AU J.T. 11(4): 209-214 (Apr.

2008)

Experiment on Triangular Copper-Strip as Impedance Matching for

2.4-GHz Helical Antenna
Virach Wongpaibool
Faculty of Engineering, Assumption University
Bangkok, Thailand
Email: <virachwng@au.edu>

Abstract
Performance of a helical antenna designed for 2.4-GHz WLAN is experimentally
investigated in terms of impedance matching. A section of the helical wire near the
antenna feed point is replaced by a simple triangular copper strip. By deploying such
technique, impedance matching can be achieved. Overall return loss is below −14 dB
across entire WLAN frequency range. Moreover, overall return loss does not strongly
depend on strip length around its optimum, allowing 15 mm of strip-length variation
around the optimum of 50 mm. Adding the triangular impedance-matching strip causes
the characteristic of the helical antenna to be similar to a parallel resonant RLC circuit,
resulting in frequency dependence of return loss for a given length of triangular
impedance-matching strip.
Keywords: Helix, transformer, WLAN antenna, resonant RLC circuit, portable
computer, axial-mode helical antenna.

possible to extend the connection range

Introduction between a WLAN-equipped computer and an
access point by utilizing high-directivity (high
In the past, computers were connected gain) antennas either at one end or both. There
together by using wireline techniques, such as are many types of antenna, which exhibits high
local area network (LAN). This seemed to be gains, and are suitable for 2.4-GHz WLAN.
inconvenient in some situations since The one that is of our interest is an axial-mode
computers must be within the range of cables. helical antenna (gain is highest along the axis
As the network technology progresses, it is of the antenna) (Kraus 2003; Balanis 1997;
possible for computers to connect together ARRL2007; Bart 2003; and Weeratumanoon
wirelessly. For example, at the present time, we 2000). The picture of a helical antenna is as
can use our computers to connect to the shown in Fig. 1.
internet by using wireless LAN (WLAN),
operating at 2.4-GHz frequency range (WLAN
2008). Nowadays the use of the Internet is
inevitable in our everyday lives. With portable
computers, and WLAN, the internet can be
accessed anywhere provided that it is within
the coverage area of a WLAN access point.
The range of 802.11g WLAN is typically at the
order of 100 m (WLAN 2008).
One of the most important parts of
WLAN is an antenna. The antenna functions as
a transducer, which converts electrical signal to
electromagnetic wave, and vice versa.
Generally, antennas, deployed in WLAN, are
low-gain omni-directional. However, it is Fig. 1. Helical antenna.

