Successful Wire

Antennas

Edited by
Ian Poole, G3YWX
and
Steve Telenius-Lowe, 9M6DXX, KH0UN

Radio Society of Great Britain

1
 Published by the Radio Society of Great Britain, 3 Abbey Court, Fraser Road,
Priory Business Park, Bedford MK44 3WH. Tel: 01234 832700. Web: www.rsgb.org

Published 2012.

Radio Society of Great Britain, 2012. All rights reserved. No part of this publica-
tion may be reproduced, stored in a retrieval system, or transmitted, in any form
or by any means, electronic, mechanical, photocopying, recording or otherwise,
without the prior written permission of the Radio Society of Great Britain.

ISBN: 9781 9050 8677 1

Cover design: Kim Meyern

Design and layout: Steve Telenius-Lowe, 9M6DXX, KH0UN

Production: Mark Allgar, M1MPA

Printed in Great Britain by Latimer Trend of Plymouth

Publishers Note:
The opinions expressed in this book are those of the author(s) and are not neces-
sarily those of the Radio Society of Great Britain. Whilst the information pre-
sented is believed to be correct, the publishers and their agents cannot accept
responsibility for consequences arising from any inaccuracies or omissions.

2
 Contents

Foreword, by John D Heys, G3BDQ ................................... 4

Preface ................................................................................. 5

Editors note .............................................................................6

1 Antenna Basics ....................................................................... 7

2 Feeders .............................................................................. 27

3 Dipoles ....................................................................................42

4 Doublets ..................................................................................... 86

5 Verticals ................................................................................115

6 Loops ............................................................................................. 164

7 End-fed wires ....................................................................181

8 Impedance Matching and Baluns ...................................200

9 Antenna Masts and Rigging .............................................227

Appendix ...........................................................................234

Index .......................................................................... 237

3
SUCCESSFUL WIRE ANTENNAS

3 Dipoles

T
HE WIRE DIPOLE is a very effective antenna that can be constructed and
installed very easily and for only a small cost. The half-wave version of the
dipole has become the standard against which other radiating systems are
judged and it remains as perhaps the most effective, yet simple, single-band an-
tenna, and one which can virtually be guaranteed to perform well even when used in
far-from-ideal situations.
As the name suggests, it contains two legs or poles. The most common form is
the half-wave dipole, which (not surprisingly) is an electrical half-wavelength long.
The basic format for a half-wave dipole along with the voltage and current waveforms
can be seen in Fig 3.1. The voltage rises to a maximum at either end and falls to a
minimum at the centre, whereas the current is at its minimum at the end and its
maximum in the centre. Its feedpoint in the centre forms a low impedance point
suitable for many sorts of feeder.
A dipole does not have to be a half-wavelength long. A three half-wavelength
version can be seen in Fig 3.2. Again the points of voltage maximum are at either end
and at a minimum in the centre. Likewise the current is at its minimum at either end
and maximum in the centre.

DIPOLE LENGTHS
A resonant half-wavelength of wire will be somewhat shorter than its name implies.
RF energy in free space (electromagnetic radiation) can travel at the speed of light,
but when moving along a conductor it travels more slowly. At HF (between 3 and
30MHz) wires exhibit skin effect, i.e. most of the RF energy flows along the outer
surface of the conductor. A practical half-wave antenna made from wire needs end
supports; each end usually being terminated with an insulator. The capacitance
between the ends of dipole and its supports, even when the supporting material is
non-metallic, gives rise to end effect. This effect additionally loads the wire capaci-
tively and contributes towards its shortening from the theoretical half-wavelength.

(Left) Fig 3.1: A basic half-wavelength dipole

antenna with the voltage and current waveforms.

(Above) Fig 3.2: A three half-wavelength dipole.

42
 CHAPTER 3 DIPOLES

Frequency Length
With insulators Without insulators
(kHz) (feet) (metres) (feet) (metres)

1850 252' 11" 77.29 258' 5" 78.75

1950 240' 0" 73.33 245' 1" 74.71
3550 131' 10" 40.28 134' 8" 41.04
3750 124' 9" 38.13 127' 5" 38.85
7050 66' 4" 20.28 67' 10" 20.66
10100 46' 4" 14.15 47' 4" 14.42
14100 33' 2" 10.14 33' 11" 10.33
14250 32' 10" 10.03 33' 6" 10.22
18100 25' 10" 7.90 26' 5" 8.04
21100 22' 2" 6.77 22' 8" 6.90
21300 21' 11" 6.71 22' 5" 6.84
24940 18' 9" 5.73 19' 2" 5.84
28100 16' 8" 5.08 17' 0" 5.18
28500 16' 5" 5.01 16' 9" 5.11
29000 16' 1" 4.93 16' 6" 5.02
29500 15' 10" 4.84 16' 2" 4.93

Table 3.1: Lengths of half-wave dipoles.

The theoretical half-wavelength may be calculated from the expression:

Theoretical half wavelength (metres) = 150 / f (MHz)
or
Theoretical half-wavelength (feet) = 492 / f (MHz)
To take account of the end effect and the use of insulators, the length may be
calculated by using either:
Antenna length (metres) = 143 / f (MHz)
or
Antenna length (feet) = 468 / f (MHz).
When using nylon rope it has been suggested that no insulators are required. In
his book HF Antennas for All Locations (published by the RSGB), Les Moxon, G6XN,
suggests that when no insulators are used a half-wavelength can be found by using
either 478 / f (MHz) feet or 145.7 / f (MHz) metres.
A further factor which influences antenna resonant length is the diameter of the
wire used for that antenna. The formulas above are for typical wire dimensions.
Typical antenna lengths for the amateur bands from 160 - 10m, both when using
insulators or nylon rope, are shown in Table 3.1 above.

DIPOLE IMPEDANCES
A half-wave transmitting antenna, when energised and resonant, will have high RF
voltages at its ends with theoretically zero RF currents there. This means that the
ends of a half-wave dipole in free space will have an infinitely high impedance, but in
practice in the real world there will always be some leakage from its ends and into
the supporting insulators. This means that in reality the impedance at the dipole
ends is close to 100,000, a value which depends upon the wire or element
thick- ness. At a distance of approximately one-sixteenth wavelength from either
end it is 1000, and at the dipole centre, where the current is greatest and the
RF voltage is low, the impedance is also low.

43
SUCCESSFUL WIRE ANTENNAS

If it were made from an infinitely thin conductor wire, our theoretical dipole in free
space would have an impedance of about 73at its centre. Such an
antenna is impossible in the material world, and a practical half-wave dipole
made from wire will have an impedance at its centre at resonance close to 65.
Antennas fabricated from tubing have lower values at their centres, of between
55 and 60. These im- pedance values also depend upon the height of the
antenna above ground, as will be shown later.
The very high values of self-impedance at the ends of a half-wave wire makes
end-feeding difficult, and this is why breaking the wire at its centre and connecting
the inner ends so formed to a low-impedance feedline makes a convenient and
efficient coupling and match. Suitable feeder is available in the form of twin-lead or
coaxial cable, which both have design impedances lying between 50 and 75.
These present a good match to dipole centres.
At exact resonance the impedance at the centre of a half-wave dipole is like a
pure resistance. At any other frequencies the same dipole will have either inductive
or capacitive reactance at its feedpoint. If the dipole is too short to be resonant the
reactance is capacitive and when it is too long the reactance becomes inductive. In
either case there will be problems in matching the 50 or 70feeder to the dipole
and if the reactances are great, there will be a high SWR on the feeder and
considerable power loss.

