You are on page 1of 83

CHAPTER10

Traveling Wave and Broadband Antennas



10.1INTRODUCTION....................................................................................................................................................................................................................................2
10.2TRAVELINGWAVEANTENNAS.............................................................................................................................................................................................................3
10.2.1LongWire................................................................................................................................................................................................................................12
10.2.2VAntenna...............................................................................................................................................................................................................................29
10.2.3RhombicAntenna...................................................................................................................................................................................................................37
10.3BROADBANDANTENNAS...................................................................................................................................................................................................................40
10.3.1HelicalAntenna.......................................................................................................................................................................................................................40
10.3.3YagiUdaArrayofLinearElements.........................................................................................................................................................................................66
PROBLEMS.................................................................................................................................................................................................................................................81

10.1INTRODUCTION
In the previous chapters we have presented the details of classical methods
that are used to analyze the radiation characteristics of some of the simplest and
most common forms of antennas (i.e., infinitely thin linear and circular wires,
broadband dipoles, and arrays). In practice there is a myriad of antenna
configurations, and it would be almost impossible to consider all of them in this
book. The general performance behavior of some of them will be presented in this
chapter with a minimum of analytical formulations.

10.2TRAVELINGWAVEANTENNAS
In Chapter 4, centerfed linear wire antennas were discussed whose
amplitude current distribution was
1. constant for infinitesimal dipoles (

/50)

2. linear (triangular) for short dipoles ( /50


3. sinusoidal for long dipoles (

/10)

/10)

In all cases the phase distribution was assumed to be constant.



The sinusoidal current distribution of long openended linear antennas is a
standing wave constructed by two waves of equal amplitude and 180o phase
difference at the open end traveling in opposite directions along its length.

The current and voltage distributions on openended wire antennas are


similar to the standing wave patterns on openended transmission lines.
Linear antennas that exhibit current and voltage standing wave
patterns formed by reflections from the open end of the wire are
referred to as standing wave or resonant antennas.

Antennas can be designed which have traveling wave (uniform) patterns in


current and voltage. This can be achieved by properly terminating the antenna
wire so that the reflections are minimized if not completely eliminated.
An example of such an antenna is a long wire that runs horizontal to the
earth, as shown in Figure 10.1.


F
Figure 10.1
1 Beverage (longwire)) antenna ab
bove ground
d

The input terrminals cconsist off the gro


ound and one end
d of the wire.
w
This
cconfiguraation is kn
nown as B
Beverage or wave antenna. There are
m
many other configu
urations o
of traveling wave a
antennas.

In general, all antennas whose current and voltage distributions can


be represented by one or more traveling waves, usually in the same
direction, are referred to as traveling wave or nonresonant antennas. A
progressive phase pattern is usually associated with the current and
voltage distributions.

Standing wave antennas, such as the dipole, can be analyzed as traveling
wave antennas with waves propagating in opposite directions and represented
by traveling wave currents

and

in Figure 10.1(a).

Besides the long wire antenna there are many examples of traveling wave
antennas such as dielectric rod, helix, and various surface wave antennas.
Aperture antennas, such as reflectors and horns, can also be treated as traveling
wave antennas. In addition, arrays of closely spaced radiators (usually less than
/2 apart) can also be analyzed as traveling wave antennas by approximating
their current or field distribution by a continuous traveling wave. YagiUda,
logperiodic, and slots and holes in a waveguide are some examples of
discreteelement traveling wave antennas.

In general, a traveling wave antenna is usually one that is associated with


radiation from a continuous source.
A traveling wave may be classified as a slow wave if its phase velocity
(

/k, : wave angular frequency, k = wave phase constant) is equal or

smaller than the velocity of light c in freespace ( /c

1).

A fast wave is one whose phase velocity is greater than the speed of light
( /c

1).

In general, there are two types of traveling wave antennas.

1
1. Surface wave a
antenna
urface w
wave antenna defin
ned as aan antenna which
h radiates
One is the su
p
power flo
ow from d
discontinu
uities in tthe structture that interrupt a bound
d wave on
n
tthe antenna surfacce.A surface wavee antenna
a is, in gen
neral, a sllow wavee structure
w
whose ph
hase veloccity of th
he travelin
ng wave is equal to or
t
lesss than thee speed of
o
llight in freeespace (

1).

For slow waave structtures rad


diation ta
akes placce only at
a nonuniiformitiess,
ccurvaturees, and discontin
nuities. D
Discontin
nuities caan be either
e
disscrete or
d
distributeed.

On
ne type of
o discrette discontinuity on a surfaace
w
wave anttenna is a
a transm
mission lin
ne termin
nated in an
u
unmatcheed load.
A distribu
uted surfface wav
ve antenna can be
aanalyzed in terms of the vaariation o
of the amplitude aand
p
phase of tthe curren
nt along its structu
ure.
In geeneral, power flow
ws parallel to the
e structu
ure,
eexcept wh
hen losses are present, and for plane
e structurres
tthe fieldss decay exponenti
e
ially away
y from th
he antenn
na.
M
Most of the surfface wav
ve antenn
nas are endfire or
n
nearendfire radiaators. Praactical co
onfigurations inclu
ude
lline, planaar surfacee, curved,, and mod
dulated sttructures..