Regular Paper 209

AU J.T. 11(4): 209-214 (Apr. 2008)
Practically, in order to utilize an antenna
efficiently, the input impedance of the antenna
itself must be matched to the characteristic
impedance of the transmission line, used to
connect the antenna to a transmitter or a
receiver. Unfortunately, this is not the case for
a helical antenna. Emphasis in this paper is the
experimental impedance matching of a helical
antenna by using a simple triangular copper
strip, which was demonstrated by Hecker
(2007) and Besten (2007). The main objective
is to experimentally investigate the length
tolerance of a triangular copper strip, deployed
for impedance matching, under the condition
that the return loss of the helical antenna is
approximately below −14 dB, which
corresponds to standing-wave ratio (SWR) of
1.5. In addition, the characteristic of the helical
antenna, utilizing the triangular copper strip as
impedance matching, is studied by means of an
equivalent parallel RLC circuit.
Helical Antenna
In general, a helical antenna can be
physically described by its circumference C ,
spacing between turns S (or pitch angle α ),
and the total number of turns N . The first two
parameters, C and S , are usually normalized
by the carrier wavelength λ , and they become
Cλ , and Sλ , respectively. Since there are 13
channels, whose carrier frequencies range from
2.412 GHz to 2.472 GHz (European Standard)
(WLAN 2008), the carrier frequency of 2.442
GHz is selected as a nominal carrier frequency
since this frequency is at the middle of
operating band for 2.4-GHz WLAN. This
corresponds to λc = 12.29 cm. For a helical
antenna to be in axial mode with optimum
performance in terms of gain, bandwidth, and
an axial ratio, the following conditions must be
met (Balanis 1997; Kraus 2003; and
Weeratumanoon 2000):
0.8 ≤ Cλ ≤ 1.2 , (1)
12o ≤ α ≤ 14o , (2)
N ≥ 4. (3)
Theoretically, the core of a helical
antenna is an air. However, an air-core helical
antenna tends not to be durable in practice
Regular Paper
especially when the helical wire size is small.
For the helical antenna used in the experiment,
the wire, having the diameter of 1.0 mm, is
supported on a polyvinyl chloride (PVC) tube
to strengthen the antenna structure. The
selected PVC tube has the outer diameter of 42
mm, which corresponds to Cλ of 1.074. The
pitch angle is selected to be 12.41° to make
S = 2.90 cm for ease of construction. The total
number of turns for this antenna is 16, and the
reflector has the diameter of 18 cm. The helical
wire is peripherally fed at the base by using
Type-N female connector. The constructed
helical antenna is shown in Fig. 1.
Due to the physical geometry of the
helical antenna, it is extremely difficult to
mathematically analyze the radiation
properties, such as gain, and input impedance.
Therefore, the investigation of radiation
properties is generally accomplished by
experiments, analytical approximations, and
numerical analyses. For a peripherally-fed
helical antenna, the empirical formula for the
input impedance is (Kraus 2003; and Balanis
1997):
150
R = (Ω). (4)
Cλ
For the helical antenna used in the
experiment, Eq. (4) yields R ≈ 140 Ω . The
standard coaxial cables deployed for WLAN
installation have nominal characteristic
impedance of 50 Ω, thus, there is the mismatch
between the antenna input impedance, and the
coaxial-cable impedance. Several techniques
have been proposed to lessen that impedance
mismatch. Tapering the last several helical
turns has been shown to help reduce the
impedance mismatch over a wide frequency
band (Wong and King 1979). Gradually
tapering helical windings at both ends, and
deploying a cone reflector could alter the input
impedance as desired (Angelakos and Kajfez
1967). Alternative approach to solve the
impedance mismatch was to deploy a short
length of tapered transmission line between a
helical-antenna feed point and the main coaxial
cable (Bart 2003; Manthur and Sinha 1988; and
Tsandoulas 1967). Similarly, gradually flatten-
ing the section of a helical wire near the feed
point also resulted in a smooth transition from
210
AU J.T. 11(4): 209-214 (Apr. 2008)
140-Ω helix input impedance down to 50-Ω
standard-coaxial-cable impedance at the feed
point (Kraus 1997, 2003; and Balanis 1997).
It should be noted that these mentioned
techniques are more difficult than the approach
demonstrated by Hecker (2007) and Besten
(2007). In stead of flattening the helical wire
near the feed point, that section of helical wire
was replaced by a thin triangular copper strip,
having the same length as that of the helical
wire being substituted as shown in Fig. 2.
However, no details were provided for the
effect of change in strip length on the
impedance mismatch. This is the main
objective of this paper, which is discussed in
the next sections.
Fig. 2. Impedance matching by replacing
section of helical wire near feed point by thin
triangular copper strip.
Experimental Results
The experiment was conducted by using
the helical antenna, constructed with geometry
described in the previous section. The height of
the 0.1-mm-thick triangular strip was
arbitrarily fixed to 15 mm, and the length of the
strip was varied from 30 mm to 70 mm with
the increment of 5 mm. In order to quantify the
impedance mismatch, a measurable parameter
related to impedance mismatch is required.
That parameter is the return loss (dB) defined
as
⎛P ⎞
Return Loss (dB) = 10 log ⎜ r ⎟ , (5)
⎜ Pf ⎟
⎝ ⎠
where Pr and Pf are reflected power and
forward or incident power at the point where
Regular Paper
return loss is measured, respectively. For this
experiment, it is at the antenna feed point. Note
that return loss is expressed as negative
quantity to conform with the results obtained
from the Hewlett-Packard network analyzer,
but most texts express it as positive. The
frequency range for return-loss measurement is
from 2.36 GHz to 2.5 GHz, which covers the
whole frequency range of WLAN.
Return loss as a function of frequency is
plotted in Fig. 3 at different strip lengths. In the
ideal case at which the input impedance of an
antenna matches with the impedance of the
transmission line, incident power is totally
absorbed by the antenna, thus, no reflected
power, and return loss being equal to −∞ dB.
However, it can be seen from Fig. 3 that
without any kinds of impedance matching,
return loss is very poor, and relatively flat
across the measured frequency band. At the
frequency around 2.48 GHz, return loss is as
poor as −7 dB. This corresponds to only 80 %
of incident power being absorbed by the
antenna.
Fig. 3. Return loss as a function of frequency
at different strip lengths.
Replacing a short section (30 mm) of
helical wire at the feed point by a triangular
copper strip having the same length does not
yield significant improvement in terms of
return loss. This implies that short length of
triangular copper strip as impedance matching
is not effective. However, as the length of the
copper strip is increased, return loss is
improved (lower).
211
AU J.T. 11(4): 209-214 (Apr. 2008)