ANTENNA Q
A half-wave antenna is something like a conventional tuned circuit where the Q, or
Quality factor, is largely determined by the resistance of the coil. Losses in the
capacitor used in the circuit are generally small and are not so significant in the
determination of Q. A high-Q tuned circuit exhibits very sharp tuning (selectivity) and
this is also the case when an antenna has a high Q.
Using thin wires lowers the bandwidth of a half-wave antenna, but not dramatically.
However, short wires that are brought into resonance will exhibit high Q. The shorter the
wire in terms of wavelength, the higher
the Q. Small changes in the transmit-
ting frequency away from the antenna
resonances will give rise to a rapid rise
in the reactance at the feedpoint.
Thicker wire will lower the Q, reduce
resistive loss and make the half-wave
dipole less frequency conscious. It is
therefore best to ensure that such an
antenna is made from the thickest pos-
sible wire consistent with such factors
as the pull on the antenna supports,
windage and sag.

DIPOLE HEIGHT
The height of a horizontal dipole above
the ground as a ratio of its design fre-
quency is important (see the stand-
ard curves of feed impedance against
height in Fig 3.3). When below about
half a wavelength high the radiation
resistance at the feedpoint will be re-
Fig 3.3: Radiation resistance of a half-wave dipole as a duced, and down at a height of just
function of height above the ground (reprinted with per- one-tenth of a wavelength it will only
mission of the American Radio Relay League). be 25. This means that a dipole
fed
 CHAPTER 3 DIPOLES

44
SUCCESSFUL WIRE ANTENNAS

THE SLOPING DIPOLE

Horizontal half-wave dipoles require two end sup-
ports and it is not always possible to provide
these in some awkward locations. In such situa-
tions a single support, preferably a non-metallic
mast or a high point on a building, will suffice,
and then the antenna can be arranged to slope
down towards the ground at an angle lying some-
where between 30 and 60 (Fig 3.14). The slop-
ing half-wave dipole should have its lower end at
least one-sixth of a wavelength above ground,
and its feeder should ideally come away from
the radiator at 90 for at least a quarter of a wave-
length. If coaxial feeder is used the braid should
connect to the lower half of the antenna.
The performance of a sloping dipole is quite
different from one of the horizontal variety and it
can be good for long distance work. The radia-
tion from a sloping dipole shows slant polarisa-
tion with both vertical and horizontal components
according to the amount of slope. Its lower angle
of radiation to the horizon can result in a little low
angle gain over a horizontal dipole. This kind of
gain is difficult to realise on the low bands in
other ways, where for most amateurs multi-ele-
ment Yagi beams are out of the question.
There is some high-angle radiation from the Fig 3.14: A half-wave sloping dipole which can be
sides of the sloping dipole but very little radiation put up in a small space and which will be useful
from its high end. An actual plan of the horizontal for long-distance working. Most of its low-angle
radiation pattern resembles a heart with a null radiation is towards the low end of the antenna
between its two upper lobes. This null corre- but there is also considerable radiation at high
sponds with the high end of the sloping dipole. A angles in other directions.
disadvantage is of course that long-distance
working will only be possible towards one direc-
tion, but this may be overcome by having a group of three or four slopers suspended
from a common central support, each with its individual feedline which may be
selectively switched to the transceiver. (There are designs which involve the unused
dipoles in such arrangements as reflectors to improve forward gain and front-to-
back ratios, but their correct adjustment can be complicated.)
Slopers are ideal in many applications where a single support is available. Many
people who have beams and towers, mount a sloper on the tower for one of the lower
frequency bands, ensuring that the direction of maximum radiation is arranged to-
wards the areas of the globe they want to contact, sometimes having two or more
around the tower.

THE VERTICAL DIPOLE

A vertical half-wave dipole will radiate vertically-polarised signals all round, and much
of the radiation will be at the low angles favourable for DX working. If a feed imped-
ance of around 70 is required, the centre of this antenna must be around
0.45above the ground and so it is usually more convenient to arrange for a vertical
quar- ter-wave antenna to be used, which can then have its feedpoint at or near
ground level. A vertical dipole cannot be hung down from a metal mast or tower, and it
should have its feeder come away from the radiator wire at right angles if the
radiation pattern is to be preserved, which may present some problems. As a
result, vertical
CHAPTER 3 DIPOLES

55
SUCCESSFUL WIRE ANTENNAS

half-waves are not often used by amateurs, although they can be practical on the
higher-frequency HF bands. A practical design for vertical dipoles is given in the
chapter on vertical antennas.

A 3 /2 DIPOLE: THE 40M DIPOLE ON 15M

The dipole, when fed with coaxial cable, is basically a single band antenna. While
this is true, there are a few ways that dipoles can be made to work on more than one
band. One method is to parallel two or more dipoles for different bands together;
another is to use traps. We will look at both of these methods later in this chapter.
The simplest way, though, is to take advantage of the fact that the low impedance at
the centre feedpoint of a dipole occurs not only when it is one half-wave long, but also
when it is three half-waves long (and in fact all odd numbers of half-waves). We can
take advantage of this where amateur bands have this same harmonic relationship,
i.e. where one band has three times the frequency (or five or seven times, etc), and
therefore one third (or one fifth or one seventh) of the wavelength of another band.
On HF, the best example of this is the relationship between 40m and 15m: the
third harmonic of 7MHz is 21MHz. What this means is that a 40m dipole should also
work on 15m. Unfortunately, life isnt quite that simple. As we have already seen, the
end effect means that a half-wave dipole is physically about 5% shorter than its
theoretical (electrical) half-wave length. However, when the same antenna is operat-
ing on its third harmonic, it becomes fifteen per cent shorter than the electrical length
of a three half-waves antenna. What this means in practice is that it is actually reso-
nant quite a bit higher in frequency than you might expect.
The way around this problem is to resonate the 40m dipole at the very bottom of
the band, 7000kHz, or even make it somewhat longer still, so that the minimum SWR
point is actually below the bottom of the band on, say, 6980kHz. On 21MHz you will
find the minimum SWR point is nevertheless at the top of the band, around 21400
or 21450kHz.
Furthermore, QST Technical Editor Joel Hallas, W1ZR, points out (in the July
2009 QST) that the resonant impedance of a 3/2 dipole is above 100, so its
not as good a match to 50coax as is the /2 case.
The good news is that an
external ATU or the internal
automatic ATU in your rig
should be able to reduce the

Fig 3.15:Comparison between the 40m (black)

and 15m (white) EZNEC SWR plots of a 66ft
high, 67.2ft long, 40m dipole (diagrams re- Fig 3.16: Comparison between the 40m (black)
printed with permission of the American Radio and 15m (grey) azimuth patterns of a 40m /2
Relay League). dipole.

56
 CHAPTER 3 DIPOLES

SWR to close to 1:1 at your operating frequency of choice in both the 40m and 15m
bands.
Fig 3.15 shows a comparison between the 40m and 15m EZNEC SWR plots of
a 66ft high, 67.2ft long, 40m dipole made of 14 gauge bare wire. In Fig 3.16 note that
the azimuth pattern of a 3/2 dipole is not the same as the usual /2 case. While
different, the pattern can be useful and provides a bit of additional gain in its prime
directions.
On HF there are a few other combinations of bands that have an odd harmonic
relationship, for example an 80m half-wave dipole cut for the lower-frequency end of
the band is five half-waves long on the 17m band, and seven half-waves long on the
12m band. (However, if the 80m dipole is cut for the SSB DX end of the band, around
3800kHz, its five and seven half-wave resonances will be well above the top end of
the 18MHz and 24MHz bands respectively, and the SWR is likely to be very high on
both bands.)

