Loop Antennas

Peter J. Massey, P. Fellows, Dariush Mirshekar-Syahkal, Arpan Pal,

and Amit Mehta

Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 724
Electrically Small Loop and Coil Antennas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 725
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 725
Radiation and Losses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 726
Receiving Antennas and Man-Made Electrical Noise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 737
Tuning Components and Quality Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 739
Proximity Effects and Detuning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 742
Matching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 746
Full-Wave Loops . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 747
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 747
Resonant Full-Wave Loop . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 749
Resonant Full-Wave Loop in Close Proximity to Metallic Structures . . . . . . . . . . . . . . . . . . . . . 750
Tuning Resonant Loop Antennas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 755
Circularly Polarizing the Resonant Loop Antenna . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 755
Dual Banding the Resonant Loop Antenna . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 758
Dual Banding and Circular Polarizing the Resonant Loop Antenna . . . . . . . . . . . . . . . . . . . . . . . 760
Resonant Loop over Planar Reflector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 760
Directive Resonant Loop Array . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 762
Dual Rhombic/Circular Loop Antenna over Planar Reflector (Biquad Antenna) . . . . . . . . . . 763
Circularly Polarized Dual Rhombic/Circular Loop Antenna over Planar Reflector . . . . . . . 764
Wideband Circularly Polarized Dual Rhombic Loop Antenna with Parasitic Loops . . . . . . 764

P.J. Massey (*) • P. Fellows

Her Majesty’s Government Communications Centre, Milton Keynes, UK
e-mail: [email protected]; [email protected]
D. Mirshekar-Syahkal
University of Essex, Colchester, UK
e-mail: [email protected]
A. Pal • A. Mehta
College of Engineering, Swansea University, Swansea, UK
e-mail: [email protected]; [email protected]

# Springer Science+Business Media Singapore 2016 723

Z.N. Chen et al. (eds.), Handbook of Antenna Technologies,
DOI 10.1007/978-981-4560-44-3_26
724 P.J. Massey et al.

Reconfigurable Multi-feed Loops and Their Arrays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 765

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 765
Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 766
Radiation Mechanism for Beam Steering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 767
Using High-Impedance Surfaces for Height Reduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 768
Incorporating Electronic RF Switching Circuitry for Electronic Beam Switching . . . . . . . . 770
Array of SLAs for High-Gain Wide-Angle Steering, Wide Null Scanning,
and Low Grating Lobes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 771
HHIS-Based SLA for Demonstrating Gain Beam Steering Using Tilted
and Axial Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 772
Conceptual Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 775
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 783
Cross-References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 784
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 784

Abstract
This chapter on loop antennas covers implementation aspects of both electrically
small and electrically large loops which have a wide range of applications. On one
side, they are still a part of the old broadcasting systems such as analog radios,
and, on the other side, they are increasingly gaining attention for the contempo-
rary high-definition video transmission and reception. The chapter begins with
electrically small loops and coil antennas and covers its different aspects of
radiation and losses, tuning, quality factor, and matching. Subsequently, resonant
full-wave loops and their polarization are presented. A technique is shown on
how loops can be placed in closed vicinity of a communicating device conducting
surface without losing efficiency. In the final part, single-element beam steering
loop antennas and their arrays for enabling devices to achieve high throughput
and high-gain wide scanning range are discussed.

Keywords
Loop antennas • Circularly polarized antennas • Coil antennas • Electrically small
antennas • Directive antennas • Ferrite antennas • Receiving antennas • Man-
made noise • Multi-feed antennas • Resonant antennas • Rhombic antennas •
Switched feeds • Reconfigurable loop antennas and their arrays • Wide scan high-
gain adaptive arrays

Introduction
A loop antenna is an antenna where the radiating conductor forms a loop. Loop
antennas have existed and have continuously evolved since the earliest days of radio.
Consequently, “loop antennas” is now a subject area that is too large for all its
features to be covered in just a chapter or even in any single book. Instead, three
particular aspects are covered in this chapter:

1. Electrically small loop antennas and their close relations: electrically small multi-
turn coils
Loop Antennas 725

2. Full-wave self-resonant loops and how they can be modified to cover different
frequency bands and to radiate different polarizations and to radiate with high
gain
3. Multi-feed loops where the radiation pattern is steered by switching between
feeds

A common feature of all these aspects is that they are ways of saving physical
space. Electrically small loop and coil antennas save space by being small. Full-
wave loops can be used to save space by sharing that space with conducting
structures already occupying the space. They can also be combined with other
loops to create compact directive radiators. And multi-feed loops save space by
reusing the same radiating element (the loop) to generate several beams.

Electrically Small Loop and Coil Antennas

Introduction

What an electrically small loop or coil antenna consists of. An electrically small
antenna is one that is much less than a wavelength across. In addition for the loops
and coil antennas discussed below, it is assumed that the distance around the loop
and along the length of the coil is short enough that the current can be approximated
as near constant.
Because of the low conversion factor between currents on a small structure and
radiation and because of the way the current contributions to radiation cancel out in
loop and coil antennas (gone into in detail in subsection “Radiation and Losses”),
electrically small antennas generally require tuning components to resonate, so that
the amount of current carried, and hence the coupling to radiation, is maximized. For
electrically small coil and loop antennas, this means tuning with one or more
capacitors across the ends of the inductive coil or loop. In addition, they require
some connection to their transceiver. There are many ways of achieving this, and
some of them are discussed in subsection “Matching.” However, other than ensuring
that the antenna is impedance and/or noise matched to the transceiver circuitry, the
method of matching usually has little effect upon the antenna performance, which is
more dependent upon the antenna geometry and upon the environment surrounding
the antenna.
Applications. The main application for electrically small loop and coil antennas
has been as the antennas in medium- and long-wave broadcast radio receivers. The
earlier antennas had the interior of the antenna filled with air. High-permeability soft
ferrite was introduced in the 1950s, and coil antennas wrapped around a ferrite core
have now become predominant as antennas for medium and long-wave reception.
(Subsection “Receiving Antennas and Man-Made Electrical Noise” gives the rea-
sons why they are widely used for this application.)
In the second half of the twentieth century, electrically small loop and coil
antennas found another major application within radio pagers, operating at up to
726 P.J. Massey et al.

UHF (ultra high frequency band; 300MHz to 3GHz). The significant advantages
over alternative antenna types are their relatively small size and their relative
immunity to detuning due to the pager being carried on a user’s body, e.g., mounted
on a belt, clipped to a shirt, or placed in a pocket (subsection “Proximity Effects and
Detuning” discusses proximity detuning).
Section overview. This section is arranged as follows. First, in subsection
“Radiation and Losses,” the radiation and losses are calculated, both for electrically
small loops and coils and also for comparison for electrically small dipoles. These
calculations reveal that electrically small loops and coils are less efficient radiators
than dipoles of comparable size. Then, it is followed by subsections “Receiving
Antennas and Man-Made Electrical Noise,” “Near Field Noise,” “Far Field Noise,”
“Tuning Components and Quality Factors,” “Quality Factors,” “Finding the Induc-
tance of an Electrically Small Loop or Coil Antenna,” and “Proximity Effects and
Detuning,” which are on the reasons why, despite their inefficiencies, loops and coils
became popular for radio reception and for near-human body antennas. These sub-
sections cover (a) the effects of man-made electrical noise on receiving antenna
performance, (b) losses due to tuning components and how to find the component
values, and (c) detuning from coupling to human bodies and other nearby objects.
The last subsection, “Matching”, discusses those matching issues that have not
already been covered in earlier sections.
There is little discussion of an antenna’s impedance bandwidth. This is because
electrically small loops and coils tend to be used for narrowband communications.
For electrically small loops and coils, the antenna bandwidth tends to be primarily
determined by the Ohmic losses of the antenna. For most designs the antenna
bandwidth has a lower limit of around 1/2 % and is often around 1–2%, which is
adequate for their applications.

Radiation and Losses

This subsection discusses radiation and losses by calculating them for a transmitting
antenna. That the efficiency of the antenna when receiving is the same follows from
the reciprocity theorem [e.g., see Collin (2001) for a discussion and derivation of the
reciprocity theorem].
Radiation in terms of current. In order to calculate the radiation from a loop or
coil antenna, one can use the Schelkunoff formula for the far field due to currents
(Schelkunoff 1939) [see also Lee (1984)]. This formula comes from the formula for
electric field in terms of magnetic vector potential A and electric scalar potential Φ.
In general, with

ððð ððð
μ ejkr 1 ejkr
A¼ J dv Φ ¼ ρ dv (1)
4π r 4πe r
v v
Loop Antennas 727

Fig. 1 Coordinates used in

calculating radiation

where J is the current density, ρ is the charge density, r is the distance between
current or charge source at point p and the point q at which A or Φ is being
evaluated, and v denotes integration over volume (Fig. 1), then the electric field
E is given by

E ¼ jωA  ∇Φ (2)

See, for instance, Collin (2001), for a derivation of the above formulae.
Schelkunoff showed that when the electric field is due to the far field, the
following simplifications apply:

ejkr ejkR jkp:r

ffi e (3)
r R

where R is the distance between the antenna’s center and q, r is the direction vector
pointing from the center of the antenna to q in the far field, and p is the position
vector of point p. Then, in an r, θ, ϕ spherical coordinate system, the contribution of
Φ to E is to cancel out with the radial component of A, and E becomes
 
E ¼ jω 0, Aθ , Aϕ (4)

where A simplifies to
ððð
μ ejkR jkpr
A ffi AFF ¼ Je dv (5)
4π R
v
728 P.J. Massey et al.

Electrically small dipole. Before applying the above formula to calculating the
radiation from a small loop or coil, it will be useful to apply it to the simpler case of a
small dipole. Suppose the electrically small dipole is oriented along the z-axis with
its terminals at the origin of the coordinate system. The current at the terminals is
denoted I0. The current’s magnitude decays toward the open ends of the arms. The
way it decays depends on the shape of the arms, but in practice for small antennas
one can approximate the current distribution as linear, with it decaying to zero at the
tip of the arms. Therefore, the current I can be modeled as
0 1
B jzjC l0 l0
I ðzÞ ¼ I 0 @1  A for  z (6)
l0 2 2
2
where the dipole’s overall length is l0 and each of the dipole arms is l0/2 long. In the
above, the magnitude of I is taken as the root-mean-square (rms) value of the current,
so as to avoid the need for factors of ½ in the formulae for radiated power and Ohmic
pffiffiffi
losses. The peak magnitude is 2 times the rms value (Harrington 2001).
The exponential part of the formula for A can be expanded as

s2 s3
e jkpr ¼ 1 þ s þ þ þ ... where s ¼ jkz cos θ (7)
2! 3!
For small distances where kz  1, only the leading term is significant. Substituting
this approximation and the formula for I(z) into the equations for the far field gives

μ ejkR I 0 l0
A¼ z (8)
4π R 2
jωμ ejkR I 0 l0 sin θ
Eθ ¼ Eϕ ¼ 0 (9)
4π R 2
where z is the unit direction vector in the z-direction. The radiated power density is
given by
 2
ωμjI 0 jl0 sin θ  
jEj2 8πR jI 0 jl0 sin θ 2
PRad dipole ¼ ¼ ¼ Z0 (10)
Z0 Z0 4λR

The total radiated power is the integration of the radiated power density over the
sphere of radius R:

ðπ    
jI 0 jl0 sin θ 2 Z0 π jI 0 jl0 2
PRadT dipole ¼ 2πR Z02
sin θ dθ ¼ (11)
4λR 6 λ
0
Loop Antennas 729

The directivity is given by

4π PRad dipole 3 sin2 θ

Ddipole ¼ ¼ (12)
PRadT dipole 2

It is useful to compare the radiated power with the power lost through resistive
losses. Assuming that the dipole is made of a wire with resistivity ρw per meter, then
the losses are given by

l0
ð2 !2
2 j zj 1
PΩ dipole ¼ 2ρw jI 0 j 1 l0
dz ¼ ρw jI 0 j2 l0 (13)
2
3
0

The radiation efficiency R.E. is given by

PRadT P:F:
R:E: ¼ ¼ (14)
PRadT þ PΩ P:F: þ 1
where the power factor P.F. is

PRadT
P:F: ¼ (15)
PΩ
When P.F.  1, then R.E.  P.F.
For the dipole,

Z0 πl0
P:F:dipole ¼ (16)
2ρw λ2

So the power factor, and therefore for small values the radiation efficiency, is
proportional to the dipole length. For example, if the dipole is 1/20th wave-
length long, P:F:dipole ffi 0:079Z0 =ðρw λÞ, and if the dipole is 1/100th wavelength
long, P:F:dipole ffi 0:016Z 0 =ðρw λÞ.
Electrically small wire loop. In order to compute radiation from the currents in a
small loop, first the case of a small rectangular loop will be considered. Then the
results will be extended to loops that are planar but otherwise of arbitrary shape.
Consider an electrically small rectangular loop whose sides are parallel to the x
and y axes. For the moment, consider only the far electric field in the plane where
ϕ = 0. Then the direction vector to the far field has the form r ¼ x sin θ þ z cos θ
where x, y, and z are the unit length direction vectors in the x-, y-, and z-directions.
Therefore, p  r ¼ px sin θ þ pz cos θ, and

e jkpr ¼ e jkðpx sin θþpz cos θÞ ¼ e jkpx sin θ e jkpz cos θ (17)
730 P.J. Massey et al.

Substituting this into the equation for AFF gives

ððð jkpx sin θ
μ ejkR ejkpz cos θ
AFF ¼ Je dv
4π R
v
0 1
xð2 yð2
jkR jkpz cos θ
μ e e B C
¼ @ dx xI 0 ejkx sin θ  xI 0 ejkx sin θ þ dy yI 0 ejkx2 sin θ  yI 0 ejkx1 sin θ A
4π R
x1 y1

(18)
where the sides parallel to the x-axis run between x1 and x2 and the sides parallel to
the y-axis run between y1 and y2, and as with the dipole derivation case, I0’s
magnitude corresponds to the rms value.
The integration with respect to x is for the two arms parallel to the x-axis, and the
contributions from these arms cancel with each other. This leaves the contribution
from the two arms parallel to the y-direction. The exponent remaining within the
integral can be evaluated using a Taylor expansion:

sx 2 sx 3
ejkpx sin θ ¼ 1 þ sx þ þ þ ... where sx ¼ jkpx sin θ (19)
2! 3!

