 Overview of microstrip antennas

 Feeding methods
 Basic principles of operation
 General characteristics
 CAD Formulas
 Input Impedance
 Radiation pattern
 Circular polarization
 Improving bandwidth

1
 Notation
c  speed of light in free space c  2.99792458  108  m/s 

0  wavelength of free space 0  c / f

k0  wavenumber of free space k0   0 0  2 / 0

k1  k0  r
k1  wavenumber of substrate

0
0  intrinsic impedance of free space 0   376.7303  
0
1  intrinsic impedance of substrate 1  0 /  r

 r  relative permtitivity (dielectric constant) of substrate

1
c
 0 0
 reff  effective relative permtitivity
(accouting for fringing of flux lines at edges) 0  4 107  H/m 
1
 eff
 complex effective relative permtitivity 0   8.854188  1012  F/m 
rc 0 c 2
(used in the cavity model to account for all losses)

2
 Outline
 Overview of microstrip antennas
 Feeding methods
 Basic principles of operation
 General characteristics
 CAD Formulas
 Input Impedance
 Radiation pattern
 Circular polarization
 Improving bandwidth

3
 Overview of Microstrip Antennas
Also called “patch antennas”

 One of the most useful antennas at microwave frequencies (f > 1 GHz).

 It usually consists of a metal “patch” on top of a grounded dielectric substrate.
 The patch may be in a variety of shapes, but rectangular and circular are the
most common.

Microstrip line feed Coax feed

4
 Overview of Microstrip Antennas
Common Shapes

Rectangular Square Circular Annular ring

Elliptical Triangular

5
 Overview of Microstrip Antennas
Advantages of Microstrip Antennas

 Low profile (can even be “conformal,” i.e. flexible to conform to a

surface).
 Easy to fabricate (use etching and photolithography).
 Easy to feed (coaxial cable, microstrip line, etc.).
 Easy to incorporate with other microstrip circuit elements and
integrate into systems.
 Patterns are somewhat hemispherical, with a moderate directivity
(about 6-8 dB is typical).
 Easy to use in an array to increase the directivity.

6
 Overview of Microstrip Antennas
Disadvantages of Microstrip Antennas

 Low bandwidth (but can be improved by a variety of techniques). Bandwidths of

a few percent are typical. Bandwidth is roughly proportional to the substrate
thickness and inversely proportional to the substrate permittivity.

 Efficiency may be lower than with other antennas. Efficiency is limited by

conductor and dielectric losses*, and by surface-wave loss**.

 Only used at microwave frequencies and above (the substrate becomes too
large at lower frequencies).

 Cannot handle extremely large amounts of power (dielectric breakdown).

* Conductor and dielectric losses become more severe for thinner

substrates.
** Surface-wave losses become more severe for thicker substrates
(unless air or foam is used).
7
 Overview of Microstrip Antennas
Applications

Applications include:

 Satellite communications

 Microwave communications

 Cell phone antennas

 GPS antennas

8
 Overview of Microstrip Antennas

Microstrip
antenna
Filter
DC supply Micro-D
connector K-connector

LNA
PD
Fiber input with
collimating lens Diplexer

Microstrip Antenna Integrated into a System: HIC Antenna Base-Station for 28-43 GHz

9
 Overview of Microstrip Antennas
Arrays

Linear array (1-D corporate feed)

22 array

2-D 8X8 corporate-fed array 4  8 corporate-fed / series-fed array

10
 Overview of Microstrip Antennas
Wraparound Array (conformal)

The substrate is so thin that it can be bent to “conform” to the surface.

11
 Overview of Microstrip Antennas
Rectangular patch
The fields and current
are approximately y
independent of y for the
dominant (1,0) mode.

W Js
x
L
h r

L is the resonant dimension (direction of current flow). The width W is

usually chosen to be larger than L (to get higher bandwidth). However,
usually W < 2L (to avoid problems with the (0,2) mode).

W = 1.5L is typical.
12
 Overview of Microstrip Antennas
Circular Patch

a x

h r

The location of the feed determines the direction of current flow and hence
the polarization of the radiated field.

13
 Outline
 Overview of microstrip antennas
 Feeding methods
 Basic principles of operation
 General characteristics
 CAD Formulas
 Input Impedance
 Radiation pattern
 Circular polarization
 Improving bandwidth

14
 Feeding Methods

Some of the more common methods for

feeding microstrip antennas are shown.

The feeding methods are illustrated for a rectangular patch,

but the principles apply for circular and other shapes as well.

15
 Feeding Methods
z Coaxial Feed

r h
x

y
Note:
A feed along the centerline at y = W/2
is the most common
Surface current (this minimizes higher-order modes
and cross-pol).