MOUNTING A WIRE DIPOLE ABOVE A ROTATOR

Most amateurs with masts and towers suspend their HF wire dipole or inverted-V
dipole from a point below the rotator, leaving the beam free to rotate above. However,
mounting a dipole on a mast extension above the beam is a much better option,
provided the inverted-V angle can be made shallow enough to clear the beam as it
rotates underneath. Another strong reason for mounting low-band dipoles above the
beam is the extra height above ground, which makes them more effective undoubt-
edly for DX, and often for more local QSOs as well. The problem with this over the
top approach is that the centre of the dipole must be able to pivot on the top of the
mast, so that the mast and beams can rotate beneath it.
One successful solution to this problem was designed by Jan Fisher, G0IVZ,
and taken up by Ian White, GM3SEK, in his In Practice column in the April 2009
RadCom. The rotating mast for the HF beam is made from scaffold tubing, extended
by a 1.5in fibreglass pole (an aluminium pole of that diameter probably couldnt
handle the bending forces). The top of the extension pole is filled by a close-fitting

80m dipole over the top of a small HF Close-up of the rotating dipole centre
beam. with balun box to the rear.

57
SUCCESSFUL WIRE ANTENNAS

hardwood plug, drilled 8mm through the centre and secured with epoxy.
G0IVZs idea was to use a ready-made centre insulator that was originally de-
signed for mounting a tubular dipole on the boom of a Yagi. The plastic moulding is
strong enough to support a much longer wire dipole, simply tied on through the fixing
holes as shown in the photograph. G0IVZ used the built-in terminal box to connect
the dipole to the coax feedline, while GM3SEK, who has adopted a similar set-up,
connected the two wire ends to the terminals on the balun box.
The main mounting hole of the centre insulator is drilled 8mm for an M8x100mm
stainless steel screw which is the pivot pin for the whole assembly. In GM3SEKs
version, the balun bracket is fixed to the bottom of the insulator with a nut, so those
two parts rotate together. A large washer is added to spread the down-thrust, and the
free end of the screw simply drops into the hole in the wooden plug. A later addition
was a piece of white PVC waste pipe, taped to the bracket to prevent the bottom edge
scratching the fibreglass. Below this fitting there has to be a rotation loop in the coax
and of course theres the usual loop around the rotator itself.

THE FOLDED DIPOLE

Another form of dipole is the folded dipole, shown in Fig 3.17. It is often used as a part
of more complex antennas such as Yagis, but it can also be a useful antenna on its
own. It has the advantages that it has a higher impedance and a wider bandwidth
than an ordinary dipole.
The 300feed impedance is an important feature of a folded dipole. The
power supplied to a folded dipole is evenly shared between the two conductors which
make up the antenna, so therefore the RF current, I, in each conductor is reduced
to I/2. This is a half of the current value (assuming that the same power is applied)
at the centre of the common half-wave dipole, so the impedance is raised. By
halving the current at the feedpoint yet still maintaining the same power level, the
impedance at that point will be four times greater. This means that a two-conductor
folded dipole will have a feed impedance of 280, which is close to the
impedance of 300twin feeder. It can therefore be satisfactorily matched and
fed with this feeder, and have a low SWR along the feedline.
If a third conductor is added to the folded dipole (Fig 3.18), the antenna current
will be evenly split three ways and the impedance at the feedpoint will be nine times
greater than the nominal 70 impedance of a simple dipole. Such a three-
wire dipole with its feed impedance
of 630 will make a
good match to a 600
feeder. This feeder may be
made from 18SWG wires
which are spaced at 75mm
(3in).

/2

Fig 3.18: A three-wire folded dipole. If each wire

is of equal diameter the total current will be
shared equally between the three wires and the
impedance at the feedpoint will be nine times that
of a conventional half-wave dipole (9 x
Fig 3.17: The basic two-wire folded dipole. 70= 630), a close match to a
600feedline.
 CHAPTER 3 DIPOLES

58
SUCCESSFUL WIRE ANTENNAS

4 Doublets

O
NE CLASS OF antenna that is not as widely used as it might be is that of
tuned feeder antennas. Using an open-wire tuned feedline as part of the
overall antenna system enables multi-band operation to be achieved, al-
though such an antenna - often called a doublet - does require the use of an ATU to
ensure that there is a good match to the transceiver.
The key to tuned feedline antennas is naturally the feeder. As discussed in
Chapter 2, these open-wire feedlines have a characteristic impedance which relates
to the diameter of the wire used and the spacing between the feed wires. This
impedance is important in many applications, but note that it is of no consequence
when considering centre-fed antennas which use tuned lines exclusively.
Tuned feedlines operate on the principle that they are really a part of the antenna
and have standing waves along their lengths. Standing waves are a feature of most
radiating wires but, if two such wires of equal length are closely spaced (in terms of
wavelength) and fed in anti-phase, in theory they will not radiate (in practice they will
radiate a very small proportion of the RF power applied).

THE BASIC DOUBLET ANTENNA

The basic doublet (Fig 4.1) is a probably the most useful simple multi-band antenna
for amateur use. It is simple and yet effective, and requires no special earth or
counterpoise arrangements. The only drawbacks are the requirement to use an ATU

Fig 4.1: The basic doublet antenna.

86
SUCCESSFUL WIRE ANTENNAS

and that the balanced feeder cannot be routed through

the house.
The doublet is essentially a balanced system and
each half of the top, plus each wire in the feedline,
must be equal in length. The antenna top is not cut to
resonate at any particular frequency (unlike the half-
wave dipole), and almost any length may be chosen
to suit an individual location.
The doublet can be used over a wide range of
frequencies although as the frequency changes so
the radiation pattern of the antenna will alter. A half-
wavelength antenna has the maximum radiation at
right angles to the axis or line of the antenna. As the
electrical length of the antenna increases the phas-
ing of the radiation from the antenna wire means that
new lobes appear and grow. Examples of polar dia-
grams of a half-wave and a three half-wave radiator
are shown in Fig 4.2.
When erecting an antenna of this nature there
are no particular precautions to observe except that,
due to possible problems with reactance making a
good match difficult to achieve, certain combinations
of feeder / top leg length should be avoided. These
are summarised in Table 4.1. The table shows that
when using doublet legs of 15.2m (50ft) together with
16.4m (54ft) of feeder there ought to be little difficulty
with reactance on most amateur bands. There are of Fig 4.2: Horizontal polar diagrams for half-
course many other combinations of top length and wave and three half-wave horizontal wires.

Band Lengths to be avoided (metres)

(MHz) (half the total top length plus feeder length)

1.8 / 1.9 56.4m 93.7m 131m

3.6 29.26m 48.8m 68.3m
7 15m 25.14m 35.2m 45.26m
10.1 10.5m 17.52m 24.53m 31.54m
14.15 7.5m 12.6m 17.6m 24.2m 27.7m 32.7m
18.1 5.9m 9.9m 13.86m 17.83m 21.8m 25.8m
29.7m 33.7m
21.2 4.9m 8.2m 11.6m 14.9m 18.1m 21.5m
24.7m 28.0m 31.4m 34.8m
24.94 4.3m 7.1m 10m 12.8m 15.6m 18.5m
21.3m 24.1m 27.1m 29.9m
29 3.7m 6.1m 8.5m 11m 13.4m 15.8m
18.3m 20.7m 23.2m 25.6m 28m 30.5m

Table 4.1: Lengths to avoid when designing multi-band doublets with tuned feeders.

