Chapter 2

Transmission Lines

ECE 130a

Z L cos b l + j Zo sin b l Z + j Zo tan b l

a f
Zin - l = Zo
Zo cos b l + j Z L sin b l
= Zo L
Zo + j Z L tan b l
Examples of Transmission Lines: (Chapter 2)
Parallel Strip Line:
metal conductors

d
w

Coaxial Line:
metal
radius b
metal
radius a

dielectric

Parallel Wire Line:

Air Line Dielectric Line
D

εr
conductor radius a

Microstrip Line:
w Dielectric
Metal
d

There is a simple way to view the guided wave on a transmission line.

z a f
I z,t

a f
V z, t I az,t f
The potential difference (voltage) between the metal conductors with equal
and opposite current flowing in them are circuit concepts, except they
depend not only on time, but also on the distance z.
So we describe the wave as voltage and current waves.
a f
i t,z
+
generator v at,zf load
-
Length L >> 2 π / β = λ
Other guiding structures:
1. Waveguides -- consist of a single hollow metal tube of various cross-
sectional geometry. An EM wave propagates longitudinally inside the
hollow structure.

The wave propagation in waveguides is not transverse (not TEM). That

is, it has longitudinal field component(s). Transverse spatial depen-
dence is fairly complicated.
The propagation constant β of a waveguide wave is not equal to that of
plane waves, and velocity of propagation thus is not the same as light.
2. Optical Fibers -- are used at optical frequencies, at infrared, and at visi-
ble wavelengths. An optical fiber consists of a very thin (50-300 mm)
dielectric circular cross section cylinder. The material is usually glass or
plastic. Inner and outer portions have different dielectric constants
(index of refraction), as shown below.
core radius a, index n1

cladding, index n2

Optical fibers do not support TEM waves, like the hollow metallic guides.
Their propagation constant and mode structures are even more compli-
cated than for hollow metallic guides.
Derivation of Transmission Line Equations (1-3)
Let us consider a length of a transmission line at location z . The circuit
model is clearly a series inductance and resistance (since the whole line
with a load at the end forms a loop) and a shunt capacitance and shunt leak-
age conductance (between the good conductors, across the dielectric which
also has a small conductance). With the above circuit parameters being
L , R , C , and G , the model is
a f
i z,t L R a f
i z +D z,t
+ +
v a z,t f C G v a z +D z,t f
− −

A straight forward application of Kirchoff’s Loop Law gives

a f a f
v z + D z,t - v z,t = - L D z
a f
∂ i z,t
- R D z i z,t , a f
∂t
and Kirchoff’s Current Law at the upper node gives

a f a f a
i z + D z , t - i z, t = - G D z v z + D z, t - C D z f a f
∂ v z + D z, t .
∂t
Dividing through by ∆z and taking the limit ∆z → 0 , we get:
∂v ∂i
= - L - R i (1-5)
∂z ∂t
∂i ∂v
= - C - G v (1-6)
∂z ∂t
Sinusoidal Analysis of Transmission Lines (using phasors)
∂v
∂z
a
= - R + jw L i f (1)

∂i
∂z
a
= - G + jw C v f (2)

Differentiate (1) with respect to z :

∂ 2v di
∂z2
= - R +aj w L
dz
f
Substitute from (2) ∂ i / ∂ z :
∂ 2v
∂z2
a fa
= G + jw C R + jw L v f
Similarly, we get:
∂ 2i
∂z2
a fa
= G + jw C R + jw L i f
The Lossless Transmission Line
R = 0 (perfect conductor) G = 0 (lossless dielectric)
This is not only instructive and simple, but also a good approximation of
many real lines which are made of very good conductors (typically, copper)
and nearly lossless dielectrics (e.g., teflon). For example, a commercial
coaxial line is a nearly lossless line.

∂ 2v
= jwC◊jwLv
∂z2
∂ 2v ∂ 2i
∂z 2 c
= - w 2L C v h Also,
∂z 2 c
= - w 2L C i h
∂ 2v ∂ 2i
2 = -b v 2 = -b i
2 2
∂z ∂z
∂V ∂I ∂I ∂V
= -L = -C
∂z ∂t ∂z ∂t
The solutions are travelling waves:
e ± jβ z (with e jω t already assumed)
e j aw t - b z f or e j aw t + b z f or sum of both.