Feed at (x0, y0) x

L
16
 Feeding Methods
Coaxial Feed
 x  z
R  Redge cos2  0 
 L 
(The resistance varies as the
square of the modal field shape.) r h
x

Advantages: y
 Simple
 Directly compatible with coaxial cables
 Easy to obtain input match by adjusting feed position  x0 , y0 
W
Disadvantages:
 Significant probe (feed) radiation for thicker substrates x
 Significant probe inductance for thicker substrates (limits L
bandwidth)
 Not easily compatible with arrays
17
 Feeding Methods
Inset Feed

Advantages:
Microstrip line
 Simple
 Allows for planar feeding
 Easy to use with arrays
 Easy to obtain input match

Disadvantages:
 Significant line radiation for thicker substrates
 For deep notches, patch current and radiation pattern may show distortion

18
 Feeding Methods
Inset Feed

An investigation has shown that the

resonant input resistance varies as:
x0
2    2 x0 
Rin  A cos    B Wf
2 L  S W

Less accurate approximation:

 x  L
R  Redge cos  0 
2

 L 

The coefficients A and B depend on the notch width S but (to a good
approximation) not on the line width Wf .

19
 Feeding Methods
Proximity-coupled Feed
(Electromagnetically-coupled Feed)
Advantages:
 Allows for planar feeding
 Less line radiation compared to microstrip feed (the line is closer to the ground plane)
 Can allow for higher bandwidth (no probe inductance, so substrate can be thicker)

Patch

Microstrip line
Microstrip
Top view line
Disadvantages:
 Requires multilayer fabrication
 Alignment is important for input match
20
 Feeding Methods
Gap-coupled Feed
Advantages:
 Allows for planar feeding
 Can allow for a match even with high edge impedances, where a notch
might be too large (e.g., when using a high permittivity substrate)

Gap Patch
Patch

Microstrip line Microstrip

Top view line

Disadvantages:
 Requires accurate gap fabrication
 Requires full-wave design

21
 Feeding Methods
Aperture-coupled Patch (ACP)

Advantages:
 Allows for planar feeding Slot

 Feed-line radiation is isolated from patch radiation

 Higher bandwidth is possible since probe inductance is
eliminated (allowing for a thick substrate), and also a
Microstrip
double-resonance can be created Top view line

 Allows for use of different substrates to optimize

antenna and feed-circuit performance
Patch

Disadvantages: Slot
 Requires multilayer fabrication
 Alignment is important for input match Microstrip line

22
 Outline
 Overview of microstrip antennas
 Feeding methods
 Basic principles of operation
 General characteristics
 CAD Formulas
 Input Impedance
 Radiation pattern
 Circular polarization
 Improving bandwidth

23
 Basic Principles of Operation

 The basic principles are illustrated here for a rectangular patch, but the
principles apply similarly for other patch shapes.

 We use the cavity model to explain the operation of the patch antenna.

h n̂

PMC

24
 Basic Principles of Operation
Main Ideas:
 The patch acts approximately as a resonant cavity (with perfect electric
conductor (PEC) walls on top and bottom, and perfect magnetic conductor
(PMC) walls on the edges).

 Radiation is accounted for by using an effective loss tangent for the substrate.

 In a cavity, only certain modes are allowed to exist, at different resonance

frequencies.

 If the antenna is excited at a resonance frequency, a strong field is set up inside

the cavity, and a strong current on the (bottom) surface of the patch. This
produces significant radiation (a good antenna).
z

PMC
h n̂

25
 Basic Principles of Operation
A microstrip antenna can radiate well, even with a thin substrate,
because of resonance.

 As the substrate gets thinner the patch current radiates less, due to image
cancellation (current and image are separated by 2h).
 However, the Q of the resonant cavity mode also increases, making the patch
currents stronger at resonance.
 These two effects cancel, allowing the patch to radiate well even for thin
substrates (though the bandwidth decreases).

z 1
Js  Q 
h

r Js h
x

26
 Basic Principles of Operation
Thin Substrate Approximation

On patch and ground plane: Et  0 E  zˆ E z

Inside the patch cavity, because of the thin substrate,

the electric field vector is approximately independent of z.

Hence E  x, y, z   zˆ Ez  x, y 

E z  x, y 
h

27
 Basic Principles of Operation
Thin Substrate Approximation

Magnetic field inside patch cavity:

1
H   E
j
1
    zEˆ z  x, y  
j
1

j
  zˆ  Ez  x, y  

28
Basic Principles of Operation
Thin Substrate Approximation

1
H  x, y  
j
 zˆ  E  x, y  
z

The magnetic field is purely horizontal.

(The mode is TMz.)

