Guided Waves

Daniel S. Weile

Department of Electrical and Computer Engineering

University of Delaware

ELEG 648—Guided Waves

D. S. Weile Guided Waves

Outline

D. S. Weile Guided Waves

Outline

D. S. Weile Guided Waves

The Circuit Model of Transmission Lines

R + jωL I(z + ∆z)

I(z) Z

V (z) G + jωC Y V (z + ∆z)

dz

All circuit quantities (R, L, G, C, Z and Y ) are per unit length.

V (z + ∆z) − V (z) = −I(z)Z ∆z

I(z + ∆z) − I(z) = −V (z)Y ∆z

D. S. Weile Guided Waves

Telegrapher’s Equations
This leads to the
Telegrapher’s Equations
dV dI
= −ZI = −YV
dz dz

D. S. Weile Guided Waves

Telegrapher’s Equations
This leads to the
Telegrapher’s Equations
dV dI
= −ZI = −YV
dz dz
which give rise to the
Transmission Line Helmholtz Equations
d2 V d2 I
− ZYV = 0 − ZYI = 0
dz 2 dz 2

D. S. Weile Guided Waves

Telegrapher’s Equations
This leads to the
Telegrapher’s Equations
dV dI
= −ZI = −YV
dz dz
which give rise to the
Transmission Line Helmholtz Equations
d2 V d2 I
− ZYV = 0 − ZYI = 0
dz 2 dz 2
Now define
√
γ = jk = ZY

For a lossless transmission line R = G = 0, so ZY < 0, and γ

is imaginary.
D. S. Weile Guided Waves
Solution of the Telegrapher’s Equations

The +z-directed voltage wave is

V + (z) = V0 e−γz

From the Telegrapher’s Equation we have

1 dV γ
I + (z) = − = V0 e−γz
Z dz Z
Thus, define the
Characteristic Impedance
r
V + (z) Z Z
Z0 = + = =
I (z) γ Y

D. S. Weile Guided Waves

Outline

D. S. Weile Guided Waves

Relationship to Maxwell’s Equations

Implicit in the transmission line model is the fact that there is no

coupling (mutual impedance or shunt current flow) between
sections. This is equivalent to the assumption that

Ez = Hz = 0

We are also interested in traveling wave solutions, so we

assume that the remaining fields have z-dependence e−γz . We
can thus make the substitution
∂
→ −γ
∂z

D. S. Weile Guided Waves

Relationship to Maxwell’s Equations

−ẑH = ∇ × E
 
 ux uy uz 
 ∂ ∂

=  ∂x ∂y −γ 

 E E 0 
x y

Thus, writing this out

γEy = −ẑHx
γEx = ẑHy
∂Ey ∂Ex
− = 0
∂x ∂y

D. S. Weile Guided Waves

Relationship to Maxwell’s Equations

The first two equations (and their duals) can be written in vector
form as

E = ηH × uz ηH = uz × E

Defining the
Transverse Laplacian
∂2 ∂2
∇2t = +
∂x 2 ∂y 2

The last equation and its dual give

Transverse Laplace Equations

∇2t E = 0 ∇2t H = 0

D. S. Weile Guided Waves

Relationship to Maxwell’s Equations

Note that the “in-plane statics” observation implies that there

must be two conductors.

D. S. Weile Guided Waves

Relationship to Maxwell’s Equations

Note that the “in-plane statics” observation implies that there

must be two conductors. Why?

D. S. Weile Guided Waves

Relationship to Maxwell’s Equations

Note that the “in-plane statics” observation implies that there

must be two conductors. Why?

