Jurnal Cantenna
Jurnal Cantenna
Jurnal Cantenna
Page 13.270.1
Abstract:
This paper describes the design and development of a large 20 dBi (decibels isotropic)
Wi-Fi antenna for a class project in the Communication Circuit Design course. This large
antenna is based on smaller Wi-Fi antennas commonly referred to as cantennas (gain of
about 10 dBi). The smaller version is made with a single can (3-4 inches) in diameter and
an appropriately placed feed probe. Our version consists of several progressively larger
cylindrical sections connected together by 34 degree flared sections (3 inches long). The
first cylindrical section has a diameter of 4 inches and the last flared section has a
diameter of about 36 inches. The overall length of the antenna is about 5 feet long. While
the antenna was designed by electronic and computer engineering technology (ECET)
students, mechanical engineering technology (MET) students took charge of building it
under the supervision of their MET instructor. The structure was built by spot welding
laser-cut pieces of sheet metal. ECET students made experimental measurements to
verify the predicted gain and functionality.
Motivation:
The design of the large Cantenna (Figure 1) was taken up by students as a class
project for the Communication Circuit Design (ECET 420) course. Chief among
the motivational factors influencing their choice seemed to be a desire to learn
more about cantennas that could potentially help them share internet access with
their friends that lived reasonably close. In the early part of the research cycle we
came upon the world record1 for "unamplified" Wi-Fi distance (125 miles). This
feat was accomplished by linking two dish antennas (10 feet and 11 feet) in
diameter. Each antenna was attached to a Zcom PCMCIA card with a built in
power of 300mW. This record influenced the design team to start thinking about
building a high-gain Wi-Fi antenna that would be able to communicate over
several miles.
Our design did not attempt to match the records for distance. We were interested
in building the largest known antenna based on the standard cantenna2 design,
using some of the same methodologies.
From the point of view of a communication design course, the construction of any
antenna is a good way to learn about impedance matching, guided wave propagation and
the radiation characteristics of antennas. Furthermore, the building of a high gain antenna
with real world application and record dimensions appealed to all involved. The paper
will also comment on student enthusiasm and teamwork in this interdisciplinary
Page 13.270.2
endeavor. The sections to follow cover all aspects of antenna development, including
design, simulation, construction, measurement and testing. A discussion of the learning
outcomes is presented just before the concluding remarks.
Design:
We employed design principles to those of the standard cantenna2 combined with the
“Theory of Small Reflections3.” The standard cantenna is essentially a cylindrical
waveguide that permits only the dominant TE11 mode4 (or electromagnetic field
distribution) to propagate at the 2.4 GHz Wi-Fi frequency. The can should have a length
that exceeds ¾ of the guide wavelength for the TE11 mode at 2.4 GHz. The feed is placed
at ¼ of a guide wavelength from the back wall of the can. This assures that the reflected
wave from the back wall and the wave traveling toward the opening are in phase and do
not destructively interfere with one another.
In designing the first section of the large antenna, we followed the design guidelines of
the standard cantenna2. With the exception of the first section (smallest of sections), all
other sections carry multiple modes and the number of modes in any section is
proportional to the diameter. The large number of modes in the last section allows a field
distribution that permits highly directive radiation patterns. Our version consists of six
different cylindrical sections connected together by six 3-inch 34˚ flared or conical
Page 13.270.3
sections (Figure 2). Flared sections were used to minimize reflections at the junctions.
The “Theory of Small Reflections3” suggests that a gradual change in waveguide
dimensions will keep the reflections small at their junctions. When we make the
transition between two cylinders the intermediary conical section keeps the reflections at
a minimum. Indeed the conical section itself can be seen to be composed of many
progressively larger cylindrical sections having infinitesimal lengths with no reflections
at their junctions. The first cylindrical section has a diameter of 4 inches and the last
flared section has a diameter of about 36 inches. The overall length of the antenna is
about 5 feet long.
Simulation:
A simulation of the above design was performed using MicroStripes5, a 3-D EM solver.
The results of simulation seemed to indicate a gain of almost 21 dBi (decibels isotropic).
The antenna is highly directive and has a 3-dB beamwidth of about 14 degrees (Figure 3).
There are no noticeable side lobes or back radiation to speak of. Figure 4 indicates the
antenna has a very large bandwidth (630 MHz).
Construction:
Figure 4. The graph above shows the simulation results of the return loss of our antenna
in the 1.5 to 2.0 GHz range.
Page 13.270.5
allows it to rotate around the axis to minimize polarization mismatch6. The dimensional
specifications seemed to be quite closely followed.
Measurements:
Bandwidth: After constructing the antenna, the bandwidth was measured by using a
spectrum analyzer in combination with a directional coupler. The tracking generator of
the spectrum analyzer inputs power into the directional coupler. The input power
propagates to the antenna hook up and is reflected. The directional coupler separates the
returned signal from the source and passes it to the input of the spectrum analyzer. Lower
levels of return loss, means that more energy is being coupled to the antenna. When the
return loss is lower than -10 dB we are considered to be in the operational part of the
bandwidth. The onset of the pass band at around 1800 MHz was verified by the spectrum
analyzer. The upper bound cannot be accurately established due to the operational
limitations of the directional coupler, which is around 2 GHz. Measurements of gain and
passively picked signals (shown later) corroborate the simulation result of 600+ MHz
bandwidth.