E z  x, y 
h

H  x, y 

29
 Basic Principles of Operation
Magnetic-wall Approximation
The patch edge acts as an approximate open circuit.

H t  0 (PMC) L

or
Js n̂
nˆ  H  x, y   0
W

tˆ
x

PMC
H edge
0
h
t
h
n̂

Actual patch PMC Model

30
 Basic Principles of Operation
Magnetic-wall Approximation
y
nˆ  H  x, y   0
L
1
H  x, y  
j
 zˆ  E  x, y  
z

Js n̂
W
Hence,
tˆ
nˆ   zˆ  Ez  x, y    0 x

nˆ   zˆ  Ez  x, y    zˆ  nˆ Ez  x, y    Ez  x, y  nˆ  zˆ 

zˆ  nˆ Ez  x, y    0

Ez h
n̂
0 (Neumann B.C.)
n PMC
31
 Basic Principles of Operation
Resonance Frequencies
y
 Ez  k E z  0
2 2
k  k1  k0  r
E z  x, y 
From separation of variables:
 m x   n y  PMC
Ez  cos   cos   W
 L   W 
(TMmn mode)
x
L
  m 2  n  2 
We then have        k1  Ez  0
2

  L   W  
We ignore the loss tangent of
the substrate for the
  m  2  n 2  calculation of the resonance
Hence        k1  0
2
frequencies.
  L   W  
32
 Basic Principles of Operation
Resonance Frequencies

We thus have y
E z  x, y 
2 2
 m   n 
k 
1
2
  
 L   W 
PMC
W
Recall that
k1  k0  r   0 0  r
x
L
  2 f
Hence
2 2
c  m   n 
f      c  1/ 0 0
2  r  L  W 

33
 Basic Principles of Operation
Resonance Frequencies

y
E z  x, y 
Hence f  f mn
(resonance frequency of (m,n) mode)
PMC
W

where
x
2 2
L
c  m   n 
f mn     
2  r  L   W 

34
 Basic Principles of Operation
Dominant (1,0) mode
y
This structure operates as a “fat planar dipole.”
Current
This mode is usually used because the
radiation pattern has a broadside beam.
c 1 W
f10   
2 r  L 
x
x  L
Ez  cos  
 L  The current is maximum in the middle of
the patch, when plotted along x.
 1    x 
H  x, y    yˆ     sin  
 j L
   L  The resonant length L is about 0.5
guided wavelengths in the x direction
 1       x  (see next slide).
J s  xˆ     sin  
 j 0   L   L 
35
 Basic Principles of Operation
Resonance Frequency of Dominant (1,0) Mode

The resonance frequency is mainly controlled by the

patch length L and the substrate permittivity.

Approximately, (assuming PMC walls)

2 2
 m   n  This is equivalent to saying that
k1  
2
   the length L is one-half of a
 L   W  wavelength in the dielectric.

0 / 2
(1,0) mode: k1 L   L  d / 2 
k1  2 / d r

A higher substrate permittivity allows for a smaller antenna (miniaturization),

but with a lower bandwidth.
36
 Basic Principles of Operation
Resonance Frequency of Dominant Mode
The resonance frequency calculation can be improved by adding a
“fringing length extension” L to each edge of the patch to get an
“effective length” Le .

Le  L  2L

c  1  L L
f10   
2  r  Le  L
x
Le

37
 Basic Principles of Operation
Resonance Frequency of Dominant Mode

Hammerstad formula:

 eff W 
 r  0.3   h  0.264 
L / h  0.412  
  eff  0.258  W  0.8  
  r   h 


1/ 2
 r  1   r 1    h 
 reff   1  12  
2  2   W 

Even though the Hammerstad formula c 1 

involves an effective permittivity, we still use f10   
the actual substrate permittivity in the 2 r  L  2 L 
resonance frequency formula.
38
Basic Principles of Operation
Resonance Frequency of Dominant Mode

L  0.5 h

39
 Outline
 Overview of microstrip antennas
 Feeding methods
 Basic principles of operation
 General characteristics
 CAD Formulas
 Input Impedance
 Radiation pattern
 Circular polarization
 Improving bandwidth

40
 General Characteristics
Bandwidth

 The bandwidth is directly proportional to substrate thickness h.

 However, if h is greater than about 0.05 0 , the probe inductance (for a

coaxial feed) becomes large enough so that matching is difficult – the
bandwidth will decrease.

 The bandwidth is inversely proportional to r (a foam substrate gives a high

bandwidth).

 The bandwidth of a rectangular patch is proportional to the patch width W

(but we need to keep W < 2L ; see the next slide).

41
 General Characteristics
Width Restriction for a Rectangular Patch

W  2L c1
f 01   
2 r  W 

c
2
 m   n 
2
c 1
f mn  f10   
    2 r  L 
2  r  L  W 

c  2
f 02   
c  1 1  2 r  W 
f 02  f10    
 r  W 2L 

fc
W
f01 f10 f02
W = 1.5 L is typical.
L
42
 General Characteristics

Some Bandwidth Observations

 For a typical substrate thickness (h /0 = 0.02), and a typical substrate

permittivity (r = 2.2) the bandwidth is about 3%.

 By using a thick foam substrate, bandwidth of about 10% can be achieved.

 By using special feeding techniques (aperture coupling) and stacked

patches, bandwidths of 100% have been achieved.

