Microstrip Antennas (MSA)

Lecture 1
MSA Configuration
Microstrip antennas are characterized by their physical parameters.

Patch Traveling slot

2
3
 Advantages
 Light weight, low volume, low profile, planar configuration, which can be made conformal
 Low fabrication cost and ease of mass production
 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.
 Linear and circular polarizations are possible
 Dual frequency antennas can be easily realized
 Feed lines and matching network can be easily integrated with antenna structure 4
 Disadvantages
1. Narrow bandwidth (1 to 5%). Bandwidth is roughly proportional to the
substrate thickness and inversely proportional to the substrate permittivity.
2. Efficiency may be lower than with other antennas. Efficiency is limited by
conductor and dielectric losses*, and by surface-wave loss**.
3. Low power handling capacity. Cannot handle extremely large amounts of
power (dielectric breakdown).
4. Practical limitation on Gain (around 30 dB)

* 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).
5
 Disadvantages cont.

5. Poor isolation between the feed and radiating elements

6. Excitation of surface waves
7. Tolerance problem requires good quality substrate, which are expensive
8. Polarization purity is difficult to achieve
9. Size is large at lower frequency

6
 Applications
1. Pagers and mobile phones
2. Doppler and other radars
3. Satellite communication
4. Radio altimeter
5. Command guidance and telemetry in missiles
6. Feed elements in complex antennas
7. Satellite navigation receiver
8. Biomedical radiator 7
Basic Properties

 𝜆 𝜆
< 𝐿<
3 2

 2.2 ≤ 𝜀 𝑟 ≤ 12

0.003 𝜆 ≤ h≤ 0.05 𝜆
 

𝑊 𝑖𝑠𝑡𝑢𝑛𝑒𝑑 𝑡𝑜 𝑔𝑒𝑡 𝑡h𝑒 𝑑𝑒𝑠𝑖𝑟𝑒𝑑 𝑖𝑚𝑝𝑒𝑑𝑎𝑛𝑐𝑒

 

8
Rectangular MSA
Note:

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.

9
Electrical Properties of Commonly Used Substrate Materials For Microstrip
Antennas
Material
Material Dielectric Constant Loss Tangent
Air 1 0
Air 1 0
Unreinforced PTFE, Cuflon 2.1 0.0004
Unreinforced PTFE, Cuflon 2.1 0.0004
0.0009
RT Duroid 5880 2.20
RT Duroid 5880 2.20 0.0009
Polyster 3.2 0.003
Polyster 3.2 0.0001
0.003
Fused Quartz 3.78
FR-4
Fused Quartz 4.4
3.78 0.0001
0.020
96% Alumina 9.40 0.0010
FR-4 4.4 0.020
Semi-Insulating 0.0020
0.0010
96% AluminaGaAs 12.9
9.40

Semi-Insulating GaAs 12.9 0.0020

10
 

Alumina is not very useful as an antenna because it is a high

leads to very less radiation.

Glass epoxy substrate also popularly known as FR4 substrate. Now this is
the commonly used substrate for all printed circuit board. The typical
dielectric constant can be 4.4, but in the reality it may be from 3.8 to about
4.6. The problem with this is that the tangent is very high which is 0.02, but
bit advantage is the cost is very low.
11
Bandwidth for VSWR < 2
is from 1.76 to 1.855 GHz
(95 MHz)
% BW ~ 5%

12
 Microwave Integrated Circuits (MIC) vs MSA

Parameters MIC MSA

Dielectric Constant (εr) Large Small
Thickness (h) Small Large
Generally Small
Width (W) Generally Large
(impedance dependent)
Minimum (small fringing Maximum (large fringing
Radiation
fields) fields)
Filters, power dividers,
Examples Antennas
couplers, amplifiers, etc.

13
 Probe Coupling Method
The inner conductor of the probe line is connected to
patch lower surface through slot in the ground plane and
substrate material.

To get perfect impedance matching we need to find

out the location of the feed point over the antenna
element.
Design simplicity and input impedance adjustment
through feed point positioning, makes this feeding
method popular.

But there are some limitations also like larger lead for thicker substrate, difficulty in soldering
for array elements etc. 14
Probe Coupling Method cont.

