Applied Electromagnetics: Early Transmission Lines Approach

Chapter 2
Transmission Lines
Applied Electromagnetics:
Early Transmission Lines Approach
Stuart M. Wentworth
Applied Electromagnetics: Early Transmission Lines Approach Chap 2-1

Wentworth 10/5/07, Wiley
Learning Objectives
• Develop equations for wave propagation on a transmission line
and define characteristic impedance and propagation
constant
• Investigate wave reflection from terminated transmission lines
and define input impedance and standing wave ratio
• Introduce the Smith chart, a graphical tool for the study of
transmission lines, and use it to develop impedance
matching networks
• Study the behavior of transient signals on a terminated
transmission line
• Investigate the dispersion of a signal pulse as it travels along a
transmission line

Ideal wires and transmission lines
Connecting wires must be considered as transmission line if
their length is a significant fraction of a wavelength

Transmission line examples

Distributed parameters for a section of T-Line
The general model for T-Line resembles twin-lead cable

Distributed parameters at high frequencies
Coaxial Cable Twin-Lead Cable
R’ (/m) 1 1 1  f 1 f
  
2  a b  c a c
L’ (H/m)  b   d 
ln   cosh 1  
2  a    2a 
G’ (S/m) 2 d  d
 
ln b
a 
cosh 1 d
2a 
C’ (F/m) 2 
 
ln b
a 
cosh 1 d
2a 

Distributed parameter model
Circuit analysis of the distributed parameter model yields
equations for wave propagation (the Telegraphist’s Equations)

Telegraphist’s Equations
i ( z, t )
Kirchhoff’s voltage law: v ( z, t )  v ( z  z, t )  i ( z , t ) R ' z  L ' z .
t
Divide both sides by z, and take the limit as z approaches zero:
v ( z , t )  v ( z  z , t ) i ( z , t )
lim  i ( z, t ) R '  L ' .
z  0 z t
The limit on the left is the definition of a derivative, so

v ( z , t ) i ( z , t )
  i ( z, t ) R ' L ' .
z t
Likewise, by Kirchhoff’s current law Telegraphist’s

v ( z  z , t ) Equations
i ( z , t )  i ( z  z , t )  v( z  z , t )G ' z  C ' z .
t
i ( z , t ) v ( z , t )
Leading to   v ( z , t )G '  C ' .
z t

Time-harmonic waves on T-Line
If v( z , t )  V ( z ) cos(t   ),
then v( z , t )  Re V ( z )e 
j t  
  Re V ( z )e jt 
   s 
where the phasor is Vs ( z )  V  z  e j (likewise I s ( z )  I  z  e j )
v( z , t ) i ( z , t )
Phasor time derivative:  jVs ( z ) and  j I s ( z )
t t
Telegraphist’s equations for time-harmonic fields are then
dVs ( z ) dI s ( z )
   R ' j L ' I ( z ) , and    G ' jC ' V ( z ) .
dz dz

The traveling wave equations
Taking the position derivative of both sides of the Vs(z)
expression leads to
d 2Vs ( z )
  2Vs ( z )  0,
dz 2
wherepropagation constant
   R ' j L '   G '  j C '     j  .
Solving the differential equation, we arrive at a general solution

Vs ( z )  Vo e  z  Vo e   z .
The instantaneous form is

v( z , t )  Vo e  z cos  t   z   Voe  z cos  t   z  . traveling
In a like manner, we find wave
i ( z , t )  I o e  z cos  t   z   I oe  z cos  t   z  . equations