43
 General Characteristics
Resonant Input Resistance
 The resonant input resistance is fairly independent of the substrate
thickness h unless h gets small (the variation is then mainly due to
dielectric and conductor loss).

 The resonant input resistance is proportional to r.

 The resonant input resistance is directly controlled by the location of the

feed point (maximum at edges x = 0 or x = L, zero at center of patch).

(x0, y0)
W

L L
x
44
 General Characteristics
Resonant Input Resistance (cont.)

The patch is usually fed along the centerline (y0 = W / 2)

to maintain symmetry and thus minimize excitation of undesirable modes
(which cause cross-pol).

y
Desired mode: (1,0)

Feed: (x0, y0)

x
L
45
 General Characteristics
Resonant Input Resistance (cont.)
For a given mode, it can be shown that the resonant input resistance is
proportional to the square of the cavity-mode field at the feed point.

y
Rin  E 2
z  x0 , y0 
(x0, y0)
For (1,0) mode:
W
  x0 
Rin  cos  2

 L  x
L

46
 General Characteristics
Resonant Input Resistance (cont.)

y
Hence, for (1,0) mode:

  x0  (x0, y0)
Rin  Redge cos  2

 L  W

x
L

The value of Redge depends strongly on the substrate permittivity

(it is proportional to the permittivity).

For a typical patch, it is often in the range of 100-200 Ohms.

47
 General Characteristics
Radiation Efficiency

 Radiation efficiency is the ratio of power radiated into

space, to the total input power.

Pr
er 
Ptot

 The radiation efficiency is less than 100% due to

 Conductor loss
 Dielectric loss
 Surface-wave excitation*
*assuming the substrate is infinite

48
General Characteristics
Radiation Efficiency (cont.)

TM0
surface wave
Js
x
cos () pattern

49
 General Characteristics
Radiation Efficiency (cont.)
Hence,

Pr Pr
er  
Ptot Pr   Pc  Pd  Psw 

Pr = radiated power Pc = power dissipated by conductors

Ptot = total input power Pd = power dissipated by dielectric

Psw = power launched into surface wave

50
 General Characteristics
Radiation Efficiency (cont.)

 Conductor and dielectric loss is more important for thinner substrates (the
Q of the cavity is higher, and thus the resonance is more seriously affected
by loss).

 Conductor loss increases with frequency (proportional to f 1/2) due to the

skin effect. It can be very serious at millimeter-wave frequencies.

 Conductor loss is usually more important than dielectric loss for typical
substrate thicknesses and loss tangents.

1 2
Rs  
  Rs is the surface resistance of the metal.
The skin depth of the metal is .
0
Rs   f
2
51
 General Characteristics
Radiation Efficiency (cont.)

 Surface-wave power is more important for thicker substrates or for

higher-substrate permittivities. (The surface-wave power can be
minimized by using a thin substrate or a foam substrate.)

 For a foam substrate, a high radiation efficiency is obtained by making the

substrate thicker (minimizing the conductor and dielectric.

 For a typical substrate such as r = 2.2, the radiation efficiency is maximum for
h / 0  0.02.

52
 General Characteristics
Radiation Pattern

y
E-plane: co-pol is E

L H-plane: co-pol is E
Js
W
x E plane
Probe

For radiation patterns, it is usually more convenient to

place the origin at the middle of the patch
(this keeps the formulas as simple as possible).
H plane

53
 General Characteristics
Radiation Patterns (cont.)

Comments on radiation patterns:

 The E-plane pattern is typically broader than the H-plane pattern.

 The truncation of the ground plane will cause edge diffraction, which
tends to degrade the pattern by introducing:

 Rippling in the forward direction

 Back-radiation

 Pattern distortion is more severe in the E-plane, due to the angle

dependence of the vertical polarization E on the ground plane.
(It varies as cos ()).

54
 General Characteristics
Radiation Patterns

Edge diffraction is the most serious in the E plane.

y E varies as cos 
Space wave
L
Js
W
E plane

H plane

55
General Characteristics
Radiation Patterns

E-plane pattern

Red: infinite substrate and ground plane

Blue: 1 meter ground plane
0

30 -30

-10

60 -20 -60

-30

-40 -30 -20 -10

90 -90

120 240

150 210

180 56
General Characteristics
Radiation Patterns

H-plane pattern

Red: infinite substrate and ground plane

Blue: 1 meter ground plane
0

45 -10 -45

-20

-30

-40 -30 -20 -10

90 -90

135 225

180
57
 General Characteristics

Directivity

The directivity is fairly insensitive to the substrate thickness.

The directivity is higher for lower permittivity, because the patch is

larger.