2
2. Leakywave an
ntenna
Anotther trav
veling waave anten
nna is a leakywav
l
ve anten
nna defineed as an
n
aantenna that cou
uples po
ower in small in
ncrementss per unit lengtth, either
ccontinuou
usly or discretely
d
y, from a traveliing wavee structu
ure to frreespacee
L
Leakywaave anten
nnas conttinuously
y lose ene
ergy due to radiaation, as shown in
n
F
Figure 10
0.2 by a slotted rectangu
ular wave
eguide. The
T
fieldss decay along the
sstructure in the dirrection off wave traavel and iincrease iin others.

Figu
ure 10.2 Le
eakywave w
waveguide
e slots; upper (broad) and side ((narrow) w
walls.

1
10.2.1 Lo
ong Wire
An aantenna is
i usually
y classifieed as a long wiree antennaa if it is a straigh
ht
cconductorr with a leength from one to many wa
avelength
hs.


The long wirre of Figgure 10.1(a), in th
he presen
nce of th
he ground, can be
aanalyzed approxim
mately b
by introd
ducing an
n image to take into acccount the
p
presence of the grround. Th
he magnittude and phase off the image are deetermined
d
u
using the reflection
n coefficieent for ho
orizontal polarizattion as giv
ven by

1
for

(4129)

0 , 180 plane

90 , 270 plane

: the reflection coefficients for parallel and perpendicular polarization

The angles
and

for

and

are angles of incidence and refraction, respectively.

are intrinsic impedance of freespace and the ground, respectively.


The height of the antenna above the ground must be chosen so that the

reflected wave is in phase with the direct wave at the angles of desired maximum
radiation. The total field can be found by multiplying the field radiated by the
wire in free space by the array factor of a twoelement array.

As tthe wavee travels along the wire from the


source to
oward thee load, it ccontinuou
usly leakss energy.
Thiss can bee repressented b
by an atttenuation
n
coefficien
nt. Therefore the current distributiion of thee
forward travelingg wave aalong thee structurre can bee
nted by
represen
z


Figure 1
10.3 Longw
wire antennaa

(101)

z : th
he propagaation coeffiicient.

The atttenuation
n factor

z can also repressent the oh


hmic lossees of the w
wire as welll

as gro
ound lossess, which arre very sm
mall and are
e neglected
d.
When the radiatting mediu
um is air, tthe loss of energy in a long wirre due to
o leakage iis
very ssmall, and iit can also be neglectted.

Therefore the current distribution of (101) can be approximated by


z


where

(101)
(101a)

is assumed to be constant. In the far field


0 (102a);
0

sin

102c
102b

is used to represent the ratio of the phase constant of the wave along the
transmission line ( ) to that of freespace ( ), or

wavelength along the transmission line (103)

Assuming a perfect electric conductor for the ground, the total field for
Figure 10.1(a) is obtained by multiplying each of (102a)(102c) by the array
factor sin
For

sin
k K

.
1 the timeaverage power density can be written as
| |

| |

1 (104)
1 (105)

From (105) it is evident that the power distribution of a wire antenna of


length is a multilobe pattern whose number of lobes depends upon its length.

Assuming that is very large such that the variations in the sine function of
(105) are more rapid than those of the cotangent, the peaks of the lobes occur
approximately when
sin

cos

1;

cos

1
2

, m

0,1,2, (106)

The angles where the peaks occur are given by


cos

1 ,

0,1,2, (107)

The angle where the maximum of the major lobe occurs is given by m = 0.

As becomes very large ( ), the angle of the maximum of the


major lobe approaches zero degrees and the structure becomes a
nearendfire array.

In finding the values of the maxima, the variations of the cotangent term in
(105) were negligible. If the effects of the cotangent term were to be included,
the values of the 2m
2

1 term in (107) should be


1

0.742, 2.93, 4.96, 6.97, 8.99, 11, 13, . .. (108)

In a similar manner, the nulls of the pattern can be found and occur when
sin

cos

0 ,

cos

The angles where the nulls occur are given by

1,2,3, (109)

1,2,3,4 1010

The total radiated power can be found by integrating (105) over a closed
sphere of radius and reduces to

where

| | 1.415

1011

is the cosine integral of (468a). The radiation resistance is then

found to be
| |

1.415

(1012)

Using (105) and (1011) the directivity can be written as


.

(1013)

A
A. Amplittude Pattterns, Ma
axima, an
nd Nulls
Fig 10.4
4(a): the 3D patteern of a trraveling w
wire anten
nna with

Fig 10.4
4(b): the 3D patteern of a sttanding w
wave wiree antenna with

5 .

The corrresponding 2D p
patterns aare shown
n in Figuree 10.5.
90
0

120
0

60

-10
-20

30

150

-30
-40 180

-30
-20

330

210

-10
0

300

240
0
270

Figure 10.4
4 Threedim
mensional freespace
e amplitude
e patterns for travelin
ng and stan
nding wave
e
wire an
ntennas of

The paattern fo
ormed by
y the fo
orward
ttraveling wave cu
urrent

has

m
maximum
m radiattion in the fo
orward
d
direction

The p
pattern formed

. when
w

by staanding

There

iis maximu
um radiaation in th
he forwarrd and
b
backward
d directions.