15
 Probe Coupling Method cont.
Coaxial Feed y
2   x0 
R  Redge cos   z
 L 
 x0 , y0 
(The resistance varies as the square r h W
of the modal field shape.) x
x
Advantages: L
 Simple
 Directly compatible with coaxial cables
Note:
 Easy to obtain input match by adjusting feed position
A feed along the centerline at
y = W/2 is the most common
Disadvantages: (this minimizes higher-order
 Significant probe (feed) radiation for thicker substrates modes and cross-pol).
 Significant probe inductance for thicker substrates (limits
bandwidth)
 Not easily compatible with arrays 16
 Microstrip Line Feed

This method is very simple to design and fabricate. But this technique suffers from some
limitations. If substrate thickness is increased in the design then the surface waves and the
spurious radiation also increases. Because of that the undesired cross polarization radiation
arises. Microstrip line feeding can be used in the conditions where performance of the antenna
is not a strict matter.
17
Microstrip Feed (contd.)

18
Microstrip Feed (contd.)

19
Microstrip Feed (contd.)
Inset Feed
Advantages:
 Simple
 Allows for planar feeding Microstrip line

 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

20
Microstrip Feed (contd.)

An investigation has shown that the resonant

input resistance varies as: y0
Wf
  𝑦𝑜
S W
𝜋
𝑅𝑖𝑛 = 𝐴 𝑐𝑜𝑠
2
( (
2 𝐿
−𝐵 ))
L
Less accurate approximation:

  𝜋 𝑦𝑜 The coefficients A and B depend on the

𝑅𝑖𝑛 ≈ 𝑅 𝑒𝑑𝑔𝑒 𝑐𝑜𝑠
2
(
𝐿 ) notch width S but (to a good
approximation) not on the line width Wf .

See equations (14.12) – 14.20a) Balanis, also example 14.2 21

Example 14.2 (Balanis)
  𝜺 𝒓 =𝟐 . 𝟐 𝒇 =𝟏𝟎 𝑮𝑯𝒛 𝒉 =𝟎 . 𝟏𝟓𝟖𝟖 𝒄𝒎 𝒁𝒐 =𝟓𝟎 𝜴

Solution:
 2
W   𝑊
  =1.186 𝑐𝑚
2 r  1

𝜀 𝑟 +1 𝜀 𝑟 − 1 1
𝜀  𝑒 =
2
+
2
(√ 1+12
h
𝑊
) 𝜀  𝑒 =1.972


Le 
 𝐿𝑒=1.0683 𝑐𝑚
2 e

∆  L=0.081 cm
 𝐿= 𝐿𝑒 −2 ∆ 𝐿=0.906 𝑐𝑚 22
Example 14.2 (Balanis) cont.

2
  𝑘𝑊

𝐼 =∫
𝜋

( sin
2 (
cos 𝜃 ) ) ( sin 𝜃 )3 d θ
G1 
I
2 𝐺
  1=0.00157 𝑠𝑖𝑒𝑚𝑒𝑛𝑠
0 cos 𝜃 120 

2
  𝑘𝑊
1
𝜋

( sin( 2
cos 𝜃 ) ) 3
𝐺 12= ∫ 𝐽 𝑜 ( 𝑘𝐿𝑠𝑖𝑛 𝜃 ) ( sin 𝜃 ) d θ   12=6.169 ×10 −4
𝐺
𝜂𝜋 0 cos 𝜃

  1
𝑅 𝑖𝑛 =   𝑅 𝑖𝑛 = 228.38 Ω
2 (𝐺 1 +𝐺 12)

 𝒙 𝒐=𝟎 . 𝟑𝟏𝟐𝟔 𝒄𝒎
  𝐿 −1 𝑍𝑜
𝑥 𝑜= 𝑐𝑜𝑠
𝜋 (√ ) 𝑅𝑖𝑛
Example 14.2 (Balanis) cont.