β = ω LC = ω µε
It will be shown later, when we consider transmission lines from the Electro-
magnetic Field’s point of view, that these voltage and current waves corre-
spond to EM waves, also. It will be shown that:
1
LC = µε = ,
v
where v is velocity of light in the medium between the conductors.
We can get the relationship of V and I by substituting back into the original
transmission line equations.
i = Ie a
j w t - b zf
v = Ve a
j w t - b zf

(i.e., only + z going wave)

dv di
= − jω L i = − jω C v
dz dz
− j β v = − jω L i − j β i = − jω C v

v ωL β
= = β = ω LC = ω µε ,
i β ωC
v+ L
= = Zo This is for + z going wave only.
i+ C
Zo is called the characteristic impedance of the line. For this lossless line,
Zo is real. It means that V and I are in phase. For example, power is prop-
agated. None is absorbed. (Contrast this with ordinary circuits, where real
impedance means power is absorbed and dissipated!)
Power Propagated on a Lossless Line
The power propagated can be calculated either from the electromagnetic
wave or from the voltage-current wave. It is usual to use voltage and cur-
rent for a transmission line. The power, of course, is an oscillating quantity.
We are mostly interested, however, in the average power flow. If
(real), then from the circuit picture:
1 1 Vo2
Pav = Re v i* =
2 2 Zo for lossless line
If we have only a − z going wave, the same kind of derivation we did for the
+ z going wave gives:
v−
= − Zo ( − z going wave)
i−

Circuits vs. Transmission Lines

1) Q: When does the “regular” circuit (E17) method hold correct; that is,
only e jω t dependence (no propagation) and when does the transmission
line method need to be used?
 A: Simple. When the wavelength is large compared to the dimensions,
we can neglect transmission line concepts:
2π
βz = z→0
λ
That is, whenever the distance scale (z) is small and/or the wavelength is
long; that is, z << λ .
2) For circuits: If is real (that is, Z = R ), Pav = 12 I 2 R . This is power
absorbed in the real impedance (resistance).
For a lossless transmission line: Zo is real. If we have only + going
wave:
2
1 v+ 1 2
Pav = = i + Zo
2 Zo 2
This is not absorbed power. It is power propagating in the + z direction
down the lossless line.
R = Zo
Power is absorbed in the load
⇒ to Load
where it does “something.”
If we have - going wave:
2
1 1 v− 1 2
Pav = Re v − i*− = − = − i − Zo
2 2 Zo 2
The negative sign means that power is propagating in the negative direc-
tion. Unless otherwise stated, we always will mean average power.
Transmission Line Circuits
Transmission lines are used to carry signals from one location to another.
Thus, they usually have a signal generator at one end and a load at another
end. Let’s now examine the behavior of a transmission line with load. We
use phasors and will assume lossless transmission line.
1) Transmission line terminated in its characteristic impedance. This
is called “matched.”
Zo Z L = Zo
Let’s assume we launched a + z going wave on this transmission line of
characteristic impedance Zo . Then, at any point z, we know that
v + / i + = Zo . At the load, v L / i L = Z L . (This is definition of Z L .) Thus, if
 Z L = Z o , then v L = v + ( Z = L ) and i L = i + ( Z = L ). Thus, there is no
reflected wave ( − z going wave). Rather, the wave continues into the
load, where it gets dissipated. (In practical situations, the load is a
whole circuit, for example, a detector, a radio, an oscilloscope. As long
as its input impedance is Z o , the foregoing discussion is true.)
2) Transmission line terminated in other impedance (1-7, text)

⇒ Zo Z L ≠ Zo

Assume again we launched a signal at some frequency. Then,

v + / i + = Zo . But at the load, v L / i L = Z L . If Z L ≠ Zo , v L cannot be v +
and i L cannot be i + . The boundary condition at the load can only be
satisfied if there is also a reflected wave v − , i − . This wave propagates
backward in the − z direction. Then, at any point on the line:
v = V+ e a f+ V e a f= v + v
j wt-bz j wt+bz
- + -
V j w t - b z f V- j aw t + b z f v−
i= + e a - e = i+ + i- i− = −
Zo Zo Zo
Note the − sign on the current of the reflected wave! Now we can satisfy
the condition at the load v / i = Z L . Only one value of V− will satisfy.
Let us define:
v−
= Γ , the reflection coefficient
v+

Then:
v + + Γ v + = v (1)
v+ v v
−Γ + = (2)
Zo Zo Z L
We solve for Γ to get: (Problem: Solve to G )
Z L − Zo
Γ=
Z L + Zo

Note: Γ is complex, in general, because Z L may be complex.

