

2:1 Vout

Vin VP VS

Primary phasor diagram Secondary phasor diagram

VP = 120 V ∠ 0o
VS = 60 V ∠ 0o
VP = 120 V ∠ 0o Vout = 60 V ∠ 0o

The result of this “buck” configuration is a Vout that is the difference between VP and VS rather
than being the sum of the two as was the case with the “boost” configuration. If the transformer
ratio were 4:1 instead of 2:1, the output of this buck autotransformer would be 90 volts (120 volts
− 120
4 volts).
468 CHAPTER 5. AC ELECTRICITY

5.9.2 Phasors in three-phase transformer circuits

In the following three-phase transformer circuit, each transformer has been labeled with a number,
and each pair of phasors representing primary and secondary winding voltages for each transformer
has that same identifying number written next to it for convenient reference. Numbering the phasors
makes it clear which phasor(s) belong to which transformer.
As usual, the primary and secondary phasors for each transformer share the same angle, but the
secondary phasors are “stacked” polarity-end-to-nonpolarity-end because that is precisely how the
secondary windings are electrically connected to each other. Assuming an ABC phase sequence with
VA being the reference phasor defining 0o phase angle:

3-phase transformer with "corner-grounded" Delta secondary

VA VX
2:1

VY
VB
2:1

VZ
VC
2:1

Primary phasor diagram Secondary phasor diagram

VC = 120 V ∠ 120o
3 VZ = 60 V ∠ 120o
VA = 120 V ∠ 0o 2 3

2 1 1 VX = 0 V
VY = 60 V ∠ 180o

VB = 120 V ∠ -120o
5.10. TRANSMISSION LINES 469

5.10 Transmission lines

Many years ago, when I was first learning about electricity, I happened to discover a length of
coaxial television cable with the rating “50 ohm” printed on the jacket. Having my own ohmmeter,
and wanting to know what the “50 ohm” label referred to, I tried measuring the resistance of this
cable to see where it would exhibit 50 ohms of resistance. First I tried measuring between the
two open-ended wires (the center conductor and the braided shield conductor), where my meter
registered infinite resistance (no continuity). This is what I expected a normal cable would read
when connected to nothing else. Next I tried connecting my meter from one end of the cable to
the other: between both ends of the center conductor my meter registered 0 ohms, and between
both ends of the shield conductor my meter also read 0 ohms. In fact, this is all I was ever able
to measure with my ohmmeter: either 0 ohms or infinite ohms (open), no matter what I tried. I
was confounded – how could this piece of cable possibly be called a “50 ohm” cable if I could not
measure 50 ohms anywhere on it?

Ohmmeter

V A

V A
OFF
/U
-58

A COM
RG

Inner
conductor
Outer
Ω
50

conductor
(wire braid)

RG-58/U coaxial cable

Insulation

It was not until years later that I came across a reference in an old book to something called
surge impedance, describing how lengths of cable responded to short-duration electrical pulses, that
I finally understood the “50 ohm” rating of that coaxial cable. What I failed to discover with my
ohmmeter – in fact, what I never could have even seen with a plain ohmmeter – is that the cable
did indeed act like a 50 ohm resistor, but only for extremely short periods of time. This is an aspect
of electrical theory I was completely unprepared to comprehend at that time, knowing only how DC
circuits functioned. This is the subject we are about to explore now.

In basic DC electrical theory, students learn that open circuits always drop the full applied
voltage and cannot have electric currents anywhere in them. Students likewise learn that short
470 CHAPTER 5. AC ELECTRICITY