Definitions
Cc
Z I
V =− E · dl I= H · dl
Co Cc

Given the relationship between

Co the transverse fields
I
1
I = η uz × E · dl
Cc
I
1
= η dl × uz · E
Cc

D. S. Weile Guided Waves

Relationship to Maxwell’s Equations
Shrinking Cc to the surface of the conductor, we can write
I
1
I=η En dl
Cc

Now, the charge per unit length on the center conductor is

I
q= En dl
Cc

from boundary conditions. Therefore, since C = q/V

I

V = En dl
C Cc

Thus

V 
Z0 = =η
I C
D. S. Weile Guided Waves
Relationship to Maxwell’s Equations

It may similarly be shown that

V L
Z0 = =η
I µ

by appeal to the magnetostatic problem. In short, in a TEM

wave,

D. S. Weile Guided Waves

Relationship to Maxwell’s Equations

It may similarly be shown that

V L
Z0 = =η
I µ

by appeal to the magnetostatic problem. In short, in a TEM

wave,
1 E, H, and uz form a right handed triplet at every point,

D. S. Weile Guided Waves

Relationship to Maxwell’s Equations

It may similarly be shown that

V L
Z0 = =η
I µ

by appeal to the magnetostatic problem. In short, in a TEM

wave,
1 E, H, and uz form a right handed triplet at every point,
2 In a transverse plane, E and H are static in nature,

D. S. Weile Guided Waves

Relationship to Maxwell’s Equations

It may similarly be shown that

V L
Z0 = =η
I µ

by appeal to the magnetostatic problem. In short, in a TEM

wave,
1 E, H, and uz form a right handed triplet at every point,
2 In a transverse plane, E and H are static in nature,
3 By computing C or L using a static solver, the entire
transmission line can be characterized.

D. S. Weile Guided Waves

Relationship to Maxwell’s Equations

It may similarly be shown that

V L
Z0 = =η
I µ

by appeal to the magnetostatic problem. In short, in a TEM

wave,
1 E, H, and uz form a right handed triplet at every point,
2 In a transverse plane, E and H are static in nature,
3 By computing C or L using a static solver, the entire
transmission line can be characterized.
4 None of this is strictly true if the dielectric is
inhomogeneous (even piecewise inhomogeneous).

D. S. Weile Guided Waves

Outline

D. S. Weile Guided Waves

Lossy Transmission Lines

This theory applies to transmission lines with lossy

dielectrics perfectly.

D. S. Weile Guided Waves

Lossy Transmission Lines

This theory applies to transmission lines with lossy

dielectrics perfectly.
Lossy conductors are more trouble, since they cannot
support TEM waves.

D. S. Weile Guided Waves

Lossy Transmission Lines

This theory applies to transmission lines with lossy

dielectrics perfectly.
Lossy conductors are more trouble, since they cannot
support TEM waves. Why?

D. S. Weile Guided Waves

Lossy Transmission Lines

This theory applies to transmission lines with lossy

dielectrics perfectly.
Lossy conductors are more trouble, since they cannot
support TEM waves. Why?
In any case, even with lossy conductors, we can assume the
fields are pretty much unchanged and just use the above
equations with
γ = α + jβ
We still have
V
V = V0 e−(α+jβ)z I=
Z0

D. S. Weile Guided Waves

Perturbation Theory

The time average power flow is

2
1 |V0 |
Pf = Re {VI ∗ } = e−2αz
2 2Re {Z0 }

Now, a power dissipated per unit length, Pd must be responsible for

the decrease:
dPf
Pd = − = 2αPf
dz

D. S. Weile Guided Waves

Perturbation Theory

The time average power flow is

2
1 |V0 |
Pf = Re {VI ∗ } = e−2αz
2 2Re {Z0 }

Now, a power dissipated per unit length, Pd must be responsible for

the decrease:
dPf
Pd = − = 2αPf
dz
Therefore, if we can approximate Pd (by, say assuming the current
distribution is unchanged by the introduction of loss), we find the

Perturbation Formula for the Dissipation Rate

Pd
α=
2Pf

D. S. Weile Guided Waves

Outline

D. S. Weile Guided Waves

The Basics

Transmission lines are two conductor systems that carry

TEM waves.
Other systems are possible. Anything that guides waves is
called a waveguide.
For now, we study the simplest of these, the rectangular
waveguide.

a x
z

D. S. Weile Guided Waves

Solution by Superposition

Consider two waves traveling at angles ±φ with respect to the

y –z plane. Assuming a y -polarization for the E-field, we can
write
 
Ey = A e−jkx sin φ + ejkx sin φ e−jkz cos φ
= −2jA sin (kx sin φ) e−jkz cos φ

Define

E0 = 2jA
kc = k sin φ
γ = jk cos φ

Of course γ 2 = kc2 − k 2 since sin2 φ + cos2 φ = 1.