Characteristic impedance Zo
Given the traveling wave equations:
v( z , t )  Vo e  z cos  t   z   Voe  z cos  t   z 
i ( z , t )  I o e  z cos  t   z   I oe  z cos  t   z 
Characteristic impedance Zo is the ratio of voltage to current wave

amplitudes in one direction, or
Vo Vo
Zo     
Io Io
Inserting the wave equations into the phasor-form Telegraphist’s

equations we find
R ' j L '
Zo 
G ' jC '
L'
For low loss line, this becomes Z o 
C'
Low-loss and lossless line
Low loss lines feature high conductivity conductors and low loss
dielectrics such that R’ << L’ and G’ << C’. For this case,
L'
  j L ' C '    j  and Z o  .
C'
phase constant    L ' C '
 1
propagation velocity u p  
 L 'C '
For low-loss coaxial cable, we find
c 60  b 
up  and Z o  ln   
r r  a 

Example 2.1 – Coaxial Cable
A 1.0 mm diameter copper wire is surrounded by a 1.0 mm
thickness of Teflon, then jacketed by copper. Assuming this
coaxial cable is lossless, we want to find the propagation velocity,
up, and the characteristic impedance, Zo.
For Teflon, r = 2.1 from Appendix E, so
m
3x108
c s  2.1x108 m
up  
r 2.1 s
For Zo, we have a = 0.50 mm and b = 1.5 mm, so
60  b  60  1.5 
Zo  ln    ln    46 .
r  a  2.1  0.5 

Power Transmission
average power for a time- 1
harmonic wave: Pave  z   Re Vs I s* 
2
A +z propagating voltage   j z  j  j  z
wave on a lossless line: Vs  Vo e  Vo e e
where Vo has a magnitude ( Vo ) and a phase ().
Is is in phase with Vs for a Vo

lossless line, so I s  I o e j e  j  z  e j e j  z
Zo
2
1   j  j  z Vo  j  j  z  1 Vo
 

thus Pave ( z )  Re  Vo e e e e 
2  Zo  2 Z o

For a lossy line, Is is no longer in phase with Vs so

2
1 Vo
Z o  Z o e j , leading to Pave

( z)  e2 z cos 
2 Zo

Example 2.2 – power transmission in coax
A copper conductor coaxial cable (a = 1.0 mm, b = 3.0 mm) with thick outer
conductor is filled with a nonmagnetic dielectric with r = 5.0 and tan = 0.00010
measured at 2.0 GHz. Find power lost per meter length of this cable.
We will find in chapter 6 that the loss tangent is related to an effective

S
conductivity, eff, of the dielectric:  eff   tan   55.6 x 106 .
m
This can be used in the Table 2.1 equation for R’.
From the table, R’=2.48 /m, L’=220nH/m, C’=253pF/m and G’=318S/m.
Then
1 Np
   R ' j L '  G ' jC '   0.0467  j 93.73 , so   0.0467
m m
We don’t know the voltage amplitude, so we calculate the power loss as a ratio:

Pave ( z  1m)

 e 2 (1m )  0.91
Pave ( z  0m)

Decibels (dB)
Decibels
- a logarithmic scale useful for expressing very large or very small ratios
P 
- gain in decibels: G ( dB )  10 log  out .
 Pin 
- The 0.91 ratio of Example 2.2 becomes G(dB) =10 log(0.91) = -0.4 dB
(note that a negative gain is an attenuation)
- To express absolute power in the decibel scale, a reference power is


needed. The dBm scale is referenced to 1 mW: G ( dBm )  10 log 
P .

 1 mW 
- Decibels are related to Nepers by 1 Np  8.686 dB .

T-Line terminated in impedance ZL
at z = 0 we have
Vs ( z  0) Vo e  (0)  Vo e   (0) Vo  Vo

ZL     (0)    (0)
  
I s ( z  0) I o e  Io e Io  Io
Vo Vo Vo  Vo

Since Z o     , then Z L  Z o 
Io Io Vo  Vo

Voltage wave incident on terminated T-Line
Z L  Zo
Rearranging, Vo  Vo   LVo
Z L  Zo
Where L is the reflection coefficient and is a measure of the impedance

mismatch at the load.