circuits drop negligible voltage while conducting full electrical current. They also learn that the
effects of electricity occur instantaneously throughout the circuit: for example, if a functioning
circuit suffers an “open” fault, current through that branch of the circuit halts immediately and
everywhere. These basic rules of electricity are extremely useful for DC circuit analysis, indeed even
essential for troubleshooting DC circuits, but they are not 100% correct. Like many of the scientific
principles one initially learns, they are only approximations of reality.
The effects of electricity do not propagate instantaneously throughout a circuit, but rather spread
at the speed of light (approximately 300,000,000 meters per second). Normally we never see any
consequences of this finite speed because everything happens so fast, just as I never saw my ohmmeter
register 50 ohms when I tried to test that coaxial cable many years ago. However, when we test
a circuit on a time scale of nano-seconds (billionths of a second!), we find some very interesting
phenomena during the time electricity propagates along the length of a circuit: open circuits can
indeed (temporarily) exhibit current, and short circuits can indeed (temporarily) exhibit voltage
drops.
When a pulse signal is applied to the beginning of a two-conductor cable, the reactive elements
of that cable (i.e. capacitance between the conductors, inductance along the cable length) begin to
store energy. This translates to a current drawn by the cable from the source of the pulse, as though
the cable were acting as a (momentarily) resistive load. If the cable under test were infinitely long,
this charging effect would never end, and the cable would indeed behave exactly like a resistor from
the perspective of the signal source. However, real cables (having finite length) do stop “charging”
after some time following the application of a voltage signal at one end, the length of that time being
a function of the cable’s physical length and the speed of light.
In honor of the cable’s capacity to behave as a temporary load to any impressed signal, we
typically refer to it as something more than just a cable. From the perspective of an electrical pulse,
measured on a time scale of nanoseconds, we refer to any cable as a transmission line. All electrical
cables act as transmission lines, but the effects just mentioned are so brief in duration that we only
notice them when the cable is exceptionally long and/or the pulse durations are exceptionally short
(i.e. high-frequency signals).
During the time when a transmission line is absorbing energy from a power source – whether
this is indefinitely for a transmission line of infinite length, or momentarily for a transmission line
of finite length – the current it draws will be in direct proportion to the voltage applied by the
source. In other words, a transmission line behaves like a resistor, at least for a moment. The
amount of “resistance” presented by a transmission line is called its characteristic impedance, or
surge impedance, symbolized in equations as Z0 . Only after the pulse signal has had time to travel
down the length of the transmission line and reflect back to the source does the cable stop acting as
a load and begin acting as a plain pair of wires.
A transmission line’s characteristic impedance is a function of its conductor geometry (wire
diameter, spacing) and the permittivity of the dielectric separating those conductors. If the line’s
design is altered to increase its bulk capacitance and/or decrease its bulk inductance (e.g. decreasing
the distance between conductors), the characteristic impedance will decrease. Conversely, if the
transmission line is altered such that its bulk capacitance decreases and/or its bulk inductance
increases, the characteristic impedance will increase. It should be noted that the length of the
transmission line has absolutely no bearing on characteristic impedance. A 10-meter length of RG-
58/U coaxial cable will have the exact same characteristic impedance as a 10000 kilometer length of
RG-58/U coaxial cable (50 ohms, in both cases). The only difference is the length of time the cable
will behave like a resistor to an applied voltage.
5.10. TRANSMISSION LINES 471

5.10.1 Open-ended transmission lines

The following sequence illustrates the propagation of a voltage pulse forward and back (reflected) on
an open-ended transmission line beginning from the time the DC voltage source is first connected
to the left-hand end:

current voltage
current
Time

current voltage
current

current
current

current voltage (no current)

current

voltage (no current)

The end result is a transmission line exhibiting the full source voltage, but no current. This is
exactly what we would expect in an open circuit. However, during the time it took for the pulse to
travel down the line’s length and back, it drew current from the source equal to the source voltage
divided by the cable’s characteristic impedance (Isurge = Vsource
Z0 ). For a brief amount of time, the
two-conductor transmission line acted as a load to the voltage source rather than an open circuit.
Only after the pulse traveled down the full length of the line and back did the line finally act as a
plain open circuit.
472 CHAPTER 5. AC ELECTRICITY

An experiment performed with a square-wave signal generator and oscilloscope46 connected to

one end of a long wire pair cable (open on the far end) shows the effect of the reflected signal:

The waveform steps up for a brief time, then steps up further to full source voltage. The first
step represents the voltage at the source during the time the pulse traveled along the cable’s length,
when the cable’s characteristic impedance acted as a load to the signal generator (making its output
voltage “sag” to a value less than its full potential). The next step represents the reflected pulse’s
return to the signal generator, when the cable’s capacitance is fully charged and is no longer drawing
current from the signal generator (making its output voltage “rise”). A two-step “fall” appears at
the trailing edge of the pulse, when the signal generator reverses polarity and sends an opposing
pulse down the cable.
The duration of the first and last “steps” on the waveform represents the time taken by the signal
to propagate down the length of the cable and return to the source. This oscilloscope’s timebase was
set to 0.5 microseconds per division for this experiment, indicating a pulse round-trip travel time
of approximately 0.2 microseconds. Assuming a velocity factor of 0.7 (70% the speed of light), the
round-trip distance calculates to be approximately 42 meters, making the cable 21 meters in length:

Distance = Velocity × Time = (70%)(3 × 108 m/s)(0.2 µs) = 42 m

1
Cable length = (Distance traveled by pulse) = (0.5)(42 m) = 21 m
2

46 The signal generator was set to a frequency of approximately 240 kHz with a Thévenin resistance of 118 ohms to

closely match the cable’s characteristic impedance of 120 ohms. The signal amplitude was just over 6 volts peak-to-
peak.
5.10. TRANSMISSION LINES 473

5.10.2 Shorted transmission lines

The following sequence illustrates the propagation of a voltage pulse forward and back (reflected)
on a shorted-end transmission line beginning from the time the DC voltage source is first connected
to the left-hand end:

current voltage
current
Time

current voltage
current

current
current

current (no voltage)

current

current (no voltage)

current

The end result is a transmission line exhibiting the full current of the source (Imax = VRsource
wire
),
but no voltage. This is exactly what we would expect in a short circuit. However, during the time
it took for the pulse to travel down the line’s length and back, it drew current from the source equal
to the source voltage divided by the cable’s characteristic impedance (Isurge = Vsource
Z0 ). For a brief
amount of time, the two-conductor transmission line acted as a moderate load to the voltage source
rather than a direct short. Only after the pulse traveled down the full length of the line and back
did the line finally act as a plain short-circuit.
474 CHAPTER 5. AC ELECTRICITY

An experiment performed with the same signal generator and oscilloscope connected to one end
of the same long wire pair cable (shorted on the far end) shows the effect of the reflected signal:

Here, the waveform steps up for a brief time, then steps down toward zero. As before, the
first step represents the voltage at the source during the time the pulse traveled along the cable’s
length, when the cable’s characteristic impedance acted as a load to the signal generator (making its
output voltage “sag” to a value less than its full potential). The step down represents the (inverted)
reflected pulse’s return to the signal generator, nearly canceling the incident voltage and causing
the signal to fall toward zero. A similar pattern appears at the trailing edge of the pulse, when the
signal generator reverses polarity and sends an opposing pulse down the cable.
Note the duration of the pulse on this waveform, compared to the first and last “steps” on the
open-circuited waveform. This pulse width represents the time taken by the signal to propagate
down the length of the cable and return to the source. This oscilloscope’s timebase remained at 0.5
microseconds per division for this experiment as well, indicating the same pulse round-trip travel
time of approximately 0.2 microseconds. This stands to reason, as the cable length was not altered
between tests; only the type of termination (short versus open).
5.10. TRANSMISSION LINES 475

5.10.3 Properly terminated transmission lines

Proper “termination” of a transmission line consists of connecting a resistance to the end(s) of
the line so that the pulse “sees” the exact same amount of impedance at the end as it did while
propagating along the line’s length. The purpose of the termination resistor is to completely absorb
the pulse’s energy so that none of it will be reflected back to the source.
The following sequence illustrates the propagation of a voltage pulse on a transmission line with
proper “termination” (i.e. a resistor matching the line’s surge impedance, connected to the far end)
beginning from the time the DC voltage source is first connected to the left-hand end:

R = Z0

current voltage
current
Time

current voltage
current

current voltage
current

From the perspective of the pulse source, this properly terminated transmission line “looks” the
same as an unterminated line of infinite length. There is no reflected pulse, and the DC voltage
source “sees” an unchanging load resistance the entire time.
476 CHAPTER 5. AC ELECTRICITY

An experiment performed with a termination resistor in place shows the near-elimination of

reflected pulses:

The pulse looks much more like the square wave it should be, now that the cable has been properly
terminated47 . With the termination resistor in place, a transmission line always presents the same
impedance to the source, no matter what the signal level or the time of signal application. Another
way to think of this is from the perspective of cable length. With the proper size of termination
resistor in place, the cable appears infinitely long from the perspective of the power source because
it never reflects any signals back to the source and it always consumes power from the source.