Gain: To determine the radiation pattern of our antenna, one would normally require a
fairly large range; alternatively one could use a smaller range with a scaled model. Our
approach was to use a software package called Network Stumbler7 running on a computer
which receives its Wi-Fi signal from an external antenna and a 2511 PLUS EXT2
Wireless Card. These cards are available from Jefatech8. Network Stumbler is capable of
picking out the signal and noise levels of different Wi-Fi sources. The following
procedure was followed in figuring out the gain of a large antenna.
1. A standard cantenna (Figure 6) of known gain (10 dBi) was placed in a location
directly in front of an access point. The signal and noise levels from this access
point were recorded.
2. The test antenna was placed at the same position in front of the same access point.
3. Take readings of power and noise levels of this particular antenna.
4. The difference of the readings of the signal level in step 3 and step 1 plus the gain
of the known antenna was calculated to be the gain of the device being tested (the
large cantenna).
Using the procedure outlined above, the gain was calculated to be about 21 dBi. It should
be noted that the gain was only measured at the maximum direction and this result
validates the maximum simulated pattern gain. The maximum gain conveys a great deal
of information for highly directive antennas with narrow beamwidths.
Testing:
Spectrum Analyzer Test: When the large antenna was hooked up directly to a spectrum
analyzer it was able to pick up the distinct Wi-Fi signals in its environment without any
amplification. Figure 7 shows the recorded maximum signal strengths centering around
Page 13.270.6
2.41 GHz. Notice that channel 1 signal (2.412 GHz) possesses the most strength. We had
previously identified the channel 1 and 6 to be the two strongest channels in that vicinity
using an omni-directional antenna, a 2511 PLUS EXT2 wireless card and the Net
Stumbler software for that room. The fact that channel 6 did not show up is not surprising
considering the directional nature of the antenna.
Outdoor Signal Test: We rolled the large Cantenna to a distance of about 200 yards from
our building (Technology Center) and hooked it up through 2511 PLUS EXT2 wireless
card to a laptop which was running Net Stumbler. We then turned the antenna to face the
Tech Center. Using our antenna we were able to pick up 8 access points with 4 of them
having a signal to noise ratio (S/N) greater than or equal to 20 dB (a threshold required
for proper access). We then replaced the large cantenna with a smaller one and we were
then able to pick up 7 access points with 2 of them having S/N ratio that is greater than
the threshold. Figure 8 shows location of our measurements with a white star.
The course outcomes for the Communication Circuits Design course are listed in column
1 of table 1. Column 2 shows which of these areas were directly impacted by the project;
column 3 shows the aspect of the project that relates to the course objective and column 4
shows TAC of ABET outcomes that have been touched upon by the activity in column 3.
Apart from addressing outcomes shown in table 1 the students have successfully
completed an interdisciplinary team project (TAC of ABET criteria e). While a complete
assessment of all direct and indirect data for the course have not yet been completed,
teaching evaluation data seem to indicate a significant improvement in the desire to learn
from the previous year.
Page 13.270.8
Page 13.270.9
Figure 8. Measurement location is marked with a red circle in the above map.
TABLE 1
Impact of project on ECET 420 course outcomes
Course Outcome Impacted by Related Activity TAC of ABET General
Project Criteria
Yes/No a thru i
Understanding the no
different components
of an RF
communication
System
Design and build RF no
amplifiers
Design and build an yes The design of the tuning d. an ability to apply
impedance matching piston arrangement creativity in the design
network of systems,
components or
processes appropriate
to program objectives
Design and build a yes The design and building of d. an ability to apply
communication the Large Cantenna creativity in the design
antenna of systems,
components or
processes appropriate
to program objectives
Understand the role of yes 1. Simulation using b. an ability to apply
software in modern Microstripes current knowledge and
wireless design 2. Use of Net adapt to emerging
Stumbler to study applications of
access point signals. mathematics, science,
engineering and
technology.
We started the project with the intention of building a standard cantenna for Wi-Fi
purposes and ended up building a high-gain cantenna with record dimensions. So what
brought about this level of overachievement? Student enthusiasm regarding cantennas
was high, for one, but the interdisciplinary nature of the project was also a big factor. One
cannot but stop and admire the synergistic relationship that had formed between ECET
and MET students. Even non-participating students from the two disciplines were excited
and eager to learn about the project during the development process. From the
instructor’s point of view, it was the question of finding common ground between course
objectives and student interest during a time of the dreaded “senioritis.”
References:
1. http://www.its.state.ut.us/productsservices/radioshop/newsletter/Wireless_08-03-05.pdf
2. http://www.turnpoint.net/wireless/cantennahowto.html
3. David M. Pozar, “Microwave Engineering,” Addison-Wesley, page 305-307, 1990
4. R. F. Harrington, “Time-Harmonic Electromagnetic Fields,” McGraw-Hill, page 204-207,
1961
5. http://www.flomerics.com/microstripes/
6. Constantine Balanis, “Antenna Theory: Analysis and Design,” John Wiley, 1997
7. http://www.stumbler.net/
8. http://www.jefatech.com/
Page 13.270.11