Voltage standing wave ratio (VSWR)
Vmax
VSWR 
Vmin
1  L

1  L

Equivalent input impedance for a terminated line
Looking into the line at z = -l, input impedance Zin is
V ( z  l ) Vo e  l  Vo e  l
Z in     l   l
Zo.
I ( z  l ) V e
o
V e
o
Z L  jZ o tan(  l )
For a lossless line, Z in  Z o
Z o  jZ L tan(  l )

Impedance values for R, L and C
Complex loads may be modeled using resistor, inductor

and capacitor lumped elements

Standing wave pattern for shorted load
 L  1
VSWR  
for lossless line

Z in  jZ o tan  l

Standing wave pattern for open load
 L  1
VSWR  
for lossless line

Z in   jZ o cot  l

The complete circuit
Z in
Vin  VSS
Z S  Z in
 V ( z  -l ).
 
V ( z )  Vo e   z  Vo e   z  Vo e   z   L e   z .
 Vin
V 
o
e   l   L e  l
VL  V ( z  0)  Vo  1   L  .

Example 2.3 circuit: find voltage across the load
ZL  Zo
L 
ZL  Zo
100  50 1
  .
100  50 3
convert source voltage to phasor: VSS  10e j 30

V.
Z L  jZ o tan(  l )
for a lossless T-Line, Z in  Z o .
Z o  jZ L tan(  l )
2 
  Z o2
Since  l   , tan  l  tan   , and Z in   25.
 4 2 2 ZL
(continue Example 2.3)
Z in  25 
Vin  Vss  10e j 30

 5e j 30
V.
Z s  Z in 25  25
For the lossless case,  = j and l = j l = j/2. So


j 30
Vin 5e

  j 60

V   7.5e V.
1 e
o  l  l
  Le  j 90
 
e e j 90
3
VL  V o

 1   L   7.5e  j 60

 1 1
3   10e  j 60

V.
Converting to instantaneous form: v L  10 cos  t  60 V .  

ML2.3: When must wires be treated as T-Line?
Compare VL vs line length for T-Line and lumped element models
Z IN 1 RL  Z o
T-Line: VL  Vo  1   L  where Vo  Vs j 2 d  j 2 d
and  L  .
Z IN  Rs e  Le RL  Z o
 j  jZ o L'
Lumped element: j L  j 2 dZ o and  where Z o 
C 2 d C'
j Z1
Letting Z1  RL and Z 2  RS  j L , VL  Vs
C Z1  Z 2

Cont. ML2.3
Our example:
RL = RS = 200 , and Zo = 50 .
 connecting wire begins to impact

load voltage, especially phase, at
less than 0.01.
 appreciable difference between

the lumped and the distributed
element model begins at about
0.10.

The Smith Chart
• “Transmission line
calculator” created in
the 1930s by Philip
H. Smith
• Most useful for

lossless T-Lines
• Used to find
impedance and
reflection coefficient
at any point along a
terminated T-Line

Equivalent terminated T-Lines
Input impedance at an
arbitrary point along a T-
Line can be used to create an
equivalent terminated T-
Line

Creating the Smith Chart
ZL
 Normalized load resistance z L 
Zo
Z L  Zo zL  1
 L  
Z L  Z o zL  1
 At a point z along the T-Line,

Z IN
o Normalized impedance z IN 
Zo
j 2 z z IN  1
o    Le    Re  j IM
z IN  1

Normalized resistance and reactance circles
2 2 2 2
 r   1   1 1
  Re  1
2
  Re     IM      IM    
2

 r 1   r 1  x x

Smith Chart from MATLAB 2.4
j1
j0.5 j2
j0.2
0
0.5 1 2 4
-j0.2
-j0.5 -j2
-j1

Using the Smith Chart – normalized load
Smith Chart is a plot of normalized impedance: zL = r + jx

Normalized load
To locate
zL = 1 + j2
on the Smith Chart,
find the intersection
of the r = 1 circle and
the jx = +j2 “circle”
Drill 2.11: locating points on Smith chart
Locate the points on the
Smith chart if Zo = 50 for
the following loads:
(a) ZL = 0
(b) ZL = ∞
(c) ZL = 100 + j100 