Figure 10
0.5 Twodim
mensional frreespace
amplitude p
pattern for ttraveling an
nd standing
wave wire anten
nnas of
5

The llobe nearr the axis of the wire in the directions of traveel is the laargest.

The traaveling wave


w
antenna is ussed
when it is dessired to radiate or
minantly from o
one
receivee predom
directio
on.
As the length off the wire increasses,
n lobe shiifts
the maxximum off the main
closer toward the axiis and tthe
number of lobess increasee.


Figu
ure 10.6 Frreespace p
pattern for traveling
wave wire a
w
antenna of
an
nd

The angles off the maxima of th


he first fou
ur lobes, computed using (108), are
p
plotted in
n Figure 10.7(a)
1
fo
or 0.5

10
0 . The correspon
c
nding anggles of the

ffirst fourr nulls, computed


c
d using (1010), are sho
own in Figure
F
10
0.7(b) for
0
0.5

10 . These curves can be used e


effectively
y to desiggn long w
wires when
n

tthe directtion of thee maximu


um or null is desire
ed.

Figu
ure 10.7 An
ngles versu
us length off wire ante
enna where
e maxima a
and nulls occcur

B
B. Inputt Impedance
For travelingg wave w
wire anten
nnas the radiation
n in the opposite direction
n
ffrom the maximum
m is suppressed by
y reducin
ng the currrent refleected from
m the end
d
o
of the wirre. This is accomplished by
In
ncreasing the diameter of th
he wire
Orr terminaating it to the groun
nd, as sho
own in Figgure 10.1.
Ideaally a com
mplete eliimination
n of the
reflection
ns (perfeect match
h) can o
only be
accompliished if the
t
anten
nna is ellevated
only at small heeights (co
ompared to the
ve the grround, an
nd it is
wavelenggth) abov
terminatted by a reesistive lo
oad.

The value of the load resistor is equal to the characteristic impedance of the
wire near the ground (which is found using image theory). For a wire with
diameter and height

above the ground, an approximate value of the

termination resistance is obtained from


138 log

(1014)

If the antenna is not properly terminated, standing wave pattern would be


created. Therefore the input impedance of the line is not equal to the load
impedance. The transmission line impedance transfer equation can be used to
calculate the impedance at the input terminals

(1015)

C. Polarization
A longwire antenna is linearly polarized, and it is always parallel to the
plane formed by the wire and radial vector from the center of the wire to the
observation point.
The direction of the linear polarization is not the same in all parts of the
pattern, but it is perpendicular to the radial vector (and parallel to the plane
formed by it and the wire). Thus the wire antenna is not an effective element for
horizontal polarization. Instead it is usually used to transmit or receive waves
that have an appreciable vector component in the vertical plane. This is what is
known as a Beverage antenna which is used more as a receiving rather than a
transmitting element because of its poor radiation efficiency due to power
absorbed in the load resistor.

D. Resonant Wires
Resonant wire antennas are formed when the load impedance of Figure
10.1(a) is not matched to the characteristic impedance of the line. This causes
reflections which with the incident wave form a standing wave. Resonant
antennas, including the dipole, were examined in Chapter 4.
Resonant antennas can also be formed by long wires. For resonant long
wires with lengths odd multiple of half wavelength (

/2, n = 1, 3, . . .), the

radiation resistance is given approximately (within 0.5 ohms)by


73

69 log

(1016)

For the same elements, the angle of maximum radiation is given by


(1017)

This formula is more accurate for small values of , although it gives good
results even for large values of . It can also be shown that the maximum
directivity is related to the radiation resistance by
(1018)






10.2.2 V A
Antenna
For ssome app
plications a single llongwire
e antenna is not practical beecause
(1) iits directiivity may
y be low
(2) iits side lo
obes may be high
(3) iits main b
beam is in
nclined att an angle
e, which iss controlled by its length.
One very practical arrray of long wires iss
ntenna fo
ormed by
y using ttwo wiress
the V an
each with
h one of iits ends cconnected
d to a feed
d
line as sh
hown in F
Figure 10..8(a).
In m
most app
plications,, the plan
ne formed
d
by the leegs of the V is paraallel to th
he ground
d,
whose p
principal polarizattion is p
parallel to
o
the groun
nd and th
he plane o
of the V.

Becaause of in
ncreased side lobees, the directivity
y of ordin
nary lineaar dipoles
b
begins to diminish
h for lengtths greateer than ab
bout 1.25 . Howev
ver by adjusting the
iincluded aangle of aa Vdipolee, its direectivity ca
an be mad
de greaterr and its side lobes
ssmaller th
han thosee of a corrrespondin
ng linear d
dipole.
Desiggns for m
maximum directivitty usually
y require smaller iincluded angles for
llonger Vss. Most V antennass are symm
metrical (

and
d

o
). Also

V
V antenn
nas can be
b design
ned to haave unidiirectionall or bidirrectional radiation
n
p
patterns, as shown
n in Figures 10.8(b
b) and (c),, respectiv
vely.