  Single slot

2
  𝑘𝑊

𝑃𝑟𝑎𝑑 =
|𝑉 𝑜|
2 𝜋

∫ ( sin( 2
cos 𝜃 ) ) ( sin 𝜃 )3 d θ
2
2 𝜂𝑜 𝜋 0 cos 𝜃

Single slot D = 10.2 dB

  Two slots
2
  𝑘𝑊
|𝑉 𝑜|
2 𝜋 𝜋

( sin( 2
cos 𝜃 ) )( 𝑘𝐿
2
𝑃𝑟𝑎𝑑 =
2 𝜂𝑜 𝜋 2 ∫∫
0 0 cos 𝜃 (
co s
2
s∈(𝜃)sin ⁡(𝜑) )) ( sin 𝜃 )3 𝑑 𝜃 𝑑 𝜑

Two slots D = 7.315 dB

24
‫‪Example 14.2 (Balanis) cont.‬‬

‫‪ ‬‬

‫𝑊𝐵‬
‫‪ ‬‬ ‫‪=6.48 %‬‬

‫‪ ‬‬ ‫‪4‬‬ ‫𝑡‬

‫= 𝑊𝐵‬
‫𝑜 𝜆 𝑟𝜀 ‪√ 2‬‬ ‫ممكن المقارنة هنا‬
‫تأخذ باالعتبار السمك ‪t‬‬

‫‪25‬‬
 Proximity Coupled Method

This method can be employed, where two or

multilayer substrate configuration is
considered. Generally in this configuration,
microstrip line will be placed on lower
substrate and the patch element will be placed
on the upper substrate. Other name for this
feeding is electromagnetically coupled feed.
Capacitive nature will appear between feed line
and patch in this case.

By choosing thin lower substrate layer and placing patch on top layer will improve
the bandwidth and reduce the spurious radiation. Fabrication of this feeding is slightly
difficult because of alignment problems in feed and patch at proper location. Peaceful thing
is soldering and related problems can be eliminated. 26
Proximity Coupled Method cont.

27
Proximity Coupled Method cont.

28
Proximity Coupled Method cont.

29
 Aperture Coupled Feed
This method employs ground plane
between two substrates. A slot will be
placed on the ground plane and feed line
will be placed on lower substrate. This will
be electromagnetically connected to patch
on the upper substrate through the ground
plane slot.

One should take care about substrate parameters and they have to choose in a way that feed
optimization and independent radiation functioning can exist. The coupling slot should be
nearly cantered so that the patch magnetic field will be maximum.
30
Aperture Coupled Feed cont.

31
Aperture Coupled Feed cont.

1. Ground plane between substrates isolates the feed from the radiating
element, minimizing interference
2. Moderate spurious radiation
3. Matching is easily done by changing feed position
4. Narrow bandwidth
5. Most difficult to fabricate
6. Independent optimization of feed and radiating element
32
 Transmission line Model
The Rectangular Patch
• 
The TL model is the simplest of all, representing the rectangular patch as two slots,
separated by a low impedance () transmission line of length L.

𝐿𝑒𝑓𝑓 =𝐿+2 ∆ 𝐿
  𝑊 𝑒𝑓𝑓 =𝑊 +2 ∆ 𝑊
 

Where
  h   2 2
∆ 𝐿=
√ 𝜀 𝑒𝑓𝑓 𝑓 𝑜=
𝑐
2 √ 𝜀 𝑒𝑓𝑓 [√ ( ) ( ) ]
𝑚
𝐿
+
𝑛
𝑊
33
RMSA – Characterization

34
RMSA: Design Equations

  𝜀 𝑟 +1 𝜀 𝑟 − 1 1
𝜀 𝑒𝑓𝑓 =
2
+
2
(√ 1+12
h
𝑊
)
  𝑊
∆ 𝐿 =0.412 h
( 𝜀 𝑒𝑓𝑓 +0.3 ) ( h
+0.264)
𝑊
( 𝜀 𝑒𝑓𝑓 − 0.258 ) ( +0. 8)
h

35
RMSA: Design Equations cont.
 
𝝀 𝟐
𝑾=
𝟐 √ 𝜺 𝒓 +𝟏
Larger W, then aperture area will increase.   𝑾 ∝ 𝑮𝒂𝒊𝒏

Sometimes we may have to take a smaller than this W because we need to fit the antenna into
a certain area. Suppose let us say you want to design an antenna for a mobile phone you can’t
use a very large W, because of the size of the mobile phone is small. So, you may end up
taking a small W which will then compromise on the bandwidth as well as on the gain.
36
RMSA: Design Equations cont.

•   𝐿= 𝐿 −2 Δ 𝐿
𝑒𝑓𝑓

A higher substrate permittivity allows for a smaller

antenna (miniaturization), but with a lower bandwidth.