(d) ZL = 100 – j100 
(e) ZL =50 

Reflection coefficient
 Smith Chart is a plot

of reflection coefficient:
   Re  j Im   e j
 At z = 1 + j2, we see from

the Smith Chart
  0.7 and    45
Moving along a T-Line
• movement along T-Line

corresponds to movement
along a constant || circle on
the Smith Chart
•One revolution of the

constant || circle corresponds
to a half wavelength travel
Impedance measurement
Probe measures electric

field amplitude
4
Short
3.5
3
Load impedance
Probe Voltage
found by 2.5
comparing its 2
field pattern with 1.5

Load
that of a short 1
termination 0.5
0
40 35 30 25 20 15 10 5 0
scale (cm)

Exp 2.4: load impedance measurement
1. Determine G and
voltage minima location
for shorted line
2. Determine VSWR and

draw || circle for load
3. Move from load minima

(a) to reference location
of load (b)
4. In the example,
zL=0.8 + j1.0, or
ZL = 40 + j50 
Exp 2.5: load impedance measurement
• Calibration short
determines length of
connection l to the
device to be tested
• moving from
measured impedance
ZinL a distance l
towards the load
finds ZL

Continue Exp 2.5 (50  T-Line)
• we measure ZINSC =+j128.
We normalize this and
locate point a
• the length of the line, l, is

from the location of a short
(b) to point (a): l = 0.191
• next, we measure ZINL = 30

–j 40  and locate (c)
• finally, we move a distance

l towards the load (d), and
find ZL = 65 + j65 .
Impedance matching
•  looking into a matched load is zero,

so all wave power enters the matching network
• if matching network only consists of reactive components (L,

C, lengths of lossless T-Line) then no power is dissipated in the
network and therefore all power is received by the load
Quarter-wave transformer
 Useful for resistive terminations
RL  jZ S tan  l
 Z IN  Z S
Z S  jRL tan  l
 At l = /4, l = /2 and tan l = ∞
Z S2
 Z IN   Z o , or ZS  Z o RL
RL
Impedance matching with Smith chart
Move from normalized load
impedance zL to the center of the
Smith chart where || = 0.
 Move along the constant ||

circle by adding sections of T-
Line

 Reactive components add ±jx

Exp. 2.6: simple matching network
Normalize load impedance:

zL = 0.22 + j0.50
Locate and draw || circle

Continue Exp.2.6
• Move along constant ||

circle a distance d from zL to
point 1 + j2
• Add -j2 = -j/(CZo)

Admittance of shunt stubs

Normalized admittance
• zL = 2 + j1
• yL = 1/zL = 0.4 – j0.20
This corresponds to a
point on the constant ||
circle opposite zL

Shorted stub admittance
• Zin = jZo tan(d)
• Any value of reactance can

be achieved by proper choice
of stub length

Shunt stub matching
(shorted stub)
• Locate zL and draw || circle

yshort
• Locate yL
• Move from yL towards
generator a distance d to reach yd
= 1 + jb circle
• Move from yshort towards
generator a distance l to 0-jb
• ytot = 1 (line is matched)
Shunt stub matching
(open ended stub)
• Locate zL and draw || circle

• Locate yL yopen
• Move from yL towards
generator a distance d to reach yd
= 1 + jb circle
• Move from yopen towards
generator a distance l to 0-jb
• ytot = 1 (line is matched)
Transients
 Transient voltages on a T-Line can be

modeled by a bounce diagram
Ro
 Initial voltage launched: Vo  VS
RS  Ro
l
 Transit time tl 
up
R  Ro R  Ro
 L  L , S  S
RL  Ro RS  Ro
 Using the bounce diagram, voltage at an
arbitrary point (say –l/2) at an arbitrary
time (say 2.5 tl ) is
V ( l 2, 2.5tl )  Vo (1   L   L  S )
Exp. 2.9
RL  Ro 125  75 1
 L   
RL  Ro 125  75 4
RS  Ro 25  75 1
 S   
RS  Ro 25  75 2
l 0.06m
tl    2ns
u p (0.1)(3 10 m / s)
8