-1 £ G £ 1
Fraction of the incident power reflected back:

Incident Power LM 1 v i* OP = Re LM 1 v / Z OP = 1 V / Z
Pi = Re
2 2

N2 Q N2 + +
Q 2 + o + o

P = Re LM v i* OP = Re LM G v G* i* OPa- 1f
1 1
Reflected Power r
N2 Q N2 - -
Q - +

= Re LM- G
1 *O 2 2

N 2 v i PQ = - G P + + i

P = Re LM v i* OP = Re LM a1 + G f v e1 - G* j i* OP
1 1
Total Power T
N2 Q N2 T T
Q + +

= Re LM e1 + G - G* - G G* j v i* OP
1
N2 Q + +

= 1− Γ Pi ≡ Transmitted power (into Load)

2

1
Note: G - G * = Im G is purely imaginary.
e j
2

3) Special Cases of Termination

A special case of the general load is Z L = 0 , that is, a short circuited trans-
mission line. (A transmission line may be shorted deliberately in order to
achieve a certain result, or it may be shorted accidentally, something that
has to be repaired.)
For this case:
Zo Z
Z
Γ = − o = −1 ZL = 0
Zo
Z=0
~
We can say here that Γ = Γ e jϕ = 1 e + j π here.
If v + = A e j aw t - b z f ,
af
v z = Ae - jb z - Ae jb z ce jw t
understood f
1
af
i z =
Zo
A e - jb z + A e j b z

Note that this time, the z dependence has two different parts. Thus, the z
dependence cannot be lumped together with e jω t dependence, as it can be
for a wave traveling in one direction only. Now, the entire V or I, including z
dependence, is the phasor:
af
v z = -2 jA sin b z
ce jw t
understood γ
2A
af
i z =
Zo
cos b z

The real voltage and current are ( for A = A e jθ )

a f m
v z ,t = Re - 2 j A e jq sin b z e jw t r
= 2 A sin b z sin w t + qa f
R2 A e
i a z ,t f = Re S
jq
cos b z e jw t
UV
T Z o W V
2 A
=
Zo
a
cos b z cos w t + q f
z
Conductor
I

Both voltage and current oscillate sinusoidally in time with different maxi-
mum values (amplitudes) at different locations. This is known as a standing
wave--“standing,” because the amplitude remains the same at each location
and the oscillating pattern is standing. Contrast this with “travelling” wave,
in which a given point on the wave form progresses in distance with time.
The amplitude of a travelling wave, however, is a constant value regardless
of location.
The instantaneous power flow is (must be calculated from real V, I ):
a f a fa f
P z,t = v z,t i z,t
2
4 A
=
Zo
a f a
sin b z cos b z sin w t + q cos w t + q f
2
A
=
Zo
sin 2 b z sin 2 w t + q a f
The average power flow (we calculate it here direction from phasors) is:
L 1 *O L 1
Re M V I P = Re M - 2 j A sin b z ◊
2 A* O
cos b z P
N 2 Q MN 2 Z o PQ
= Re 2 j A sin β z cos β z = 0
2

(purely imaginary)
This makes physical sense, since no power flows into the short circuit.
Thus, all the power must be reflected back, giving net power flow equal to
zero (average). The same result, of course, could be derived from the field
point of view. The short circuit is a perfectly conducting plate closing off the
line. No fields penetrate into it. Thus,
Pav = ez P+ − ez P− = 0
The Concept of Input Impedance ( Zin )
Characteristic impedance Zo is the ratio of voltage to current of one wave
direction.
v+ v
Zo = = −
i+ i−
Since transmission lines are used in conjunction with lumped circuits, we
must be able to treat them as regular circuits as well. The input impedance
of any circuit is Zin = V / I at the terminals. Thus, for a transmission line, at
any point on it:
vT
Zin = ,
iT
where v T and i T are the total phasor fields at that point.
Important: Since Zin = v T / i T , this is now the same as a regular circuit con-
cept. So we can always replace a transmission line circuit with its input
impedance!
If there is only a single wave propagating in one direction (transmission line
terminated in Z L = Zo ), then, of course, Zin = Zo . But, in general, Zin ≠ Zo ,
and Zin is a function of both the load and of the position z on the line.
Let’s measure as follows: Z = 0 at the load. Therefore, z is negative on
the line z = − l .
For the short circuited lossless transmission line:

a f viaa--llff = -22Aj A sin b a- lf

Zin - l = z
cos b a- l f
Zo z = −l z=0
2p 2p f
a f
Zin - l = j Zo tan b l = j Zo tan
l
l = j Zo tan
v
l (for a shorted line)