37
Rectangular patch and equivalent circuit

a) Rectangular patch b) Transmission model equivalent

38
conductance and input impedance

39
 conductance and input impedance cont.

At resonance

This formula does not take into account mutual effects between the slots.

40
 conductance and input impedance cont.

  To take into account mutual effects between the slots, we introduce .

Example 14.2
41
 RMSA: Design Example
• 
Design a RMSA for Wi-Fi application (2.400 to 2.483 GHz)

Chose Substrate: , h = 0.16 cm and tan δ= 0.001

  3 ×10 10
𝜆= 9
=𝟏𝟐 . 𝟑 𝒄𝒎
2.44 ×10

 
𝜆 2 12.3 2
𝑊=
√
2 𝜀 𝑟 +1
=
2 √
2.32+1
=𝟒 .𝟕𝟕 𝐜𝐦
42
 RMSA: Design Example cont.

  𝜀 𝑟 +1 𝜀 𝑟 − 1 1   2.32+1 2.32 − 1 1
𝜀 𝑒𝑓𝑓 =
2
+
2
(√ 1+12
h
𝑊
) ¿
2
+
2
(√ 1+12
0.16
4.77
) = 2.22

𝑊 4.77
 
∆ 𝐿 =0.412 h
( 𝜀 𝑒𝑓𝑓 +0.3 ) ( h )
+0.264  
¿ 0.412 ( 0.16 )
( 2.22+ 0.3 ) ( 0.16 )
+0.264
=0.083
𝑊 4.77
( 𝜀 𝑒𝑓𝑓 − 0.258 ) ( +0. 8) ( 2.22 − 0.258 ) ( + 0.8 )
h 0.16

  𝜆   12.3
𝐿𝑒𝑓𝑓 = ¿ =4.13 𝑐𝑚  𝐿= 𝐿𝑒𝑓𝑓 −2 Δ 𝐿  𝐿=4.13 −2 ( 0.083 ) =3.96 𝑐𝑚
2 √ 𝜀 𝑒𝑓𝑓 2 √ 2.22 43
RMSA: Design Example –Simulation using IE3D
• L  = 3.9 cm, W = 4.7 cm, x = 0.7 cm = 2.32, h = 0.16 cm and tan δ= 0.001

Zin= 54Ωat f = 2.414 GHz  BW for is from 2.395 to 2.435 GHz = 40 MHz
Designed f = 2.4415 and Simulated f = 2.414 GHz
error = 1.1%. Also, BW is small.
SOLUTION: Increase h and reduce L 44
Radiation Pattern, Input Impedance and Bandwidth
1) The radiation patterns in the two principal planes are

a) E-plane (φ = 0◦): b) H-plane (φ = 90◦):

  𝑘𝑊
 
𝐸𝜃 =Eo cos(𝑘𝐿
)
sin (𝜃)
𝐸𝜑 =− Eo s inc (
2 )
sin(𝜃) co s (𝜃)
2

2) Input Impedance and Bandwidth

  𝜀 2𝑟 𝐿
2   𝜀 𝑟 −1 h 𝑊 2

𝑍 𝑎 ≈ 90 ( )
𝜀 𝑟− 1 𝑊
𝐵 𝑊 =3.771
𝜀 2𝑟 𝜆 𝐿 ( )
45
 Example:

Design a rectangular patch. RT/Duroid 5880 substrate (ε = 2.2 and h =

r

1.588 mm) is to be used to make a resonant rectangular patch antenna of

linear polarization.
a) Design such an antenna to work at 2.45 GHz for Bluetooth applications.
b) Estimate its directivity.
c) If it is to be connected to a 50 ohm microstrip using the same PCB board,
design the feed to this antenna.
d) Find the fractional bandwidth for VSWR < 2.