Data communication cables for digital instruments behave as transmission lines, and must be
terminated at both ends to prevent signal reflections. Reflected signals (or “echoes”) may cause
errors in received data in a communications network, which is why proper termination can be so
important. For point-to-point networks (networks formed by exactly two electronic devices, one at
either end of a single cable), the proper termination resistance is often designed into the transmission
and receiving circuitry, and so no external resistors need be connected. For “multi-drop” networks
where multiple electronic devices tap into the same electrical cable, excessive signal loading would
occur if each and every device had its own built-in termination resistance, and so the devices are
built with no internal termination, and the installer must place two termination resistors in the
network (one at each far end of the cable).

47 The termination shown here is imperfect, as evidenced by the irregular amplitude of the square wave. The cable

used for this experiment was a length of twin-lead speaker cable, with a characteristic impedance of approximately 120
ohms. I used a 120 ohm (± 5%) resistor to terminate the cable, which apparently was not close enough to eliminate
all reflections.
5.10. TRANSMISSION LINES 477

5.10.4 Discontinuities
A transmission line’s characteristic impedance will be constant throughout its length so long as its
conductor geometry and dielectric properties are consistent throughout its length. Abrupt changes
in either of these parameters, however, will create a discontinuity in the cable capable of producing
signal reflections. This is why transmission lines must never be sharply bent, crimped, pinched,
twisted, or otherwise deformed.
The probe for a guided-wave radar (GWR) level transmitter is another example of a transmission
line, one where the vapor/liquid interface creates a discontinuity: there will be an abrupt change
in characteristic impedance between the transmission line in vapor space versus the transmission
line submerged in a liquid due to the differing dielectric permittivities of the two substances. This
sudden change in characteristic impedance sends a reflected signal back to the transmitter. The
time delay measured between the signal’s transmission and the signal’s reception by the transmitter
represents the vapor space distance, or ullage.
For more detail on the theory and function of radar level measurement, see section 20.5.2
beginning on page 1419.

5.10.5 Velocity factor

The speed at which an electrical signal propagates down a transmission line is never as fast as the
speed of light in a vacuum. A value called the velocity factor expresses the propagation velocity as
a ratio to light, and its value is always less than one:
v
Velocity factor =
c
Where,
v = Propagation velocity of signal traveling along the transmission line
c = Speed of light in a vacuum (≈ 3.0 × 108 meters per second)

Velocity factor is a function of dielectric constant, but not conductor geometry. A greater
permittivity value results in a slower velocity (lesser velocity factor).

5.10.6 Cable losses

Ideally, a transmission line is a perfectly loss-less conduit for electrical energy. That is, every watt
of signal power entering the transmission line is available at the end where the load is connected. In
reality, though, this is never the case. Conductor resistance, as well as losses within the dielectric
(insulating) materials of the cable, rob the signal of energy.
For transmission lines, power loss is typically expressed in units of decibels per 100 feet or per
100 meters. A “decibel,” as you may recall, is ten times the logarithm of a power ratio:
 
P
Gain or Loss in dB = 10 log
Pref
Thus, if a transmission line receives 25 milliwatts of signal power at one end, but only conveys
0.018
18 milliwatts to the far end, it has suffered a 1.427 dB loss (10 log 0.025 = −1.427 dB) from end to
end. Power loss in cables is strongly dependent on frequency: the greater the signal frequency, the
more severe the power loss per unit cable length.
478 CHAPTER 5. AC ELECTRICITY

5.11 Antennas
Capacitors store energy in electric fields, proportional to the square of voltage. Inductors store
energy in magnetic fields, proportional to the square of current. If capacitors and inductors are
connected together, their complementary energy storage modes create a condition where electrical
energy transfers back and forth between the capacitance and the inductance: voltage and current
both oscillating sinusoidally. We refer to this cyclic exchange of energy as resonance. The simplest
resonant circuit possible is the so-called tank circuit, comprised of a single inductor connected to a
single capacitor:

Tank circuit

C L

The natural frequency at which a tank circuit oscillates is given by the formula fr = 2π√1LC ,
where fr is the resonant frequency in Hertz, C is the capacitance in Farads, and L is the inductance
in Henrys.
A perfect tank circuit – devoid of electrical resistance and any other energy-dissipating
characteristics (e.g. capacitor dielectric losses, inductor hysteresis losses) – would continue to
oscillate forever once initially stimulated. That is, if initially “charged” with an impulse of DC
voltage, the tank circuit would continue to produce AC voltage and current oscillations at its resonant
frequency, at constant peak amplitude, forever:

Pushbutton switch Ideal tank circuit behavior

DC power +
source − C L

Time
Switch pressed
and released
5.11. ANTENNAS 479

Since real capacitors and inductors are not lossless, real tank circuits exhibit decaying-amplitude
oscillations after initial “charging,” until no energy is stored in either the capacitor or the inductor:

Pushbutton switch Real tank circuit behavior

DC power +
source − C L

Time
Switch pressed
and released

Capacitive losses take the form of heat loss in the dielectric substance separating the capacitor
plates. The electric field between the capacitor’s plates imparts forces on any polar48 molecules
within that substance, thereby doing “work” on those molecules by forcing them back and forth
as the electric field alternates. Though these forces and motions are extremely small, they are
nevertheless capable of draining considerable energy from the capacitor, dissipating it in the form
of heat.
Inductive losses are similar, but in the form of work done on ferromagnetic molecules in the core
material as the magnetic field alternates polarity. Like dielectric heating, magnetic losses also drain
energy from the inductor, dissipating it in the form of heat.
Of course, both capacitors and inductors also contain ohmic resistance in the metals used to form
the plates and wire coils, respectively. Resistance naturally dissipates energy in the form of heat,
constituting another energy-loss mechanism for both capacitors and inductors (albeit much more
significant in inductors than in capacitors!).
The combined effect of all these energy-loss mechanisms is that the oscillations of an unpowered
tank circuit decay over time, until they cease completely. This is similar in principle to a pendulum
gradually coming to a halt after being set in motion with a single push: if not for air resistance and
other forms of friction, the pendulum should swing forever! With air friction and mechanical friction
in the pendulum’s bearing, though, a pendulum’s oscillations gradually diminish in amplitude until
all its energy has been lost – swing by swing – to heat.

48 A “polar” molecule is one where the constituent atoms are bound together in such a way that there is a definite

electrical polarity from one end of the molecule to the other. Water (H2 O) is an example of a polar molecule: the
positively charged hydrogen atoms are bound to the negatively charged oxygen atom in a “V” shape, so the molecule
as a whole has a positive side and a negative side which allows the molecule to be influenced by external electric
fields. Carbon dioxide (CO2 ) is an example of a non-polar molecule whose constituent atoms lie in a straight line
with no apparent electrical poles. Interestingly, microwave ovens exploit the fact of water molecules’ polarization by
subjecting food containing water to a strong oscillating electric field (microwave energy in the gigahertz frequency
range) which causes the water molecules to rotate as they continuously orient themselves to the changing field polarity.
This oscillatory rotation manifests itself as heat within the food.
480 CHAPTER 5. AC ELECTRICITY

Capacitance and inductance, however, are not limited to capacitors and inductors: any pair of
conductors separated by an insulating medium will exhibit capacitance, and any electrical conductor
will exhibit inductance along its length. Even a two-conductor cable (a transmission line) has
distributed capacitance and inductance capable of storing energy. If a long, unterminated (no load
connected to the far end) cable is “charged” by a pulse of applied DC voltage, it will sustain a series
of echoing pulses at a period dependent on the cable’s length and velocity factor:

Pushbutton switch

Transmission line behavior

DC power +
−
incident pulse reflected pulse
source

Ideal behavior Real (lossy) behavior

Time Time
Switch pressed Switch pressed
and released and released

The ability for a transmission line to support periodic signal “echoes” means it may also resonate
when energized by an AC power source, just like a tank circuit. Unlike a tank circuit, however, a
transmission line is able to resonate at more than one frequency: a “fundamental” frequency, or any
whole-number multiple of that fundamental frequency called a harmonic frequency. For example,
an unterminated transmission line with a length of 1 kilometer and a velocity factor of 0.7 has a
round-trip echo time (period) of 9.53 microseconds, equivalent to a resonant frequency of 104.9 kHz.
However, it will resonate equally well at exactly twice this fundamental frequency (209.8 kHz – the
second harmonic of 104.9 kHz) as well as three times the fundamental frequency (314.8 kHz – the
third harmonic of 104.9 kHz), etc. A simple LC tank circuit, by contrast, will only resonate at a
single frequency.
This “poly-resonant” behavior of transmission lines has an analogue in the world of music.
“Wind” instruments such as trombones, flutes, trumpets, saxophones, clarinets, recorders, etc., are
really nothing more than tubes with at least one open end. These tubes will acoustically resonate
at specific frequencies when “excited” by turbulent air flow at one end. The lowest frequency such a
tube will resonate at is its “fundamental” frequency, but increasing the turbulence of the incoming
air flow will cause the tone to “jump” to some harmonic of that fundamental frequency. The
fundamental frequency of the tube may be altered by changing the length of the tube (e.g. as in a
trombone or trumpet) or by opening ports along the tube’s length to alter its effective length (flute,
saxophone, clarinet, recorder, etc.).
5.11. ANTENNAS 481

If we were to alter our transmission line test circuit, splaying the two conductors apart from each
other rather than running them alongside one another, it would also form another type of resonant
circuit, owing to the distributed capacitance and inductance along the wires’ lengths:

Pushbutton switch

DC power +
source −

This special case of resonant circuit has some very interesting properties. First, its resonant
frequency is quite high, because the distributed inductance and capacitance values are extremely
small compared to the values of discrete components such as inductors and capacitors. Second, it is
a very “lossy” circuit despite having no appreciable electrical resistance to dissipate energy, no solid
insulating medium to incur dielectric losses, and no ferromagnetic core to exhibit hysteresis losses.
This special form of resonant circuit loses energy not to heat, but rather to electromagnetic radiation.
In other words, the energy in the electric and magnetic fields leave the circuit and propagate through
space in the form of electro-magnetic waves, or what we more commonly refer to now as radio waves:
a series of electric and magnetic fields oscillating as they radiate away from the wires at the speed
of light.
Tank circuits and transmission lines do not radiate energy because they are intentionally designed
to contain their fields: capacitors are designed to fully contain their electric fields, inductors are
designed to fully contain their magnetic fields, and the fields within transmission lines are likewise
constrained. Our two-wire resonant circuit, by contrast, does just the opposite: its electric and
magnetic fields are exposed to open space with no containment whatsoever. What we have built
here is not a tank circuit nor a transmission line, but rather an antenna 49 : a structure designed to
radiate electric and magnetic fields as waves into surrounding space.

49 An older term used by radio pioneers to describe antennas is radiator, which I personally find very descriptive.

The word “antenna” does an admirable job describing the physical appearance of the structure – like antennae on an
insect – but the word “radiator” actually describes its function, which is a far more useful principle for our purposes.
482 CHAPTER 5. AC ELECTRICITY

5.11.1 Maxwell and Hertz

An interesting historical footnote is that this phenomenon of electromagnetic waves propagating
through space was predicted theoretically before it was demonstrated experimentally. A Scottish
physicist named James Clerk Maxwell made an astonishing theoretical prediction which he published
in 1873, expressed in these four equations:
Q
I
E · dA =
ǫ0
I
B · dA = 0

dΦB
I
E · ds = −
dt
dΦE
I
B · ds = µ0 I + µ0 ǫ0
dt
The last two equations hold the most interest to us with respect to electromagnetic waves. The
third equation states that an electric field (E) will be produced in open space by a changing magnetic
flux dΦ


B
dt . The fourth equation states than a magnetic field (B) will be produced in open space
either by an electric current (I) or by a changing electric flux dΦ


dt . Given this complementary
E

relationship, Maxwell reasoned, it was possible for a changing electric field to create a changing
magnetic field which would then create another changing electric field, and so on. This cause-and-
effect cycle could continue, ad infinitum, with fast-changing electric and magnetic fields radiating off
into open space without needing wires to carry or guide them. In other words, the complementary
fields would be self-sustaining as they traveled.
The Prussian Academy of Science offered a reward to anyone who could experimentally validate
Maxwell’s prediction, and this challenge was met by Professor Heinrich Hertz at the Engineering
College in Karlsruhe, Germany in 1887, eight years after Maxwell’s death. Hertz constructed and
tested a pair of devices: a “radiator” to produce the electromagnetic waves, and a “resonator” to
receive them.
5.11. ANTENNAS 483

