You are on page 1of 10

Sheet 1 of 10

Micro-strip Patch Antenna Primer


J P Silver E-mail: john@rfic.co.uk

1 ABSTRACT
This paper discusses the theory and design of microstrip patch antennas. The pertinent design parameters are given together with their relevant equations to allow basic hand calculations before simulation is attempted. Finally a worked example is given to highlight the design steps required and CAD simulations are also described.

position of feed Dual frequency use possible Solid-state devices easily integrated Feed lines and matching fabricated with antenna

Low power handling

Table 1 Advantages and disadvantages of patch antennas.

2 INTRODUCTION
Proposed by Deschamps in 1953, and the first patch antenna was fabricated in the early 1970s. A microstrip patch consists of a radiating patch of any planar geometry (eg Circle, square, Ellipse, ring and rectangle) on one side of a dielectric material substrate backed by a ground plane on the other side. A typical patch antenna is shown in Figure 1.
Radiating Patch Dielectric Substrate

3 FEED ARRANGEMENTS
Figure 2 shows various ways to feed patch antennas, capacitive-coupling of the various design is popular.

4 RADIATION MECHANISM
The important characteristic of microstrip antennas is their inherent ability to radiate efficiently despite their low profile. The primary source of this radiation is the electric fringing fields between the edges of the conductor element and the ground-plane behind it. Lewis[1] first analysed this in the form of a transmission line discontinuity and he discovered that the Q (quality factor) of the dielectric cavity formed by two short circuit walls and four open circuit walls depended on several parameters. The parameters are dielectric constant (r), height (h) of the substrate, the patch dimensions and the frequency. The results showed that at high frequency, radiation loss as opposed to conductive or dielectric loss is the main source of energy dissipation. The electric field patterns of a square patch antenna are shown in Figure 3.

Conducting Ground Plane

Figure 1 Basic features of a patch antenna

Generally, a substrate with a low dielectric constant (r) is used (typically ~ 2.5) to reduce the fringing fields, but in loss critical applications, Alumina (r =10) must be used. Table 1 highlights the advantages and disadvantages of patch antennas the main disadvantage being the narrow bandwidth. Advantages Lightweight, low volume Easy to mount Low fabrication cost Aerodynamic (good for fitting on missiles and ships Linear and circular polarization easy to implement by Disadvantages Narrow bandwidth Loss Poor end-fire radiation characteristics Very limited maximum gain Poor isolation between feed and antenna

Sheet 2 of 10

~/2

Gap Coupled Feed

Non-radiating edge feed

r
h

/4

Quarter-wave Match Feed


Patch
Patch

Inset Feed

Dielectric Substrate (r) Radiating Patch

Feedline

Feed

Coaxial Connector

Probe Coupling Feed


Substrates

Aperture

~/2

h Conducting Ground Plane

Radiating Patch

Ground Plane

Aperture Coupling

Feedline

Figure 3 Rectangular microstrip patch antenna electric field patterns.

Substrates

The E field is considered to be constant along W and varies sinusoidally with L.


EM Coupling

Figure 2 Patch antenna feed arrangements

The fringing fields at the two edges add in phase in the far field on boresight (the vector normal to the plane of patch) and cancel along the broadside. The radiation patterns for the TM mode are shown in Figure 4. For a/b = 1.5, r = 2.32 and 9.8. These patterns are of particular interest since they provide maximum radiation on bore sight.