D. S. Weile Guided Waves

Solution by Superposition

Our superposition is of the form

Modal Electric Field
Ey = E0 sin (kc x) e−γz

Now, for this to be a solution, we need

D. S. Weile Guided Waves

Solution by Superposition

Our superposition is of the form

Modal Electric Field
Ey = E0 sin (kc x) e−γz

Now, for this to be a solution, we need

Ey (x = 0) = Ey (x = a) = 0

D. S. Weile Guided Waves

Solution by Superposition

Our superposition is of the form

Modal Electric Field
Ey = E0 sin (kc x) e−γz

Now, for this to be a solution, we need

Ey (x = 0) = Ey (x = a) = 0

The first condition is satisfied automatically; for the second we

find
The Cutoff Wavenumber
nπ
kc = n = 1, 2, 3 . . .
a

D. S. Weile Guided Waves

Modes

Each choice of n is called a mode. It represents a solution

that can exist without a source (i.e. an eigenfunction).
This mode is called transverse electric TE because the
electric field is normal to the direction of propapagation.
There are transverse magnetic (TM) modes, in which
Hz = 0.
The particular modes we have found here are called the
TEn0 modes. The reason for this nomenclature will be
clear when we solve the problem in general.

D. S. Weile Guided Waves

Outline

D. S. Weile Guided Waves

Propagation Constant

We have seen that γ 2 = kc2 − k 2 . This means we have (for

k ∈ R)
The Propagation Constant
 q
 jβ = j k 2 − nπ
2 for k ≥ nπ
a a
γ= q
nπ 2 nπ


 α =
a − k 2 for k ≤ a

D. S. Weile Guided Waves

Propagation Constant

We have seen that γ 2 = kc2 − k 2 . This means we have (for

k ∈ R)
The Propagation Constant
 q
 jβ = j k 2 − nπ
2 for k ≥ nπ
a a
γ= q
nπ 2 nπ


 α =
a − k 2 for k ≤ a

When γ ∈ jR, the wave propagates in the z-direction.

When γ ∈ R, the wave evanesces in the z-direction.
The transition from evanescence to propagation occurs at
cutoff, when
nπ
k = kc =
a

D. S. Weile Guided Waves

Cutoff

Associated with the cutoff wavenumber, there is a

Cutoff Frequency
kc n
fc = √ = √
2π µ 2π µ

D. S. Weile Guided Waves

Cutoff

Associated with the cutoff wavenumber, there is a

Cutoff Frequency
kc n
fc = √ = √
2π µ 2π µ

Cutoff Wavelength
2π 2b
λc = =
kc n

D. S. Weile Guided Waves

Cutoff

Associated with the cutoff wavenumber, there is a

Cutoff Frequency
kc n
fc = √ = √
2π µ 2π µ

Cutoff Wavelength
2π 2b
λc = =
kc n

Below the cutoff frequency, waves evanesce. Above it, they

propagate.