Continue Exp. 2.9
The reflections eventually settle out, and in

RL
this example: V  VS  3.33V
RL  RS

Pulse input circuit

Bounce diagram for pulse response

Exp. 2.10: 4V pulse on circuit of Exp. 2.9

Triangular pulse
• Bounce diagram is
difficult to use for pulses that have sloped edges
• MATLAB 2.5 traces voltage at an arbitrary point on the line
for
a triangular pulse
• Program
Applied may be modified
Electromagnetics: for other
Early Transmission pulse shapes
Lines Approach Chap 2-59
Practical Application: Schottky-diode termination
 Impedance mismatch in digital logic is often countered by using
Schottky-diode terminations
 Signal reflections are reduced by clamping action of the diodes
 D1 clamps to Vcc: reflections causing voltage greater than Vcc
will be shunted to Vcc
 D2 clamps to ground: reflections causing negative voltage will be
shunted to ground

Reactive loads
 vL (t )   Voi  Voi (t )  U ( )
0 for   0
where U( )   and   t  tl
1 for   0
diL (t )
 vL (t )  L
dt
diL (t ) 2Voi
 L
dt
 Ro iL (t )  2Voi  iL (t ) 
Ro

1  e  Ro L
 U ( )
Inductive load
2.5
voltage at source end

2
1.5
0.5
 
VS (t )  Voi  Voi 2e  Ro L  1 U ( ) 0
0 1 2 3 4 5 6 7 8
time (ns)
2
vL (t )   Voi  Voi (t )  U ( )
voltage at load end

1.5
0.5
0
0 1 2 3 4 5 6 7 8
time (ns)

Capacitive load
2.5
voltage at load end 1.5
0.5
0
0 1 2 3 4 5 6 7 8
time (ns)

vL (t )  2Voi 1  e  Ro C
 U ( )
Time domain reflectometry (TDR)
Used to determine location and nature of a load or discontinuity

TDR

TDR

Exp. 2.11: TDR response
Given a Zo = 50 line with up = 0.6c

1
 Location of discontinuity: l  (0.6)(3 108 m / s )   (24 109 s)  2.2m
2
Vtot  V  1.50  1 1  0.5
 Nature of discontinuity:     0.5 and RL  50   150
V 1 1  0.5

Dispersion
Ideal pulse example
• Digital signals may be represented as a Fourier Series

• In realistic media, propagation velocity is a function of
frequency
• Different frequency components travel at different speeds, so the
pulse spreads as it propagates (dispersion)

Fourier series review
For even function symmetry (as in our example pulse)

N
f (t )  ao   an cos  not 
n 1
To / 2 To / 2
2 4
where ao   f (t ) dt and an   f (t ) cos  not  dt
To 0
To 0
12  n 
Our example pulse: ao = 1.2 V, an  sin  V
n  5 

Dispersion
more harmonics gives better model of the ideal pulse
7
6 N = 10
N = 100
5 N = 1000
4
volts
-1
0 5 10 15 20 25 30 35 40
time (ns)

Dispersion
 f 
For a dielectric with  r  6   10 
, we can generate an  diagram.
 5  10 
2500
beta (radians/m) 2000
1500
1000
500
0
0 50 100 150 200 250 300 350
omega (Gigacycles/sec)

Dispersion
• original pulse (dashed) moving along line characterized by the
 diagram of the previous slide
• pulse is distorted by dispersion
8
7
6
5
4
Volts
3
2
1
0
-1
-2
0 5 10 15 20 25 30 35 40
time (ns)