Sheet 3 of 10

E
Er = 9.8 Er = 2.32

Q=

Pr + Pd + Pc + Psw

1 2.f .W . r t h

Where f = resonant frequency. r 1 W . = stored energy in the cavity. t h Pr = energy lost through radiation. Pd = Dielectric Loss. Pc = Conductor Loss. and Psw = surface wave loss (very small)

-10

-10

dB

E
Er = 9.8 Er = 2.32

The antenna efficiency is given by:

(% ) =
0
-10 -10 dB

Pr x 100% Pd + Pc + Pr

Typically can be between 80 to 90% for a patch antenna with a dielectric constant of 2.3 at 10GHz. Losses reduce the Q factor, which results in an increased bandwidth and reduced efficiency. Typical Q factors are in the range 20-200 and the bandwidth is given by:
B.W = 100(s - 1) Q s

Figure 4 Radiation Patterns

Where s = VSWR

Looking at the radiation patterns it is clear that the efficiency of the patch antenna (ie the directivity) is poor. When designing these antennas, emphasis is placed on trying to maximize efficiency at the expense of gain or radiation pattern.

6 MODELLING
Two models commonly used for analyzing patch antennas are the Transmission Line model and the Cavity model. (1) Transmission Line Model The structure is considered as a transmission line along the line joining the radiating apertures loaded by impedances at its radiating edges. The radiating wall may be characterised by an equivalent admittance, the susceptance being due to the fringing field and the conductance due to the radiation loss. The transmission line model is shown in Figure 5.

5 LOSSES
The antenna is a resonant circuit therefore, energy will be stored in the system. This energy stored is inversely proportional to the dielectric height h. We can also describe the energy stored by the parameter Q (known as Quality Factor). Losses in the antenna will allow energy to leak away and such an antenna will have a lower Q factor. The Q factor can be calculated from the following equation:

Sheet 4 of 10

/2

2.Fr. . o o Where
G

W=

2 +1 r

jB

jB

Figure 5 Transmission line model of a patch antenna

(2) Cavity Model The region between the patch and the ground plane may be treated as a cavity, bounded by magnetic walls on the sides and electric walls on the top and bottom. The modes in this cavity are given by:
2 Ez 2 Ez + + . . 2 .Ez = 0 (Helmholtz equation) x 2 y 2 Applying boundary conditions, Ez = 0 at boundary walls Hx = - j Ez j Ez . . and Hy = . y . x

Fr = Resonant Frequency = Relative dielectric constant r = 8.9 x10 12 C 2 /Nm 2 o = 4.x107 Tm/A o W is usually chosen to be in the range 1 <= L <= 2. Referring to Figure 6, the length of the resonant element is then obtained from:

L=

c 2.Fr. . . eff o o

2l

For a dielectric substrate of thickness h, an antenna operating frequency of Fr, and for an effective radiator, a practical width W is:

l = 0.412 h

( (

eff

eff

W + 0.3) + 0.264 h W 0.258) + 0.8 h

x y Ez = Eo. cos m. . cos m. . 0 x a and 0 y b a b where n and m are postive integers n 2 m2 = . 2 + 2 . b a

~ /2

L l l

Figure 6 Length of resonant element

7 DESIGN PROCEDURE
(1) Element Width The first design step is to choose a suitable dielectric substrate of appropriate thickness. The three most common substrate materials used are rexolite (r = 2.6), RT Duroid (r = 2.32) and Alumina (r = 9.8). Element width is given by the following equation:

(2) Radiation Pattern For two slots spaced by a distance L apart, the E and H plane radiation patterns may be plotted using the equations:
Ko.W sin cos 2 sin F( ) = Ko.h cos 2 2

(H - Plane) Ko =

Ko.h sin cos 2 cos Ko.h cos (E - Plane) F( ) = Ko.h 2 cos 2

(3) Slot Susceptance

Sheet 5 of 10

The slot susceptance can be represented by a capacitance C:


C= l eff Where Zo is the characteristic impedance C.Zo

For W o Rr = 90. o 2 W2 o W 2

(4) Input Admittance The input impedance or admittance should be accurately known so that a good match between the element and the feed can be designed. Referring to Figure 7. The distance of the feed entry point to the edge of the patch (z) is given by:
G 2 + B2 B Yin(z) = 2G cos 2 ( .z ) + sin 2 (z ) sin (2 .z ) 2 Yo Yo Where 1 Yo = Zo 1 = 2 eff ; B = k.l eff ; G = Rr o Zo 1

W o Rr = 120.