87
 CHAPTER 4 DOUBLETS

Band (metres) 80 40 30 20 17 15 12 10 6 2

L1 end (2) 168 89.4 63.2 45 35.3 30 25.6 22.4 12.7 4.37
L2 centre (2) 65 34.7 24.6 17.5 13.7 11.7 9.97 8.72 4.95 1.70
L3 total size 467 248 176 125 98 83.3 71.2 62.3 35.4 12.2
L4 stubs 48.6 25.8 18.3 13 10.2 8.66 7.4 6.47 3.68 1.26
L5 height 120 64 45 32 25 21 18 16 10 10
Inductor (H) 25.9 11.7 7.3 4.9 3.4 2.8 2.2 1.9 0.85 0.13
Gain (dBi) 11.4 11.4 11.3 11.2 11.1 11.0 11.0 11.0 11.4 10.9
Freq (MHz) 3.8 7.15 10.1 14.2 18.12 21.3 24.93 28.5 50.2 146

Table 4.3: Lengths (in feet) of an HGSW beam for 10 amateur bands.

insulators as shown. The lower ends of the two lines should be stripped and bent
over and soldered together. The resultant active line length must be 13ft. The dis-
tance from the centre insulator to the ladder line should be 17.5ft. If you have a lot of
wind in your area you might want to tie a 1oz lead fishing sinker to the bottom of each
of the phasing lines. Alternately a string can be attached and tied to some secure
point below the antenna. AL7KK says that he has had no problem with his phasing
lines except that they curl slightly, which is not ordinarily a serious difficulty.
The antenna is completed by winding five turns of coax near the feedpoint into a
6in diameter coil and securing them with tie wraps. This acts as a cheap but effective
choke balun.
EZNEC modelling results indicate that with the antenna at 2 high (32.8ft on
20m), the gain will be about 11.2dBi with a peak of the elevation lobe at 29. Calcu-
lated azimuth, elevation and SWR plots at 2 height are shown in Figs 4.14, 4.15 and
4.16 respectively. Even more gain is available, and more importantly lower elevation
angles of the main lobe, with greater heights. For example at 3/4-, the peak eleva-
tion drops to 20, and to 15 at 1.
Table 4.3 shows the lengths necessary to build an HGSW beam for all bands
from 80 to 2 metres. The dimensions were scaled from the 20m model that was built
and tested, while EZNEC was used to calculate the gain and inductor values.

THE G5RV ANTENNA

Louis Varney, G5RV, designed his famous G5RV antenna in 1946, but it was not until
1958 that he wrote about it in An Effective Multi-band Aerial of Simple Construction
(RSGB Bulletin, July 1958). He described it in greater detail, again in the RSGB
Bulletin, in November 1966 (The G5RV Aerial - Some Notes on Theory and Operation).
Finally, he wrote a further article, G5RV Multiband Antenna . . .
Up-to-Date, published in Radio Communication in July 1984.
The G5RV antenna has achieved almost iconic status
during the last half century, so it is perhaps worth looking at it in
some detail. In this chapter we first look at the design and how
it evolved, using Louis Varneys own words and the original
diagrams which accompanied his articles. Then we take a 21st
century look at the design, using computer analysis, a luxury
that obviously Varney did not have in 1946, or even in 1984.
Louis Varneys original design, as published in 1958, is
shown in the diagram opposite. Then, he wrote, The aerial
consists essentially of a 102ft flat-top split in the centre where a
Pyrex type insulator is inserted, a 34ft long open-wire stub (spac-
ing is unimportant) and sufficient length of 72 ohm coax or twin
Louis Varney, G5RV, in 1998. feeder to reach the transmitter. Alternatively, open-wire feeder

98
SUCCESSFUL WIRE ANTENNAS

may be employed from the centre of the aerial

right back to the transmitter output or ATU.
These two methods of feeding the antenna
are shown in Fig 4.17 (a) and (b) respec-
tively. It will be noted that what is shown in
Fig 4.17(b) is simply a basic doublet with a
102ft top and open-wire feeder.
Describing the antennas performance
on each band, G5RV wrote that on 20m: ...the
aerial really comes into its own. On this band
it functions as a three half-wavelength (a)
aerial... Since the impedance at the centre is
about 100 ohms, a satisfactory match to the
72 ohm feeder is obtained via the 34ft of half-
wave stub. . . By making the height a half-
wave or a full wave above ground at 14 Mc/s
and then raising or lowering the aerial a bit
at a time while observing the standing-wave
ratio on the 72 ohm twin-lead or coax feeder
by means of an SWR bridge, an excellent
impedance match may be obtained on this
band. The technique of matching an an-
tenna by raising or lowering it seems to have
been lost over the years!
By 1966, 300 ribbon feeder had
be- come more widely available, and
G5RV wrote, A word about the matching
stub is in order. If this is of open wire feeder
construc- tion (preferred because of lower (b)
losses, es- pecially on 21 and 28 Mc/s) its
length should be 34ft... but if 300 ohm Fig 4.17: The two methods of feeding the G5RV
ribbon is used, al- lowance must be made antenna, as described by Lois Varney in his original
for the velocity factor of this type of twinlead. 1958 article.
Since this is approxi- mately 0.88, the actual
physical length of the
300 ohm ribbon stub should be 29ft 6in. It should be born in mind that this matching
stub is intended to resonate as a half-wave impedance transformer at 14 Mc/s, which
was chosen as the design centre frequency for the G5RV aerial, thus giving a very good
impedance match for a 75 to 100 ohm twin-lead or coaxial cable connected to the base
of the stub. Thus it is clear that, although Louis Varney described the G5RV as a multi-
band antenna, he optimised it for use on 20m.
G5RV went on to say, An alternative arrangement to that of the matching stub and
twin-lead or coaxial cable feeder is to use an 83ft length of open-wire feeder measured
from the centre of the flat top to the terminals of the ATU. The specific length of 83ft
(modified to 84ft in G5RVs 1984 article) was chosen because it permits parallel
tuning of the ATU on all bands from 3.5 to 28 Mc/s with very low feeder losses.
G5RVs 1966 article gave current distribution diagrams for the antenna on each of
the five HF bands then allocated to amateurs. He also described the ATU designed for
use with the antenna.
The problem of currents flowing on the outer of the coax was recognised by G5RV,
for he wrote: Although it may be very convenient to use a length of, say, up to 100ft of
coax direct from the transmitter to the base of the matching stub, it must be remem-
bered that such an arrangement will tend to produce currents which will flow in the outer
conductor of the coax, causing unwanted radiation from the coaxial feeder. This may be
avoided by the use of either 75 ohm twin-lead and a suitable ATU or the open-wire
CHAPTER 4 DOUBLETS