D. S. Weile Guided Waves

Alternative Formula for Cutoff

Note that
kc fc
=
k f

D. S. Weile Guided Waves

Alternative Formula for Cutoff

Note that
kc fc
=
k f
Thus, we can write
s
 2
fc
q
k 2 − kc2 = k 1−
f

and
The Propagation Constant
 r  2
 jβ = jk 1 − ffc

 for f ≥ fc
γ= r  2
 α = kc 1 − ff


c
for f ≤ fc

D. S. Weile Guided Waves

Guide Wavelength and Phase Velocity

The guide wavelength λg is the distance over which the phase

of propagation in the z-direction changes by 2π. Thus
Guide Wavelength
2π 2π λ
λg = = r =r
β   2  2
k 1 − ffc 1 − ffc

By the same token, the guided phase velocity is the speed at

which a constant phase plane moves down the guide, i.e.
Guided Phase Velocity
fλ vp
vg = f λ g = r  2 = r  2
1 − ffc 1 − ffc

D. S. Weile Guided Waves

A Word About Phase Velocity

Note that the guided phase velocity is faster than the speed of
light! Is this a violation of Einstein’s theory of relativity?

D. S. Weile Guided Waves

A Word About Phase Velocity

Note that the guided phase velocity is faster than the speed of
light! Is this a violation of Einstein’s theory of relativity?
Can we use this observation to make a faster-than-light
communication system?

D. S. Weile Guided Waves

A Word About Phase Velocity

Note that the guided phase velocity is faster than the speed of
light! Is this a violation of Einstein’s theory of relativity?
Can we use this observation to make a faster-than-light
communication system?
Recall our derivation....

Phase fronts

λg
λ
φ k
φ z

D. S. Weile Guided Waves

A Word About Phase Velocity

From the figure,

λ
λg =
cos φ
This demonstrates that the wave is merely traveling obliquely,
and that the larger wavelength is merely a geometrical artifact.
It also demonstrates that the artifact is controlled by a
Geometrical Factor
s
 2
fc
cos φ = 1−
f

D. S. Weile Guided Waves

Magnetic Field

We can compute the magnetic field from Maxwell’s Equation

1
H=− ∇×E
jωµ

This immediately gives

The Magnetic Field

γ
Hx = − E0 sin (kc x) e−γz
jωµ
kc
Hz = E0 cos (kc x) e−γz
jωµ

D. S. Weile Guided Waves

Mode Impedance

Definition of Mode Impedance

Ey jωµ
Z0 = Zz = − =
Hx γ

Why is this the relevant definition? Why the minus sign?

D. S. Weile Guided Waves

Mode Impedance

Definition of Mode Impedance

Ey jωµ
Z0 = Zz = − =
Hx γ

Why is this the relevant definition? Why the minus sign?

We call it Z0 by analogy with transmission lines. Plugging in for
γ we have the frequently more useful expression for
Modal Impedance
 η
 r  2 for f ≥ fc
1− ffc


Z0 =
 r jη for f ≤ fc
 2
f
−1

f c

D. S. Weile Guided Waves

The Dominant Mode

The mode with the lowest cutoff frequency in a particular

waveguide is called the dominant mode.
For two-conductor, homogeneously-filled waveguiding
systems, this is the TEM mode.
For the rectangular waveguide, it is the TE10 mode.
The waveguide should be used between the cutoff of its
dominant mode and that of its first higher-order mode. Why?

D. S. Weile Guided Waves

The Dominant Mode

The mode with the lowest cutoff frequency in a particular

waveguide is called the dominant mode.
For two-conductor, homogeneously-filled waveguiding
systems, this is the TEM mode.
For the rectangular waveguide, it is the TE10 mode.
The waveguide should be used between the cutoff of its
dominant mode and that of its first higher-order mode. Why?
Will other modes exist in the guide in this operating bandwidth?
Where?

D. S. Weile Guided Waves

Power Transmission

The transverse fields can be written

Ey = E0 sin (kc x) e−γz

E0
Hx = − sin (kc x) e−γz
Zz
Above cutoff, we can compute

Modal Power

Z a Z b
1
Pf = − Ey Hx∗ dy dx
2 0 0
2 a b
|E0 |
Z Z  πx 
= sin2 dy dx =
2Z0 0 0 a

D. S. Weile Guided Waves

Power Transmission

The transverse fields can be written

Ey = E0 sin (kc x) e−γz

E0
Hx = − sin (kc x) e−γz
Zz
Above cutoff, we can compute

Modal Power

Z a Z b
1
Pf = − Ey Hx∗ dy dx
2 0 0
2 a b 2
|E0 | |E0 |
Z Z  πx 
= sin2 dy dx = ab
2Z0 0 0 a 4Z0

How does this change below cutoff?