It can be seen from Fig. 3 that at the

length of around 50 mm, return loss is
approximately below −15 dB across the whole
WLAN frequency band (2.410 GHz to 2.475
GHz). This value of return loss corresponds to
97% of incident power being absorbed by the
antenna. If the antenna ohmic loss is negligible,
97% of total electrical power fed to the antenna
would be converted to the power of radiated
electromagnetic wave. Only a small fraction of
power incident to the antenna is reflected back.
Moreover, at the frequency around 2.448 GHz,
return loss is as low as −36 dB, which
corresponds to 99.97% of power absorption. Fig. 4. Return losses at different WLAN
That is, the input impedance of the helical channels as a function of strip length.
antenna is extremely close to 50 Ω at that
frequency. mm. It is clearly seen that all the curves look
However, when the copper strip is too nearly identical. This implies that moderate
long (70 mm for example), return loss becomes variation of strip length around the optimum
poorer, compared with that of 50-mm copper does not strongly affect the return loss in terms
strip. This suggests that there exists the of the worst-channel return loss. That is, the
optimum length of triangular copper strip, length of triangular copper strip, used as
which performs impedance transformation. At impedance transformer, is not critical as long
the short length of triangular copper strip, the as it does not overly differ from the optimum.
impedance transformer is not effective. On the
other hand, the antenna input impedance is
overly transformed when the triangular copper
strip is excessively long.
In order to thoroughly investigate the
length tolerance of impedance matching
obtained from this technique, return loss as a
function of strip length is plotted in Fig. 4 at
three different WLAN channels (2.412, 2,442,
and 2.472 GHz). For the strip length ranging
from 45 to 60 mm, return losses for all three
channels are in the region below −14 dB. In
terms of the worst-channel return loss, the
optimum performance is found to be at the strip
length of 50 mm. At 50-mm strip length, return Fig. 5. Return loss as a function of WLAN
losses for all 3 channels considered are below frequency range at different strip lengths
−15 dB. around the optimum.

Discussions When all the plots in Fig. 5 are carefully

With the purpose of gaining more
understanding on how the strip length affects
impedance matching, return loss as a function
of frequency is plotted in Fig. 5 at different
strip lengths around the optimum length of 50
examined, it can be seen that for a given length
of triangular strip, the return loss is not
relatively flat across the entire frequency range
considered, and has a deep dip at a certain
frequency. This implies that return loss is
frequency-selective for a given triangular-strip
length. This is similar to the behavior of a
resonant circuit, for which its reactant becomes