The input impedance of a short circuited lossless line is purely imaginary (as
it should be, since it consists of distributed capacitance and inductance.)
The values of the input impedance vary with the distance if the frequency is
fixed or varies with frequency if the length is constant.

Reactance can be either capacitive or inductive, and its absolute value var-
ies from zero to infinity. This property can be utilized in a number of ways.
(a) Easily adjustable values of reactance.
(b) Location of an accidental short on a line. (Problem)
(c) Matching of a load not equal to Zo (will study later)

Open Circuited Lossless Line

In practice, we try to avoid open circuiting a transmission line. Reason:
Theoretically, open circuit (i.e., open ends) means infinite load impedance
( R L = ∞ ). This would mean that:
∞ − Zo
Γ= = +1
∞ + Zo
That is, all of the wave is reflected, resulting in standing waves.
However, in practice, this is not really true. The open end is not a truly infi-
nite impedance. While no current flows beyond the open end, some of the
electromagnetic wave goes out the open end and becomes a free space
wave. Only a small fraction does, since the open circuited transmission line
is “poorly matched” to space (i.e., it is a poor antenna). Yet, if not all the
power is reflected, we obviously do not have infinite impedance.

Lossless transmission line terminated in a real load RLπ Z01

A
Z01 Z02 = RL RL

Z01 RL
A
The above two cases are identical as far as T.L. 1 is concerned, since the
impedance at point A is RL for both.
We can always replace any transmission line by its equivalent input
impedance. The fields up to that input point will be unchanged.

For the above, since Z02 ≠ Z01 :

RL − Z 01
Γ=
RL + Z 01
Γ is real and its value varies from −1 to +1 .
1 A 2
Z02 = RL Zin A = RL

Line 1 has both + z and − z going wave.

Line 2 has only + z going wave, since it is properly terminated.
Just to the left of junction A: v = v1+ + v1−
Just to the right of junction A: v = v 2 +
These are obviously equal. So we replace at A by RL :
1

v 2 + v1+ + v1− v
τ= = = 1 + 1− = 1 + Γ
v1+ v1+ v1+
τ is called the transmission coefficient. Thus:
2 ZL
v 2 + = τ v1+ τ = 1+ Γ =
Z L + Zo
τ can be greater than unity, but that does not mean more than 100% of
power transmitted.
Power balance: Pr + Pt = Pi 0£ t £2
v12− v2
Pr = = Γ 2 1+ = Γ 2 Pi
2 Zo 2 Zo
v 22+ τ 2 v12+
= τ 2 1+ ⋅ o
v2 Z
Pt = =
2 ZL 2 ZL 2 Zo Z L
Substituting for Γ and τ and writing the sum of the reflected and transmit-
ted power = incident:

v12+ LMF Z - Zo IJ + 4 Z FG Z IJ OP = v
2
2 2

MNGH Z
L L o 1+
Zo L + Zo K b Z + Z g H Z K PQ Z
L o
2
L o

ZL2 + Zo2 - 2 ZL Zo + 4 ZL Zo
=1
b
ZL + Zo
2
g
1 = 1 Q.E.D.
Example:
Zin = 300 Ω
z=0
50 Ω 300 Ω Z L = 300 Ω