D. S. Weile Guided Waves

Outline

D. S. Weile Guided Waves

Dielectric Losses

Dielectric loss essentially changes nothing above:

γ is complex (rather than pure real or imaginary) since
k 6∈ R
Z0 is complex (rather than pure real or imaginary)
Either one can be simply computed from the formulas
given above.
The only issue is that there is no real cutoff in a guide with
losses. Why?

D. S. Weile Guided Waves

Dielectric Losses

Dielectric loss essentially changes nothing above:

γ is complex (rather than pure real or imaginary) since
k 6∈ R
Z0 is complex (rather than pure real or imaginary)
Either one can be simply computed from the formulas
given above.
The only issue is that there is no real cutoff in a guide with
losses. Why?
If the losses are small, we can ignore the issue and define a
real fc . What happens around this frequency?

D. S. Weile Guided Waves

Conductive Losses

If there are losses in the conductor, our solution is not strictly

correct since the boundary condition has changed.
Nonetheless, we can use our current solution with the
perturbation theory we discussed for transmission lines.
Perturbation Formula for the Dissipation Rate
Pd
α=
2Pf

We have already computed Pf . How can we compute Pd ?

D. S. Weile Guided Waves

Conductive Losses

If there are losses in the conductor, our solution is not strictly

correct since the boundary condition has changed.
Nonetheless, we can use our current solution with the
perturbation theory we discussed for transmission lines.
Perturbation Formula for the Dissipation Rate
Pd
α=
2Pf

We have already computed Pf . How can we compute Pd ?

J = un × H and we know the surface resistance!

D. S. Weile Guided Waves

The Left Wall

On the x = 0 wall we have

b
R
Z
|Hz |2 dy

Pd x=0 =
2 0
R  kc E0 2 b
  Z
= dy
2  jωµ  0
bR kc 2
 
= |E0 |2
2 ηk
bR fc 2
 
= |E0 |2
2 ηf

The same amount of power is dissipated on the x = b wall.

D. S. Weile Guided Waves

The Bottom Wall

On the y = 0 wall we have

R a
Z
|Hz |2 + |Hx |2 dx

Pd y =0 =

2 0
Z b" 2  2 #
R sin (πx/a) fc
 πx 
= |E0 |2 2
+ cos2 dx
2 0 Z 0
ηf a
"  2 #
Ra 2 1 fc
= |E0 | 2
+
4 Z0 ηf

The same expression is obtained for y = b.

D. S. Weile Guided Waves

The Dissipation Rate

Adding it all together, we have

"  2 #
R |E0 |2 a fc
Pd = + (2b + a)
2 Z02 ηf

Therefore, we have the

Dissipation Constant
"  2 #
RZ0 a fc
α= + (2b + a)
ab Z02 ηf

D. S. Weile Guided Waves

The Dissipation Rate

Adding it all together, we have

"  2 #
R |E0 |2 a fc
Pd = + (2b + a)
2 Z02 ηf

Therefore, we have the

Dissipation Constant
"  2 #
RZ0 a fc
α= + (2b + a)
ab Z02 ηf

This is so obvious as to hardly require comment. (Cough!)

D. S. Weile Guided Waves

The Dissipation Rate

Adding it all together, we have

"  2 #
R |E0 |2 a fc
Pd = + (2b + a)
2 Z02 ηf

Therefore, we have the

Dissipation Constant
"  2 #
RZ0 a fc
α= + (2b + a)
ab Z02 ηf

This is so obvious as to hardly require comment. (Cough!)

What are the units of α?