The first term of the expansion cancels in the integrand. This leaves the second term
as the most significant, and the integration reduces to

yð2 yð2
jkx2 sin θ jkx1 sin θ
dy I 0 e  I0 e ffi dyI0 jkðx2  x1 Þ sin θ
y1 y1

¼ I 0 jk sin θðy2  y1 Þðx2  x1 Þ (20)

ðy2  y1 Þðx2  x1 Þ is the product of the length and width of the rectangular loop and
is equivalent to the loop’s area Arectangular loop. So AFF simplifies to

jkμ ejkR ejkpz cos θ

AFF ¼ I 0 sin θArectangular loop y^ (21)
4π R

Substituting this back into the equation for the electric far field gives

ωμk ejkR ejkpz cos θ

Eθ ¼ 0 Eϕ ¼ I 0 sin θArectangular loop
4π R
¼ crect I 0 sin θArectangular loop (22)
Loop Antennas 731

where crect represents the factors that are independent of I0, θ, and loop dimensions.
It can be rearranged by expressing ω and k in terms of the speed of light c and
wavelength λ and expressing c μ in terms of Z0:

ωμk 4π 2 cμ πZ 0
¼ ¼ 2 (23)
4π 4πλ2 λ

ejkR ejkpz cos θ πZ 0

crect ¼ (24)
R λ2

Now consider the far field at a point q in the x-z plane from a loop LA of arbitrary
outline in a plane parallel to the x-y plane with a current I0 circulating around it in the
anticlockwise direction. To calculate the far field, one can conduct a thought
experiment where the area enclosed by the loop is filled with small rectangular
loops all of which have I0 circulating around them in the anticlockwise direction. On
the boundaries between adjacent rectangular loops, the currents cancel, and so the
net current within the interior of the tessellation of rectangular loops is 0, the same as
for LA. The only places where the current does not cancel out are at the exterior
boundary of the tessellation, where here it is I0, the same as on the arbitrary outlined
loop. Therefore, the radiated field from the arbitrary outlined loop is the same as
from the tessellation of rectangular loops. The radiation from the tessellation of
rectangular loops is given by summing the fields of the individual loops, which for
Eϕ is the sum of crect sinθ times their areas, or

Eϕ ¼ crect I 0 ALA sin θ (25)

where ALA is the total area enclosed by the loop.

The above was calculated for far field points where ϕ = 0. But one can argue that
the result applies for any angle of ϕ by noting that as the shape of the loop in the x-y
plane is arbitrary, one could rotate the coordinate system about the z-axis by ϕ and
arrive at the same result. So the above equation applies for all ϕ.
The radiated power density is given by
 
jEj2 jcrect j2 ðjI 0 jALA sin θÞ2 π jI 0 jALA sin θ 2
PRad loop ¼ ¼ ¼ Z0 (26)
Z0 Z0 λ2 R
The total power radiated is given by

ðπ   
π jI 0 jALA sin θ 2 8Z0 π 3 jI 0 jALA 2
PRadT loop ¼ 2πR Z 0
2
sin θ dθ ¼ (27)
λ2 R 3 λ2
0
732 P.J. Massey et al.

The Ohmic losses are given by

þ
PΩloop ¼ ρw dl jI 0 j2 ¼ ρw pLA jI 0 j2 (28)

where pLA is the perimeter of the loop.

The power factor is given by
 
PRadT loop 8Z0 π 3 ALA 2
P:F:wire loop ¼ ¼ (29)
PΩloop 3ρw pLA λ2

To compare with the result for a dipole, consider a square loop, with sides of length
l0. If l0 = λ/20, then P:F:loop ffi 0:0026 Z 0 =ðρw λÞ, and if l0 = λ/100, P:F:loop ffi 2:1
105 Z 0 =ðρw λÞ . So for antennas made out of the same wire and a 20th of a
wavelength across, the dipole’s power factor is 30 times better than the loop’s, and
for antennas a hundredth of a wavelength across the dipole’s power factor is
760 times larger than that of the loop.
For later comparison with the strip metal loop case, it is useful to consider when
the wire diameter is much greater than the skin depth, and therefore the relation ρw
¼ ρs =ðπdÞ applies, where ρs is the surface resistivity. Then
 
8Z0 π 4 d ALA 2
P:F:wire loop ¼ (30)
3ρs pLA λ2

Electrically small loop of strip metal. Some electrically small loop antennas are
constructed from strips of metal rather than wires, in order to reduce the Ohmic
losses of the loop. The computation of the radiated field can be thought of as a
summation of the field from a collection of electrically small wire loops stacked
along the z-axis. So the field calculation proceeds as for the calculation of a small
wire loop above, except that there is an integration over the current density with
z. That is, Eq. 25 is modified to

ðz2
Eϕ ¼ dz crect ðzÞ I ðzÞALA ðzÞ sin θ (31)
z1

The z dependant factor in crect is ejkz cosθ

, which can be expanded by Taylor
expansion:

s2z s3z
ejkz cos θ ¼ 1 þ sz þ þ þ ... where sz ¼ jkz cos θ (32)
2! 3!
and for small kz the leading constant term dominates and the other terms can be
neglected. Also usually the loop has constant cross-sectional area ALA. Then the
Loop Antennas 733

right-hand side of Eq. 31 becomes the same as Eq. 25, and the formulae for the
radiated power and total power radiated are as given in Eqs. 26 and 27.
The formula for the Ohmic losses is
þ ðð ðð
PΩ strip loop ¼ dl dSρ jJ j2 ¼ ρ pLA dSρ jJ j2 (33)
S S

where ρ is the bulk resistivity, J is the current density, pLA is the perimeter of the
loop, and the double integral over S denotes integration across a cross section of the
strip. The variation in current density depends on the loop geometry; because of
induction, current tends to be pulled towards the inner side of the strip. It also
depends upon z as the current concentrates near the corners of the strip. Therefore,
the double integral is difficult to evaluate. However, for many applications, the
thickness of the strip is significantly greater than the skin depth, and then the formula
can be written as
0z 1
ð2  ð
 xouter  
¼ ρS pLA @ dz jJ S inner j þ jJ S outer j þ dx J S top þ jJ S bottom j A
2 2 2 2
PΩ strip loop
z1 xinner

(34)
where ρs is the surface resistivity in Ohms per square, Js is the surface current
density, “inner” and “outer” signify the surfaces to the inside x = xinner and outside
x = xouter of the strip, and “top” and “bottom” signify the surfaces at the top z = z2
and bottom z = z1 of the strip. The above integrals are difficult to estimate analyt-
ically but, for a wide range of cases applicable to VHF (very high frequency band;
30 to 300MHz) and UHF body-borne loops (e.g., for radio pagers), have been
examined numerically. These numerical studies have shown that the following
approximation can be made:
0z 1
ð2
jI 0 j2
PΩ strip loop ¼ ρS pLA β@ dz 2jJ S j2 A ¼ 2ρS pLA β lz jJ S j2 ¼ ρS pLA β (35)
2 lz
z1

where lz = (z2  z1) is the strip’s width. In other words, one can approximate the
power lost by calculating the Ohmic losses assuming that the current density is
constant and along the inner and outer sides. The additional losses due to the current
preferring the inner side and collecting at the edges of the strip and due to the current
at the top and bottom surfaces are taken care of by introducing a factor β where 1 <β
<~1.6. Combining this with Eq. 27 gives
 2
16 lz Z 0 π 3 ALA
λ2
P:F:strip loop ¼ (36)
3 ρS pLA β
734 P.J. Massey et al.

Fig. 2 Dimensions of a rectangular loop. While the tuning capacitor is shown in its often preferred
position in one of the shorter sides (see subsection “Proximity Effects and Detuning”), its location is
immaterial to the validity of Eq. 39

The difference between this and Eq. 30 is

P:F:strip loop 2lz

¼ (37)
P:F:wire loop dπβ

In radio pagers, there is often enough space available to use a 5 mm width strip
metal, or one could use a relatively thick wire (e.g., 1 mm diameter) to save material
costs. The power factor improvement from using a 5 mm wide strip instead of a
1 mm diameter wire would be 10/(πβ) which (depending on the value of β) is
between ~2 and ~3 or between 3 and 4.8 dB. And because of Eq. 14 for small
values of P.F., this gives a similar difference in the radiation efficiency. This may
seem a very useful improvement, however as discussed in subsection “Far Field
Noise,” often such differences in antenna performance make little difference to the
overall performance of a receiver.
For body-borne VHF and UHF loop applications, a rectangular loop is
particularly of interest (see subsection “Proximity Effects and Detuning”). Then as
pLA ¼ 2 lx þ ly where lx is the length across the loop and ly is the width across the
loop (see Fig. 2), the formula for Ohmic losses simplifies to

jI 0 j2
PΩ rectangular strip loop ¼ lx þ ly ρS β (38)
lz
Loop Antennas 735

Combining this with Eq. 26 gives

 
8Z0 π 2 jI 0 jlx ly 2
1:48  1023 f 2 lz lx ly
7 2
3 λ2
P:F: rectangular strip loop ¼ ¼ pffiffiffiffiffi (39)
jI 0 j2 β ρr lx þ ly
lx þ ly ρS β
lz
where f is the frequency (in hertz); lx, ly, and lz are the loop dimensions in meters; ρr is
the bulk resistivity relative to that of annealed copper (1.72  108 Ω m); and the
formula for surface resistivity in terms of bulk resistivity ρ has been applied:
pffiffiffiffiffiffiffiffiffiffi
ρS ¼ πf ρμ (40)

Equation 39 has been used in industry for early evaluation of design choices in pager
receiver design. In this context, it has been found that the losses associated with the
tuning capacitors can also be accounted for by increasing the estimate for β to around
3 (Massey 2001).
Electrically small coils. The electric field from coils can be approximated by
considering them as a collection of small loops spaced along the coil axis parallel to
the z-axis. For an N-turn coil, the radiated field from the coil is given by

Eϕcoil ¼ Ncrect I 0 ALA sin θ (41)

and
 
Nπ jI 0 jALA sin θ 2
PRad coil ¼ (42)
λ2 R
The total power radiated is given by
 
8Z 0 π 3 NI 0 ALA 2
PRadT coil ¼ (43)
3 λ2
The directivity is given by

4π PRad coil 3 sin2 θ

Dcoil ¼ ¼ (44)
PRadT coil 2
This is the same as the formula for the directivity of an electrically small dipole. As
the above formula applies for all N, including N = 1, a loop has the same directivity
as a coil. And as for loops and coils, PRad and PRadT depend only on the total current
I0 and are independent of conducting wire or strip’s cross section; the directivity is
also independent of wire or strip size and cross section.
736 P.J. Massey et al.

The Ohmic losses are

PΩ coil ¼ Nρw pLA jI 0 j2 (45)

So,
 
8NZ 0 π 3 ALA 2
P:F:coil ¼ (46)
3ρw pLA λ2

At first sight, Eq. 46 appears to state that coils have the advantage over loops that their
power factor is improved by a factor of N, where N is the number of turns. While this is
true if one was restricted to only using a particular wire, to see that this is not generally
true, consider if instead of using a loop with one strand of wire in the turn, one had
N strands. The total current I0 stays the same, and Eq. 27 holds. However, assuming
that the wire diameter is significantly more than the skin depth, then the Ohmic losses
are proportional to the inverse of the perimeter of the conductor cross section and are
reduced by a factor of N. So an N-turn coil made with one strand of wire has no
performance advantage over a loop made from N strands of wire.
In practice loop antennas are rarely made from multiple strands of wire. Using
metal strips often makes better use of the space that would be occupied by the
multiple wire strands.
The reason that coils are popular is that their impedance is higher than loops. This
reason was particularly important for wireless receiver applications when valves were
used. The alternative to using a coil antenna for those applications was to include an
impedance transformer, which is itself made using coils and is relatively expensive.
Because the radiation from an electrically small loop or coil is proportional to the
loop or coil cross section’s area squared, while the radiation from a dipole is
proportional to the dipole length squared, electrically small loops rapidly become
more inefficient than dipole antennas as the antenna dimensions are reduced. How-
ever, despite the better efficiency of the dipole-type structure, loop antennas find uses
because they have a number of advantages. These advantages are related to:

• Man-made electrical noise: The near fields of human-generated interference are

primarily electrical. Therefore, antennas which are insensitive to electrical near
fields pick up less noise.
• Quality factors of tuning components: Electrically small loop antennas require
capacitors for tuning components. Generally these have higher quality factors
than the inductors used by electrically small dipole antennas and therefore
introduce less loss.
• Proximity detuning: Nearby objects such as metal plates and human bodies can
detune an electrically small antenna. The amount of detuning depends upon the type
and orientation of the antenna, and the shape and preferred orientations of some loop
antennas are better for use in many VHF and UHF body-mounted applications.

The following sections go into these topics in more detail.

Loop Antennas 737

Receiving Antennas and Man-Made Electrical Noise

When Marconi made his first successful transatlantic communications in 1901, for
the receiving antenna he used a long wire antenna (effectively an electric monopole)
suspended from a kite. Yet by the 1930s, loop antennas were used in practically all
consumer radio receiver sets. The reason for this change in antenna type was that in
the interim three decades, man-made electrical noise had increased enormously.
When this noise is from nearby sources and the receiver is in the near field of the
sources, the noise usually has a much higher electric field, and loop antennas have a
clear advantage because of their insensitivity to electric fields – see subsection “Near
Field Noise.” Noise sources further away reduce the performance advantages of
larger and less lossy antennas – see subsection “Far Field Noise.”