To aachieve th
he unidirrectional characte
eristics, th
he wires of the V
V antennaa
m
must be n
nonresonaant. The rreflected w
waves can be redu
uced by
M
Make the inclined wires of tthe V rela
atively thiick
P
Properly terminatte the opeen ends off the V
One way to terminatte the V antenna
a is to
aattach a lo
oad, usuaally a resisstor equaal in value
e to the
o
open end
d characteristic im
mpedancee of the Vwire
V
ttransmisssion line, as shown
n in Figuree 10.9(a)..
The terminatiing resisttance can also be d
divided
iin half aand each half con
nnected to the ground
g
lleading to
o the term
mination o
of Figure 10.9(b).


Figure 10.9 Termiinated V
antennas..

If the length of each leg of the V is very long ( typically

5 ), there will

be sufficient leakage of the field along each leg that when the wave reaches the
end of the V it will be sufficiently reduced that there will not necessarily be a need
for a termination.
The patterns of the individual wires of the V antenna are conical and inclined
at an angle from their corresponding axes. The angle of inclination is determined
by the length of each wire.
The patterns of each leg of a symmetrical V antenna will add in the direction
of the line bisecting the angle of the V and form one major lobe, the total included
angle 2 of the V should be equal to 2

, which is twice the angle that the cone

of maximum radiation of each wire makes with its axis. When this is done, beams
2 and 3 of Figure 10.8(b) are aligned and add constructively.


Similarly forr Figure 10.8(c),, beams 2 and 3 are aligned and add
d
cconstructtively in the forward directiion, while
e beams 5
5 and 8 are aligned and add
d
cconstructtively in th
he backw
ward direcction.
If 2

2
2

thee main lob


be is split into two distinct b
beams.

If (2

2 ), the maximum of the single major


2
m
lobe is still along the

p
plane that bisects the V butt it is tilteed upwarrd from th
he plane of the V d
due to the
eexistence of GND.

For a symmetrical V antenna with legs each of length , there is an optimum


included angle which leads to the largest directivity. The polynomials for
optimum included angles and maximum directivities are given by

149 /

603.4 /

13.39 /

78.27 /

2.94 /

1.15,

809.5 /
443.6 10
0.5
/
1.5
169.77 10
1.5
/
3
1.5

19a
19b

3 (1020)

The corresponding input impedances of the Vs are slightly smaller than


those of straight dipoles.

Anotther form
m of a V an
ntenna is shown in
n Figure 1
10.11(a). T
The V is fformed by
y
aa monopo
ole wire, bent at aan angle o
over a gro
ound plan
ne, and b
by its imaage shown
n
d
dashed. T
The includ
ded angle as well as the leng
gth can bee used to tune the antenna.
For 2

p
primarily
y

120 , the an
ntenna eexhibits
verticcal

polaarization

with

rradiation pattern
ns almosst identiical to
tthose of straight diipoles.
As 2

120 , a horizontally po
olarized

ffield com
mponent iss excited which teends to
ffill the pattern toward the horrizontal
d
direction,, makingg it a v
very atttractive
ccommunication an
ntenna forr aircraft.

The computted impeedance o


of the ground plane and
a
freespace V
V
cconfiguraations obttained by the MoM is shown
n plotted iin Figure 10.11(a).
Anotther practtical form
m of a dipo
ole antenna, particcularly useful for airplane or
gground pllane applications, is the 90
0 bent w
wire configguration of Figuree 10.11(b)).
T
The comp
puted imp
pedance o
of the anteenna is sh
hown plottted in Figgure 10.1
11(b).
This antennaa can b
be tuned by
aadjusting its perpeendicularr and parallel
llengths

and

.Th
he radiaation

p
pattern in
n the plaane of the antenn
na is
n
nearly om
mnidirecttional fo
or h
F
For h

0.1 thee pattern


n approacches

tthat of vertical /2
2 dipole.

0
0.1 .

1
10.2.3 Rh
hombic A
Antenna
A
A. Geome
etry and Radiatio
on Characcteristicss
Two V antenn
nas can b
be conneccted at th
heir open ends to fform a diamond or
rrhombic aantenna, as shown
n in Figurre 10.12((a). To acchieve thee single main
m
lobee,
b
beams 2, 3, 6, and 7 are aligned and
d add con
nstructively. The otther end is used to
o
ffeed the aantenna.

The anttenna is u
usually teerminated
d at one e
end in a reesistor off 600800
0 ohms, in
n
order tto reduce if not elim
minate reeflections..
If each leg is lon
ng enough
h (>5) su
ufficient le
eakage occcurs alon
ng each leeg that the
hat reach
hes the farr end of th
he rhomb
bus is suffficiently rreduced th
hat it may
y
wave th
not be n
necessary
y to terminate the rhombuss.
Anotther conffiguration
n of a rh
hombus is that off Figure 10.12(b) which is
fformed by
y an inverrted V and its imagge (shown
n dashed)).