A simplified diagram showing Hertz’s experimental device is shown here:

Metal plate

Small
spark gap

Radiator
A few meters’
Spark gap
distance
Wire loop

Resonator
Induction coil

Metal plate

An “induction coil” (a buzzing device constructed of a self-interrupting relay and step-up

transformer winding to generate a continuous pulsing waveform of high voltage) provided an
extremely noisy (i.e. frequency-rich) AC signal to the radiator, while a spark gap at the resonator
provided indication that the electromagnetic waves were captured and converted into voltage by the
resonator wire.
Both the radiator and the resonator are what we would now call antennas. The purpose of the
transmitting antenna (radiator) is to take high-frequency AC power and radiate that power in the
form of electromagnetic waves: self-sustaining electric and magnetic fields propagating out into open
space. The purpose of the receiving antenna is to capture those electromagnetic waves and convert
them into an AC signal. All antennas – from historical to the most modern – behave in fundamentally
the same way: energize them with high-frequency AC power, and they will radiate electromagnetic
waves at that frequency; expose them to electromagnetic waves, and they will produce a very small
AC signal at the same frequency as the incident radiation.
484 CHAPTER 5. AC ELECTRICITY

5.11.2 Antenna size

Earlier it was mentioned that antennas are fundamentally resonant elements: they “prefer” to
electrically oscillate at a fundamental frequency (and at whole-number multiples of that fundamental
frequency). An antenna will behave most efficiently – whether transmitting or receiving – when
operated in a condition of resonance. The relationship between ideal frequency and antenna length
is inverse: the longer the antenna, the lower its fundamental frequency, and vice-versa. This is the
same mathematical relationship we see between frequency and wavelength (λ) for any moving wave:

v = λf
The prototypical antenna shown earlier – with two wires oriented 180o from each other – operates
best at a wavelength twice as long as the total length of wire.

"Half-wave" dipole antenna "Half-wave" dipole antenna

operating in transmission mode operating in reception mode

AC power v AC signal
source λ/2 λ= λ/2 V receiver
f (e.g. RF ammeter)

A “half-wave” dipole antenna with a total length of 5 meters50 will radiate optimally at a
frequency of 30 MHz, given that the velocity of electromagnetic waves in open space is approximately
3 × 108 meters per second. This same antenna will also effectively resonate at any harmonic (integer
multiple) of 30 MHz (e.g. 60 MHz, 90 MHz, 120 MHz, etc.) just as a transmission line is able to
resonate at a fundamental frequency as well as any harmonic thereof.

50 In practice, the ideal length of a dipole antenna turns out to be just a bit shorter than theoretical, due to lumped-

capacitive effects at the wire ends. Thus, a resonant 30 MHz half-wave dipole antenna should actually be about 4.75
meters in length rather than exactly 5 meters in length.
5.11. ANTENNAS 485

A popular variation on the theme of the half-wave dipole antenna is the so-called quarter-wave
or “whip” antenna, which is exactly what you might think it is: one-half of a half-wave antenna.
Instead of two wires pointed away from each other, we substitute an earth-ground connection for
one of the wires:

"Quarter-wave" antenna "Quarter-wave" antenna

operating in transmission mode operating in reception mode

v
λ/4 λ= λ/4
f
AC power AC signal
source V receiver
(e.g. RF ammeter)

Quarter-wave antennas tend to be less effective than half-wave antennas, but are usually much
more convenient to construct for real applications.

5.11.3 Antenna orientation and directionality

Another factor affecting antenna performance is the orientation of the receiving antenna with respect
to the transmitting antenna. Antenna conductors should be parallel to each other in order to
maximize reception, in order that the electric and magnetic fields emanating from the wires of the
transmitting antenna will “link” properly with the wires of the receiving antenna(s). If the goal is
optimum communication in any direction (omnidirectionality), dipole and whip antennas should be
arranged vertically so that all antenna conductors will be parallel to each other regardless of their
geographic location.