99
SUCCESSFUL WIRE ANTENNAS

feeder and ATU as already mentioned. However, the use of a wide-band balun... would
be preferable if coaxial cable is to be used. Nevertheless, in practice very satisfactory
operation can be achieved by the simple use of coax direct from the transmitter to the
base of the matching stub even though the VSWR may reach 10 to 1 or more on 3.5
Mc/s. This figure may be reduced to about 5 to 1 on 3.5 Mc/s by pruning the coax. On the
higher frequency bands the VSWR on the coax lies between 5 to 1 and 1.5 to 1, the latter
figure applying to 14 Mc/s where, as explained above, the matching is very good.
This suggestion of using a balun was reversed in Louis Varneys 1984 article, in
which he wrote: In the original article describing the G5RV antenna, published in the,
then, RSGB Bulletin November 1966 [Varney himself appears to have forgotten about
the earlier 1958 article - Ed], it was suggested that if a coaxial cable feeder was used,
a balun might be employed to provide the necessary unbalanced-to-balanced transfor-
mation at the base of the matching section. This was because the antenna and its
matching section constitute a balanced system, whereas a coaxial cable is an unbal-
anced type of feeder. However, later experiments and a better understanding of the
theory of operation of the balun indicated that such a device was unsuitable because of
the highly reactive load it would see at the base of the matching or make-up section
on most HF bands.
It is now known that if a balun is connected to a reactivbe load presenting a VSWR
of more than about 2:1, its internal losses increase, resulting in heating of the windings
and saturation of the core (if used). In extreme cases, with relatively high power opera-
tion, the heat generated due to the power dissipated in the device can cause it to burn
out. However, the main reason for not employing a blaun in the case of the G5RV
antenna is that, unlike an astu [ATU] which employs a tuned circuit, the balun cannot
compensate for the reactive load condition presented to it by the antenna on most of the
HF bands, whereas a suitable type of astu can do this most effectively and efficiently.
(Louis Varney used the term ATU in 1958 and 1966, but in the August 1983 Radio
Communication he had had an article published in which he argued the case that the
device ought more properly be called an Antenna System Tuning Unit, or astu. More
accurate or not, the name did not catch on.)
Instead, he recommended the use of an HF choke, a device which these days is
often referred to as a common-mode choke balun: Under certain conditions, either
due to the inherent unbalanced-to-balanced effect caused by the direct connection of
a coaxial feeder to the base of the (balanced) matching section, or to pick-up of energy

3.5MHz Flat top plus about 17ft (5.18m) of the matching section forms a /2 dipole partially
folded up at the centre. Reactive load.
7MHz Flat top plus 16ft (4.87m) of the matching section functiuons as a partially folded-up
collinear array with two half-waves in phase. Reactive load.
10MHz Collinear array with two half-waves in phase. Reactive load.
14MHz 3/2 centre-fed long wire. Matching section functions as a 1:1 impedance transformer.
Resistive load, approx 90
18MHz Two full-wave antennas, slightly folded up at the centre, fed in phase. High imped-
ance load, slightly reactive.
21MHz 5/2 long wire. High impedance load, virtually non-reactive.
24MHz 5/2 long wire with low resistive load of approx 90 - 100.
28MHz Two x 3/2 long wires fed in phase. High impedance load, slightly reactive.

Table 4.4: G5RV antenna theory of operation on each of the HF bands (Source: G5RV Multiband
Antenna... Up-to-Date, by G5RV, July 1984.)

100
 CHAPTER 4 DOUBLETS

3.5MHz 7MHz

10MHz 14MHz

18MHz 21MHz

24MHz 28MHz

Fig 4.18: Current standing-wave distribution on the G5RV antenna and matching section on each of the
HF bands. (Source: G5RV Multiband Antenna... Up-to-Date, by G5RV, July 1984. )

radiated by the antenna, a current may flow on the outside of the coaxial outer conduc-
tor. This is an undesirable condition and may increase chances of TVI to nearby TV
receivers. This effect may be considerably reduced, or eliminated, by winding the co-
axial feeder into a coil of 8 to 10 turns about 6in in diameter immediately below the
point of connection of the coaxial cable to the base of the matching section.
By 1984, radio amateurs had been allocated additional bands at 10.1, 18.0 and
24.8MHz, and in his article G5RV Multiband Antenna . . . Up-to-Date, Louis Varney
described the theory of operation on each of the HF bands, including the three new
ones (Table 4.4). The current distribution on each band is shown in Fig 4.18.

101
SUCCESSFUL WIRE ANTENNAS

For use in restricted spaces, Louis Varney wrote that, because the most useful
radiation from a horizontal or inverted-V resonant antenna takes place from the cen-
tre two-thirds of its total length, up to one-sixth of this total length at each end of the
antenna may be dropped vertically, semi-vertically, or bent at some convenient angle
to the main body of the antenna without significant loss of effective radiation effi-
ciency. This would imply that the full-size G5RV could be fitted into a space just 68ft
(20.72m) long, if 17ft of wire were to be dropped vertically at either end of the antenna.

HALF-SIZE G5RV
The Half-Size G5RV is shown in Fig 4.19.
Writing in 1966, Louis Varney, G5RV, stated
51ft (15.5m)
that, It is quite possible to scale all wire
dimensions (including that of the stub) down
Matching stub to exactly half-size and the resulting aerial
17ft (5.18m) open-wire feeder o r
15ft (4.56m) 300ribbon will work from 7 to 28 Mc/s. Optimum per-
Common-mode choke balun formance and impedance matching will
(8 - 10 turns coax, 6in diameter) occur on 28 Mc/s, where the operating con-
-------

ditions will be as for the full-size version at

50coax to 14 Mc/s.
ATU
In 1984 he added that by strapping the
Wide-range
To transceiver
station end of the feeder (either balanced or
ATU coaxial) and feeding it via a suitable ATU
using a good earth connection or a counter-
poise wire, the half-size version may also
Fig 4.19: The Half-Size G5RV.
be used on the 3.5 and 1.8MHz bands.

THE G5RV ANALYSED

Writing in the August 2010 QST, Joel Hallas, W1ZR, commented on three questions
often asked about the G5RV antenna:
What is the function of the usual 34ft sec-
tion of window line or twinlead between
the antenna and the coax?
Should there be a balun or choke at the
transition from the balanced line to coax?
Should the SWR on the coax be a matter
of concern?
G5RV wanted an antenna that would work
well in certain directions on 20m, and that
could also be used on all the HF bands, at
that time just 80, 40, 20, 15 and 10m. EZNEC
computer analysis, which of course was not
available to G5RV when he designed the an-
tenna, allows us to see the radiation pattern
on 20m (Fig 4.20).
The section of balanced line, 34ft of open
wire in Varneys article, transforms whatever
the antenna impedance is to a different im-
Fig 4.20: Azimuth pattern of G5RV antenna on pedance at its bottom. To say that it provides a
20m. At a typical height of 40ft, the peak take-off good match on all bands may be wishful think-
elevation is 24. Note the sharp lobes perpen- ing. On 20m it is a half-wave long and thus
dicular to the antenna, as well the broad lobes repeats the antenna impedance, as it does
at other potentially useful angles. The gain at on 10m. The half-wave window line or twinlead
each is within a dB or two of the dual lobes of a that he used as a transforming section resulted
half-wave dipole at the same height. in an impedance at the bottom on 20m of

102
 CHAPTER 4 DOUBLETS

around 100with some reactance.