The inverted V is connected to the ground through a resistor. As with the V


antennas, the pattern of rhombic antennas can be controlled by varying the
element lengths, angles between elements, and the plane of the rhombus.
Rhombic antennas are usually preferred over Vs for nonresonant and
unidirectional pattern applications because they are less difficult to terminate.
Additional directivity and reduction in side lobes can be obtained by
stacking, vertically or horizontally, a number of rhombic and/or V antennas to
form arrays.



10.3BROADBANDANTENNAS
In Chapter 9 broadband dipole antennas were discussed. There are
numerous other antenna designs that exhibit greater broadband characteristics
than those of the dipoles. Some of these antennas can also provide circular
polarization, a desired extra feature for many applications.
10.3.1 Helical Antenna
Another basic, simple, and practical configuration of an electromagnetic
radiator is that of a conducting wire wound in the form of a screw thread forming
a helix.

In most cases the helix is used with a ground plane. The ground plane can

take different forms.


O
One is for the groun
nd to be fflat, as sho
own
in Figuree 10.13. Typically
y the diam
meter of the
ground p
plane shou
uld be at least 3 //4.
The groun
nd plane ccan also b
be cuppe
ed in
ndrical caavity or in
n the form
m of
the form of a cylin
a frustrum cavity.
The helix is usually cconnected
d to the
onductor of a coaxxial transm
mission
center co
line at the feed
d point w
with thee outer
or of thee line aattached to the
conducto
ground p
plane.

The geometrical configuration of a helix consists usually of

: turns,

: diameter

: space between each turn.

: The total length of the antenna


: the total length of the wire

is the circumference of the helix.

Another important parameter is the pitch angle


(1024)
The radiation characteristics of the antenna can be varied by controlling the
size of its geometrical properties compared to the wavelength.
The input impedance is critically dependent upon

and the size of the

conducting wire, and it can be adjusted by controlling their values.

The geeneral po
olarization
n of the antenna is elliptiical. How
wever circular and
d
linear p
polarizatiions can b
be achieveed over d
different fr
frequency
y ranges.
The helical an
ntenna caan operatte in many modes;; howeverr the two
o principaal
o
ones are tthe normal (broad
dside) and
d the axiall (endfiree) modess.

Figure 10
0.14 Threed
dimensional normalized amplitude linear powerr patterns for normal and
d endfire
mode
es helical designs.

Figure 10.14(a), representing the normal mode, has its maximum in a


plane normal to the axis and is nearly null along the axis. The pattern is similar in
shape to that of a small dipole or circular loop.
Figure 10.14(b), representative of the axial mode, has its maximum along
the axis of the helix, and it is similar to that of an endfire array. The axial
(endfire) mode is usually the most practical because it can achieve circular
polarization over a wider bandwidth (usually 2:1) and it is more efficient.
A helix can always receive a signal transmitted from a rotating linearly
polarized antenna. Therefore helices are usually positioned on the ground for
space telemetry applications of satellites, space probes, and ballistic missiles to
transmit or receive signals.

A. Normal Mode
To achieve the normal mode of operation, the dimensions of the helix are
usually small compared to the wavelength (i.e.,

to a loop of diameter

to a linear wire of length when

when the pitch angle

). The helix reduces


0

90 .

Since the limiting geometries of the helix are a loop and a dipole, the far field
radiated by a small helix in the normal mode can be described in terms of
components of the dipole and loop, respectively.

and

In the normal mode, the helix of Figure 10.15(a) can be simulated


approximately by

small loops and

short dipoles connected together in

series as shown in Figure 10.15(b). The fields are obtained by superposition of


the fields from these elemental radiators.

Sincee in the normal mode tthe helix


d
dimension
ns are sm
mall,
the currrent throughout its length
h can be
assumeed to be constant
its reelative farfield
f
pattern to be
independent of the num
mber of lo
oops and
dipoles.
short d


Figure 1
10.15 Norma
al (broadside
e) mode for
helical antenna and its equivalent.

Thuss its operration can


n be desccribed by
y the sum
m of the fiields radiiated by a
a
ssmall loop
p of radiu
us

and
d a short d
dipole of length , with its
,
axis perp
pendiculaar

tto the plane of the loop, and


d each witth the sam
me constaant curren
nt distribution.

The farzone electric field


constant current

radiated by a short dipole of length and

is
(426a/1025)

The electric field

radiated by a loop is
/

(10-26)

A comparison of (1025) and (1026) indicates that the two components are in
timephase quadrature, a necessary but not sufficient condition for circular or
elliptical polarization.
The ratio of the magnitudes of the
axial ratio (AR), and it is given by

and

components is defined as the

By varying the

(1027)

and/or the axial ratio attains values of 0

AR = 0 occurs when

AR

0 leading to a linearly polarized wave of horizontal

polarization (the helix is a loop).


AR

0 and the radiated wave is linearly polarized with vertical

polarization (the helix is a vertical dipole).


AR = 1, the radiated field is circularly polarized in all directions other than

0 where the fields vanish.