46
 Example:
  𝜀 2𝑟 𝐿
2
𝑍 𝑎 ≈ 90
𝜀 𝑟− 1 𝑊( ) =254 Ω
 
𝝀 𝟐
𝑾=
√
𝟐 𝜺 𝒓 +𝟏
=¿ 48.40 𝑚𝑚 ¿
which does not match well with a 50 standard
microstrip and therefore a quarter wavelength
transformer is used to connect them. The
  ε r +1 ε r − 1 1
ε eff =
2
+
2
(√ 1+12
h
W
)
=2.108 characteristic
impedance of the transition section should be

𝑊 𝑍
  𝑇 = √ 50 𝑥 254=122.69 Ω
 
∆ 𝐿 =0.412 h
(
( 𝜀 𝑒𝑓𝑓 +0.3 ) h +0.264 ) =0.84 𝑚𝑚
𝑊
( 𝜀 𝑒𝑓𝑓 − 0.258 ) ( +0. 8)
h 50- transmission lines become inconveniently wide on
low-dielectric-constant substrates.
𝑳=𝟒𝟎 . 𝟒𝟗 𝒎𝒎
 
47
 Example:
Hence, the length of the transition should be:

  60 8h 𝑊𝑇
𝑍𝑜=
√𝜀𝑟
𝑙𝑛
𝑊 𝑇
+
(4 h
=122.69
)   𝜆
=23.44 𝑚𝑚
4 √ 𝜀 𝑒𝑓𝑓
𝑾
  𝑻 =𝟎 . 𝟔𝟏𝟓 𝒎𝒎

  120 𝜋
𝑍𝑜= = 50
To obtain its length, we need to calculate the 𝑊𝑚 𝑊𝑚
relative effective permittivity for the line
√𝜀𝑟 ( h
+1.393+0.6667 𝑙𝑛(h
+1.44 ))
 
𝐿 −1 50
 
ε eff =
ε r +1 ε r − 1
2
+
2
(√
1

1+12
h
W𝑇 )
=1.706 𝑥 𝑜= 𝑐𝑜𝑠
𝜋 (√ )
122.69
=11.32 𝑚𝑚

48
Example:

  𝜀 𝑟 −1 h 𝐿 ❑
𝐵 𝑊 =3.771 2
𝜀𝑟 𝜆 𝑊 ( ) =1.01 %
ANTENNAS FROM THEORY TO PRACTICE, Yi Huang, Kevin Boyle, 1st ed., 2008. 49
Effect of Feed Point Location (x)

With increase in x, input impedance plot shifts right towards higher impedance values.
50
 Effect of Width (W)

 With increase in W, aperture area, and fringing fields increase, hence frequency decreases and
input impedance plot shifts towards lower impedance values.
51
 Effect of Thickness (h)

As h increases, fringing fields and probe inductance increase, frequency decreases and
input impedance plot shifts upward.
  However, to reduce surface waves
52
 Effect of Probe
Diameter

As probe diameter decreases, its inductance increases, so resonance frequency decreases and
input impedance locus moves upward to the inductive region. 53
 Effect of Loss Tangent
(tanδ)

With increase in tan δ, dielectric losses increase, so input impedance locus moves left
towards lower impedance value. BW increases but efficiency and gain decrease.
54
  
RMSA –Pattern for Different

 
With decrease in , both L and W increase, which increases fringing fields and aperture area,
hence both BW and Gain increase. 55
 
RMSA –Pattern for Different

 
With increase in , size of the antenna
decreases for same resonance frequency.
Hence, gain decreases and HPBW
increases.

56
 MSA – BW Variation with hand
f
8 cm
.2 0. 31
  = 2 h =
m
5 9c
0.1
h=

  =1
0 7 0 cm
h =0.0

  (a) Substrate thickmess (b) Frequency (GHz)

57
 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).
58
 General Characteristics
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.
59
 General Characteristics y
Resonant Input Resistance
(x0, y0)
W

L L
x

 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).
60
 General Characteristics
Resonant Input Resistance (cont.)
y

Hence, for (1,0) mode: (x0, y0)

W
  x0 
Rin  Redge cos 
2
 x
 L  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.

See slide 23 and example 14.2 (text book) 61

 General Characteristics
Radiation Efficiency

 Radiation efficiency is the ratio of power radiated into space, to the total
input power.
  𝑃𝑟
𝑒 𝑟=
𝑅𝑡

 The radiation efficiency is less than 100% due to

 Conductor loss
 Dielectric loss
 Surface-wave excitation 62
 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

63
 General Characteristics
Radiation Efficiency (cont.)
Some observations:

 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  𝑅 𝑠 ∝ √ 𝑓 𝑜 ) 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  
𝜔 𝜇𝑜 𝜋 𝑓 𝑜 𝜇 𝑜
√ √
 
𝑅 𝑠= Rs is the surface resistance of the metal.
𝜎𝛿 ¿ =
2 𝜎  𝜎 The skin depth of the metal is .
2
𝑊h𝑒𝑟𝑒𝛿=
√
𝜔𝜇 𝜎 64
 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 losses). There is
no surface-wave power to worry about.