Near Field Noise

For noise sources closer than the order of a wavelength away, the noise is predom-
inantly electric field. Therefore, a loop antenna, being more sensitive to the magnetic
field, picks up less of the near field noise (Blok and Rietveld 1955). The excess of
electric field in near field noise is particularly strong at long and medium wave-
lengths and is sometimes called the “aerial effect.”
In suppressing near field noise, it is very important to ensure that the antenna is
truly behaving as a loop antenna and that it is not inadvertently coupling to nearby
conductors in such a way to form a dipole antenna. This issue is discussed in
subsection “Proximity Effects and Detuning.”

Far Field Noise

Electrically small loop and coil antennas are generally less efficient (more lossy) than
long wire antennas. However, by the 1930s, the high levels of man-made environ-
mental noise were making the performance reductions due to antenna losses much less
significant. A full analysis of this would require a discussion of noise figures and their
application to system noise analysis [see, e.g., ITU (2001)], which is beyond the scope
of this chapter. However, the mechanism by which high levels of environmental noise
reduce the effects of antenna losses can be explained with some examples.
The ability of a receiver to accurately detect a signal transmission is limited by how
much the signal’s power level PSO is at the receiver input compared to the random
noise power level PNO at the input. If the antenna were lossless, then these powers
would correspond to the signal power PSI and the noise power PNI picked up by the
antenna. However, antenna losses degrade the ratio PSO/PNO. Therefore, it is of interest
to examine how the signal-to-noise changes with antenna losses. One is interested in
PSI
PNI PSI PNO
fa ¼ ¼ (47)
PSO PSO PNI
PNO
738 P.J. Massey et al.

(If the input and output powers of a two-port circuit were being discussed, the ratio fa
would be called the noise factor, and its value in dB would usually be called the noise
figure. However, the term “antenna noise figure” is normally applied to different
quantity that describes the excess of environmental noise over thermal noise, and so
to avoid confusion, the terms “antenna noise factor” and “antenna noise figure” will
not be used in the remainder of the subsection.)
With a perfect antenna, the signal incident upon the antenna would be transferred
from space to the antenna terminals, and no additional noise would be introduced:
PSO = PSI, PNO = PNI, and fa = 1. Unfortunately, real antennas have attenuation due
to resistive losses. This degrades both the signal and the noise from the atmosphere.
It also introduces noise that is proportional to the temperature of the antenna.
Consequently, for real antennas, fa > 1. If the attenuation is represented by g <
1, then PSO = gPSI, and the noise power at the output transferred from space is =
gPNI. The total noise power at the output is that transferred from space plus the noise
created by the antenna resistance: PNO ¼ gPNI þ ð1  gÞkT 0 B where k is
Boltzmann’s constant, T0 is the temperature of the antenna structure, and B is the
bandwidth across which the power is being measured. Substituting into the formula
for fa gives

1 gPNI þ ð1  gÞkT 0 B ð1  gÞkT 0 B

fa ¼ ¼1þ (48)
g PNI gPNI

Suppose a well-matched receiving antenna is lossy – e.g., it is only 10 % efficient.

Then g = 0.1 and fa ¼ 1 þ 9kT 0 B=PNI . If the incident environmental noise power is
comparable to the antenna structure’s noise temperature: PNI = kT0B, then fa = 1 +
9 = 10, or expressed in dB, the antenna losses reduce the signal-to-noise ratio by
10 dB. So for this low environmental noise example, the signal is attenuated by
10 dB, and the 10 dB attenuation of the environmental noise attenuation is made up
by the noise introduced by the antenna resistance.
Now consider the same receiving antenna in a high environmental noise envi-
ronment. If the environmental noise level PNI ¼ 9kT 0 B then fa ¼ 1 þ 9  ð1=9Þ ¼ 2.
Expressed in dB, the signal-to-noise level has degraded by 3 dB. The presence of
high environmental noise reduces the increase in signal-to-noise level due to the
antenna. In order to achieve the same 3 dB degradation level with PNI = kT0B, the
antenna efficiency would have to be 50 %.
In both urban and rural areas throughout the populated parts of North America,
Europe, and in many other regions, man-made electromagnetic noise has reached
significant levels. At 100 MHz in these regions, the noise power that an antenna
captures is often a hundred times or 20 dB above kT0B. For a 10 % efficient antenna
in this environment, the signal-to-noise ratio degrades by just 0.37 dB. In order to
achieve this limited degradation with PNI = kT0B, the antenna efficiency would have
to be 92 %.
For the reason explained in the next paragraph, man-made noise power contrib-
uting to PNI increases as the frequency reduces. So the contribution of man-made
noise to reducing the performance degradation caused by low antenna efficiency is
Loop Antennas 739

more marked in the lower frequency bands. Conversely, for the upper UHF band and
higher frequencies, man-made noise power is so low that the performance of a
receiving antenna is strongly affected by its antenna efficiency – i.e., an ndB antenna
loss increases the signal-to-noise level by ndB.
Atmospheric man-made radio frequency noise has a power density spectrum
(measured in watt/m2/Hz) with a frequency dependency that varies with location.
However, the amount of man-made noise power (measured in watt/Hz) picked up by
a receiving antenna is the product of both the power density spectrum and the
antenna’s capture area, and PNI is the noise power received before the antenna losses
are accounted for. The reciprocity relations between a transmitting and a receiving
antenna and the Friis equation show that a lossless receiving antenna’s capture area is
proportional to the wavelength squared or equivalently to the inverse of the fre-
quency squared. This inverse square dependency attenuates PNI with increasing
frequency faster than any growth in power density spectrum with frequency and
means that regardless of location, the amount of man-made noise power PNI
increases quickly as the frequency is reduced. See Wagstaff and Merricks (2003)
for a range of examples of how PNI varies with frequency.

Tuning Components and Quality Factors

This subsection is about quality factors of components and circuits and why they are
significant in discussing the performances of electrically small dipoles and loops. It
also contains a discussion of how the inductance of the electrically small loop
antenna and its tuning capacitance is found.

Quality Factors
Inductors are required to tune electrically small dipoles, and capacitors are required
to tune electrically small loop antennas. The resistive losses associated with these
tuning components have a detrimental effect upon the antenna performance. Com-
ponent manufacturers specify the losses in terms of quality factors as follows. The
quality factors Qc of a capacitor and QL of an inductor are defined as:

2ωEC 1 2ωEL ωL
QC ¼ ¼ , QL ¼ ¼ (49)
PC ωCRSC PL RSL
where EC is the time-averaged energy stored in the capacitor, PC is the time-average
power dissipated by the capacitor, C is the capacitance, RSC is the resistance that
dissipates power losses PC when put in series with a lossless capacitance C, EL is the
time-averaged energy stored in the inductor, PL is the time-average power dissipated
by the inductor, L is the inductance, and RSL is the resistance that dissipates power PL
when put in series with an ideal lossless inductance L.
The quality factor of a circuit is the ratio of its stored energy to the energy
dissipated per cycle. Since for a series or parallel LC circuit the total power
dissipated PT is
740 P.J. Massey et al.

PT ¼ P C þ PL (50)

Then

PT 1 1 1
¼ ¼ þ (51)
ω E T QT QC QL

where ET is the energy stored within the circuit, and as this energy moves between
the inductor and capacitor, ET = 2EC = 2EL.
The LC circuit is of significance because:

• An electrically small dipole behaves like a capacitor, which has Ohmic and
radiation loss resistances associated with it. It is tuned using an inductor.
• An electrically small loop or coil behaves like an inductor, which has Ohmic and
radiation loss resistances associated with it. It is tuned using a capacitor.

With this in mind, Eq. 50 can be written as

PT ¼ Pantenna þ Ptune ¼ PΩ þ PRadT þ Ptune (52)

where Ptune is the Ohmic losses of the tuning component, and as in subsection
“Radiation and Losses,” PΩ is the Ohmic loss of the antenna and PRadT is the radiated
power from the antenna. Then, Eq. 52 can be rewritten as

1 1 1 1 1 1
¼ þ ¼ þ þ (53)
QT Qantenna Qtune QΩ QRadT Qtune

Then, the ratio of the increase in Ohmic losses due to the tuning component to the
Ohmic losses with a perfect tuning component is
 
PΩ þ Ptune 1 1 Q
¼ QΩ þ ¼1þ Ω (54)
PΩ QΩ Qtune Qtune

The quality factors of most commercially available capacitors and inductors are
quoted in the manufacturers’ data sheets, and commercially available capacitors
typically have quality factors of a few thousand. Commercially available inductors
typically have quality factors ranging from 20 to around 200. Because electrically
small antennas are usually significantly bigger than tuning components, the currents
are spread across larger conducting surfaces, and so the Ohmic loss quality factor QΩ
of electrically small dipole is usually a little better than QC of a capacitor, and factor
QΩ of electrically small loop or coil is usually a little better than QL of an inductor. So
one finds that for an electrically small dipole,

QΩ 2000 PΩ þ Ptune
 and > 11 (55)
Qtune 20 to 200 PΩ
Loop Antennas 741

And for an electrically small loop,

QΩ 20 to 200 PΩ þ Ptune
 and 1 (56)
Qtune 2000 PΩ

So the capacitor used to tune an electrically small loop or coil usually increases the
losses much less than the inductor used to tune an electrically small dipole.

Finding the Inductance of an Electrically Small Loop or Coil Antenna

There are several methods for finding inductance. There are formulae for a variety
of loop shapes and coil configurations, e.g., see Balanis (2005) for a simple
circular cross section wire loop and Grover (2009) for a large selection of loop
and coil configurations. In recent years, electromagnetic simulators have been
useful and often more accurate than formulae in calculating inductance and can
model realistic loops in their environment, e.g., noncircular shapes and
varying cross sections that fit within the space available within their receiver,
accounting for the effect of nearby metalwork within the receiver. However,
often in practice, none of these theoretical methods account for all the contribu-
tions to the loop/coil’s inductance, which may include manufacturing tolerances
and construction variations. Consequently, measurement is the most reliable
method of determining the inductance. At radio and microwave frequencies, a
practical method of determining the inductance of the antenna loop/coil is to place
tuning capacitance across the ends of the coil and search using a network analyzer
for the lowest frequency resonance of the antenna. The inductance L is then
calculated from the resonant frequency f0 and the known capacitance value of
the tuning capacitor(s) C:

1
L¼ (57)
ð2πf0 Þ2 C

The way in which the network analyzer is used to determine the resonant frequency
depends on what connections are available on loop/coil antenna. If the network
analyzer is connected across the ends of the loop/coil, then a search is done for the
lowest frequency at which the magnitude of the impedance peaks. N.B. in this
situation, it is very important that the connection between loop/coil and network
analyzer uses a balanced line or balun; otherwise, the results will be contaminated by
the currents traveling down the outside conducting surfaces of an unbalanced feed.
Alternatively, if a matching circuit is already in place (see subsection “Matching”
for discussion of matching circuit types), then the network analyzer can be used to
find the lowest frequency at which there is a peak in the insertion loss.
The accuracy of the resonance search method described above to determine the
antenna inductance is often limited by the tolerances on the value of the tuning
capacitor(s). These vary with capacitor type and series but typically lie between 2 %
and 10 %. These tolerances also limit the accuracy to which the operating frequency
of an electrically small loop/coil antenna can be preset using off-the-shelf fixed
742 P.J. Massey et al.

capacitors. For fixed frequency antennas, it is a good practice to use an adjustable

trimmer capacitor in parallel with a fixed capacitor, so that the antenna’s frequency
can be adjusted prior to deployment.

Proximity Effects and Detuning

In this section, proximity detuning for a dipole is discussed first in order to set the
scene for explaining the benefits of a loop antenna.
For both electrically small dipoles and for electrically small loops and coil
antennas, detuning can occur through:

1. Capacitive coupling to nearby objects – This additional capacitance increases the

tuning capacitance.
2. Inductive coupling to nearby objects – Nearby metal surfaces support currents
that move in opposition to currents on the loop and act to reduce the inductance.

So capacitive proximity coupling lowers the resonant frequency, and inductive

proximity coupling raises the resonant frequency. Some examples are described
below.
All the following examples show a conducting sheet in their diagrams. These are
included to represent the surface of a conductor, which could be a thin sheet
(as shown) or a thick conducting body such as a block of metal. As the human
body is composed of high dielectric lossy conducting fluids and tissues, it also has a
similar effect upon the antennas.
Figure 3 shows a dipole perpendicular to the surface of a conducting body. The
body couples capacitively to the dipole via the high electric fields at the dipole’s
lower tip. However, this capacitance has little effect upon the tuning as the
conducting body is relatively far away from the upper arm.
In contrast, Fig. 4 shows the situation when a dipole is placed parallel to the
surface of a conducting body. The capacitance between each of the tips and the
conducting body is much higher than the capacitance between the tips. Therefore,
the presence of the conducting body provides a high capacitance path between the
tips that detunes the antenna. Moreover, the current excited in the conducting body
flows in the opposite direction to the current in the dipole and so cancels out its
radiation.
The consequence of the above is that dipole-type antennas are suitable for use
near conducting bodies, providing that they are oriented perpendicular to the surface
of the body. For human body-mounted devices, this would mean that the antenna
sticks out, which is generally considered undesirable. As is discussed below, loop
antennas offer an alternative that gives a low profile above the body.
Figures 5 and 6 show some examples of proximity coupling in electrically small
loops. The figures use a loop that is long and relatively narrow, which is a common
shape as it is suited for fitting in devices such as pagers that fit in pockets next to a user’s
body. See Fig. 9 for an example on how this type of loop is installed in a radio pager.
Loop Antennas 743

Fig. 3 Dipole perpendicular to the surface of a conducting body. The black lines show the
approximate route of field lines between the tips and from the tips to the conducting surface. The
distance along the field lines across space from one tip to the other tip via the conducting surface
(the path on the left + the path in the lower center) is greater than the distance across space running
directly between the tips. This indicates that the additional capacitance between the tips due to the
presence of the conducting surface is less than the capacitance between the tips in free space

When mounting loop antennas in body-worn devices, to avoid the detrimental

effects of the inductively induced currents discussed in Fig. 5, it is important that the
plane of the loop is perpendicular to the body’s surface. And to avoid the capacitive
detuning seen in the arrangement shown in Fig. 6, it is important that the tuning
capacitor is mounted at the end of the long side of the loop. Figure 7 shows the
preferred location for the tuning capacitor and for the orientation of the loop when
next to a conducting body.
Besides the issues of detuning and of reduction in radiation, proximity coupling
can also increase the coupling to noise. This can occur in two ways. The first is if the
nearby structures are carrying noise. Mains cables are a common example; besides
carrying 50–60 Hz mains signals, their currents can be contaminated with radio
frequency noise picked up from connected equipment. The second way is if capac-
itive coupling adds a dipole-type antenna structure to the loop/coil. Then the dipole
structure will pick up electric near field, and the near field noise-reducing advantage
of using a loop (section “Near Field Noise”) is diminished.
The capacitance available for capacitive proximity coupling is proportional to the
size of the loop/coil. The effect of the proximity capacitance is dependent upon the
inductance of the loop/coil and its tuning capacitance. If it is a large loop or a coil
with a large number of turns, then the tuning capacitance will be less, and the
proximity capacitance will be more significant. The large amount of metal wire
associated with the high number of turns also increases the proximity capacitance
compared with a few turns or a single turn of the same gauge of wire.
744 P.J. Massey et al.