2

/ 2

To achieve the normal mode of operation, it has been assumed that the
current throughout the length of the helix is of constant magnitude and phase.

This is satisfied to a large extent provided


The total length of the helix wire
wavelength (

is very small compared to the

Its end is terminated properly to reduce multiple reflections.


Because of the critical dependence of its radiation characteristics on its
geometrical dimensions, which must be very small compared to the wavelength,
this mode of operation is very narrow in bandwidth and its radiation efficiency is
very small. Practically this mode of operation is limited, and it is seldom utilized.


B. Axial Mode
A more practical mode of operation, which can be generated with great ease,
is the axial or endfire mode. In this mode of operation,
There is only one major lobe and its maximum radiation intensity is along
the axis of the helix.
The minor lobes are at oblique angles to the axis.
To excite this mode, the diameter

and spacing must be large fractions

of the wavelength. To achieve circular polarization, primarily in the major lobe,


the circumference of the helix must be in the
near optimum), and the spacing about S
12

14 .

range (with / = 1
/4. The pitch angle is usually

Most often the antenna is used in conjunction with a ground plane, whose
diameter is at least

/2, and it is fed by a coaxial line. However, other types of

feeds (such as waveguides and dielectric rods) are possible, especially at


microwave frequencies.
The dimensions of the helix for this mode of operation are not as
critical, thus resulting in a greater bandwidth.

C. Design Procedure
The terminal impedance of a helix radiating in the axial mode is nearly
resistive with values between 100 and 200 ohms. Smaller values, even near 50
ohms, can be obtained by properly designing the feed. Empirical expressions,
based on a large number of measurements, have been derived. The input
impedance (purely resistive) is obtained by
140

(1030)

which is accurate to about 20%, the halfpower beamwidth by


HPBW degree
the beamwidth between nulls by

(1031)

FNBW degree

dimensionless

15

(1032)
(1033)

the axial ratio (for the condition of increased directivity) by


(1034)
and the normalized farfield pattern by

/
/

cos

1035

/ (1035a)

For ordinary endfire radiation (1035b)

For HansenWoodyard endfire radiation (1035c)

All these relations are approximately valid provided 12


/

14 , 3/4

4/3 and N >3.

The farfield pattern of the helix, as given by (1035), has been developed by
assuming that the helix consists of an array of

identical turns, a uniform

spacing between them, and the elements are placed along the zaxis.
The cos term in (1035) represents the field pattern of a single turn,
The last term

/
/

in (1035) is the array factor of a uniform array of

elements.
The total field is obtained by multiplying the field from one turn with the array
factor.

The value of p is the ratio of the velocity with which the wave travels along
the helix wire, and it is selected according to (1035b) for ordinary endfire
radiation or (1035c) for HansenWoodyard endfire radiation.
(1) For ordinary endfire
The relative phase among the various turns of the helix (elements of the
array) is given by (67a), or
(1036)
where

is the spacing between the turns.

For an endfire design, the radiation from each one of the turns along

must be in phase. Since the wave along the helix wire between turns travels a
distance

with a wave velocity

( <1 where

is the wave velocity

in free space) and the desired maximum radiation is along

0 , then (1036)

for ordinary endfire radiation is equal to


cos

0,1,2, (1037)

Solving (1037) for leads to


/

(1038)

For
0 and
1, 0


. This corresponds to a straight
1, and it
wire ( 90 ), and not a helix. Therefore the next value is
corresponds to the first transmission mode for a helix. Substituting

1 in

(1038) leads to
/

(1038a)

(2) for HansenWoodyard endfire radiation


In a similar manner, it can be shown that for HansenWoodyard endfire
radiation (1037) is equal to
cos

0,1,2, (1039)

which when solved for leads to


/

(1040)

Example 10.1
Design a 10turn helix to operate in the axial mode. For an optimum design,
1. Determine the:
a. Circumference (in

), pitch angle (in degrees), and separation between turns (in

b. Relative (to free space) wave velocity along the wire of the helix for:
i. Ordinary endfire design
ii. HansenWoodyard endfire design
c. Halfpower beamwidth of the mainlobe (in degrees)
d. Directivity (in dB) using:
i. A formula

ii. The computer program Directivity of Chapter 2


e. Axial ratio (dimensionless and in dB)
2. Plot the normalized threedimensional linear power pattern for the ordinary and
HansenWoodyard designs.

Solution:
1. a. For an optimum design
Condition: 12

14 , 3/4

13 S

Ctan

tan13

b. The length of a single turn is


Therefore the relative wave velocity is:


i.

Ordinary endfire:

4/3 and N >3

1.0263

0.231

ii.

1.0263
0.231 1

/
/ 1
HansenWoodyard endfire:
/

.
.

0.8337

=0.8012

c. The halfpower beamwidth according to (1031) is


52

HPBW degree

52
10 0.231

34.21o

d. The directivity is:


i.