 For a typical substrate such as r = 2.2, the radiation efficiency is maximum

for h / 0  0.02.
65
 General Characteristics
Results: Efficiency (Conductor and dielectric losses are neglected.)
100
2.2
2.2
80
EFFICIENCY (%)

60
 r10.8
 10.8
40
exact
CAD
20

0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
hh // 
 00

r = 2.2 or 10.8 W/L = 1.5 Note: CAD plot uses the Pozar formula. 66
 General Characteristics
Results: Efficiency (All losses are accounted for.)

100
2.2
2.2
80
EFFICIENCY (%)

exact
60
CAD
 rr = 10.8
10.8
40

tan   0.001
20   3.0  107 [S/m]

0
0 0.02 0.04 0.06 0.08 0.1
hh // 00

r = 2.2 or 10.8 W/L = 1.5 Note: CAD plot uses the Pozar formula 67
 General Characteristics
Radiation Pattern
y
E-plane: co-pol is E
L
H-plane: co-pol is E
Js
W
x E plane
Probe

Note:
For radiation patterns, it is usually more convenient
H plane to place the origin at the middle of the patch
(this keeps the formulas as simple as possible).

68
 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 ()).
69
 General Characteristics
Directivity

The directivity is fairly insensitive to the substrate thickness.

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

70
 CAD Formulas

 Radiation efficiency

 Bandwidth (Q)

 Resonant input resistance

 Directivity

 D. R. Jackson, “Microstrip Antennas,” Chapter 7 of Antenna Engineering Handbook, J. L. Volakis, Editor,

McGraw Hill, 2007.

71
 CAD Formulas cont.

• 
This section presents CAD formulas for the rectangular patch.

All of the formulas are independent of the feed except for the input
resistance formula, which assumes a coaxial feed.

All of these formulas assume that the patch is operating at the

resonance of the mode, which is the usual mode of broadside
operation.
72
 CAD Formulas
Radiation Efficiency

The constants are defined as follows:

  𝟏 𝟎.𝟒 𝒄
  𝟐=− 𝟎 .𝟎𝟗𝟏𝟒𝟏𝟓𝟑
𝒄 𝟏=𝟏 − + 𝟐
𝜺𝒓 𝜺𝒓

𝒂
  𝟐 =−𝟎 . 𝟏𝟔𝟔𝟎𝟓  𝒂 𝟒= 𝟎 .𝟎𝟎𝟕𝟔𝟏

  𝒂𝟐 𝟑 𝟏
𝒑=𝟏+
𝟏𝟎
𝟐
( 𝒌 𝒐 𝑾 ) + ( 𝒂 𝟐𝟐+𝟐 𝒂 𝟒 )( )
𝟓𝟔𝟎
𝟒 𝟐
( )
( 𝒌 𝒐 𝑾 ) + 𝟎 .𝟐 𝒄 𝟐 ( 𝒌 𝒐 𝑳 ) + 𝒂𝟐 𝒄 𝟐
𝟕𝟎
𝟐
(𝒌 𝒐 𝑾 ) (𝒌 𝒐 𝑳)
𝟐

73
 CAD Formulas
Radiation Efficiency (cont.)

Hence, we have

  h𝑒𝑑 1
𝑒𝑟 = 3
1 1
1+ 0.75 𝜋 ( 𝑘 𝑜 h )
𝑐1 ( )( 1−
𝜀𝑟 )

Physically, this term is the radiation efficiency of a horizontal electric dipole (hed) on top of the substrate.
74
 CAD Formulas
Radiation Efficiency
  𝑒h𝑒𝑑
𝑟
𝑒 𝑟=
𝑎𝑣𝑒
1+𝑒
h𝑒𝑑
𝑟
[ 𝑡𝑎𝑛 𝛿 +
𝑅𝑠
(
𝜋 𝜂𝑜 )( )][( )(
𝜆𝑜
h
3
16
𝜀𝑟
𝑝 𝑐1 )( )( )]
𝐿
𝑊
𝜆𝑜
h

Comment:
The efficiency becomes small as the substrate gets thin,
if there is dielectric or conductor loss.