58
 Outline
 Overview of microstrip antennas
 Feeding methods
 Basic principles of operation
 General characteristics
 CAD Formulas
 Input Impedance
 Radiation pattern
 Circular polarization
 Improving bandwidth

59
 CAD Formulas
CAD formulas for the important properties of the
rectangular microstrip antenna will be shown.

 Radiation efficiency
 Bandwidth (Q)
 Resonant input resistance
 Directivity

60
 CAD Formulas
Radiation Efficiency
erhed
er 
  Rsave   1    3    r   L   1 
1 e hed
 d           
r
  
 0  h / 0    16   1  W
p c   h / 0  

The efficiency becomes small as the substrate gets thin,

where if there is dielectric or conductor loss.

 d  tan   loss tangent of substrate

1 
Rs  surface resistance of metal   Rsave   Rspatch  Rsground  / 2
 2

“hed” refers to a unit-amplitude horizontal electric dipole.

61
 CAD Formulas
Radiation Efficiency (cont.)

Psphed 1
erhed  
Psphed  Pswhed Pswhed
1  hed
Psp
where
1
P hed
sp  2  0 
0
kh
2
 c1 
80 2

 1 
3
1 3
 2  k0 h  60 1   
3
Pswhed
0    r  

“hed” refers to a unit-amplitude horizontal electric dipole.

62
 CAD Formulas
Radiation Efficiency (cont.)

Hence, we have

1
ehed
r  3
3  1  1
1    k0 h   1  
4  c1   r 

Physically, this term is the radiation efficiency of a

horizontal electric dipole (hed) on top of the substrate.

63
 CAD Formulas
Radiation Efficiency (cont.)

The constants are defined as follows:

1 2/5
c1  1  
r  r2

 k0 W    a22  2a4     k0 W   c2    k0 L 
a2 2 3 4 1 2
p  1
10  560  5
 1 
 a2 c 2    k 0 W   k 0 L 
2 2

 70 

c2   0.0914153

a2   0.16605
a4  0.00761
64
 CAD Formulas
Improved formula for HED surface-wave power (due to Pozar)

 r  x  1
2 3/2
k2
0
P hed
 0 0 Note: x0 in this formula is
8  r 1  x1   (k0 h) x02  1 1   r2 x1 
sw not the feed location!

x02  1  r2   01   r  r2  2 01   02

x1  x0  1 
 r  x02  r2  12

 0  s tan  k0 h  s 
1
1    tan  k0 h  s  
 k0 h  s 

s  cos  k0 h  s  
2

s   r 1

65
 CAD Formulas
Bandwidth

1   Rsave  1   16   p c1   h   W  1 
BW   d          hed 
2    0  h / 0   3    r   0   L   er  

1
Q
2 BW For a lossless patch, the bandwidth is approximately
proportional to the patch width and to the substrate thickness.
It is inversely proportional to the substrate permittivity.
For very thin substrates the bandwidth will increase for a
lossy patch, but as the expense of efficiency.

BW is defined from the frequency limits f1 and f2 at which SWR = 2.0.

f 2  f1
BW  (multiply by 100 if you want to get %)
f0
66
 CAD Formulas
Quality Factor Q

Us
Q  0 U s  energy stored in patch cavity
P
P  power that is radiated and dissipated by patch
1 P

Q 0U s

P  Pd  Pc  Psp  Psw

1 1 1 1 1
   
Q Qd Qc Qsp Qsw

67
 CAD Formulas
Q Components

Qd  1 / tan 

  0   ( k0 h )  Rsave   Rspatch  Rsground  / 2

Qc     ave 
 2   Rs 
3  r  L  1 
Qsp      The constants p and c1 were defined previously.
16  pc1   W   h / 0 

 erhed  1
Qsw  Qsp  erhed 
hed 
3
 1  e  3  1  1
r
1    k0 h    1  
4  c1   r 

68
 CAD Formulas
Resonant Input Resistance
Probe-feed Patch

  x0 
R R max
in  Redge cos  2

 L 

 40   L   h 
   
    W   0 
Redge 
 Rs   1   16   p c1   W  h  1 
d       L     hed 
 
 0  h / 0   3 
 r    0   er 

For a lossless patch, the resonant resistance is approximately independent of

the substrate thickness. For a lossy patch it tends to zero as the substrate gets
very thin. For a lossless patch it is inversely proportional to the square of the
patch width and it is proportional to the substrate permittivity.

69
 CAD Formulas
Approximate CAD formula for probe (feed) reactance (in Ohms)

a = probe radius h = probe height

0   2 
Xp   k0 h      ln   
2    r  k0 a   
This is based on an infinite parallel-plate model.

r 2a h
X p   Lp

  0.577216 (Euler’s constant)

0  0 /  0  376.7303 
70
 CAD Formulas
Observations:

 Feed (probe) reactance increases proportionally with substrate

thickness h.

 Feed reactance increases for smaller probe radius.

0   2 
Xp   k0 h   ln  
2   r  k0 a 

Important point:

If the substrate gets too thick, the probe reactance will make it difficult
to get an input match, and the bandwidth will suffer.