Using (1033):
dimensionless

15

15 10 0.231

34.65

15.397

e. The axial ratio according to (1034) is:


AR

2N

1 /2N

21/20

1.05 dimensionless

0.21 dB

2. The threeedimensional lineaar power p


patterns fo
or the two endfire d
designs, orrdinary and
d
HansenWoodyard
d, are show
wn in Figure 10.16


Figure 10
0.16 Threed
dimensional normalized amplitude liinear powerr patterns forr helical ord
dinary (p =
HansenWo
0
0.8337) and
oodyard (p = 0.8012) end
dfire design
ns

D. Feed Design
The nominal impedance of a helical antenna operating in the axial mode,
computed using (1030)
140

is 100~200. However, many practical transmission lines have characteristic


impedance of about 50. The input impedance of the helix must be reduced to
near that value. There may be a number of ways by which this can be
accomplished. One way to properly design the first 1/4 turn of the helix which is
next to the feed.
To bring the input impedance of the helix from nearly 150 down to 50,

tthe wire o
of the first 1/4 turn
n should be flat in the form of a strip
p and the transition
n
iinto a heliix should be very ggradual.

14

140
150

50

This is accom
mplished b
by makin
ng the wirre from th
he feed, aat the begginning o
of

the formation of the helix, in the form of a strip of width

by flattening it and

nearly touching the ground plane which is covered with a dielectric slab of height
(1041)

where
= width of strip conductor of the helix starting at the feed
= dielectric constant of the dielectric slab covering the ground plane
= characteristic impedance of the input transmission line

Typically the strip configuration of the helix transitions from the strip to the
regular circular wire and the designed pitch angle of the helix very gradually
within the first 1/41/2 turn.

This modification decreases the characteristic impedance of the


conductorground plane effective transmission line, and it provides a lower
impedance over a substantial but reduced bandwidth.

For example, a 50 helix has a VSWR of less than 2:1 over a 40%
bandwidth compared to a 70% bandwidth for a 140 helix.
In addition, the 50 helix has a VSWR of less than 1.2:1 over a 12%
bandwidth as contrasted to a 20% bandwidth for one of 140 .
A simple and effective way of increasing the thickness of the conductor near the
feed point will be to bond a thin metal strip to the helix conductor. For example, a
metal strip 70mm wide was used to provide a 50 impedance in a helix whose
conducting wire was 13mm in diameter and it was operating at 230.77 MHz.

1
10.3.3 Ya
agiUda A
Array of L
Linear Ellements
Anotther very
y practical radiatorr in the HF
H (330 MHz), VHF (303
300 MHz)),
aand UHF ((3003,000 MHz) ranges is the YagiUda anteenna.
This antennaa consistts of a n
number of
o
llinear dip
pole elem
ments, ass shown in Figurre
1
10.19, one of whicch is enerrgized dirrectly by a
ffeed transsmission line whille the oth
hers act as
a
p
parasitic radiato
ors or reflectorr, whosse
ccurrents aare inducced by mu
utual coup
pling.

This radiator is an end


dfire arrray, Yagi d
designateed the row
w of direectors as aa
wave can
nal.


T
To achiev
ve the end
dfire beam
m formattion,
The parasiticc elementts in the direction
n of the beam
b
aree smaller in length
h
tthan the ffeed elem
ment.
The d
driven eleement is resonant with its llength sligghtly lesss than //2
The lengths of
o the dirrectors sh
hould be about 0.4~0.45. The direectors are
n
not necessarily of tthe same length an
nd/or dia
ameter.

The separation between the directors is 0.3~0.4 , and it is not necessarily


uniform.
The gain was independent of the radii of the directors up to ~0.024 .
The length of the reflector is greater than that of the feed. In addition, the
separation is smaller than the spacing between the driven element and the
nearest director, and it is ~0.25 .
Since the length of each director is smaller than its corresponding resonant
length, the impedance of each is capacitive and its current leads the induced emf.
Similarly the impedance of the reflector is inductive and the phases of the
currents lag those of the induced emfs.
The phase of the currents in the directors and reflectors is not determined
solely by their lengths but also by their spacing to the adjacent elements.

Thus, properly spaced elements with lengths slightly less than their
corresponding resonant lengths (less than /2) act as directors because they
form an array with currents approximately equal in magnitude and with equal
progressive phase shifts which will reinforce the field of the energized element
toward the directors.
Similarly, a properly spaced element with a length of /2 or slightly
greater will act as a reflector.
Thus a YagiUda array may be regarded as a structure supporting a traveling
wave. Higher resonances are available near lengths of , 3 /2, and so forth, but
are seldom used.


Figure 10.23 Directivity
y and frontttoback ratio
o, as a Figu
ure 10.24 Directivity and
d fronttoba
ack ratio, as a
function of director spa
ffunction of reflector spaccing, of a 15element Yag
giUda
acing, for 15element
a
array.
YagiUda array.

1. The total field represented

1065


1065a

2. Input Impedance and Matching Techniques


The input impedance of a YagiUda array, measured at the center of the
driven element,
Usually small
Strongly influenced by the spacing between the reflector and feed
element.