  Where
Rsave   Rspatch  Rsground  / 2

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

75
 CAD Formulas
Bandwidth
The bandwidth of the patch may be defined from the frequency limits at which the
standing wave ratio (SWR) reaches a maximum threshold, assuming that the feeding
transmission line that connects to the patch is perfectly matched at the resonance frequency

  𝑎𝑣𝑒

𝐵𝑊 =
1
√2 [
𝑡𝑎𝑛 𝛿 +
𝑅𝑠
𝜋 𝜂𝑜 ( 𝜆𝑜
h
+
16
)( ) ( )(
3
𝑝 𝑐1
𝜀𝑟
𝑊
)( )( )(
𝐿
h
𝜆𝑜
1
𝑒h𝑒𝑑
𝑟
)]
Comments:
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.
76
 CAD Formulas
Bandwidth cont.
f 2  f1
BW  (multiply by 100 if you want to get %)
f0

 where is the impedance resonance frequency of the patch, and and are the low and
high frequencies on either side of the resonance frequency at which SWR = S, with S
being a prescribed value. The bandwidth is then

𝑠 −1   1
 
BW = BW = 𝑓𝑜𝑟 𝑠=2
√𝑠𝑄 √ 2𝑄
BW is defined from the frequency limits f1 and f2 at which SWR = 2.0.

77
 CAD Formulas
Q Components

1   𝜂0 𝑘 𝑜 h
 
𝑄𝑑 =
tan 𝛿
𝑄𝑐 =
2 𝑅 𝑎𝑣𝑒
𝑠
( )
h𝑒𝑑
  3 𝜀𝑟 𝐿 𝜆𝑜   𝑒𝑟
𝑄 𝑠𝑝 ≈
( )( )( )
16 𝑝 𝑐 1 𝑊 h
𝑄 𝑠𝑤 =𝑄 𝑠𝑝
( 1− 𝑒 h𝑒𝑑
𝑟
)
Rsave   Rspatch  Rsground  / 2

78
 Quality Factor Q

Us U s  energy stored in patch cavity

Q  0
P P  power that is radiated and dissipated by patch

P  Pd  Pc  Psp  Psw

1 1 1 1 1
   
Q Qd Qc Qsp Qsw

79
 CAD Formulas
Resonant Input Resistance
Probe-feed Patch

  𝜋 𝑦𝑜
𝑅 = 𝑅 𝑒𝑑𝑔𝑒 𝑐𝑜𝑠
2
( 𝐿 )
  4 𝜂𝑜 𝐿 h

𝑅𝑒𝑑𝑔𝑒 =
( 𝜋 )( )( )
𝑊 𝜆𝑜
𝑅𝑠 𝜆𝑜 16 𝑝 𝑐1 𝑊 h 1
𝑡𝑎𝑛 𝛿 +
( 𝜋 𝜂𝑜 )( ) ( )(
h
+
3 𝜀𝑟 )( )( )( )
𝐿 𝜆𝑜 𝑒 h𝑒𝑑
𝑟

Comments:
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.
80
 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

𝛾  =0.577216 ( 𝐸𝑢𝑙𝑒 𝑟 ′ 𝑠 𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡 )

0  0 /  0  376.7303 
81
 CAD Formulas
Observations:
 Feed (probe) reactance increases proportionally with substrate thickness h.

 Feed reactance increases for smaller probe radius.

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.

82
 CAD Formulas
Directivity

  3 𝜀𝑟
𝐷=
(
𝑝 𝑐1 )[ 𝜀 +𝑡𝑎𝑛 ( 𝑘 h ) ] 𝑡𝑛𝑎𝑐 ( 𝑘 h )
𝑟
2
1
2
1

where
 𝑘 1 =𝑘 𝑜 √𝜀𝑟
  tan ⁡(𝑥 )
𝑡𝑎𝑛𝑐 ( 𝑥 ) =
𝑥
The constants p and c1 were defined previously.