Fig. 4 Dipole parallel to the surface of a conducting body. The black lines show the approximate
route of field lines between the tips and from the tips to the conducting surface. The distance along
the field lines across space from one tip to the other tip via the conducting surface (the path on the
left + the path on the right) is much shorter than the distance across space running directly between
the tips. This indicates that the additional capacitance between the tips due to the presence of the
conducting surface is much greater than the capacitance between the tips in free space

Fig. 5 Inductive proximity coupling to loops. The presence of a conducting sheet parallel to the
plane of the loop (or perpendicular to the axis of a coil) reduces the loop’s (or coil’s) inductance
because currents are excited in the sheet that travel in the opposite direction to those in the loop.
Therefore, the antenna’s resonant frequency increases
Loop Antennas 745

Fig. 6 Capacitive proximity coupling to loops. Capacitance between the loop and the conducting
sheet adds to that of the tuning capacitor, reducing the resonant frequency. Placing the tuning
capacitor in the middle of a long side of a loop antenna is the worst location for it, because of the
propensity of the antenna to suffer from detuning due to capacitive coupling, by the same
mechanism as shown in Fig. 4. This can even happen if the sheet is placed next to the long side
without the tuning capacitor. For this reason, the tuning capacitor is normally placed at or near a
narrow side

Fig. 7 Preferred configuration and orientation of electrically small loop next to conducting body

In the early twentieth century, multi-turn coils were used to provide the maximum
voltage and high impedance across the tuning capacitor in order to directly drive the
grid of the valve of a broadcast radio receiver. It was known that using a single-turn
loop would reduce the effects of the proximity capacitance and reduce the amount of
near field noise. However, in order to meet the impedance requirements for the input
stage of the valve, a transformer would have been required, which would have added
cost to the consumer product [see Blok and Rietveld (1955)]. The incorporation of a
soft ferrite high-permeability core permitted a reduction in the diameter of the coil,
which reduced the amount of proximity capacitance.
Advice on the use of soft ferrite cores can be found in Blok and Rietveld (1955)
and Snelling (1988).
746 P.J. Massey et al.

Soft ferrite has been successfully used to boost the performance of coil and loop
antennas at frequencies up to VHF. However, the permeability of soft ferrite reduces,
and the losses increase with increasing frequency (van der Zaag 1999), and beyond
VHF air-cored loops and coils give the best performance.

Matching

Figure 8 shows some possible matching arrangements. It can be shown that most of
these have very similar properties [see Massey (2009)], and therefore the choice of
matching arrangement is determined by considerations other than which gives the
best radio frequency (RF) performance. As examples:

• The direct connection across the tuning capacitor shown in Fig. 8a gives the
highest impedance.
• The capacitor bridge shown in Fig. 8b is easily implemented as a circuit on a
printed circuit board (PCB) to which the end of the loops is attached. See Fig. 9.
• Directly tapping across the loop or coil in Fig. 8c is very practical for experimen-
tal work, as the impedance match can be adjusted by moving the tap points. See
Fig. 10.
• Inductively coupling into the loop using a small coupling loop in Fig. 8d builds
the balun into the coupling structure. This is usually achieved at the cost of a small
loss (~1 dB or less typically) due to the imperfect trapping of the magnetic flux
between the antenna and the coupling loop.

The first three of these examples require either that they are connected to a
balanced port or that a balun is used.

Fig. 8 Some matching arrangements for a loop or coil antenna

Loop Antennas 747

Fig. 9 Implementation of a capacitor bridge circuit in a pager receiver. The picture shows a radio
pager with its casing and battery removed. The loop antenna is silver plated for maximum
conductivity. The matching circuit is basically that shown in Fig. 8b, and the receiver has a balanced
input port. However, in order to allow the tuning capacitance and matching capacitors to be
mounted on the upper PCB, the ends of the loop are soldered to tracks on the PCB. The tuning
capacitance consists of fixed capacitors in parallel with the trimmer capacitor. This is done because
a fixed capacitors usually have greater Q than adjustable capacitors, and, b as most of the tuning
capacitance is supplied by the fixed capacitors, the tuning range just covers the range necessary for
adjustment (rather than a range of perhaps twice the center frequency, which would the case if the
trimming capacitor was relied on to provide all the tuning capacitance). Therefore, the change in
tuning frequency with rotation angle of the trimmer is less, and it is easier to bring the antenna into
tune. The vibrator motor and its weight to the right of the PCB are mounted with the weight nearest
the loop. This is because the permanent magnet in the motor introduces significant losses if placed
close to the loop antenna

Full-Wave Loops

Introduction

The electrically small antennas of the last section always have to be tuned with a
capacitor to make them work at the required frequency. This section discusses loops
that are large enough to be self-resonant, i.e., have a finite resistance and very low
reactance between terminals that are across a gap in the loop. As the loop size is
increased, self-resonance first occurs when the perimeter of the loop is around one
wavelength long, i.e., the full-wave loop. In contrast to electrically small loops, full-
wave loops generally have good efficiency and can be used for efficient transmission
as well as for reception.
748 P.J. Massey et al.

Fig. 10 Directly tapping across a loop using a balanced feed. The balanced line consists of two
50 Ω coaxial semirigid cables soldered together. The inners of the cables are fed from a 3 dB 180
splitter that converts an unbalanced 50 Ω feed into two 50 Ω signals that are 180 out of phase with
each other, giving a balanced impedance of 50 + 50 = 100 Ω. To enable easy connection and
disconnection between line and antenna, some sockets have been soldered onto the antenna, and
wires which fit into the sockets have been soldered to the ends of the inner conductors of the coaxial
cables. This arrangement is used for measuring the performance of different shapes of loops and for
measuring the effects of bringing components close to the loop. The single adjustable tuning
capacitor gives a very large tuning range but is very sensitive. For mass production, tuning using
fixed capacitors in parallel with an adjustable capacitor is preferred. See Fig. 9

This section is arranged as follows. Subsection “Resonant Full-Wave Loop”

discusses the basic full-wave loop structure, emphasizing that its radiation resistance
is typically ~200 or more Ohms. The subsequent subsections detail how other
features can be added to optimize full-wave loop behavior and tune the performance.
Subsection “Resonant Full-Wave Loop in Close Proximity to Metallic Structures”
describes how the full-wave loop’s natural resonant impedance can be reduced to
values suitable for directly connecting to low-impedance (e.g., 50 Ω) ports by
placing the loop around a conducting structure. To demonstrate the principle, a flat
conducting structure, such as the electronics of a tablet computer, is used as an
example. Then methods for frequency tuning (subsection “Tuning Resonant Loop
Antennas”) and creating circular polarization (subsection “Circularly Polarizing the
Resonant Loop Antenna”) are demonstrated.
Next, parasitic loops are introduced to modify the frequency coverage (subsection
“Dual Banding the Resonant Loop Antenna”). These can be combined with the
method for creating circular polarization as described in subsection “Dual Banding
and Circular Polarizing the Resonant Loop Antenna.”
The remaining subsections are devoted to describing methods for creating direc-
tive antennas with loops. Firstly, a loop over a ground plane is considered in
subsection “Resonant Loop Over Planar Reflector.” Subsection “Directive Resonant
Loop Array” discusses the spacing of parasitic elements to create a Yagi-Uda like
array with high directivity.
Another method of creating an efficient and relatively compact directive antenna
is to mount shared feed loops about a quarter wavelength above a ground plane. Both
linear (also known as biquad, described in subsection “Dual Rhombic/Circular Loop
Loop Antennas 749

Antenna Over Planar Reflector (Biquad Antenna)”) and circular polarized versions
(described in subsection “Circularly Polarized Dual Rhombic/Circular Loop
Antenna Over Planar Reflector”) are detailed. Finally, methods for bandwidth
extension are described in subsection “Wideband Circularly Polarized Dual Rhom-
bic Loop Antenna with Parasitic Loops.”

Resonant Full-Wave Loop

A full-wave loop structure can be almost any shape, for example, circular, square,
rhombic, and triangular, as illustrated in Fig. 11. The full-wave loop can be
constructed in three dimensions (e.g., from wire or a tube); it can have a planar
structure, for example, being manufactured from a sheet of metal or etched onto the
conductor of a printed circuit board. The full-wave resonant loop antenna is often
approximated by a pair of (often bent) half-wave dipoles connected in parallel, as
their current and voltage distributions are comparable to the full-wave loop. The
dipole pairs resonate in phase and are connected at their tips where the current is
almost zero. In Fig. 11, these points are denoted by red dots and labeled as Imin.

Fig. 11 Some examples of full-wave loop antennas. Around resonance, the radiation resistances of
the circular and square loop antennas are approximately 200 Ω, and the radiation resistance of the
folded half-wave dipole is approximately 300 Ω. The directivity is usually at a maximum perpen-
dicular to the antenna plane (perpendicular to the plane of the figure). For all the antennas depicted,
the radiation pattern is doughnut shaped, with its axis of minimum radiation running horizontally
across the figure. The radiation varies slightly with shape, and consequently the maximum direc-
tivities are around 1.5 (1.76 dBi) for the circular loop, 1.7 (2.3 dBi) for the square loop, and 1.64
(2.15 dBi) for the folded half-wave dipole
750 P.J. Massey et al.

Resonant Full-Wave Loop in Close Proximity to Metallic Structures

The moderately large characteristic impedance of the resonant loop at its feed point
(denoted as the loop characteristic impedance from here onward) can be exploited to
create an efficient radiator of significant bandwidth and low volumetric profile, by
intentionally coupling the loop to surrounding structures such as circuit boards or
metallic cases. To explain how this works, first consider a full-wave large loop
antenna in isolation designed to operate at 500 MHz, fed from a 50 Ω feed. The top
left diagram in Fig. 12 shows a rectangular loop example. In the figure, the loop is a
track shown in yellow against an FR4 printed circuit board shown in green, and the
location of the antenna feed is shown in red. The printed circuit board is 200 mm
long by 100 mm high.
At 500 MHz, it can be seen that the feed impedance is over 150 Ω, the return loss
is > 5 dB, and the realized gain is suboptimal at about +1 dBi. The large mismatch
between the antenna impedance and the 50 Ω impedance of the feed accounts for a
significant proportion of the low realized gain. (Realized gain is defined as radiated

Fig. 12 Basic rectangular full-wave loop on FR4 sheet. The figure shows, in clockwise order from
top left, the antenna (see text for description), the return loss response, the realized gain perpen-
dicular to the antenna’s board, and the reflection coefficient response between 400 and 600 MHz on
a Smith chart whose center is normalized to 50 Ω. In the Smith chart, a marker has been put at
500 MHz
Loop Antennas 751

Fig. 13 Basic full-wave loop on FR4 sheet with impedance modification. For key to diagrams, see
text below Fig. 12

power in a given direction compared to the power radiated by an ideal perfectly

matched ideal isotropic antenna fed with the same signal.) It should also be noted
that the size of the antenna at 500 MHz is such that it takes up a significant area.
Now consider the same antenna operating in the presence of other components. As
an example, Fig. 13 shows a copper ground plane (represented by the large yellow
rectangle) filling much of the antenna’s interior, which in this case represents a circuit
board. However, this structure could also be a metallic case or other component part
which is conductive in nature. The coupling which occurs between the loop and in this
case the copper ground plane results in modification of the loop characteristic imped-
ance. It is therefore possible to use this additional structure to improve the loop
antenna’s performance; hence, it is considered to be a cooperative structure.
Considering the example of Fig. 13 and comparing the responses to Fig. 12, it can
be seen that the addition of the ground plane has led to a significant modification in
loop characteristic impedance. The resulting impedance is now much closer to the
desired 50 Ω of the feed, and thus the cooperative structure results in an improve-
ment in both match and overall performance of the antenna. Examining the
remaining results of Fig. 13 shows that at the design frequency of 500 MHz, the
resonance is significantly stronger resulting in a return loss of < 20 dB. The
realized gain has also increased to over 2.5 dBi.
752 P.J. Massey et al.