Some of these values are low for matching to a 50, 78, or 300ohm
transmission lines.
There are many techniques that can be used to match a YagiUda array to a
transmission line and eventually to the receiver, which in many cases is a
television set which has a large impedance (on the order of 300 ohms). Two
common matching techniques are the use of the folded dipole, of Section 9.5, as a
driven element and simultaneously as an impedance transformer, and the
Gammamatch of Section 9.7.4.
3. Design Procedure
A stepbystep design procedure has been established in determining the
geometrical parameters of a YagiUda array for a desired directivity. The included
graphs can only be used to design arrays with overall lengths (from reflector

element to last director) of 0.4, 0.8, 1.2, 2.2, 3.2, and 4.2 with corresponding
directivities of 7.1, 9.2, 10.2, 12.25, 13.4, and 14.2 dB, respectively, and with a
diametertowavelength ratio of 0.001

0.04.

Assume the driven element a /2 folded dipole. The procedure is identical


for all other designs at frequencies where included data can accommodate the
specifications.
The basis of the design is the data included in
1. Table 10.6 which represents optimized antenna parameters for six
different lengths and for a /

0.0085

2. Figure 10.27 which represents uncompensated director and reflector


lengths for 0.001

0.04

Example 10.3
Designa YagiUda array with a directivity (relative to a /2 dipole at
the same height above ground) of 9.2 dB at 50.1
. The desired
diameter of the parasitic elements is 2.54 cm and of the metal supporting
boom 5.1 cm. Find the element spacings, lengths, and total array length.
Solution:
a. At

50.1

the wavelength is

2.54/598.8

4.24

10 ;

598.8
/

5.1/598.8

8.52

10

b. From Table 10.6, the array with desired gain would have five elements.
For a / = 0.0085 ratio the optimum uncompensated lengths would be
0.428 ,

0.424 ,

0.482 .

The spacing between directors=0.2 . The reflector spacing 0.2 .The overall
antenna length=0.8 .

It is now dessired to fiind the op


ptimum llengths o
of the parasitic eleements
f
for a
/ = 0.004
424.
c Plot th
c.
he optim
mized len
ngths fro
om Table 10.6 (
0
0.424
,

0.428 ,

0.482 ) on Figurre 10.27 aand mark


k them by
y a dot ().

d. In Figu
d
ure 10.27
7 draw a vertical lline throu
ugh / = 0.0042
24 interssecting
c
curves
(
(B)
at director
d
uncomp
pensated lengths
0.442 and
r
reflector
length
0.485
5 . Mark these points by an
n x.

ee. With a
a dividerr, measu
ure the distance
d
( ) alo
ong direcctor curv
ve (B)
b
between
points
0.428 and
0.4
424 . Transpose
T
e this
d
distance
ownward along
from thee point
0.442 on curvee (B), do
t
the curve
e and deteermine th
he uncom
mpensated length
0.43
38 .

Thus the boom


u
uncompe
ensated lengths
l
o
of
thee array
y at
50..1
are
a
0.44
42
0.438
0.485

f. Correcct the elem


ment lengths to compensaate for the boom d
diameter.. From
Figure 10.28, a boom diaametertowaveleength ratiio of 0.00
0852 requ
uires a
fractional lengtth increaase in eaach elemeent of ab
bout 0.00
05. Thu
us the
final leengths of the elements shou
uld be
0.442

5
0.005

0.4
447 ,
0.485

0.005

0.438
8

0.005
5

0.490
0

0.443

PROBLEMS
10.6. It is desired to place the first maximum of a long wire traveling wave
antenna at an angle of 25 from the axis of the wire. For the wire
antenna, find the
(a) exact required length
(b) radiation resistance
(c) directivity (in dB)
The wire is radiating into free space.
10.7. Compute the directivity of a long wire with lengths of

2 and 3 .

10.8. A long wire of diameter d is placed (in the air) at a height h above the
ground.
(a) Find its characteristic impedance assuming
(b) Compare this value with (1014).

10.12. Design a symmetrical V antenna so that its optimum directivity is 8


dB. Find the lengths of each leg (in ) and the total included angle of the V
(in degrees).
10.17. Design a fiveturn helical antenna which at 400 MHz operates in the
normal mode. The spacing between turns is /50. It is desired that the
antenna possesses circular polarization. Determine the
(a) circumference of the helix (in
(b) length of a single turn (in

and in meters)

and in meters)

(c) overall length of the entire helix (in

and in meters)

(d) pitch angle (in degrees)


10.20. Design a nineturn helical antenna operating in the axial mode so
that the input impedance is about 110 ohms. The required directivity is 10
dB (above isotropic). For the helix, determine the approximate:
(a) circumference (in

).

(b) spacing between the turns (in

).

(c) halfpower beamwidth (in degrees).


10.36. Design a YagiUda array of linear dipoles to cover all the VHF TV
channels. Perform the design at 216 MHz. Since the gain is not
affected appreciably at
, as Figure 10.26 indicates, this design should
accommodate all frequencies below 216 MHz. The gain of the antenna
should be 14.4 dB (above isotropic). The elements and the supporting boom
should be made of aluminum tubing with outside diameters of 38 in.( 0.95
cm) and 34 in.(1.90 cm), respectively. Find the number of elements, their
lengths and spacings, and the total length of the array (in , meters, and
feet).