A
Commercial coaxial line (Radio Shack) is 50 Ω , flat antenna line is 300 Ω .
The above situation would happen if you connect them together without a
matching transformer. (We shall learn later what a matching transformer is.)
a f
At the junction A z = 0 :
Zin − Zo 300 − 50
Γ= = = 0.714
Zin + Zo 300 + 50
τ = 1 + Γ = 1714
.
The fraction of power reflected is Γ 2 = 0.51 .
The fraction of power transmitted to 300 Ω line and on to the load is
1 − 0.51 = 0. 49 .
General Transmission Line
Now let’s return to a general case and find the total voltage and current on
the line as a function of distance:
c
v 1 = V1+ e - jb z + G V1+ e + jb z = Vo e - jb z 1 + G e 2 jb z h (Let V1+ = Vo )
V1 + - jb z V V
i1 =
Zo
e
Zo Zo
c
- G 1 + e + jb z = o e - jb z 1 - G e 2 jb z h
e jω t variation understood! We have dropped e jω t .
The real fields, of course, vary sinusoidally in time. v1 and i1 are the com-
plex phasor amplitudes. Since, for a complex voltage v = Ae jω t , the maxi-
mum value (or sinusoidal amplitude) is A , and the sinusoidal amplitude of
V1 along the line is:
a
v 1 = Vo e - jb z 1 + G e 2 jb z = Vo 1 + G cos 2 b z + j sin 2 b z f (for G real)
= Vo a1 + G cos 2b zf + G
2 2
sin 2 2b z = Vo 1 + G 2 + 2 G cos 2b z
Remember that Γ may be positive or negative.
4π
2β z = z
λ
The minimum value of this is (if Γ is positive):
v1 min = Vo 1 + G 2 - 2 G = Vo 1 - G a f
For a general Γ : v1 min = Vo 1 - G b g
Similarly: v1 = Vo 1 + G 2 + 2G = Vo 1 + G
max
b g
The location of consecutive minima and maxima are λ /2 apart. At the load,
we have either a maximum or a minimum (max. for Γ positive). The shape
of the function vs. z is not a pure sinusoid. The minima are not infinitely
sharp (as they are for a short circuit).

Load

z
An A.C. voltmeter, of course, does not measure the time variation, but only
the amplitude (top solid line). (Actually the RMS value)
This is called a partial standing wave. An important measurable quantity is
the Standing Wave Ratio (SWR) = S.

Vmax 1+ Γ
S= =
Vmin 1− Γ
S can range from 1 to • .
Example: For the previous example,
1. 714
S= =6
0. 286
We note also that for Γ positive ( Z L > Zo , Z L real),
Z L − Zo
1+
Z L + Zo 2 Z L Z
S= = = L
Z −Z
1 − L o 2 Zo Zo
Z L + Zo

and for Γ negative ( Z L < Zo , Z L real),

Zo − Z L
1+
Z o + Z L 2 Zo Z
S= = = o
Z −Z
1 − o L 2 ZL ZL
Zo + Z L
This also agrees with our example: 6 = 300 /50
If a transmission line is connected to an unknown load, the SWR can be
measured and, from this, the load value found. But to fully understand how,
we must consider a more general load.
Lossless Transmission Line Terminated in a General Complex Load ZL
Most practical circuits/devices (whether the device be a transmitting
antenna, a TV, a receiver, a microwave amplifier, an oscilloscope, etc.) have
complex input impedance. Thus, if such a device is fed by a transmission
line, the transmission line sees a complex load. Without something special
being done, the line is not matched to the load. There is a reflected wave.
 Z L − Zo
Γ= (complex, in general)
Z L + Zo
v T = V+ e − jβ z + Γ V+ e + jβ z
V+ − jβ z V
iT = e − Γ + e + jβ z
Zo Zo
Assume for simplicity that v + = Vo ∠0° and that the load is at Z = 0 . There-
fore, on the transmission line, z = − l . Then,
v T = Vo e jβ l + Γ Vo e − jβ l
Vo jβ l V
iT = e − Γ o e − jβ l ,
Zo Zo
where l is the distance measured backwards from the load.
The input impedance at a point l distance back from the load is:

e jβ l + Γ e − j β l
Zin = Zo jβ l
e − Γ e − jβ l
1+ Γ e − 2 j β l
= Zo
1 − Γ e − 2 jβ l
Substituting the value of Γ and rearranging, we get the alternate forms:
Z L cos b l + j Z o sin b l Z + j Z o tan b l
a f
Z in - l = Z o
Z o cos b l + j Z L sin b l
= Zo L
Z o + j Z L tan b l
When Z L = Zo , we get Zin = Zo .
Now returning to the voltage and current:

jβ l − 2 jβ l 1 + Γ e − 2 jβ l
v = Vo e 1 + Γ e Zin = Zo
1 − Γ e − 2 jβ l
V
i = o e jβ l 1 − Γ e − 2 jβ l another form
Zo
This is the same expression as before ( z = − l ), except now Γ is complex
Γ = Γ ∠θ r . It is easiest to see how this changes the standing wave by con-
sidering phasor diagrams. The amplitude of the quantity in square brackets
determines the amplitude of the standing wave, since e jβ l = 1 .
Definition of Generalized Reflection Coefficient
V- (l) V- (0) e - jb l
G(l) = = = G(0) e - 2 jb l z = -l
V+ (l) V+ (0) e + jb l
af
In my notation G = G 0 , reflection coefficient at the load (complex quantity
~
in general) is equivalent to the book’s Γr .

Note:
1) Γ is real if load is real and line is lossless. Γ is complex if load is com-
plex.
af
2) G l is always complex, except at certain points. These points have spe-
cial significance!
3) Gal f ∫ G , that is, moving about on the line only changes the phase
angle of Galf .
Im
At l = 0 (moving back,
away from load)
1+ Γ
Vmin.
Γ Vmax.
Re
Max.
−Γ
1− Γ

It is obvious from this phasor diagram that the cases of real load and com-
plex load are not fundamentally different. θ r = 0 or 180° for a real load, but
can be anything for a complex load. The phasor rotates as l increases,
making a complete cycle for 2β l = 2π ; that is, 4π l / λ = 2π , or l = λ /2 .
Vmax is still 1+ Γ and Vmin = 1− Γ , where Γ = Γ . Obviously, the shape of the
standing wave is also the same as for a real load, except the maximum or
minimum no longer occurs at the load.

Location of maxima and minima: Maximum occurs at:

θ r − 2β lmax = 0, ± 2π , ± 4π , etc.
Similarly, θ r − 2β lmin = ± π , ± 3π , etc.
Example: An RF signal is sent on a 300 Ω parallel wire transmission line,
which has ε r = 5 , to a receiver. A VSWR of S = 3 is measured on the line.
The distance between two minima is 0 . 67 m. and the nearest minimum to
the load minimum is 0 . 45 m. away. Find the frequency of the signal and the
The’vnenin equivalent circuit of the receiver input.
0. 67 m. 0. 45 m.

Zo = 300 Ω
ZL
S=3
Solution:
λ m / 2 = 0. 67 λ m = 1. 34
λ = λ m ε r = 1. 34 ⋅ 5 = 3 m.
c 3 × 10 8
f = = = 10 8 = 100 MHz
λ 3
2π 2π
β= = = 1. 49 π m − 1
λ m 1. 34
− 2 β lmin + θ r = π ∴ 2. 98π lm + π = θ r
2. 34π = θ r = 0 . 34 π = 61. 2°
1+ Γ S −1 2
S= ∴Γ = = = 0.5
1− Γ S +1 4
Γ = 0 . 5 ∠61. 2° = 0 . 24 + j 0 . 44
Z −Z Z − 300
Γ= L o = l
Z L + Zo Z L + 300
a f a f
ZL 1 - G = 300 1 + G
a f a
ZL 0 . 76 - j 0 . 44 = 300 1. 24 + j 0 . 44 f
1. 24 + j 0. 44 1. 32 ∠19. 5°
Z L = 300 = 300 = 450 ∠ 49. 5°
0 . 76 − j 0 . 44 0. 88 ∠− 30 °
342
= 292 + j 342 Ω = R + jω L L= = 0. 54 × 10 − 6 h
2π × 10 8

Load is R = 292Ω L = 0. 54 µ h ANS.

Looking at the definition of Zin and at the rotating phasor diagram of the pre-
vious page, we also see that:
At the location of a voltage maximum:
1+ Γ
Zin = Zo = Zo S
1− Γ
At the location of a voltage minimum:
1- G Z
Zin = Zo = o
1+ G S

Zin = Zo S at voltage maximum

Zo
Zin = at voltage minimum
S
This is an important observation. It will be used soon for matching.
Matching a transmission line to an unmatched load
Reflections on a transmission line are undesirable for a number of reasons.
For example, digital signals from PCM reflected pulses can bounce back
again from the sending end if that end is also unmatched. Then false sig-
nals can be detected.
Q: How do we match an unmatched load?
A: There are several methods. We now study the first one.