Fig. 14 Resonant loop antenna around the screen of a tablet computer. This diagram is taken from
CAD file. The tablet’s rear cover has been removed to show the loop, which is highlighted in yellow,
and its width has been exaggerated (doubled) to aid visibility. The loop travels around the interior of
the perimeter of the tablet’s PCB but is hidden from view along most of the two nearest sides by the
lip of the tablet’s front cover. The feed location is shown in red. A silver-colored rectangular sheet
represents the aluminum can covering the screen driver circuitry. The tablet’s screen cannot be seen
as it is on the far side of the view. See Fig. 15 for a more detailed CAD model with the main board
populated with circuitry and battery and Fig. 16 for the real-life demonstration piece and its
measured results

Application Example
Applying the principle of what is illustrated in Fig. 13 to a real-life example, consider
a small tablet computer. For this example, a single-turn resonant loop antenna has
been inserted around the perimeter of the tablet. For this instance, the loop antenna is
designed to operate at around 400 MHz. The loop has been positioned to closely
couple to the remaining structures in the tablet, in order to modify the impedance
appropriately. A simplified computer aided design (CAD) view of the antenna
incorporated into the tablet computer is illustrated in Fig. 14, where the loop is
highlighted in yellow and closely follows the perimeter of the tablet computer’s case.
For this demonstration example, the loop has been designed to operate in the
presence of the touch screen, battery, circuit board, and the rear case containing
Wi-Fi antennas and a large NFC (Near Field Communication) antenna.
The top left-hand corner in Fig. 15 shows a more detailed CAD image of the
tablet computer with the main board populated with circuitry and battery. The
“chassis” component shown in Fig. 15 was in its original unmodified incarnation
manufactured in a lightweight metallic alloy; however, to implement the perimeter
loop antenna, a structural plastic has been used for this component, which prevents
shorting out of the loop antenna’s fields. The remainder of Fig. 15 shows the
simulated results obtained from the installed antenna.
In this demonstration example, and without additional matching, the impedance
of the antenna is approximately 25 Ω at the center frequency (410 MHz), and the
realized gain is 0.7 dBi, with a 3 dB gain bandwidth in excess of 50 MHz. Four
hundred and ten megahertz corresponds to the natural resonant frequency for this
Loop Antennas 753

Fig. 15 Simulation results of a full-wave antenna around the screen of a tablet computer. The
figure shows, in clockwise order from top left, i) the CAD model of the tablet computer with its back
removed, (ii) the realized gain pattern in the plane perpendicular to the plane of the tablet which cuts
the tablet along its long axis, (iii) the realized gain in the direction perpendicular to the tablet plane
versus frequency, and (iv) the S11 response between 380 and 440 MHz plotted on a Smith chart
whose center is normalized to 50 Ω. In the Smith chart, a marker has been at 410 MHz, which
corresponds to the resonant frequency realized by this antenna configuration

particular antenna. Tuning to other frequencies is possible and is discussed in

subsection “Tuning Resonant Loop Antennas.”

Tested Example and Range Plots

A real-life demonstration example was manufactured consisting of a tablet computer
with a retro-fitted loop antenna, as illustrated in Fig. 16a. The tablet computer is
shown with its back cover and NFC antenna removed for clarity. The realized gain of
the installed antenna configuration is measured and illustrated in Fig. 16b.
Figure 16a also indicates the orientation of the axes used in the measurement. For
the purposes of the measurement, the antenna is mounted “upside down” compared
to the image in Fig. 16a so that its test cable hangs down and runs to the chamber’s
feed cable without crossing the loop antenna.
754 P.J. Massey et al.

Fig. 16 Tablet computer and its radiation measurement results. See text for description of the
figure
Loop Antennas 755

The actual resonant frequency of this structure occurs at 425 MHz which is
slightly higher in frequency than predicted. This is because several simplifications
have been made in the simulation, in order to be able to realize an initial design.
Much of the detail, which is present in practice, has not been replicated in simulation.
However, the measured magnitudes of the return loss and gain responses (not
shown) are similar to those of the simulated model. For the purposes of demonstrat-
ing the principle of cooperative structures, the realized gain patterns at 425 MHz are
illustrated in Fig. 16b.
It can be seen from both the simulation results in Fig. 15 and the measurement
results in Fig. 16 that the full-wave loop structure with its naturally high impedance
is ideally suited to applications where an antenna is asked to coexist with metallic
structures. These metallic structures can be used in a cooperative manner with loop
structures to enable a good installed antenna performance. For other more conven-
tional antennas such as patches or printed inverted F antennas (PIFAs), the presence
of metallic structures has a more detrimental impact on the installed antenna
performance.

Tuning Resonant Loop Antennas

It has already been shown that a resonant loop antenna can be considered as
behaving like a pair of connected dipole antennas (see subsection “Resonant Full-
Wave Loop”). This concept can be thought of as the motivation for the following
strategy to reduce the resonant frequency of a large loop antenna. The two high-
current nodes of the loop antenna are in the centers of its constituent dipole antennas.
As illustrated in Fig. 17, i.e., they are at the feed and at a point about halfway around
the antenna from the feed.
Placing an inductor at the high-current node halfway around the loop from the
feed increases the electrical length of the antenna and lowers its resonant frequency.
An example of this arrangement is shown in Fig. 18. If the inductor had been placed
closer to the left or right sides of the example in Fig. 18, its effects upon the current
distribution and electrical length would have been less.
The effect of inductor values upon the return loss of the loop is illustrated in
Fig. 19. A change in the resonant frequency of the antenna is clearly seen.
It can be seen in Fig. 19 that the impedance modified full-wave loop structure
responds well to inductive loading, retaining useful matched bandwidth at all
illustrated tuned frequencies. Of course, in a real-life example, the loop to structure
coupling would be set with a best match at the center of the tunable frequency range
in order to ensure the best possible match at all operating frequencies.

Circularly Polarizing the Resonant Loop Antenna

The resonant loop antenna can be further adapted to include a circular polarization
response by the inclusion of capacitive gaps in the tracking as illustrated in the
756 P.J. Massey et al.

Fig. 17 High voltage and current points on full-wave dipole antennas

Fig. 18 Location of
inductive load at high-
current node

example of Fig. 20. These capacitive breaks in the loop structure give rise to a
traveling wave current distribution of constant amplitude and linearly changing
phase. Circular polarization is normally achieved by breaking the tracking in one
place in the loop. For the case shown in Fig. 20, the single break version would retain
the break that is closest to the feed point. However, Fig. 20 shows a variation with a
Loop Antennas 757

Fig. 19 Tuning response of a resonant loop antenna with switched inductors. Key, 0.1 nH response
is in red, 10 nH response is in orange, 32 nH response is in green, and 47 nH response is in blue

Fig. 20 Capacitor placement

for a circularly polarized loop
antenna

pair of breaks as this allows the designer greater scope for adjustment. In this
uncluttered example, the antenna’s loop is almost exactly one wavelength in cir-
cumference, and the capacitors’ respective positions are typically at 1/6 λ and +1/3
λ from the feed point, i.e., λ/2 apart. It is important to note that these locations will be
significantly influenced by the coupling of the loop to the structure and also by the
impact of the substrate and surrounding materials, so should be taken as guidance
only.
It is possible to flip the polarization between right and left handed by mirroring of
the capacitive breaks in the loop structure from the lower to the upper half of the loop
and vice versa. The circularly polarized response can be tuned simply by varying the
value of the capacitance and is achieved electronically using varactor diodes or
758 P.J. Massey et al.

Fig. 21 Circularly polarized response of the capacitive loaded resonant loop antenna. The figure
shows, in clockwise order from top left, (i) a three-dimensional image of the antenna that is shown
in plan view in Fig. 20, with the cut plane for the radiation patterns in (iv) shown in gray; (ii) a plot
of realized gain in the z-direction versus frequency; (iii) the reflection coefficient response plotted
from 450 to 550 MHz on a Smith chart whose center is normalized to 50 Ω; and (iv) the right- and
left-hand circularly polarized components of realized gain at 510 MHz in the cut plane shown in
gray in i). Blue represents left-hand circularly polarized and red represents right-hand circularly
polarized. In (ii) the total realized gain is shown in black. In the Smith chart, the curve runs from
450 to 550 MHz, and a marker is placed at 510 MHz

similar devices. The gain and reflection coefficient plots for the circularly polarized
loop illustrated in Fig. 20 are shown in Fig. 21.

Dual Banding the Resonant Loop Antenna

One method for dual banding the resonant loop antenna is to proximity couple one or
sometimes more secondary loops to the primary driven loop. Secondary loops can be
placed either inside, outside, or both inside and outside the driven loop and effec-
tively act as parasitic resonant structures. The example shown below builds on the
cooperative structure-loaded loop described at the start of subsection “Resonant
Full-Wave Loop in Close Proximity to Metallic Structures” and uses a secondary
loop outside the primary loop that has been inductively loaded to further reduce the
frequency of the secondary band.
Loop Antennas 759

Fig. 22 Dual banding the

resonant loop antenna

Fig. 23 Realized gain of single and dual loops showing the two resonances realized by the dual
loop configuration shown in Fig. 22. The single-loop response is taken from the results shown in
Fig. 13

Note that the secondary loop is coupled predominantly along the high-current
sections of the antenna, with the high-voltage (low current) sections to the driven
loop more widely separated from the secondary loop; this prevents the collapse of
the electric field at the high-voltage nodes of the primary loop. Figure 23 illustrates
the realized gain of the dual loop structure, compared with the single-loop response
taken from Fig. 13.
760 P.J. Massey et al.

Fig. 24 Dual banding and

circularly polarizing the
resonant loop antenna

It can be seen in the gain versus frequency plot in Fig. 23 that the lower-frequency
response is typically of a higher quality factor (has a narrower peak) than that of the
higher frequency band. This is due not only to a reduction in the antenna’s electrical
size (size in free-space wavelengths) at lower frequencies but is also a function of the
coupling action between the loops.

Dual Banding and Circular Polarizing the Resonant Loop Antenna

It is possible to combine two or more of the methods described in the preceding

subsections, namely, inductive loading, dual banding, and circularly polarizing. An
example is illustrated in Fig. 24. It should be noted here that the parasitic loop has not
been capacitively loaded, as the circularly polarized wave from the primary loop
couples strongly into the secondary (parasitic) loop. In other instances to obtain
circular polarization where the coupling is less strong, it will be necessary to
capacitively load the parasitic loop.
Figure 25 illustrates the gain versus frequency response of the antenna illustrated
in Fig. 24. The gain is reported for the direction out of the plane of the antenna in
Fig. 24 towards the viewer. If the plot were done for the opposite direction into the
plane of the antenna, the red and blue traces would be interchanged.

Resonant Loop over Planar Reflector

The full-wave resonant loop element is often used over a ground plane in order to
increase its performance in both directivity and gain. The moderately large real input
impedance of the loop element in free space is useful, as the proximity of the
reflector can be used to adjust the feed impedance of the antenna without adjusting
Loop Antennas 761

Fig. 25 Dual band circularly polarized response of the primary and secondary loops. Key, black =
total gain, blue = left-hand circular polarization, and red = right-hand circular polarization

Fig. 26 Directivity and realized gain of resonant loop over infinite ground plane and over planar
reflector. See text for the key to the plots and for details of the antenna and reflecting plane

the loop circumference. This allows easy matching to 50 Ω, and thus to being fed
from a standard coaxial cable. Figure 26 shows the gain and directivity perpendicular
to the reflecting plane, where the radiation is plotted for the loop side of the reflecting
plane, in the direction perpendicular to the plane. The directivity is depicted using
dashed lines and the realized gain with solid lines. As can be seen in Fig. 26, the
single-turn resonant loop has a significantly improved directivity when placed over a
reflecting surface.
762 P.J. Massey et al.

In Fig. 26, the response over an infinite ground plane is shown in black, and the
response over a 2λ/3 long square planar reflector is shown in green. The loop is a flat
ring with interior diameter of 0.28λ and exterior diameter of λ/3. The loop and
reflecting surface have parallel planes, which are separated by the separation dis-
tance, shown on the figure’s horizontal axis. Some interesting phenomena can be
seen in Fig. 26:

• The loop is best matched to 50 Ω when it is around 0.11 wavelengths from the
reflecting surface.
• At small separations, the realized gains and the directivities of the infinite plane
and finite reflector cases are very similar, but for a given separation the infinite
plane values are slightly higher.
• At half-wavelength separation, the radiation vanishes for the infinite ground
plane.
• The radiation of the finite ground case has a minimum at a slightly higher
separation but never completely vanishes.
• With increasing separation, the next maximum in gains and directivities is at just
over 0.7λ separation. For the infinite ground plane, its magnitude is similar to that
of the 0.11 wavelength peak, but in the finite reflector case, it is somewhat
reduced.

Directive Resonant Loop Array

So far, only coupled loops lying in the same plane have been discussed, such as that
shown on the left-hand side in Fig. 27, which shows a central loop as the driven
element, with the outer loop being the low-frequency parasitic element and the inner
loop being the higher-frequency parasitic element. Moving the parasitic loops away
from the driven loop, as shown on the right-hand side in Fig. 27, gives opportunities
for enhanced gain. In a manner similar to parasitic elements in conventional

Fig. 27 In-plane coupled loops and the directive out-of-plane derivative

Loop Antennas 763

Yagi-Uda arrays of dipoles, the lower-frequency loops become reflectors and the
higher-frequency loops act as directors. The reflector can also take the form of a plate
(as illustrated in Fig. 26).
It is possible to give some typical rule of thumb sizes for the circumferences of the
different elements in order to achieve a directive resonant loop array. They are:

• Reflector: 1.05 wavelengths

• Driven element: 1 wavelength
• Directors: 0.95 wavelengths

The reflector to driven element spacing and driven element to director spacing are
highly dependent on the number of elements, and the spacing follows closely the
pattern set for the design of a conventional Yagi-Uda antenna of linear elements. In
the Yagi-Uda configuration, the reflector is separated from the driven element by
approximately 0.15 wavelengths, and there is approximately 0.2 wavelength sepa-
ration between both adjacent directors and also between the driven loop and its
nearest director.

Dual Rhombic/Circular Loop Antenna over Planar Reflector (Biquad

Antenna)

The biquad takes several forms, the most notable of which are the parallel- and
series-fed variety, shown in Fig. 28. These antennas have a higher gain than the
equivalent area patch antenna, and, because of their depth (i.e., distance between the
loops and their ground reflector, which is set to optimize the feed impedance), they

Fig. 28 Series- and parallel-fed versions of the biquad antenna

764 P.J. Massey et al.

have significant bandwidths. This coupled with their ease of construction has
contributed to their widespread adoption by the hobbyist radio community for the
Wi-Fi band. Each loop is typically one wavelength in circumference (measured
around the outside of the loop) at the lowest design frequency, with the inner
circumference setting the higher part of the resonance band.
Figure 28 also shows two possible feed configurations: a directly connected feed
as shown in the series configuration and a capacitive feed as illustrated in the
expansion of the parallel configuration. In the capacitive feed arrangement, the
center pin of the feed coax is connected to a small capacitive plate, which is located
on the opposite side of the board to the metallization which forms the loop (under-
side of the PCB in Fig. 28). Thus the loop is fed by capacitive coupling to this plate.
This configuration provides additional capacitance at the feed point which can be
useful for canceling out the additional inductance which results from the exposed
center conductor of the feed cable. This additional inductance can be particularly
substantial at higher frequencies; thus the capacitive feed variant is highly applicable
here. It should be noted that either feed configuration can be used with both the series
and parallel versions of the biquad.