Regular Paper 212

AU J.T. 11(4): 209-214 (Apr. 2008)
zero at the resonance frequency. That is, the
frequency, at which lowest return loss occurs,
is the resonance frequency of the antenna. At
that frequency, the reactance of the antenna
input impedance becomes zero, and the input
impedance is purely resistive.
This can be observed from the smith
chart plotted in Fig. 6 for the case of strip
length of 60 mm. At the frequency of 2.438
GHz, the input impedance of the antenna is
very close to 50 Ω (near Marker 3 in Fig.6);
thus, the corresponding return loss is
tremendously low (−40 dB). This corresponds
to perfect impedance match between the helical
antenna and the 50-Ω coaxial cable. It should
be noted, however, that return loss is lowest at
the strip length of 60 mm compared with others
as seen from Fig. 5; however, the strip length
of 50 mm performs better in terms of the
worst-channel return loss (overall perform-
ance), which can be seen from Fig. 4.
By inserting a triangular copper strip,
which is close to the reflector (ground plane),
the strip itself behaves like a capacitor while
the helical antenna acts like a parallel RL
circuit. A complete parallel RLC circuit is
realized at the antenna feed point when the
impedance matching is applied. When the
frequency is below the resonance frequency,
the circuit input impedance consists of
resistance plus inductive reactance, which can
be observed at Marker 1 in Fig. 6. At the
frequency equal to 2.412 GHz, the antenna
input impedance is equal to 37 − j 4 Ω. On the
other hand, the reactance of the input
impedance becomes capacitive when the
frequency is higher than the resonance
frequency. From Fig.6, at the frequency of
2.472 GHz (Marker2), the antenna input
impedance is equal to 35 − j11 Ω.
Another point, which can be observed
from Fig. 5, is that the antenna-resonance
frequency, at which the return loss is lowest,
shifts to lower frequency orderly as the length
of the triangular strip increases. This is not
surprising since the resonance frequency of a
parallel RLC circuit is given by
L matching strip around its optimum (15 mm of
f = , (6)
C
Regular Paper
Fig. 6. Smith chart of measured helical
antenna at strip length of 60 mm.
where L and C are inductance and
capacitance of the parallel RLC circuit,
respectively. The area of the triangular copper
strip is enlarged with its length; thus, the
increase in its capacitance and the decrease in
the resonance frequency result.
Lastly, it should be pointed out that
although the resistive input impedance of the
antenna is purely resistive at resonance
frequency, it can be a value other than 50 Ω.
For example, this can be seen from the Smith
chart plotted in Fig. 7 when the strip length is
equal to 70 mm. Within the WLAN frequency
range, the helical antenna resonates at the
frequency of 2.436 GHz (see Fig. 1). However,
the point on Smith chart at that frequency (the
point, at which the curve connecting Marker1
and Marker2, crosses the horizontal line) is
below 50 Ω. As a result, the corresponding
return loss does not exhibit a deep dip due to
impedance mismatch. The return loss is merely
equal to −24.5 dB at that resonance frequency,
which is poorer than those around the
optimum.
Conclusions
It is found that overall return loss loosely
depends on the length of triangular impedance-
variation around the optimum of 50 mm).
Deviation from the optimum value dos not
result in significant return-loss degradation.
213
AU J.T. 11(4): 209-214 (Apr. 2008)
Although the return loss is not flat across the
WLAN frequency range, overall performance
with the help of this simple impedance-
matching strip is satisfactory, excluding the
simplicity of impedance-matching section.
Fig. 7. Smith chart of measured helical
antenna at strip length of 70 mm.
Return loss is kept below −14 dB across
the WLAN frequency range for the strip length
ranging from 45 to 60 mm. In addition, the
helical antenna incorporated with impedance-
matching strip behaves like a parallel resonant
RLC circuit, for which its capacitance comes
from the triangular impedance-matching strip
replacing the helical wire near the feed point.
References
Angelakos, D.J.; and Kajfez, D. 1967.
Modifications on the axial-mode helical
antenna. Proc. IEEE 55: 558-9.
ARRL. 2007. Antenna Book, 21st ed. American
Radio Relay League, Newington, CT, USA.
Regular Paper
Balanis, C.A. 1997. Antenna Theory, Analysis
and Design, 2nd ed. John Wiley, New York,
NY, USA.
Bart, R.M. 2003. The Stub Loaded Helix: A
Reduced Size Helical Antenna. Ph.D. Diss.,
Virignia Polytechnic Institute and State
University, Blacksburg, VA, USA.
Besten, R.D. 2007 Helical/Helix Antenna
Cookbook Recipe for 2.4 GHz Wavelans
and/or WiFi Applications.
<http://helix.remco.tk>.
Hecker, J. 2007. How to Make a Simple
2.425GHz Helical Aerial for Wireless ISM
Band Devices.
<http://www.wireless.org.au/~jhecker/helix>
Kraus, J.D. 1977. A 50-ohm input impedance
for helical beam antennas. IEEE Trans.
Antennas Propagat., AP-25: 913.
Kraus, J.D. 2003. Antennas for All
Applications, 3rd ed. McGraw-Hill, New
York, NY, USA.
Manthur, S.P.; and Sinha, A.K. 1988. Design
of microstrip exponentially tapered lines to
match helical antennas to standard coaxial
transmission lines. Proc. IEE 135, Pt. H, 4:
272-4.
Tsandoulas, G.N. 1967. The linearly tapered
transmission line as a matching section-high
and low-frequency behavior. Proc. IEEE 55:
1658-9.
Weeratumanoon, E. 2000. Helical Antennas
with Truncated Spherical Geometry.
Master’s Thesis, Virginia Polytechnic
Institute and State University, Blacksburg,
VA, USA.
WLAN. 2008. <http://en.wikipedia.org/wiki/
Wireless_LAN>
Wong, J.L.; and King, H.E. 1979. Broadband
quasi-taper helical antennas. IEEE Trans.
Antennas Propagat. AP-27: 72-8.
214