71
 CAD Formulas
Directivity

 3  r 
D   k1h  
2
 tanc
 pc1    r  tan  k1h  
2

k1  k0  r

where

tanc  x   tan  x  / x

The constants p and c1 were defined previously.

72
 CAD Formulas
Directivity (cont.)

For thin substrates:

3
D
p c1

(The directivity is essentially independent of the substrate thickness.)

73
 Outline
 Overview of microstrip antennas
 Feeding methods
 Basic principles of operation
 General characteristics
 CAD Formulas
 Input Impedance
 Radiation pattern
 Circular polarization
 Improving bandwidth

74
 Input Impedance
Various models have been proposed over the years for calculating the
input impedance of a microstrip patch antenna.

 Transmission line model

 The first model introduced
 Very simple

 Cavity model (eigenfunction expansion)

 Simple yet accurate for thin substrates
 Gives physical insight into operation

 CAD circuit model

 Extremely simple and almost as accurate as the cavity model

 Spectral-domain method
 More challenging to implement
 Accounts rigorously for both radiation and surface-wave excitation

 Commercial software
 Very accurate
 Can be time consuming

75
 Input Impedance

CAD Circuit Model for Input Impedance

The circuit model discussed assumes a probe feed.

Other circuit models exist for other types of feeds.

The mathematical justification of the CAD circuit model comes from a

cavity-model eigenfunction analysis.

76
 Input Impedance
Probe-fed Patch
 Near the resonance frequency, the patch cavity can be approximately modeled
as a resonant RLC circuit.
 The resistance R accounts for radiation and losses.
 A probe inductance Lp is added in series, to account for the “probe inductance”
of a probe feed.

Probe Patch cavity

L
Lp
Zin R C

77
 Input Impedance
R
Z in  j Lp 
 f f0 
1  jQ   
 f0 f 
R 1
Q BW  BW is defined here by SWR < 2.0 when the
0 L 2Q RLC circuit is fed by a matched line (Z0 = R).

1
0  2 f 0 
LC

L
Lp
Zin R C
Z in  Rin  jX in

78
 Input Impedance
R
Rin  2
  f f0  Rinmax  Rin R
1  Q     f  f0
  f0 f 
R is the input resistance at the resonance of the patch cavity
(the frequency that maximizes Rin).

f  f0 (resonance of RLC circuit)

Lp

L
max C
R in R
f  f0

79
 Input Impedance
R
Z in  j Lp 
 f f0 
1  jQ   
 f0 f 
The input resistance is determined once we know four parameters:
 f0: the resonance frequency of the patch cavity
 R: the input resistance at the cavity resonance frequency f0
 Q: the quality factor of the patch cavity
 Lp: the probe inductance

Lp (R, f0, Q)

L
C
Zin R

80
 Input Impedance
Typical plot of input impedance

Z RLC
RRLC
RRLC / 2 Without probe inductance

f
f0 X RLC
 RRLC / 2

R
X p  RRLC / 2
Rinres With probe inductance

Xp
f
X p  RRLC / 2 f0 X
f res
81
 Input Impedance

Optimization to get exactly 50  at the desired resonance frequency:

 Vary the length L first until find the value that gives an input reactance
of zero at the desired frequency.

 Then adjust the feed position x0 to make the real part of the input
impedance 50  at this frequency.

82
 Design Example
Design a probe-fed rectangular patch antenna on a substrate having a relative permittivity of 2.33 and
a thickness of 62 mils (0.1575 cm). (This is Rogers RT Duroid 5870.) Choose an aspect ratio of W / L =
1.5. The patch should resonate at the operating frequency of 1.575 GHz (the GPS L1 frequency).
Ignore the probe inductance in your design, but account for fringing at the patch edges when you
determine the dimensions. At the operating frequency the input impedance should be 50  (ignoring
the probe inductance). Assume an SMA connector is used to feed the patch along the centerline (at y
= W / 2), and that the inner conductor of the SMA connector has a radius of 0.635 mm. The copper
patch and ground plane have a conductivity of  = 3.0 107 S/m and the dielectric substrate has a loss
tangent of tan = 0.001.

1) Calculate the following:

 The final patch dimensions L and W (in cm)
 The feed location x0 (distance of the feed from the closest patch edge, in cm)
 The bandwidth of the antenna (SWR < 2 definition, expressed in percent)
 The radiation efficiency of the antenna (accounting for conductor, dielectric, and surface-
wave loss, and expressed in percent)
 The probe reactance Xp at the operating frequency (in )
 The expected complex input impedance (in ) at the operating frequency, accounting for the
probe inductance
 Directivity
 Gain

83
 Design Example
Continued

2) Find (f0, R, Xp, and Q) and plot the input impedance vs. frequency using the CAD
circuit model.

3) Keep W/L = 1.5, but now vary the length L of the patch and the feed position x0 until
you find the value that makes the input impedance exactly 50+j(0)  at 1.575 GHz.