Computer analysis also allows us to plot
the 75SWR from 3.5 to 30MHz (Fig
4.21). It shows that there are indeed
multiple reson- ances; however, not many of
them line up well with amateur bands.
The dimensions given by G5RV, and with a
height of 40ft above ground, produce a 75
SWR of 6.5:1 on 3.7, 5.6:1 on 7.1, 2.4:1 on 14.2,
4.6:1
on 21.2 and 2.1:1 on 24.9MHz. Other bands are
higher, typically at least 10:1.
A fundamental limitation of the design is that
there are only three adjustments - the flat-top
length, the height and the transforming section Fig 4.21: 75SWR plot of Varneys
length. With those variables, you can probably original G5RV antenna design using a 34ft
find dimensions that will work on multiple, but section of air-dielectric open wire line as
not all, bands. Unlike trimming a half-wave di- transforming section. The resonances
pole, the direction to go with each change is not almost line up with some amateur bands,
obvious. W1ZR says he has never found a set of with a 1.7:1 SWR at the top of 20m.
dimensions that resulted in acceptable SWR on
all, or even most, bands.
The question about the importance of the SWR depends on the length and loss
of the coax used, as well as the tuning range of the ATU used. A high SWR on the
higher bands will result in significant loss for typical coax lengths. This makes the
SWR at the radio look better than it really is, since the loss reduces the power that
gets to the antenna and further reduces the reflected signal. This may explain why
many think it has better SWR on multiple bands than it really does.
As with any balanced load to unbalanced line transition, the need for a balun
depends on the amount of current that flows on the outside of the shield. This in turn
depends on the ground impedance and the electrical length of the coax. Considering
its use on multiple bands, it is likely that there will be some bands that have high
shield currents and thus could benefit from a balun. At least one commercial manu-
facturer just slips multiple ferrite beads on the coax just below the transition with
good results.

THE ZS6BKW ANTENNA

In 1982 Dr Brian Austin, ZS6BKW (now G0GSF), used some early computer model-
ling to optimise the G5RV antenna. By then, UK amateurs had access to three addi-
tional bands at 10.1, 18 and 24.9MHz, which were
not available when G5RV designed his antenna,
so operation was also considered on these bands.
In 2007 G0GSF re-computed his design and came
up with new dimensions for an antenna that
presents a better than 2:1 SWR without the use of
an ATU in the 40, 20, 17, 12 and 10m bands. It can
also be used with an ATU on 80, 30 and 15m.
He wrote: The configuration of the ZS6BKW
is shown in Fig 4.22. The dipole radiator is L1, the
series section impedance matching transformer
(to give its formal name), with characteristic im-
pedance of Z2, is L2 spaced twin-wire. The lower
end of L2 presents an impedance, Z3, to the co-
axial cable, Z4 (50as is standard practice Fig 4.22: The ZS6BKW antenna as computed
in all modern radio systems). A computer-based in 2007 by G0GSF (ex-ZS6BKW).
pre-
SUCCESSFUL WIRE ANTENNAS

103
 CHAPTER
CHAPTER 54 VERTICALS
DOUBLETS

5 Ver ticals

A
FTER THE DIPOLE, the vertical antenna in its various guises is probably the
second most widely-used HF antenna today. Like the dipole, the basic quar-
ter-wave vertical is simple to make and can almost be guaranteed to work
with minimal pruning required, provided it is made well and certain guidelines are
followed. However, while a horizontal dipole is often easy to mount in the clear, a
vertical, ground mounted in a typical garden for example, is liable to be screened by
nearby objects such as buildings and trees. As a result its performance in typical
urban or suburban locations can sometimes be disappointing. Furthermore, a quar-
ter-wave vertical needs a ground plane, usually in the form of radial wires, to work
properly, and a less than adequate ground connection can also lead to disappoint-
ing results. Nevertheless, a simple quarter-wave vertical wire can work well, and in
certain circumstances extremely well, as we shall discuss later in this chapter.
There is a tendency to think that because a vertical wire takes up virtually no
space at all, it is an ideal antenna for those with very limited space. Unfortunately, this
is usually not the case. Because quarter-wave verticals require radial earth wires, a
quarter-wave vertical antenna system can take up at least as much space as a
horizontal dipole for the same frequency band. In the ideal case, quarter-wave long
radials will extend in all directions and the vertical radiator would therefore be in the
centre of a square a half-wavelength long by a half-wavelength wide. Nevertheless,
it is possible to make certain compromises without affecting the performance too
greatly and, provided you are prepared to put in the ground work (literally), verticals
can be very effective antennas, even for those with limited space for antennas.

THE QUARTER-WAVE VERTICAL

The most basic vertical antenna is the quarter-wave. In this configuration one con-
nection from the feeder is taken to the quarter-wave vertical radiating element, and
the other is taken to ground. In this way the ground provides the image, or other half
of the antenna, as shown in Fig 5.1(a). As such the ground connection is an integral
part of the antenna system as a whole, and upon its effectiveness rests the efficiency
of the whole antenna. In fact this is true for any antenna of this nature that uses the
ground for one of its connections.
In view of the fact that one of the connections from the feeder is taken to ground,
this type of antenna is an unbalanced antenna. Accordingly it can be fed directly
using unbalanced feeder, such as coax, without the need for a balun.
The impedance at the point where a resonant quarter-wavelength vertical con-
ductor meets the ground is about 36- half of the feed impedance at the centre
of a resonant half-wave dipole. The current along the quarter-wave vertical antenna
is at its maximum at its base and therefore the greatest radiation will take place at
this point - see Fig 5.1(b). The radiation will be vertically polarised and in the
example illustrated will have equal field-strength levels in all directions.
Much of its radiation will be at low angles to the horizon when above a good
ground, and this makes the vertical antenna very attractive for both short-distance
(ground wave) and long-distance communications on the lower-frequency bands.

115
SUCCESSFUL WIRE ANTENNAS

(a) (b)

Fig 5.1: (a) The basic quarter-wave vertical antenna positioned over perfect ground, showing its
earth image. Most of the earth return currents flow through the ground in the vicinity of the antenna.
(b) A representation of a quarter-wave vertical antenna over perfect ground which is energised by
a signal with a base current of 1A. The RF current at 10 points along its length is shown and also
the impedance at these points. There is a rapid fall in current towards the top of the antenna and
the impedance therefore rises greatly there. It is interesting to note that the fall in current over the
final 30 of this antenna is almost linear.

The polarisation of an antenna when used for long-distance work does not matter,
for the effects of refraction in the ionosphere etc will inevitably induce changes in
polarisation.
In order to be able to gain the most from a vertical antenna, the ground system
that is used with it must be efficient. One solution is a mat of buried wire extending to
at least a quarter-wavelength and possibly a half-wavelength from the base of the
antenna but for most practical situations this may not be possible. The antenna will
still work with several buried radial (the more the better). Ground systems were
discussed in Chapter 1 and there is more on wire radial systems later in this chap-
ter.
As an alternative to the ground-mounted vertical it is possible to elevate the
antenna and use a ground plane system, in which case the ground plane wires
should be resonant, a quarter-wave long. Raising the antenna in height allows it to
take advantage of the height gain available.
Fig 5.1(a) represents a simplified and ideal quarter-wave vertical antenna. The
ground is shown to be a perfect conducting medium, a condition which can only be
realised when it is replaced by a sheet of metal which has dimensions that are large
relative to the length of the antenna or by a large body of salt water. The ground, if it is
a perfect conductor, will behave like an electrostatic shield and provide an image
antenna a quarter-wave below the radiator. This image completes the missing half
of a half-wave antenna, and earth return currents will be induced in the ground.