D. S. Weile Guided Waves

Resonators

A resonator is a box with conducting walls, that is, a

microwave cavity.
It works because if the walls are made of perfect conductor,
only discrete frequencies can exist in the resonator.
We can use the same superposition approach to find the
modes in a cavity.
To do this, we will assume that the cavity is simply our
waveguide with metal shorting plates at z = 0 and z = c
(not to be confused with the speed of light!)
For simplicity, we work with the fundamental mode.

D. S. Weile Guided Waves

Finding the Resonant Frequency
Superposing waves traveling in the ±z directions we get
 πx   
Ey = Ey+ + ey− = A sin e−jβz − ejβz
aπx 
= E0 sin sin βz
a
For this to vanish at z = c, we need βc = π. This gives
s
 2
2πf fc
π=√ 1−
µ f

Solving this gives the

Resonant Frequency
s
1 a2 + c 2
fr =
2ac µ

D. S. Weile Guided Waves

Cavity Fields

This is the fundamental mode of the cavity, the TE101 mode. Its
magnetic field can be computed from Faraday’s Law giving

πx πz
Ey = E0 sin sin
a c
1 ∂Ey πE0 /c πx πz
Hx = = sin cos
jωµ ∂z jk η a c
aE πx πz
= √ 0 sin cos
2
jη a + c 2 a c
1 ∂Ey πE0 /a πx πz
Hz = − = cos sin
jωµ ∂x jk η a c
cE πx πz
= √ 0 cos sin
2
jη a + c 2 a c

D. S. Weile Guided Waves

Energy Storage

Poynting’s Theorem in simple media is

Ps = Pf + Pd + 2jω(Wm − We )

In an empty lossless cavity, there is no

D. S. Weile Guided Waves

Energy Storage

Poynting’s Theorem in simple media is

Ps = Pf + Pd + 2jω(Wm − We )

In an empty lossless cavity, there is no

1 power supplied,

D. S. Weile Guided Waves

Energy Storage

Poynting’s Theorem in simple media is

Ps = Pf + Pd + 2jω(Wm − We )

In an empty lossless cavity, there is no

1 power supplied,
2 power flow, or

D. S. Weile Guided Waves

Energy Storage

Poynting’s Theorem in simple media is

Ps = Pf + Pd + 2jω(Wm − We )

In an empty lossless cavity, there is no

1 power supplied,
2 power flow, or
3 power dissipation.
Therefore, one way to think of resonance is
Equality of Electric and Magnetic Energy Storage

Wm = We

D. S. Weile Guided Waves

Energy Storage
This energy is the

Average Field Energy in the Cavity

Za Zb Zc
  πx   πz  
Wm = We = |E0 |2 sin2 sin2 dz dy dx = |E0 |2 abc
4 a c 16
0 0 0

Because energy is conserved, at any time, the sum of the electric and
magnetic energy must be constant.
Thus, when the magnetic energy goes to zero, the electric energy must
be all of the energy.
Since the whole process is sinusoidal, the maximum must be twice the
average.

Total Energy Stored


W = 2We = |E0 |2 abc
8

D. S. Weile Guided Waves

Quality Factor

Total Energy Stored


W = 2We = |E0 |2 abc
8
The quality factor of a resonant system is defined as
Quality Factor
ωW
Q=
Pd

It is shown in any sophomore circuit theory book that

the quality factor is the ratio of the bandwidth to the center
frequency
the response of a resonant system is proportional to Q.

D. S. Weile Guided Waves

Quality Factors

The quality factor due to a lossy dielectric is easily computed:

Dielectric Q
ω00 |E0 |2 dv 00
RRR
Qd = =
|E0 |2 dv 0
RRR
ω0

Conductor Q is tougher, but can be computed using

perturbation. For our mode it is
Conductor Q

3
πη b a2 + c 2 2
Qc =


2R ac a2 + c 2 + 2b a3 + c 3

D. S. Weile Guided Waves

Total Quality Factor

Since
1 Pd
= ,
Q ωW
and since the dissipation is due to both the conductor and the
dielectric, we have
Total Quality Factor
1 1 1
= +
Q Qc Qd

D. S. Weile Guided Waves