Circularly Polarized Dual Rhombic/Circular Loop Antenna over

Planar Reflector

The rhombic version of the full-wave loop antenna has four linear sides each
measuring about 0.3 wavelengths long. To induce circular polarization, capacitive
breaks are introduced into the loop structure, similar to those discussed in subsection
“Circularly Polarizing the Resonant Loop Antenna,” and a traveling wave current
distribution of constant amplitude and linearly changing phase is realized. Capacitor
position is typically at 1/6th loop length from the feed point (depending on the
polarization required). The capacitance can be realized by breaks in tracking or
wiring or by using lumped components depending upon construction techniques
being employed.
Two circularly polarized rhombic loops are combined to form a dual rhombic
loop. The antenna is illustrated in Fig. 29, which shows two possible feed configu-
rations. For clarity, the examples are depicted without showing the ground plane that
lies below the loops.
In Morishita et al. (1998), axial ratio bandwidths (<2 dB) of more than 20 % are
achieved for both parallel and series feed antennas, with gains of 10 dBi for the
parallel-fed version and 11 dBi for the series-fed antenna.

Wideband Circularly Polarized Dual Rhombic Loop Antenna

with Parasitic Loops

This antenna design expands on the concept of the circularly polarized dual rhombic
loop discussed above and introduces a set of smaller parasitic elements within the
Loop Antennas 765

Fig. 29 Circularly polarized biquad concept. See text for description

Fig. 30 Wideband circularly

polarized resonant loop
antennas

main driven loops, shown in Fig. 30 (see Li et al. 2005). The parallel-fed structure
has a significantly wider circular polarization bandwidth (measured as the 2 dB
axial ratio) of 50 %, compared to 25 % for the series-fed arrangement.

Reconfigurable Multi-feed Loops and Their Arrays

Introduction

The previous section on full-wave loop antennas and their derivatives described how
loops can be fitted around or near the conducting structures of a device to increase
efficiency while adding very little extra volume to the device. This section describes
another way in which space can be saved, by reusing the radiating structure of a loop
antenna for several different radiation patterns by switching between feeds. In order
766 P.J. Massey et al.

to have a directive high-gain steering performance, the loop is placed above a ground
plane. Further, this section discusses how high-impedance surfaces can be used to
reduce the distance between the loop and the ground plane, while keeping good
efficiency and controlling the beam shape. Finally, the latest research on array of
beam steering loop antennas for achieving high-gain scanning range will be
presented.

Background

As mobile communications migrate up from 4G systems to 5G systems and beyond

with a promise of enabling consumer end devices (phones, tablets, etc.) to work in
excess of 1 giga bit per second, there is a tremendous focus on the development of
beam steering antenna technology in order to solve some of the channel problems.
The pattern reconfigurable antennas have the ability to generate multiple radiation
patterns and are capable of directing the beam toward the intended direction of signal
arrival and evade the noise sources. From a transceiver perspective, this means a
higher signal-to-noise system capable of offering a higher data throughput capacity.
The follow-on benefits include mitigation of multipath effects, reduction in trans-
mitted power levels, longer battery life, range enhancement, reduction in co-channel
interference (jamming avoidance), and faster handover. Single-element
reconfigurable antennas as presented here are of special significance as they do not
require multiple antenna elements with complex phase shifters as in the case of a
conventional phased array antenna (Liberti and Rappaport 1999; Pozar 1986). For a
near equivalent performance scenario, a single-element adaptive antenna compared
with a conventional phased array antenna normally is 75 % smaller, ten times faster,
and 95 % cheaper.
Early research and development on single-element beam steering/beam switched/
pattern reconfigurable antennas started around 2003–2004 (Huff et al. 2003; Mehta
and Mirshekar-Syahkal 2004). Various architectures (Huff and Bernhard 2006; Jung
et al. 2006; Mehta et al. 2006), mainly traveling wave types, to realize the single-
element steerable antennas were explored. These antennas employed multiple
switches to vary the current distribution along the peripheral length for achieving
reconfigurable radiation patterns. These early designs had three major restrictions.
Firstly, due to the different current distributions, the antenna polarization differs from
one switch configuration to another, which in effect causes polarization randomness.
In situations where a link is polarization sensitive, the polarization randomness can
simply make beam steering futile. Secondly, some configurations had many switches
which introduced significant RF loss. And finally, none of these designs offered
radiation uniformity from one pattern to another.
To overcome these limitations, another form of traveling antenna called symmet-
ric planar square loop antenna (SLA) for switched beam steering applications
(Mehta and Mirshekar-Syahkal 2007; Pal et al. 2008) was proposed (Fig. 31). Its
structural symmetry offered reconfigurable radiation pattern without any polariza-
tion and pattern variation from one beam to another. Since then, that particular SLA
Loop Antennas 767

Fig. 31 Square loop antenna with four feeding ports. The simulated antenna is shown to the left,
and the experimental prototype is shown on the right. (a) is a top view of the simulated antenna,
(b) is a side view of the simulated antenna, (c) is a top view of the experimental antenna, and (d) is a
bottom view of the experimental antenna. (d) shows one active SMA feed and three other open
circuited feeds

has evolved into a much more compact and efficient array configuration. A step-by-
step evolution of the antenna for achieving beam steering is presented next.

Radiation Mechanism for Beam Steering

On a ground backed square loop antenna (SLA) of a periphery length greater than
2  guide wavelength (λg), the traveling wave phenomenon dominates, and the
antenna radiates a tilted beam (θmax  32 ) in the direction of flow of current. The
larger is the periphery in terms of guide wavelength, λg, the higher is the beam tilt.
For current flowing on a strip backed by substrate of relative permittivity er on one
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
side and air on the other, λg ¼ λ0 = ð1 þ er Þ=2 where λ0 is the free-space wave-
length. Figure 31 shows the SLA with four feeding points designed for beam steering
applications for the 5 GHz Wi-Fi band. The center and test frequency of the loop is
5.2 GHz. The metal square loop has a periphery length of 120 mm and is etched on
768 P.J. Massey et al.

Fig. 32 Frequency response of reflection coefficient of the SLA and radiation pattern at 5.2 GHz

Rogers 4350 B substrate whose relative permittivity er is 3.48. The antenna is backed
by a ground plane which is at a distance of 9.6 mm. For the center (test) frequency of
5.2 GHz, λ0 = 57.6 mm, and for current flowing on 120 mm metal loop, λg = 38.7
mm. Thus, the loop periphery is 3.1λg. The antenna has a height of 9.6 mm with
planar dimensions of 60.3  60.3 mm.
For achieving steering at one time, the antenna is excited at one of the four points,
labeled as Ae, and the other three ports (Bo, Co, and Do) are left open circuited. This
feeding arrangement creates an asymmetric traveling wave current distribution along
the square loop with respect to the whole antenna structure and generates a tilted
beam directing to the opposite space quadrant (0  ϕ  90 ) of the excited feed.
This is shown in Fig. 32 which also shows that the impedance bandwidth of the
antenna is over 12 % (610 MHz). This particular property of tilted beam generation
is exploited for beam steering. Since the four feeds are symmetric to each other, by
exciting and switching one feed at a time, the tilted beam can be moved in four
quadrants of the space in front of the antenna. This concept is depicted in Fig. 33.

Using High-Impedance Surfaces for Height Reduction

For many modern devices, antennas must be low profile. However, with a height of
9.6 mm, the example presented above is quite thick for implementation in most of
the modern portable wireless devices. To overcome this problem, a hybrid high-
impedance surface (HHIS) is used for reducing the height by a factor of approxi-
mately 50 % (Deo et al. 2010). HHIS is essentially an EBG (electromagnetic
bandgap) structure that allows the ground plane to be kept close to the square
metal loop without affecting the antenna radiation performance. For this, a
two-layer stratified model of substrate is used. The top layer has a height 1.5 mm
and is of Rogers’s substrate which has er = 3.48 and has the square loop printed on it.
Loop Antennas 769

Fig. 33 Demonstration of beam steering by four feeds of SLA. Beam points away from the
excited feed

This layer is placed over a HHIS substrate layer. That HHIS substrate layer has a
height of 3.2 mm and is composed of a 6  6 array of square metal patches on its
top surface. Each square has a side length of 8.8 mm with a gap of 1.5 mm between
two neighboring patches. The squares on the outer periphery of the HHIS structure
are shorted to the ground by metal vias, each having a diameter of 3 mm. The vias
are inserted only to the outermost patches to reduce the side lobes emanating from
the sides which can cause reduction in directivity. The HHIS layer is then backed
by the ground plane and from there four feeding pins go vertically upward toward
the loop strip at the top for excitation. This architecture not only reduces the height
of the original SLA from 9.6 mm down to 4.7 mm, but it also provides extra
bandwidth and higher directivity of 8.8 dBi at θmax = 32 and ϕmax = 45 . This is
shown in Fig. 34.
770 P.J. Massey et al.

Fig. 34 Reflection coefficient and radiation pattern (at 5.2 GHz) of the HHIS-based SLA with a
height of 4.7 mm

Incorporating Electronic RF Switching Circuitry for Electronic Beam

Switching

Figure 35 shows the experimental and simulated model for the HHIS-based SLA
with the RF switching network. Experimentally, a PIN diode-based RF switch (Deo
et al. 2011) is used to connect RF main input to any of the four ports (J1–J4).
The selection of one of the four SMA (subminiature version A) connector ports
generally can be done via binary control logic from microcontroller providing
switching DC voltages linked with a desktop computer running the c algorithm/
logic code. For multipath mitigation, the algorithm for the port selection could be
fast scan all four ports, measure the received signal/interference (S/I) on all the four
ports, compare all four S/I and lock on the port with the highest S/I, and keep
monitoring. As mentioned above, by selecting one feed at a time, the SLA can steer
its tilted radiation beam in four different space quadrants (0 ϕ90 ,
90 ϕ180 , 180 ϕ270 , and 270 ϕ360 ). With the current developments
in PIN diodes, there are PIN diodes available that have: switching speeds in the
tens of nanoseconds, low insertion loss (1–2 dB), and high power handling
capabilities.
In addition to the four tilted beams in four quadrants, the SLA can also generate a
circularly polarized axial (Pal et al. 2013) and a semi-doughnut beam (Pal
et al. 2011). For axial, the four ports of the SLA need to be simultaneously excited
with quadrature phase shifts. The axial beam would be useful for communicating
with vertically located communication systems such as satellites, unmanned aerial
vehicles (UAVs), planes, etc. The semi-doughnut can be generated when the four
ports of the SLA are simultaneously excited with the same phases. The semi-
doughnut pattern is suitable for applications where signal is coming from all the
directions but from the top and bottom. Examples for this include receiving radio and
digital television from terrestrial station in a moving car/vehicle.
Loop Antennas 771

Fig. 35 Simulated and experimental model of the HHIS-based SLA integrated with RF switch
(Deo et al. 2011)

Array of SLAs for High-Gain Wide-Angle Steering, Wide Null

Scanning, and Low Grating Lobes

In a phased array antenna system, multiple antenna elements are excited coherently
with variable phase to steer the beam in a particular direction and to suppress
radiation in undesired directions. They have been traditionally used in several
military applications for many years; however, recent advances in wireless commu-
nication are drawing attention for utilizing array antenna systems for commercial
applications such as 5G, 60 GHz, satellite TV, and cellular base stations as well. The
conventional phased array antenna system deploys a grid of ground backed patch or
dipole antennas with a fixed radiation pattern, generally axial with typically broad-
side directivities of 6–7 dBi for patch antennas. According to the phased array
theory, the net pattern of an array system is given by multiplication of the unit
pattern with the array factor. Thus, the scanning angle and radiation gain of phased
array systems are determined by the unit radiation pattern of the single antenna
element. The radiation gain of the array decreases by at least 3 dB when the scanning
angle of the array increases beyond the half power beam width (HPBW) of the unit
pattern. Due to this phenomenon, for such axial systems, the typical 3 dB scanning
range is confined within the range from 40 to +40 from the zenith (including a
grating lobe < 10 dB criteria). This limited scanning range inhibits the extensive
commercial and defense applications of phased array systems which need high-gain
signals at off-boresight (tilted) angles as well. This problem can be solved by
772 P.J. Massey et al.

deploying a pattern reconfigurable SLA as a unit element, which is capable of

producing a tilted steerable beam as its fundamental pattern. In essence, the SLA
array consisting of pattern reconfigurable unit elements can provide an extra degree
of freedom for net pattern formation. Recently, some good work has been published
to exploit pattern adaptive behavior of unit element for various applications (Roach
et al. 2007; Wu et al. 2010).
As seen from Figs. 32b and 34a, the SLA main radiation is tilted in one direction
(high gain on one side and low gain in opposite direction); therefore, its array system
will offer a much higher gain in off boresight. Further, in arrays grating lobes start to
come into play as the system steers the beam away from the unit pattern. For a patch
antenna with axial unit pattern, the grating lobes start to get dominant for steering
beyond θmax > 40 . However, in the SLA-based array, its antenna elements have
tilted beams to begin with: θmax  32 , and hence the array’s grating lobes would
only appear when the beam is steered beyond θmax > 60 . Besides providing high
gain in tilted direction, the SLA can also adapt for boresight steering. As mentioned
earlier, Pal et al. (2013) show that the SLA can change its fundamental pattern from
tilted to a circularly polarized axial beam by sequential phasing of the four ports.
Hence, by combining the SLA tilted beam scanning and axial beam, a complete
high-gain scanning range can be achieved, which is not possible with conventional
patch or dipole antenna elements. This concept is elaborated next using array of
HHIS-based SLA.

HHIS-Based SLA for Demonstrating Gain Beam Steering Using Tilted

and Axial Modes

A similar configuration to that described below is presented in Pal et al. (2014).