For thin substrates The directivity is essentially independent of the substrate thickness.
83
 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
Z in  Rin  jX in

84
�  
𝑄=
𝑅
𝜔𝑜 𝐿
 
𝐵𝑊 =
1
√ 2𝑄
Input Impedance
Z in  j Lp 

 
❑
R
 f
1  jQ   

 
f0 
 f0 f 

BW is defined here by SWR < 2.0 when the RLC ❑circuit is fed by a matched line (Z0 = R).

0  2 f 0 
1
LC

85
 Input Impedance
R
Rin  2
  f f 0  R max
 Rin R
1  Q     in f  f0
  f 0 f 

R is the input resistance at the resonance of the patch cavity

(the frequency that maximizes Rin).
f  f 0 (resonance of RLC circuit)
Lp

L
max C
R in R
f  f0

86
 Input Impedance
R
Z in  j Lp 
 f f0 
1  jQ   
 f0 f 
The input resistance is determined once we know four parameters:

CAD formulas for  f0: the resonance frequency of the patch cavity
all of these four
 R: the input resistance at the cavity resonance frequency f0
parameters have
been given  Q: the quality factor of the patch cavity
earlier.
 Lp: the probe inductance
Lp (R, f0, Q)

L
C
Zin R

87
 Input Impedance
Typical plot of input impedance

Z RLC
RRLC Without probe inductance

f
f0
X RLC

R
Rinres With probe inductance

Xp
f
f0 X
f res 88
 Input Impedance
Results: Input Resistance vs. Frequency
Note: “exact” means the cavity model will all infinite modes.
80
Rectangular patch
70

60
CAD
50 exact
Rin ( )

40
Frequency where
30 the input resistance
20
is maximum (f0):
Rin = R
10

0
4 4.5 5 5.5 6
FREQUENCY (GHz)

r = 2.2 W/L = 1.5 L = 3.0 cm 89

 Input Impedance
Results: Input Reactance vs. Frequency
Note: “exact” means the cavity model will all infinite modes.
Frequency80 where the input
resistance is maximum (f0) Rectangular patch
60
CAD
exact
Xin ( ) 40

20
Xp
0

-20
Shift due to probe reactance

-40
4 4.5 5 5.5 6
FREQUENCY (GHz)
Frequency where the
input impedance is real

r = 2.2 W/L = 1.5 L = 3.0 cm

90
 Input Impedance

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

Vary the length L first until you 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.

91
 Design Example
Design a probe-fed rectangular patch antenna on a substrate having a relative
permittivity of 2.33 and a thickness of 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.

92
 Design Example
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
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 until you find the value
that makes the input impedance exactly 50+j(0)  at 1.575 GHz.
93
 Design Example
solution
Part 1
Results from the CAD formulas: y
Feed at (x0, y0)
1) L = 6.071 cm, W = 9.106 cm
2) x0 = 1.832 cm y0 = W/2
3) BW = 1.23% W
4) er = 82.9%
5) Xp = 11.1  x
L
6) Zin = 50.0 + j(11.1) 
7) D = 5.85 (7.67 dB)
8) G = (D)(er) = 4.85 (6.86 dB)
94
 Design Example
solution
Part 2 Z in  jX p 
R
 f f 
1  jQ   0 
 f0 f 
Results from the CAD formulas:
f0 = 1.575 109 Hz 50.255
60

R = 50  50
Rin
Q = 57.5 40

Xp = 11.1   30

Rin( fghz)
20
Xin( fghz)

10
Xin

- 10

- 13.937
- 20
1.5 1.525 1.55 1.575 1.6 1.625 1.65
1.5
f (GHz)
fghz 1.65
95
 Design Example
Part 3 60
60

After optimization: 50

40
L = 6.083 cm 
x0 = 1.800 cm 30

Rin ( fghz)

Z in  50  j  0  
20
Xin ( fghz)

10

y 0

Feed at (x0, y0) - 10

- 20
- 20
1.5 1.525 1.55 1.575 1.6 1.625 1.65
y0 = W/2 1.5 fghz 1.65
f (GHz)
W

x
L
96
Square MSA in Air –Radiation
Pattern

 RadiationPattern
at 1.8 GHz
F/B = 15 dB
Cross Polar

97
 CMSA: Resonance Frequency

T ( f )
For Fundamental Mode:
 100c ( f )  d (f)  r (f )

 where is the mth root of the where a and h are in cm and

derivative of the Bessel Design Equation:
function of order n

Choose feed-point x between 0.3a to 0.5a 98