84
 Outline
 Overview of microstrip antennas
 Feeding methods
 Basic principles of operation
 General characteristics
 CAD Formulas
 Input Impedance
 Radiation pattern
 Circular polarization
 Improving bandwidth

85
 Radiation Pattern
There are two models often used for calculating the radiation pattern:
 Electric current model
 Magnetic current model

The origin is placed at the center of the patch,

at the top of the substrate, for the pattern calculations.

z
Patch
x
h Probe r

Coax feed

86
 Radiation Pattern
Electric current model:
We keep the physical currents flowing on the patch (and feed).

Patch
x
h Probe r

Coax feed

J spatch  J top
s  J bot
s J spatch

x
h J sprobe r

87
 Radiation Pattern
Magnetic current model:
We apply the equivalence principle and invoke the (approximate) PMC condition
at the edges.
Equivalence surface Patch
x
J se  nˆ  H
h Probe r
M se   nˆ  E
Coax feed

The equivalent
surface current is
approximately zero
on the top surface
x
(weak fields) and h M se M se r
the sides (PMC).
We can ignore it on
the ground plane (it
does not radiate). M se   nˆ  E
88
 Radiation Pattern
Theorem
The electric and magnetic models yield identical patterns
at the resonance frequency of the cavity mode.

Assumption:
The electric and magnetic current models are based on the fields of a
single cavity mode, corresponding to an ideal cavity with PMC walls.

89
 Radiation Pattern
Comments on the Substrate Effects

 The substrate can be neglected to simplify the far-field calculation.

 When considering the substrate, it is most convenient to assume an infinite
substrate (in order to obtain a closed-form solution).
 Reciprocity can be used to calculate the far-field pattern of electric or
magnetic current sources inside of an infinite layered structure.
 When an infinite substrate is assumed, the far-field pattern always goes to
zero at the horizon.

90
 Radiation Pattern
Comments on the Two Models

 For the rectangular patch, the electric current model is the simplest since
there is only one electric surface current (as opposed to four edges).
 For the rectangular patch, the magnetic current model allows us to classify
the “radiating” and “nonradiating” edges.

M se   nˆ  E
y
x 
J s  xˆ A10 cos  
 L  “Radiating edges” x
Ez   sin  
 L 

Js On the nonradiating edges, the

W x magnetic currents are in opposite
directions across the centerline (x = 0).

M se
L “Nonradiating edges”
91
 Radiation Pattern
Rectangular Patch Pattern Formula
(The formula is based on the electric current model.)

L
x
h εr
Infinite ground plane and substrate
H-plane
The origin is at the
center of the patch. y

(1,0) mode
W x E-plane
x
J s  x cos 
ˆ 
 L  The probe is on the x axis.
L
92
 Radiation Pattern
The far-field pattern can be determined by reciprocity.

  ky W    kx L  
 sin     cos   
  WL    2    2  
Ei (r , ,  )  Eihex  r , ,    
 2   ky W     2  k L 2 
     x  
 2   2   2  
y
i   or 
k x  k0 sin  cos 
k y  k0 sin  sin  Js
W x
The “hex” pattern is for a
horizontal electric dipole in the x direction,
sitting on top of the substrate.
L

93
 Radiation Pattern
Ehex  r , ,     E0 sin  F  
Ehex  r , ,    E0 cos  G  

where
  j 0   jk0 r
E0   e
 4 r 

2 tan  k0 h N   
F   1   TE
  
tan  k0 h N     j N   sec 

2 tan  k0 h N    cos 
G    cos  1  TM    
r
tan  k0 h N     j cos 
N  

N     r  sin 2   Note: To account for lossy substrate, use

 r   rc   r 1  j tan  
94
 Circular Polarization

Three main techniques:

1) Single feed with “nearly degenerate” eigenmodes (compact but small

CP bandwidth).

2) Dual feed with delay line or 90o hybrid phase shifter (broader CP
bandwidth but uses more space).

3) Synchronous subarray technique (produces high-quality CP due to

cancellation effect, but requires even more space).

The techniques will be illustrated with a rectangular patch.

95
 Circular Polarization
Single Feed Method

(0,1)
The feed is on the diagonal.
The patch is nearly
(but not exactly) square.
(1,0)
W
L W
L

Basic principle: The two dominant modes (1,0) and (0,1) are
excited with equal amplitude, but with a 45o phase.