End of Chapter 2


Applied Electromagnetics: Early Transmission Lines Approach

Uploaded by

Copyright:

Available Formats

Applied Electromagnetics: Early Transmission Lines Approach

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Applied Electromagnetics: Early Transmission Lines Approach

Uploaded by

Copyright:

Available Formats

Chapter 2

Applied Electromagnetics: Early Transmission Lines Approach Chap 2-1

Applied Electromagnetics: Early Transmission Lines Approach Chap 2-2

Applied Electromagnetics: Early Transmission Lines Approach Chap 2-3

Applied Electromagnetics: Early Transmission Lines Approach Chap 2-4

Applied Electromagnetics: Early Transmission Lines Approach Chap 2-5

Applied Electromagnetics: Early Transmission Lines Approach Chap 2-6

Applied Electromagnetics: Early Transmission Lines Approach Chap 2-7

The limit on the left is the definition of a derivative, so

Likewise, by Kirchhoff’s current law Telegraphist’s

Applied Electromagnetics: Early Transmission Lines Approach Chap 2-8

where the phasor is Vs ( z )  V  z  e j (likewise I s ( z )  I  z  e j )

Telegraphist’s equations for time-harmonic fields are then

Applied Electromagnetics: Early Transmission Lines Approach Chap 2-9

Solving the differential equation, we arrive at a general solution

The instantaneous form is

Applied Electromagnetics: Early Transmission Lines Approach Chap 2-10

Characteristic impedance Zo is the ratio of voltage to current wave

Inserting the wave equations into the phasor-form Telegraphist’s

phase constant    L ' C '

For low-loss coaxial cable, we find

Applied Electromagnetics: Early Transmission Lines Approach Chap 2-12

For Teflon, r = 2.1 from Appendix E, so

For Zo, we have a = 0.50 mm and b = 1.5 mm, so

Applied Electromagnetics: Early Transmission Lines Approach Chap 2-13

Is is in phase with Vs for a Vo

For a lossy line, Is is no longer in phase with Vs so

Applied Electromagnetics: Early Transmission Lines Approach Chap 2-14

We will find in chapter 6 that the loss tangent is related to an effective

Applied Electromagnetics: Early Transmission Lines Approach Chap 2-15

- To express absolute power in the decibel scale, a reference power is

Applied Electromagnetics: Early Transmission Lines Approach Chap 2-16

Vs ( z  0) Vo e  (0)  Vo e   (0) Vo  Vo

Vo Vo Vo  Vo

Applied Electromagnetics: Early Transmission Lines Approach Chap 2-17

Where L is the reflection coefficient and is a measure of the impedance

Applied Electromagnetics: Early Transmission Lines Approach Chap 2-18

Applied Electromagnetics: Early Transmission Lines Approach Chap 2-19

Looking into the line at z = -l, input impedance Zin is

Applied Electromagnetics: Early Transmission Lines Approach Chap 2-20

Complex loads may be modeled using resistor, inductor

Applied Electromagnetics: Early Transmission Lines Approach Chap 2-21

for lossless line

Applied Electromagnetics: Early Transmission Lines Approach Chap 2-22

for lossless line

Applied Electromagnetics: Early Transmission Lines Approach Chap 2-23

Applied Electromagnetics: Early Transmission Lines Approach Chap 2-24

convert source voltage to phasor: VSS  10e j 30

Vin  Vss  10e j 30

For the lossless case,  = j and l = j l = j/2. So

Converting to instantaneous form: v L  10 cos  t  60 V .  

Compare VL vs line length for T-Line and lumped element models

Applied Electromagnetics: Early Transmission Lines Approach Chap 2-27

 connecting wire begins to impact

 appreciable difference between

Applied Electromagnetics: Early Transmission Lines Approach Chap 2-28

• Most useful for

Applied Electromagnetics: Early Transmission Lines Approach Chap 2-29

Applied Electromagnetics: Early Transmission Lines Approach Chap 2-30