I. Quarter Wave Matching ( λ /4 “Transformer”)

Suppose we have a transmission line terminated in a real load Z L ∫ RL b g
Zin ⇒ Z02 Z02 is intrinsic impedance.
ZL
Z022
If the length of this line is exactly l = λ /4 , then Zin = , since
ZL
2π λ π
sin β l = sin ⋅ = sin = 1
λ 4 2
and cos β l = 0
This then offers a method to match this load to any lossless transmission
line Z01 , intrinsic impendance Z0 to be matched to ZL . Z L real RL b g
A
2
Z 01 1 Z 02 ZL
λ
4
We insert a transmission line of λ /4 , length ( 2 ), and having Z 02 = Z 01⋅ Z L .
Then, at ( A ):
Z 022 Z 01 Z L
Zin = = = Z 01 !
ZL ZL
Since Zin A = Z 01 , we are matched. No reflection, since at point A, we could
replace the whole combination by Zin A .

Z 01 Zin = Z 01

How is this possible physically? We won’t go into a detailed analysis of

what really happens to the right of point A. But briefly, we do have (in actual
physical fact) reflection of waves at point A and at ZL. The two reflections,
however, cancel each other out on the line 1 .
Q: This obviously works if the load is real, since Z01 is also real, resulting in
real Z02 . But if the load is complex, then what to do?
A: Recall that the input impedance at a voltage minimum or maximum is
real Zo S and Zo / S . Thus, these are the steps for matching a complex load
ZL to a line Z01 .
1) Add a section of Z 01 line to the load. The length should be lmax or lmin of
the standing wave that results. Say, we use lmax :
lmax

ZL

This is equivalent to a load of Z L = S Z01 .

2) Now use λ /4 , matching on this equivalent load. For example, insert a
line:
Z 02 = Z L Z 01 = S Z 01 Z 01 = S Z 01, length λ /4

Z 01 B A Z 01 Zo
λ /4
lmax
The impedance at point B then is:
S Z 012 S Z 012
Zin B = 1 = = Z 01
ZL SZ 01
Thus we are matched. No reflection going back on line 1. Same com-
ment applies as before. Actually, there are reflections at points A and B, but
they are 180∞ out of phase and cancel out.
The 50W - 300W “transformer” supplied for TV and video components is
a lumped circuit simulation of a λ /4 transformer!
2π λ π
Important Note: l = λ /4 means β l = ⋅ =
λ 4 2
2π f 2π
The matching is only valid at one frequency! Since β = = , if f is
ν λ
changed, the wavelength is changed and β l ≠ π /2 for the same l.
Example: (a) Match a 50 coaxial cable to a 300W real input impedance
device at a frequency of 180 MHz. Assume all transmission lines used
have ε r = 6 . (b) Find the reflection coefficient of the matched circuit at 90
MHz.
(a) At 180 MHz:
c 3 × 10 8
λo = = = 167
. m.
f 18 . × 10 8
λ
λ on the T.L., λ m = o = 0.68 m.
6
T.L. impedance needed Z 02 = 50 ⋅ 300 = 122.5 Ω ans.

0.68
length needed = = 0.17 m. ans.
4
50 Ω 122.5 Ω ZL = 300 Ω

λ = 0.17 m.
4
We cannot readily buy a 122.5W transmission line. However, any transmis-
sion line at these R.F. frequencies can be simulated by a proper lumped cir-
cuit network. (In this class, we won’t study how.)
(b) If the frequency is changed to 90 mHz:

λ m = 0.68 × 2 = 1. 36 2π π
βl = × 0.17 =
1. 36 4
We have:
π
122.5Ω βl =
50Ω 300Ω 4
tan β l = 1
l=λ
8
300 + j 122.5 × 1 324 ∠22.2°
Z in A = 122.5 = 122.5
122.5 + j 300 × 1 324 ∠90 − 22
= 122.5 ∠ − 45.6° = 85.8 − j 87.5
Reflection coefficient at A is:
85.8 − j 87.5 − 50 33.5 − j 87.5
Γ= =
85.8 − j 87.5 + 50 135.8 − j 87.5
937. ∠− 69°
= = 0.58 ∠− 36.2° ans.
161.5 ∠− 32.8°
Fraction of power reflected back at A is Γ 2 = 0. 34 .