116
 CHAPTER 5 VERTICALS

SHORTENED VERTICALS
It is seldom possible or convenient to erect a full-sized quarter-wave vertical for the
lower-frequency bands, although such antennas are often used on the higher fre-
quencies. For the lower frequency bands it is often necessary to look at ways of
physically reducing their length. In Fig 5.2(a) the full quarter-wave is in the vertical
plane and is shown to be bottom fed (impedance 36). Figs 5.2(b), (c) and (d)
show reducing lengths of the vertical antenna sections and corresponding
increases in the lengths of the horizontal components. The total height of the
antenna is therefore lowered and in (d), where only 25% of the quarter-wave is
vertical, the antenna is only 0.06-wavelength above ground.
The three bent quarter-wave antennas shown in (b), (c) and (d) are called in-
verted-L antennas, and they are very popular arrangements when mast height is
limited. As the vertical part of an inverted-L is reduced in length, the proportion of the
radiated power at low angles and in the vertical plane also diminishes. The horizon-
tal top section will then contribute more of the total radiation, this radiation being
horizontally polarised and at high angles to the horizon. This high-angle radiation is
a result of the antenna being close to the ground.
An inverted-L similar to that shown at (c), where the vertical and horizontal por-
tions are equal in length, should give useful vertically-polarised radiation at low
angles for both DX work and also local working within the ground wave range. The
high-angle radiation from its top horizontal half will be effective for short range com-
munications.
In Fig 5.2(e) the top half of the quarter-wave is dropped down towards the ground.

Fig 5.2: The vertical quarter-wave can have a proportion of its length bent horizontally as shown in
(b), (c) and (d). When this is done the antenna is called an 'inverted-L'. As the proportion of the
vertical section falls the vertically polarised radiation at low angles also falls, the horizontal top
giving horizontally polarised high-angle radiation. The example shown at (d) will have most of its
radiation at very high angles and will only be suitable for short to medium distance working. It will
also have a much reduced ground wave. Bending the top of the inverted-L down (e) will mean that
the antenna currents in the two sections will then tend to be out of phase and begin to cancel. At (f)
the sloping wire will behave almost like a length of unterminated open-wire feeder.

117
SUCCESSFUL WIRE ANTENNAS

It doesnt matter if the vertical section is not fully vertical; for a given support
height it may be advantageous to have a longer semi-vertical section by
sloping it slightly away from the truly vertical;
Finally, and perhaps most obviously, ensure the vertical section is as high as
possible.

The results could not have been more different between the two inverted-Ls and
are perhaps counter-intuitive, with the shorter antenna working better on 160m than
80m and the longer one working better on 80m than 160m.

FISHING FOR DX: G0GBI FISHING ROD INVERTED-L

This antenna idea was described by Glenn Loake, G0GBI, in the July 2010 RadCom.
It builds on the popular use of a fishing rod as a support for a wire vertical or inverted-
L by utilising fishing reel and line as well, as shown in Fig 5.15. The basic principle
is to use a fishing rod and reel to loft a weighted, non-conductive leader line to the top
of a tree. The weight should cause the line to hook over a branch and head for ground
level. The other end of the line is permanently attached to a 132ft (40.2m) length of

Fig 5.15: General arrangement of the fishing rod antenna system.

135
 CHAPTER 5 VERTICALS

Left: The rod support: note how the tube protrudes an

inch or so above the support pole.

Below: Attaching the feed (the antenna was not de-

ployed at this point, which is why the antenna wire is
not visible).

wire that acts as the antenna element. A stand for the

fishing rod plus a couple of guys and some wiring
completes the set-up.
The parts are quite easy to obtain. When the an-
tenna is dismantled it fits easily into a car. The only
long piece is the aluminium support tube, which could
be cut in half and then sleeve joined. It should take
less than 10 minutes to erect once a suitable tree
has been selected.
None of the parts are terribly critical, so the fol-
lowing is just a guide. You will need a beach caster
type fishing rod. G0GBI used a telescopic one, but
any kind about 8 - 10ft long will do (though you should
avoid the conductive carbon-fibre types). You will also
need a centre pin type plastic or wooden fishing reel;
a commercial fly reel will be fine. Make a connection
point in the side of the reel by putting a bolt through it
or remove the reel winder knob and put a bolt in its
place. Use a solder tag to connect one end of a 132ft
piece of thin wire; any wire can be used provided it is
thin enough to fit comfortably on the spool with space
to spare. Next, attach about a 60ft (20m) length of light
string or nylon fishing line to the other end of the wire.
A solder tag at the end of the wire provides a handy
attachment point. Wind the string on to the reel on top
of the wire.
The rod support is constructed from a length of
aluminium tube, approximately 8ft (2.5m) long and
The completed fishing rod inverted-L, 1.5in (37mm) diameter, although the dimensions are
showing the support pole, sleeve, reel not at all critical. Attached to the support is a piece of
and rod. PVC pipe of a suitable diameter to take the bottom of

136
SUCCESSFUL WIRE ANTENNAS

Length
Band (m) (ft in)

160m 39.5m 129'6"

80m 20.5m 67'5"
40m 10.5m 34'8"
30m 7.4m 24'4"
20m 5.3m 17'4"
17m 4.1m 13'7"
15m 3.5m 11'7"
12m 3.0m 9'10"
10m 2.7m 8'9"

Table 5.5: Suggested counterpoise lengths.

the fishing rod. The pipe is about 16in (40cm) long and does not need to be a tight fit
to the rod. The plastic pipe can be attached to the support rod using two jubilee clips
(hose clamps), as shown in the photo. A piece of steel reinforcing bar about 4 - 5ft
long is used as a ground stake.
The final parts to make are the feed and counterpoise. The prototype used a 10ft
(3m) length of RG58 coax with a PL259 plug on one end. The other end had an
alligator clip on the centre conductor to connect to the end of the antenna wire. The
counterpoise length in metres is calculated as 75/frequency (MHz), which allows a
bit of extra length for trimming. Table 5.5 gives suggested values for the mid-point of
the HF bands, though you may well find that trimming these by 5% or so will be better.
A single counterpoise wire per band was used, although more would probably be
better.
To deploy the antenna you need to know how to beach cast a fishing rod. If you
dont, please find someone to teach you otherwise you could injure yourself or oth-
ers. Select a suitable tree and make sure that there are no people or animals nearby
that could be hurt when you cast the leader. Trees beside footpaths are particularly
prone to people walking near them, and folk tend to get upset if you hit them with
flying lead. Respect the wildlife that may be in the tree - after all its their home!
Thread the leader through the rod loops (just like a fishing line) and attach the
weight to the end of the leader. Let a good bit of slack off the reel, ensuring it doesn't
tangle. Don't try to cast straight off the reel or a bird's nest (tangle) will result. Beach
cast towards the top of the tree. With luck the weight will carry the leader over a high
branch and fall to the ground. Pull the leader over the branch so that the end of the
antenna wire is several feet from the leaf canopy. Tie off the leader at the base of the
tree. Go back to the rod and pay out the antenna wire as you walk away from the tree.
When the wire is fully extended, set up the ground stake, slip the rod support over it
and then put the rod in the top of the support. If the antenna wire is a bit saggy then
you can go back to the tree and tighten it by pulling on the leader.
Depending on the stoutness of the ground stake and the weight of the antenna
wire, you may find it necessary to use some guys to keep the rod support upright.
Finally, connect the feed to the bolt on the reel and arrange your counterpoise.
Another method of feeding the antenna is to put an automatic ATU on the ground
at the base of the antenna with a wire connected to the driven element. The ATU earth
can then be connected to the counterpoise or even just to an earth stake. G0GBI says
that if it is windy the SWR will vary alarmingly as the tree sways about, but in practice
he has not had any real problems. He uses a small LDG auto tuner.