W o Gr =

1 W . 90 o 1 W . 120 o

W o Gr =

(6) Probe position[1] The equations given for calculating the feed position are very complicated but can be simplified to give a starting point for computer simulation.

R IN =

1 Where Gr is that from section 5 2(Gr G12 )

Patch

Feedline

z
Figure 7 Feed position on patch antenna

(5) Radiation Resistance/Conductance

Sheet 6 of 10

However G12 is complicated and is given by:


KoW sin 2 cos Jo(Ko.L sin )sin 3 d cos

Rc = 0.00027 Fr .

L .Qr 2 (Fr in GHz) W

1 G12 = 120 2

Rd =

o 30.tan .Qr 2 .h. L.W Er

Where Jo is the Bessel function of the first kind of order zero.

and the total quality factor Qt is Qt = Qr.Rt Rr Where Rt = Rr + Rd + Rc

Luckily G12 << Gr and so in most cases can be ignored. We find that Rin is given by:

Rin( y = yo) = as Gr >> G12 Rin( y = yo) =


2

1 cos2 . yo 2(Gr G12 ) L

1 cos2 . yo 2.Gr L

(8) Antenna Efficiency The efficiency (which increases as h increases) may be expressed in terms of the equivalent resistances.

cos ( x) = 1 sin 2( x) 1 Rin( y = yo) = 1 sin 2 L . yo 2.Gr Where yo = position of feed from patch edge L= Length of patch Gr = RadiatingConductanc e Note Calculatein radians (7) Q Factor and Losses The quality factor (Qr) associated with the radiation resistance is:

% =

Rr * 100 Rt

(9) Bandwidth The bandwidth may be increased by increasing the inductance of the radiators by cutting holes or slots in them or by adding reactive components to improve the match of the radiator to the feedline. (10) Directivity and Gain The directivity (D) of the antenna is defined as the ratio of the maximum power density in the main beam to the average radiated power density. The gain G is given by:

G= .D
The gain increases with resonant frequency and h and decreases with increase in r. The directivity is independent of the resonant frequency. (11) Beamwidth The half power beamwidth is equal to the angular width between directions where the gain decreases by 3dB, or the radiated field reduces to of the maximum value. The half power beamwidths in the H and E planes are given by:

Qr =

(C.

(4.Fr.h )

eff

The equivalent resistances for the conductor loss Rc and dielectric loss Rd are then:

Sheet 7 of 10

BH = 2. cos 1

1 1 + k o .W 2 2

7.03 BE = cos 1 3.k 2 L2 + k 2 h 2 o o

Thus beamwidth can be increased by choosing a smaller element, thus reducing W and L. For a given resonant frequency these dimensions may be changed, by selecting a substrate having a higher relative permittivity. As beamwidth increases, element gain and directivity decrease, however, efficiency is unaffected.

Sheet 8 of 10

8 EXAMPLE
For this example we shall design a patch antenna for reception of Meteosat WEEFAX at 1691MHz Channel 1 and 1694.5MHz for channel 2 . The patch will be used as a feed for a TV offset feed dish and therefore, we require a patch that will have a wide beamwidth to provide optimum illumination of the dish. The frequencies of 1691MHz and 1694.5MHz which equates to a very narrow bandwidth of 0.002%, and therefore, we can optimize the design to produce as much gain as possible. Before we analyse the design using Agilent Momentum (part of ADS) we need to perform some hand calculations.
Design considerations:

(2) Patch Length (L) c 3x108 L= = = 6.43cm 2.Fr. eff 2.1.691x109. 1.898
However L will be slightly shortened due to the end effect:

l = 0.412 h

( (

eff

eff

W + 0.3) + 0.264 h W 0.258) + 0.8 h

l = 0.412 h

(1.858 + 0.3)