Figure 36 shows a 2  2 array for four HHIS-based square loops. Each SLA loop
has the same specification as that of shown in Fig. 34. Again, for side lobe

Fig. 36 Radiation pattern of 2  2 array of SLA and reflection coefficient for a unit antenna with
port A1 excited and others open circuited
Loop Antennas 773

Fig. 37 Block diagram of the feeding network of 2  2 array

suppression, vias of 3 mm run around the four sides and through the middle of the
array for shorting the HHIS patches to the ground. In addition, there are also two via
lines that go through the middle of the structure which separates the four individual
SLAs. The net planar size of the array is 111  111 mm with height staying the same
at 4.7 mm. Furthermore, with four feeding ports for each SLA, the net array system
has a total of 16 ports which are fed from the bottom side. These ports are labeled as
(A1, B1, C1, D1), (A2, B2, C2, D2), (A3, B3, C3, D3), and (A4, B4, C4, D4). The
antenna array operates over 5 GHz Wi-Fi band, and, for port A1 excited, it has a
reflection coefficient bandwidth of 800 MHz (16 %), as shown in Fig. 36b.
This array has two modes of operation: one is for producing steerable high-gain
off-boresight tilted beams and the other is for producing a steerable high-gain axial
beam. For tilted beam steering, depending upon the quadrant, only four ports are
excited at a time and others are left open circuited. For example, for generating a
steerable beam in the first quadrant of 0 < ϕ <90 , the four A ports (A1–A4) need to
be excited. For axial beam mode, all of the 16 ports need to be excited at the same
time, with condition that the four ports of every individual SLA are in phase
quadrature with each other. The selection of both these modes can be achieved
using a feeding network shown below (Fig. 37). For axial mode, all of the 16 SPST
(single pole single throw) RF switches should be on, and for tilted mode, only four
SPST should be on and others open circuited. The automated selection of the switch
settings and the phase shifts can be done using computer control.
774 P.J. Massey et al.

For this 2  2 configuration, the antenna array factor is given by Balanis (2005):
8 98 9
>
< >
= >
< >
=
1 sin ðψ x Þ sin ψ
AFðθ, ϕÞ ¼ ψ  y
ψ  (58)
: sin x >
4> ;>: sin y > ;
2 2

Where:

ψ x ¼ kdx sin θ cos ϕ þ βx

ψ y ¼ kdy sin θ sin ϕ þ βy
k ¼ 2π=λ
d x ¼ Distance between two elements in x direction
dy ¼ Distance between two elements in y direction
βx ¼ Phase shift between two elements in x direction
βy ¼ Phase shift between two elements in y direction
λ ¼ Wavelength

The maximum radiation beam and grating lobes are located at

ðkd x sin θ cos ϕ þ βx Þ ¼ 2 mπ m ¼ 0, 1, 2, . . .

(59)
kd y sin θ sin ϕ þ βy ¼ 2 nπ n ¼ 0, 1, 2, . . .

To direct the main beam along θ = θmax and ϕ = ϕmax, the phase shift between the
two elements in x- and y-directions must be equal to

βx ¼ kd x sin θmax cos ϕmax

(60)
βy ¼ kd y sin θmax sin ϕmax

For the 2  2 SLA array, shown in Fig. 36, dx = dy = 51 mm = 0.9λ0. Using

the equations above, both the tilted and axial beam maximums were steered in
the space for various maximums at different θmax, and the results for elevation cut
at ϕmax = 45 are shown in Fig. 38. It was found that, for tilted beam configuration, a
maximum directivity of 14.7 dBi was achieved at θmax = 32 , and, for an axial beam
configuration, a directivity of 13.6 dBi was achieved in the boresight. Hence,
together both tilted and axial mode can provide a near constant high-gain
(14.7–12.7 dBi) scanning range from 60 to +60 (for pattern having 10 dB
grating lobes). This high-gain wide scanning range is extremely difficult to achieve
from a non-reconfigurable phased array system. For example, an array using patch
antenna elements (6dBi gain) would need approximately 9 elements to provide a
Loop Antennas 775

Fig. 38 Block diagram of the scanning of the radiation beam at ϕmax = 45

similar boresight gain but still will not be able to provide 14dBi gains at tilted angles.
Hence, the proposed system here is of a huge significance when the systems are
deployed for covering wide sky, especially for airborne and marine applications.

Conceptual Examples

Problem 1 The concept of radiation resistance Rr is very useful in various applica-

tions especially in matching wire antennas including dipole and wire loop antennas
to transmitter or receiver circuit. It is defined by the following expression:

PRadT
Rr ¼ (61)
jI 0 j2

where I0 is the rms current and PRadT is the power radiated by the antenna, as denoted
in subsection “Radiation and Losses.”

(a) Considering that the total power radiated by a single-turn small loop antenna is
given by
 
8Z0 π 3 jI 0 jALA 2
PRadTloop (62)
3 λ2
show that the radiation resistance of the loop antenna can be calculated using
776 P.J. Massey et al.

 2
ALA
Rrloop ¼ 31171 (63)
λ2
or using
 4
C
Rrloop ¼ 20π 2 (64)
λ

where C = 2πa is the perimeter of the loop of radius a.

What would be the expression for the radiation resistance of an n-turn small loop
antenna?
(b) Considering that the radiation efficiency of an antenna (Eq. 14) is defined as

PRadT
R:E: ¼ (65)
PRadT þ PΩ
show that, for the loop antenna, it reduces to

Rrloop
R:E: ¼ (66)
Rrloop þ RΩloop

where RΩloop is known as the Ohmic loss resistance of the loop antenna.
(c) Show that the Ohmic loss resistance of a single-turn loop antenna of radius a and
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
wire diameter d larger than the current skin depth δ (i.e., d >δ where 2=ðωμ0 σ Þ)
is given by
a
RΩloop ¼ ρ (67)
d S
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
where ρS ¼ ωμ0 =ð2σ Þ is the surface resistance. Note σ is the conductivity of
the loop metal/alloy and related to the resistivity ρ through σ = 1/ρ.
(d) Calculate the radiation resistance, the loss resistance, the efficiency, the Q factor,
and the input impedance of a single-turn loop antenna made of a silver wire of
conductivity σ = 6.3  107S/m. This loop antenna has a radius of a = λ/30 and a
wire radius of d = λ/3000, and is designed to operate at 200 MHz.
(e) In order to operate (resonate) the loop antenna at 200 MHz, a capacitor is
connected in parallel across the two terminals of the loop. Calculate the value
of the capacitor and the input impedance of the antenna at resonance. What
would happen to the resonant frequency and the input impedance of the antenna
if the capacitor is connected in series with the loop?

Solutions
2
(a) Inserting PRadTloop ¼ 8Z 0 π 3 jI 0 jALA =λ2 =3 in the radiation resistance definition
leads to
Loop Antennas 777

 
PRadT 8Z 0 π 3 ALA 2
Rr ¼ ¼ (68)
jI 0 j2 3 λ2

i. Since the intrinsic impedance of air Z0 ¼ 120π ¼ 377Ω, the expression for Rr
2
transforms to the following equation: Rr ¼ 8  120π  π 3 ALA =λ2 =3 ¼
2
31171 ALA =λ2 .
ii. Since the loop area ALA = πa2, the expression for Rr can be written as
follows:
 2  
8Z0 π 3 πa2 120π  π 2πa 4
Rr ¼ ¼ (69)
3 λ2 23 λ

Substituting for the perimeter, C = 2πa, in the above equation gives

Rr ¼ 20π 2 ðC=λÞ4
When the loop antenna is n-turn, the current can be assumed to have weighted
by a factor of n, i.e., I0 ! nI0. Therefore, the radiated power increases by a
factor of n2 (see the expression for radiated power). This means that Rr of an
n-turn loop antenna is n2 of that of the single-turn loop antenna.
(b) Using the definition of the radiation resistance Rr ¼ PRadT =jI 0 j2, the total power
radiated by the loop antenna is

PRadT ¼ Rrloop jI 0 j2 (70)

Similarly, the total Ohmic power loss born by the single loop antenna is

PΩ ¼ RΩloop jI 0 j2 (71)

Inserting these equations in the expression for radiation efficiency, R:E: ¼ PRadT =
ðPRadT þ PΩ Þ, leads to

PRadT Rrloop jI 0 j2 Rrloop

R:E: ¼ ¼ ¼ (72)
PRadT þ PΩ Rrloop jI 0 j þ RΩloop jI 0 j
2 2 Rrloop þ RΩloop

(c) The surface resistance, ρS, is in Ohms per square (Omega/sq). As the current
flows along the loop, the loss resistance of the loop is obtained as follows:

ρS a
RΩloop ¼ 2πa ¼ ρS (73)
2πd d

It should be noted that the loss resistance of a loop antenna with n turns is not
simply n times of that of its single loop. This is because there is an interaction
between the fields of the loops, changing the current distribution and hence the
778 P.J. Massey et al.

value of the loss resistance. However, if the loops are sufficiently separated, the
n-time approximation can be used.
(d) To find the radiation resistance of the antenna, the radius of the loop, a = λ/30, is
used to find the area of the antenna:

 2
λ
ALA ¼ πa ¼ π
2
(74)
30

Then, this area is inserted in the expression for the radiation resistance:

  " #2  2
λ 2
ALA 2 π 30 π
Rrloop ¼ 31171 ¼ 31171 ¼ 31171 ¼ 0:38 Ω (75)
λ 2
λ 2
302

To calculate the loss resistance, first the surface resistivity is determined:

rffiffiffiffiffiffiffiffi sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi
ωμ0 π  200  106  4π  107
ρS ¼ ¼ ¼ 3:54  103 Ω per square: (76)
2σ 6:3  107

Then, the surface resistance; the loop radius, a = λ/30; and the loop wire radius,
d = λ/3000, are inserted in the expression for the loss resistance:
a
RΩloop ¼ ρ ¼ ½ðλ=30Þ=ðλ=3000Þ  3:54  103 ¼ 0:35Ω (77)
d S
Therefore, the radiation efficiency of the antenna is

Rrloop 0:38
R:E: ¼ ¼ ¼ 52% (78)
Rrloop þ RΩloop 0:38 þ 0:35

To calculate the Q factor of the antenna, the value of the inductance of the loop is
required. This inductance, L, is the sum of the external inductance, Lextloop, and
the internal inductance, Lintloop, of the loop:

L ¼ Lextloop þ Lintloop (79)

The external inductance, associated with the storage of the magnetic field
external to the loop wire, is obtained using an approximate expression (Johnson
and Jasik 1993):
 
8a
Lextloop ¼ μ0 a ln 2
d
     
7 1:5 λ λ
¼ 4π  10 ln =  2 ¼ 2:94  107 H (80)
30 30 3000
Loop Antennas 779

The internal inductance, associated with the storage of the magnetic field within
the loop wire, is obtained using

1 a 1 0:35
Lintloop ¼ ρ ¼ RΩloop ¼ ¼ 2:8  1010 H (81)
ω d S ω 2  π  200  106
As can be seen, Lextloop Lintloop and this is normally the case and hence Lintloop
can be ignored. Therefore, the Q factor of the loop antenna is

ωL 2π  200  106  2:94  107

Q¼ ffi ¼ 506 (82)
Rrloop þ RΩloop 0:38 þ 0:35

To calculate the input impedance of the antenna, the equivalent circuit of the
loop antenna is used. In this circuit, the radiation and loss resistance are in series
with the inductance of the loop antenna. Therefore,

Z inloop ¼ Rrloop þ RΩloop þ jωL ffi 0:38 þ 0:35 þ j2π  200  106  2:94  107
¼ 0:73 þ j369:5Ω
(83)
As expected, the input impedance is very inductive.
(e) To find the capacitance C of the capacitor, the input admittance of the antenna is
formed as the capacitor is connected in parallel with the loop:

1
Y in ¼ jCω þ (84)
R þ jLω

where R ¼ Rrloop þ RΩloop and L ¼ Lextloop þ Lintloop . This can be written as

follows:
n h i o
R þ j Cω R2 þ ðLωÞ2  Lω
Y in ¼ (85)
R2 þ ðLωÞ2

At resonance, the imaginary part of the input admittance vanishes. Therefore,

h i
C R2 þ ðLωÞ2  L ¼ 0 (86)

from which the value of capacitance is obtained:

L 1 1
C¼h iffi ¼ 2
Lω 2 7
R þ ðLωÞ
2 2
2:94  10  2π  200  106
1
¼ ¼ 2:15pF (87)
0:464  1012
780 P.J. Massey et al.

In the above and following calculations, the resistance is ignored as it is much

smaller than the reactance.
The input impedance of the antenna at resonance is found using the input
admittance at resonance:

R
Y in ¼ (88)
R þ ðLωÞ2
2

2
R2 þ ðLωÞ2 ðLωÞ2 2:94  107  2π  200  106
Z in ¼ ffi ¼ ¼ 187 kΩ (89)
R R 0:38 þ 0:35
When the capacitor is connected in series with the loop, the resonant frequency
of the antenna stays nearly the same as before (200 MHz). However, the input
impedance at resonance changes drastically to a small value and is equal to

R ¼ Rrloop þ RΩloop ¼ 0:38 þ 0:35 ¼ 0:73Ω (90)

Comments
1. Capacitors have quality factors QC that are typically in the range of ~200 (for
low-cost variable capacitors) to ~2,000 for high-quality capacitors. For the 2.15
pF capacitor of this example, using the relation QC ¼ 1=ðωC RSc Þ where RSC is
the effective series resistance of the capacitor, gives RSC~2 Ω (for QC~200) to
RSC~0.2 Ω (for QC~2,000). These values are comparable to those of Rrloop and
RΩloop of the loop antenna and shows that in practice the losses associated with
nonideal capacitors should be considered in the design calculations. The details of
the mathematics for this are beyond the scope of these examples. However, by an
extension of the analysis described above, it can be shown that, in general, the
effect of the finite quality factor of a real-life capacitor is to reduce the resonant
impedance of the loop tuned with a capacitor in parallel and to increase the
resonant impedance of the loop tuned with a capacitor in series.
2. Transmitters and receivers generally require impedances that lie between the two
extremes calculated above for series and parallel capacitor tuning. This observa-
tion can be thought of as an inspiration for the combined series and parallel
capacitor tuning arrangement shown in Fig. 8b, which is capable of giving any
impedance between the very low-impedance case with series capacitors only
(parallel capacitance vanishes) to the very high-impedance case with the parallel
capacitor providing practically all of the tuning and the series capacitors having
such high capacitances that they appear to have negligible reactance.