96
 Circular Polarization
Design equations: y
fx  f y
f CP 
2
(0,1)
The optimum CP frequency is the
1
average of the x and y resonance BW 
frequencies. 2Q W
(1,0)
(SWR < 2 )
 1  x0  y0
f x  f CP  1   
 2Q   Top sign for LHCP,
x
 bottom sign for RHCP. L
 1  
f y  f CP  1   
 2Q 

The frequency fCP is also the resonance frequency: Zin  Rin  Rx  Ry

The resonant input resistance of the CP patch at fCP is the same as
what a linearly-polarized patch fed at the same position would be.
97
 Circular Polarization
Other Variations
Note: Diagonal modes are used as degenerate modes

y y

L L

x x
L L

Patch with slot Patch with truncated corners

98
 Circular Polarization
compare bandwidths (impedance and axial-ratio):

1
Linearly-polarized (LP) patch: BW LP
SWR  (SWR  2)
2Q

Circularly-polarized (CP) single-feed patch:

BW CP
SWR 
Q
2
(SWR  2)
CP
BWAR 
0.348
Q
 AR  2 (3dB) 

The axial-ratio bandwidth is small when using the single-feed method.

W. L. Langston and D. R. Jackson, “Impedance, Axial-Ratio, and Receive-Power Bandwidths of

Microstrip Antennas,” IEEE Trans. Antennas and Propagation, vol. 52, pp. 2769-2773, Oct. 2004.

99
 Circular Polarization
Dual-Feed Method
y RHCP
Phase shift realized with delay line: L

P
L

P+g/4
100
 Circular Polarization
Phase shift realized with 90o quadrature hybrid (branchline coupler)

RHCP

Z0 Z0 / 2 Z0
Feed

g/4 Z0

50 Ohm load
g/4

This gives us a higher bandwidth than the simple power divider,

but requires a load resistor.

101
 Circular Polarization
Synchronous Rotation
Multiple elements are rotated in space and fed with phase shifts.

-180o

-90o

-270o

0o
Because of symmetry, radiation from higher-order modes (or probes)
tends to be reduced, resulting in good cross-pol.

102
 Outline
 Overview of microstrip antennas
 Feeding methods
 Basic principles of operation
 General characteristics
 CAD Formulas
 Input Impedance
 Radiation pattern
 Circular polarization
 Circular patch
 Improving bandwidth
 Miniaturization
 Reducing surface waves and lateral radiation
103
 Outline
 Overview of microstrip antennas
 Feeding methods
 Basic principles of operation
 General characteristics
 CAD Formulas
 Input Impedance
 Radiation pattern
 Circular polarization
Improving bandwidth

104
 Improving Bandwidth
Probe Compensation
L-shaped probe: As the substrate
thickness increases the
probe inductance limits
the bandwidth – so we
compensate for it.

Top view

Capacitive “top hat” on probe:

105
 Improving Bandwidth
SSFIP: Strip Slot Foam Inverted Patch (a version of the ACP).
 Bandwidths greater than 25% have been achieved.
 Increased bandwidth is due to the thick foam substrate and
also a dual-tuned resonance (patch+slot).

Note: There is no probe inductance to worry about here.

Patch substrate Patch

Foam
Microstrip
substrate

Microstrip line Slot

106
 Improving Bandwidth
Stacked Patches

 Bandwidth increase is due to thick low-permittivity antenna

substrates and a dual or triple-tuned resonance.
 Bandwidths of 25% have been achieved using a probe feed.
 Bandwidths of 100% have been achieved using an ACP feed.

Patch substrates Coupling patch

Top Patch
Microstrip
substrate

Microstrip line
Slot

107
Improving Bandwidth
Parasitic Patches

Radiating Edges Gap Coupled

Microstrip Antennas
(REGCOMA).

Non-Radiating Edges Gap

Coupled Microstrip Antennas
(NEGCOMA)

Four-Edges Gap Coupled

Microstrip Antennas
(FEGCOMA)

Bandwidth improvement factor:

REGCOMA: 3.0, NEGCOMA: 3.0, FEGCOMA: 5.0?
108
Improving Bandwidth
Direct-Coupled Patches

Radiating Edges Direct

Coupled Microstrip Antennas
(REDCOMA).

Non-Radiating Edges Direct

Coupled Microstrip Antennas
(NEDCOMA)

Four-Edges Direct Coupled

Microstrip Antennas
(FEDCOMA)

Bandwidth improvement factor:

REDCOMA: 5.0, NEDCOMA: 5.0, FEDCOMA: 7.0
109
 Improving Bandwidth
U-Shaped Slot

The introduction of a U-shaped slot can give a

significant bandwidth (10%-40%).

(This is due to a double resonance effect, with two different modes.)

110
Improving Bandwidth
Double U-Slot

A 44% bandwidth was achieved.

111
 Improving Bandwidth
E Patch

A modification of the U-slot patch.

A bandwidth of 34% was achieved (40% using a capacitive “washer” to

compensate for the probe inductance).

112
 Multi-Band Antennas

A multi-band antenna is sometimes more desirable than a broadband

antenna, if multiple narrow-band channels are to be covered.

General Principle:

Introduce multiple resonance paths into the antenna.

113
 Multi-Band Antennas

Low-band

Low-band

Feed High-band Feed

Low-band

High-band

Dual-band E patch Dual-band patch with parasitic strip

114