0.06 + 0.264 3 3.175x10 0.06 (1.858 0.258) + 0.8 3 3.175x10

l = 0.412(1.312) l = 0.412(1.312).3.175x10 3 h l = 1.72mm Therefore, modified L= 64.3 - 2(1.72) = 6.086cm

We want a cheap design so we could design a patch using an air dielectric or use inexpensive PTFE RT Duroid. To achieve high gain we need to use a low dielectric constant (r (air) =1 or r=2.3 (RT Duroid 5880) and large h (RT Duroid 5880 is available with a maximum thickness of 3.185mm). The completed antenna will feed an LNA designed to work with a 50-ohm input impedance, so the feed design could be a quarter-wave transformer, insert feed or a probe coupling.
For this design RT Duroid 5880 was used and we select the channel 1 frequency of 1691MHz. Eeff ~ 1.898 (1) Patch Width (W)

(3) Approximate position of probe/feedpoint (yo) The very approximate relationship between input resistance (Rin) and probe position (yo) is given by the following:

W=

1 2.Fr. . o o

2 +1 r 1 2 2.2 + 1

W=

2.1.691x10 . 8.9x10 .4.x10

12

W = 6.98cm Where Fr = Resonant Frequency = Relative dielectric constant r = 8.9x10 12 C 2 /Nm 2 o = 4 .x10 7 Tm/A o

Sheet 9 of 10

For W o G1 = 1 W 2 . 90 o 1 W . 120 o 1 W . 90 o 2 2 =

Substrate layer Name: RT5880 Thickness: 3.175mm, Permittivity: 2.2, Loss Tangent: 0.001, You should see FREESPACE RT5880 //////GND//////// Under Metallisation layers click on the dotted layer under freespace and select strip. You should see FreeSpace ------Strip Upper Conductor RT5880 //////GND//////// Finally, precompute (3) Port definition Open the Port Properties Editor and select the port on the layout. Select 50-ohms, Internal port and select OK. (4) Mesh Select setup mesh and enter the highest simulation frequency used eg 1.7GHz. Then pre-compute mesh which, should appear as shown in Figure 8.

W o G1 =

As W o G1 =

1 6.98E -2 G1 = . 90 3E 8 1.691E 9 Rin

= 1.719E -3

1 cos 2 .yo G1 L

1 1 1 sin .2 .yo Note calculate in radians G1 2 L Assume yo = 21mm =


1 .21 1 sin .2 2.1.719E 3 60.8 Rin 50

These values were then simulated and optimized (for input match) using Momentum within Agilent ADS.

9 MOMENTUM ANALYSIS
Now we have a basic design we can check it using an EM simulator and for this we will use ADS Momentum. The following steps are required to enter the layout of the patch antenna, the substrate details and the frequency of operation. (1) Layout Using the layout window, draw the patch antenna on the upper conductor layer. Add the lower conductor (ie the boundary of the patch PCB) and add a port to the point where the feed will be (on the upper conductor layer). (2) Substrate details Open the create/modify substrate dialogue box and under substrate layers enter the following data: For the layer between Freespace and ground enter

Figure 8 Computed mesh and port. (Patch size L = 5.83cm, W = 5.89cm optimized)

Sheet 10 of 10

(5) Frequencies Select simulation S-parameters and add start and stop frequencies 1.65 and 1.7GHz. Select update, apply and finally simulate. (6) Results The position of the test port was moved from the center of the patch antenna to the edge, until a good input return loss was obtained as shown in Figure 9. The center frequency, was optimized for 1.691MHz, by adjusting L and W to be 5.83cm and 5.89cm respectively.

Figure 9 Input return loss prediction

Momentum can also produce the radiating fields in 2D and 3D views. The 2D view of gain and efficiency is shown in Figure 10.

Figure 10 Predicted gain (dB) and efficiency for the patch antenna.

[1] Antenna Theory, C.Balanis, Wiley, 2nd edition (1997), Chapter 14. ISBN 0-471-59268-4.

10 REFERENCES

You might also like