Omnidirectionality may seem like a universally good trait for any antenna: to be able to transmit
and receive electromagnetic waves equally well in any direction. However, there are good reasons
for wanting directionality in an antenna. One reason is for greater security: perhaps you have an
application where you do not wish to broadcast information in all directions, where anyone at all
could receive that information. In that case, the best antennas to use would be those that work
best in one direction and one direction only, with transmitting and receiving antenna pairs pointed
directly at each other.
Another reason for antenna directionality is improved reception. As noted before, the AC signal
received at an antenna is very small, typically on the order of microvolts. Since electromagnetic
radiation tends to “spread” as it travels, becoming weaker with distance from the transmitting
antenna, long-range radio communication benefits from increased sensitivity. A transmitting antenna
that is directional will focus more of its power in one direction than in others, resulting in less
“wasted” power radiated in unwanted directions. Likewise, a receiving antenna that is directional
does a better job of “collecting” the electromagnetic energy it receives from that direction (as well
as ignoring electromagnetic waves coming from other directions).
486 CHAPTER 5. AC ELECTRICITY

A simple yet highly effective antenna design for directional transmission and reception is the
Yagi, named after its inventor. A Yagi is based on a half-wave dipole element surrounded by one
or more wires longer than λ/2 to the rear (called “reflectors”) and one or more wires shorter than
λ/2 to the front (called “directors”). The terms “reflector” and “director” are quite apt, describing
their interaction with electromagnetic waves from the perspective of the dipole: reflectors reflect the
waves, while directors direct the waves. The result is an antenna array that is much more directional
than a simple dipole:

"Yagi" antenna operating "Yagi" antenna operating

in transmission mode in reception mode

Directors Directors

Reflectors Reflectors
5.11. ANTENNAS 487

An example of a Yagi antenna used as part of a SCADA system is shown in this photograph,
the antenna of course being the multi-element array in the upper-right corner:
488 CHAPTER 5. AC ELECTRICITY

Another example of a highly directional antenna design is the parabolic dish, often used in
microwave and satellite communications. This photograph shows two “dish” antennas, one open
to the weather (right) and the other covered with a weather-proof diaphragm to keep the antenna
elements protected (left):

Some dish antennas use mesh or metal-tube reflectors rather than a solid parabolic reflector, as
is the case in this next photograph:
5.11. ANTENNAS 489

References
Blackburn, J. Lewis and Domin, Thomas J., Protective Relaying Principles and Applications, Third
Edition, CRC Press, Taylor & Francis Group, Boca Raton, FL, 2007.

Boylestad, Robert L., Introductory Circuit Analysis, 9th Edition, Prentice Hall, Upper Saddle River,
NJ, 2000.

Dorf, Richard C., Modern Control Systems, Fifth Edition, Addison-Wesley Publishing Company,
Reading, MA, 1989.

Eckman, Donald P., Automatic Process Control, John Wiley & Sons, Inc., New York, NY, 1958.

Field Antenna Handbook, U.S. Marine Corps document MCRP 6-22D, 1999.

Giancoli, Douglas C., Physics for Scientists & Engineers, Third Edition, Prentice Hall, Upper Saddle
River, NJ, 2000.

Harrison, Cecil A., Transform Methods in Circuit Analysis, Saunders College Publishing,
Philadelphia, PA, 1990.

Jenkins, John D., Where Discovery Sparks Imagination, American Museum of Radio and Electricity,
Bellingham, WA, 2009.

Kaplan, Wilfred, Advanced Mathematics for Engineers, Addison-Wesley Publishing Company,

Reading, MA, 1981.

Mileaf, Harry, Electronics One-Seven, Hayden Book Company, 1976.

Nilsson, James W., Electric Circuits, Addison-Wesley Publishing Company, Reading, MA, 1983.

Palm, William J., Control Systems Engineering, John Wiley & Sons, Inc., New York, NY, 1986.

Smith, Steven W., The Scientist and Engineer’s Guide to Digital Signal Processing, California
Technical Publishing, San Diego, CA, 1997.

Steinmetz, Charles P., Theoretical Elements of Electrical Engineering, Third Edition, McGraw-Hill
Book Company, New York, NY, 1909.

Steinmetz, Charles P., Theory and Calculation of Alternating Current Phenomena, Third Edition,
McGraw Publishing Company, New York, NY, 1900.

The ARRL Antenna Book, Eleventh Edition, The American Radio Relay League, Inc., Newington,
CT, 1968.

The ARRL Handbook For Radio Amateurs, 2001 Edition, ARRL – the national association for
Amateur Radio, Newington, CT, 2001.
490 CHAPTER 5. AC ELECTRICITY