Problem 2 Consider a small loop antenna horizontally placed at a distance h over

an infinite size perfect conductor (Fig. 39). Assuming that the radiation pattern of
Loop Antennas 781

Fig. 39 Loop antenna

parallel to and above an
infinite perfectly conducting
plane, discussed in problem 2

this antenna in the absence of the ground plane varies according to the following
expression:

Eϕ ¼ k sin ðθÞ (91)

where k is constant, find an expression for the radiation pattern of this antenna when
the ground plane is present.

Solution As the ground is a perfect conductor, it can be replaced by the image of the
loop antenna. This image is another loop antenna at distance h from the surface of the
ground plane, and it carries a current opposite to that of the original loop antenna. In
other words, there is a phase shift of π between the currents of the two elements. The
system can be assumed as a linear array of two loop antennas. In un-normalized
form, the array factor for a linear array of N elements is
 
N
sin Ψ
2
AF ¼   (92)
1
sin Ψ
2

where Ψ = βs cosθ + α. In this case, the separation between the elements s = 2h, the
number of elements N = 2, and the progressive phase a = π. Therefore, the array
factor simplifies to
 
2 2π 2π
sin 2h cos ðθÞ þ π sin 2h cos ðθÞ þ π
2 λ λ
AF ¼   ¼  
1 2π 1 2π
sin 2h cos ðθÞ þ π sin 2h cos ðθÞ þ π
2 λ 2 λ
 
1 2π
¼ 2 cos 2h cos ðθÞ þ π (93)
2 λ
782 P.J. Massey et al.

Fig. 40 Current distribution on a square loop antenna

Then, the overall radiation pattern is found using the pattern multiplication rule:

EϕT ¼ antenna ðelementÞ pattern  array factor

 
h 1
¼ 2k sin ðθÞ cos 2 cos ðθÞ þ π (94)
λ 2

From this expression, several conclusions can be made:

i. EϕT = 0 along the loop axis (θ = 0).

ii. EϕT = 0 over the ground plane (θ = π/2).
iii. EϕT has beam direction and directivity which varies with h/λ.

Problem 3 Assume that a square loop antenna fed as shown in Fig. 40 lays in the
x-y plane with its center at the origin of the coordinate system. Sketch the radiation
pattern of the antenna in the y-z plane for the two cases of small square loop antenna
and one-wavelength square loop antenna.

Solution Since the current is approximately uniform on the wire of the small square
loop antenna (Fig. 40; left), the radiation pattern is similar to that of the small circular
loop antenna. Therefore, the pattern is omnidirectional with its yz-cut shown in
Fig. 41 (left).
The pattern of the one-wavelength square loop antenna is more complex if
one considers the current distribution of this antenna at the first resonance as
shown in Fig. 40 (right). Since the currents in the left and right arms of the
antenna are opposite, they contribute little to the radiation. The currents in the
top and bottom arms are the main contributors. These arms can be considered as
the array of two dipole antennas separated by a distance of quarter wavelength. In
this case, the yz-cut of the pattern has the maximum radiation occurring along the
Loop Antennas 783

Fig. 41 The yz-cut of the radiation pattern of square loop antenna

z-axis (due to the constructive interference of the fields of the two dipoles) (Fig. 41;
right). In the y-direction, the radiation strength is about half of that in the
z-direction (due to the vector sum of the two fields of the dipoles with 90 phase
shift) (Fig. 41; right). The cuts of the radiation pattern in the two other
principle planes can also be predicted. In the x-z plane, the radiation has nulls in
the x-directions and is maximum in the z-directions. In the x-y plane, the
radiation has nulls along the x-axis and has a value about half of that along the
z-axis (as explained earlier). Also, see Fig. 16 of subsection “Tested Example and
Range Plots” for the radiation pattern of an antenna example similar to that in this
problem.

Conclusion

This chapter describes three classes of loop antennas in order of their size in
wavelengths: electrically small loop (and coil) antennas, full-wave loop antennas,
and reconfigurable multi-feed loops. This ordering also reflects the historical devel-
opment of loop antennas, with the electrically small loop antennas coming early in
the history of wireless development and reconfigurable multi-feed loops still being a
current research topic.
Throughout loop antenna history, their ubiquity has inevitably led them being
combined with other innovations. For example, electronically small loop antennas
were among the first to be used in a commercial product with antenna diversity – by
the early 1950s, radios were sold with a pair of loops angled away from each other,
and the user could switch between loops for the best reception (Blok and Rietveld
1955). After World War II, electrically small loop antennas were combined with the
then very new soft ferrite material to give smaller antennas with similar sensitivity
and less noise reception through reducing the aerial effect. Today, reconfigurable
multi-feed loops are placed above high-impedance surfaces, to reduce the distance
between the loop and its backing ground plane, and also because the high-impedance
surface is found to improve isolation between a loop’s feeds.
784 P.J. Massey et al.

Loop antenna research continues to be a very active field. A search of the IEEE
Xplore database for articles published between the beginning of 2014 and June 2015
(the date when the search was conducted) revealed 123 papers with “loop antenna”
in their titles, and over three times this number in which loop antennas played a
significant role. Over the last few years, popular research and application areas have
included:

1. Creating loops from metamaterial structures (e.g., Nakano et al. 2013a; Zhang
et al. 2015)
2. Loop antennas in medical on-body and implant applications (e.g., Alrawashdeh
et al. 2014; Lee and Jung 2015)
3. Loop antennas in handheld multiband mobile phone devices (e.g., Ban
et al. 2015)
4. Using loops as circularly polarized elements in arrays (e.g., Nakano et al. 2013b;
Hirose et al. 2012, 2015a, b)

In view of the recent nature of the newer publications, it is difficult to predict how
these topics would develop and which developments would prove to be the most
significant. However, what is certain is that electrically small loop antennas will
continue to have a significant role in antennas for products designed to be worn
against the user’s body and in other portable devices. In addition, the unique ability
of traveling wave loops to provide broad impedance bandwidth and adaptive antenna
element pattern will enable the realization of wideband reconfigurable arrays for
communications and sensor networks.

Cross-References

▶ Impedance Matching and Baluns

▶ Low-Profile Antennas
▶ Metamaterials and Antennas
▶ Omnidirectional Antennas
▶ Phased Arrays
▶ Radiation Efficiency Measurements of Small Antennas

References
Alrawashdeh R, Huang Y, Sajak AAB (2014) A flexible loop antenna for biomedical bone implants.
In: 8th European conference on antennas and propagation, pp 861–864
Balanis CA (2005) Antenna theory, analysis and design, 3rd edn. Wiley, Hoboken, Chapter 5: Loop
antennas. ISBN 0-471-66782-X
Ban Y-L, Qiang Y-F, Chen Z, Kang K, Guo J-H (2015) A dual-loop antenna design for hepta-band
WWAN/LTE metal-rimmed smartphone applications. IEEE Trans Antennas Propag 63(1):48–58
Blok H, Reitveld JJ (1955) Inductive aerials in modern broadcast receivers. Philips Tech Rev
16(7):181–212
Loop Antennas 785

Collin RE (2001) Foundations for microwave engineering. Wiley, New Jersey, Chapter 2
Deo P, Mehta A, Mirshekar-Syahkal D, Massey PJ, Nakano H (2010) Thickness reduction and
performance enhancement of steerable square loop antenna using hybrid high impedance
surface. IEEE Trans Antennas Propag 58(5):1477–1485
Deo P, Pant M, Mehta A, Mirshekar-Syahkal D, Nakano H (2011) Implementation and simulation
of commercial rf switch integration with steerable square loop antenna. Electron Lett
47(12):686–687
Grover FW (2009) Inductance calculations, working formulas and tables. Dover, New York
Harrington RF (2001) Time harmonic electromagnetic fields. Wiley, New York
Hirose K, Shibasaki T, Nakano H (2012) Fundamental study on novel loop-line antennas radiating a
circularly polarized wave. IEEE Antennas Wirel Propag Lett 11:476–479
Hirose K, Shinozaki K, Nakano H (2015a) A loop antenna with parallel wires for circular
polarization—its application to two types of microstrip-line antennas. IEEE Antennas Wirel
Propag Lett 14:538–586
Hirose K, Shinozaki K, Nakano H (2015b) A comb-line antenna modified for wideband circular
polarization. IEEE Antennas Wirel Propag Lett 14:1113–1116
Huff GH, Bernhard JT (2006) Integration of packaged RF MEMS switches with radiation pattern
reconfigurable square spiral microstrip antennas. IEEE Trans Antennas Propag 54(2):464–469
Huff GH, Feng J, Zhang S, Bernhard JT (2003) A novel radiation pattern and frequency
reconfigurable single turn square spiral microstrip antenna. IEEE Microw Wirel Component
Lett 13(2):57–59
ITU-R (2001) Recommendation, P.372-7, “Radio noise”
Johnson RC, Jasik H (eds) (1993) Antenna engineering handbook, 3rd edn. McGraw-Hill,
New York, Smith GS, Chapter 5: Loop antennas. ISBN 1063-665X
Jung CW, Lee M, Li GP, Flaviis FD (2006) Reconfigurable scan-beam single-arm spiral antenna
integrated with RF-MEMS switches. IEEE Trans Antennas Propag 54(2):455–463
Lee KF (1984) Principles of antenna theory. Wiley, Chichester
Lee C, Jung C (2015) Radiation-pattern-reconfigurable antenna using monopole-loop for fitbit flex
wristband. IEEE Antennas Wirel Propag Lett 14:269–272
Li R, DeJean G, Laskar J, Tentzeris MM (2005) Investigations of circularly polarized loop antennas
with a parasitic element for bandwidth enhancement. IEEE Trans Antennas Propag
53(12):3930–3939
Liberti JC, Rappaport TS (1999) Smart antennas for wireless communications: IS-95 and third
generation CDMA applications. Prentice-Hall, Englewood Cliffs
Massey PJ (2001) New formulae for practical pager design. In: 11th international conference on
antennas and propagation (ICAP), Manchester, vol 1, pp 265–268
Massey PJ (2009) Single tuned electrically small antennas. In: Loughborough antennas and
propagation conference, 16–17 Nov, pp 497–500
Mehta A, Mirshekar-Syahkal D (2004) Spiral antenna with adaptive radiation pattern under
electronic control. In: Antennas and Propagation Society international symposium, IEEE, vol
1, 20–25 June, pp 843–846
Mehta A, Mirshekar-Syahkal D (2007) Pattern steerable square loop antenna. Electron Lett
43(9):491–493
Mehta A, Mirshekar-Syahkal D, Nakano H (2006) Beam adaptive single arm rectangular spiral
antenna with switches. IEE Proc Microw Antennas Propag 153(1):13–18
Morishita H, Hirasawa K, Nagao T (1998) Circularly polarised wire antenna with a dual rhombic
loop. IEE Proc Microw Antennas Propag 145(3):219–224
Nakano H, Yoshida K, Yamauchi J (2013a) Radiation characteristics of a metaloop antenna. IEEE
Antennas Wirel Propag Lett 12:861–863
Nakano H, Iitsuka Y, Yamauchi J (2013b) Loop-based circularly polarized grid array antenna with
edge excitation. IEEE Trans Antennas Propag 61(8):4045–4053
Pal A, Mehta A, Mirshekar-Syahkal D, Massey P (2008) Short-circuited feed terminations on beam
steering square loop antennas. Electron Lett 44(24):1389–1390
786 P.J. Massey et al.

Pal A, Mehta A, Mirshekar-Syahkal D, Nakano H (2011) A square-loop antenna with 4-port feeding
network generating semi-doughnut pattern for vehicular and wireless applications. IEEE Anten-
nas Wirel Propag Lett 10:338–341
Pal A, Mehta A, Marhic ME (2013) Generating a pure circularly polarised axial beam from a pattern
reconfigurable square loop antenna. IET Microw Antennas Propag 7(3):208–213
Pal A, Mehta A, Lewis R, Clow N (2014) Phased array system consisting of unit pattern
reconfigurable square loop antennas. In: Antennas and Propagation Society international sym-
posium (APSURSI), IEEE, pp 1658–1659, 6–11 Jul
Pozar DM (1986) Finite phased arrays of rectangular microstrip patches. IEEE Trans Antennas
Propag AP-34(5):658–665
Roach TL, Huff GH, Bernhard JT (2007) On the applications for a radiation reconfigurable antenna.
In: Second NASA/ESA conference on adaptive hardware and systems (AHS), Aug, pp 7–13
Schelkunoff SA (1939) A general radiation formula. In: Proceedings of the I.R.E., Oct 1939,
pp 660–666
Snelling EC (1988) Soft ferrites – properties and applications, 2nd edn. Butterworth, London. ISBN
0-408-02760-6
van der Zaag PJ (1999) New views on the dissipation in soft magnetic ferrites. J Magn Magn Mater
196–197(1999):315–319
Wagstaff AJ and Merricks N (2003) Man-Made Noise Measurement Programme (AY4119) final
report. Issue 2, Mass Consultants Limited, Sept 2003. This report was downloadable from www.
ofcom.org.uk. Last accessed 30 Dec 2014. In particular see section 5.1
Wu JC, Chang CC, Chin TY, Huang SY, Chang SF (2010) Sidelobe level reduction in wide-angle
scanning array system using pattern-reconfigurable antennas. Microwave Symposium Digest
(MTT), IEEE MTT-S, 23–28 May, pp 1274–1277
Zhang Y, Wei K, Zhang Z, Li Y, Feng Z (2015) A compact dual-mode metamaterial-based loop
antenna for pattern diversity. IEEE Antennas Propag Lett 14:394–397