137
 CHAPTER 5 VERTICALS

REMOTELY TUNED INVERTED-L

If an automatic ATU is available, a practical remotely tuned multiband inverted-L is a
possibility. See Fig 5.16. As with any inverted-L, the vertical section should be as long
as possible. A good overall length to aim for would be about 86ft, although this is not
critical since the antenna system is brought to resonance with the ATU. For multiband
use, a length of around 65ft should be avoided as this would be close to a half-wave
on 40m and thus would present a high impedance to the ATU and might therefore be
difficult to match.

ti

Fig 5.16: Remotely tuned multi-band inverted-L.

160M INVERTED-L PERFORMANCE

Amateurs who wish to operate on the 160m band, but only have a limited space for
antennas, often use an inverted-L. It is arguably the best compromise antenna for
160m if you are short of space. But just how well does it work? In the May 2008 QST,
Al Christman, K3LC, described computer simulations of several different configura-
tions of 160m inverted-L antennas. The height of the vertical section of the radiator
and the length of the radials are varied in 20ft intervals from 30 to 90ft. The complete
QST article includes calculated data for input resistance, peak forward gain, SWR
bandwidth, efficiency, front-to-back ratio and front-to-side ratio for each case. Here,
we present an edited summary of this study.
For those who do not have access to a support high enough to hold up a full-size
160m monopole the choice is straightforward - either use a shortened monopole
with base, centre or distributed loading, or use a full-size /4 antenna with the vertical
portion going to the top of an available support and the rest extended horizontally to
a second support, as shown in Fig 5.17. While either technique can be used, the
second provides for efficiency and bandwidth approaching that of a full-size monopole.
The loaded antenna has lower radiation resistance, resulting in more of the transmit
power being lost in the resistance of an imperfect ground, and generally has nar-

138
SUCCESSFUL WIRE ANTENNAS

Fig 5.18: This inverted-L antenna uses

a ground screen composed of 60 bur-
ied radials, each of which is 50ft long.
The height of the vertical section of
Fig 5.17: The configuration of an inverted-L antenna (dia- the radiator is 50ft, and the horizontal
grams reprinted with permission of the American Radio portion has a length of 84.428ft, which
Relay League). resonates the antenna at 1830kHz.

rower bandwidth (unless the losses are so high that it starts acting like a dummy
load). The only downside of the inverted-L is that it requires a second support and
has some directivity - but perhaps that can be used to advantage.
All of the antennas described here were modeled using EZNEC/4 with a double
precision calculating engine. Unlike other EZNEC versions, this version of the soft-
ware is capable of simulating vertical antennas with radials buried in real ground.
The soil was assumed to be average (conductivity of 0.005 siemens per metre,
dielectric constant of 13). The radiator (both vertical and horizontal portions) is made
from 12 gauge copper wire and the radials modelled with 16 gauge copper wire. The
number of radials was fixed at 60 because it is well known that it is important for a
vertical antenna to have a good ground system. Tapered segment lengths were
used for all wires, in accordance with the most conservative NEC modelling guide-
lines. The inner segment of each radial is about 1ft long, and slopes downwards
from the base of the vertical element (at exactly H = 0) to its ultimate burial depth of
3in; the remaining length of the radial is completely horizontal.
The height of the vertical section of the inverted-L radiator was initially set at 30ft,
using 60 buried radials that were also 30ft long. The length of the horizontal portion
of the wire was then adjusted in order to resonate the antenna at a frequency of
1830kHz, after which all of the important performance data was collected. Next the
length of the vertical section was progressively increased to 50, 70 and finally 90ft,
with the tuning and measurement process being repeated each time. This entire
sequence was then carried out again, as the length of the buried radials was in-
creased in succession from 30 to 50 to 70 to 90ft.
Fig 5.18 shows what the antenna looks like when the height of the vertical sec-
tion is 50ft, and the 60 buried radials are also 50ft long. In this case, the horizontal
portion of the inverted-L had to be cut to a length of 84.428ft to achieve resonance
(input reactance close to zero) at 1830kHz.

Modelling results
The resulting elevation-plane radiation pattern, in the plane containing the inverted-
L, is given in Fig 5.19. Notice that maximum gain is actually directed opposite to that
of the horizontal section of the radiating element; if you want to beam the strongest

139
 CHAPTER 5 VERTICALS

signal to the north-east, the horizontal portion of the

L must extend towards the south-west. However,
the front-to-back and front-to-side ratios are mod-
est for all of the designs studied here, and are al-
ways less than 3dB.
Table 5.6 displays the input resistance at reso-
nance for the various inverted-L configurations. We
can see that, if the length of the radials is held con-
stant, making the antenna taller will increase the
magnitude of the input resistance. In contrast, if the
height of the radiator is fixed, making the radials
longer causes the input resistance to fall because
the ground loss resistance is lower.
Table 5.7 shows how the maximum gain of the
antenna and its corresponding take-off angle (TOA)
vary with the length of the radials and the height of
the vertical section. Notice that, for 30 and 50ft
radials, making the antenna taller always yields
Fig 5.19: This is the elevation-plane radiation more gain, at least for the range of heights dis-
pattern of the antenna shown in Fig 5.18, in the cussed here. However, with 70 and 90ft radials, an
plane containing the inverted-L wire. The front element with a height of 70ft is actually a bit better
of the main lobe is directed towards the left, than one that is 90ft tall.
opposite to the position of the horizontal sec- When it comes to elevation angles, increasing
tion of the radiator element. The peak gain is the height of the antenna always produces a lower
0.27dBi at 30.1 take-off angle, and the front- TOA, for radials of any particular length. For exam-
to-back ratio is 1.42dB. ple, an element height of 30ft generates maximum
gain at an elevation angle of 38.4 (on average)
versus a typical TOA of 30.5 for a height of 50ft. Increasing the antenna height to 70ft
drops the TOA to about 27.1, while a height of 90ft results in a peak elevation angle
of 25.
The bandwidth capability of each antenna is listed in Table 5.8. These values
were determined by calculating the standing wave ratio as a function of frequency,
with the input resistance at resonance used as the reference impedance. If the radial
length is fixed, making the antenna taller always increases the SWR bandwidth. On
the other hand, for a given radiator height, making the radials longer always reduces
the bandwidth.

Height (feet) 30 50 70 90

Radial
Length Input Resistance ()
(feet)

30 19.15 25.43 33.52 40.99

50 14.72 21.35 29.43 36.81
70 12.29 19.16 27.17 34.43
90 11.02 17.97 25.86 32.99

Table 5.6: Input resistance for inverted-L antennas at resonance, as a function of

radial length and antenna height. In each case, the ground screen is composed of
60 radials in average soil (see text). The horizontal portion of the wire radiator is
trimmed to resonate the antenna at 1830kHz.

140
SUCCESSFUL WIRE ANTENNAS

Height (ft) 30 50 70 90

Radial Gain (dBi) and Take-off Angle (Degrees)

Length (ft)
30 1.77 @ 38.4 0.41 @ 30.6 0.01 @ 27.2 +0.11 @ 25.0
50 0.82 @ 39.3 +0.27 @ 30.1 +0.51 @ 27.2 +0.55 @ 24.7
70 0.21 @ 37.9 +0.66 @ 30.1 +0.82 @ 26.5 +0.81 @ 25.2
90 +0.16 @ 38.0 +0.89 @ 31.3 +1.01 @ 27.4 +0.99 @ 24.9

Table 5.7: Peak forward gain and corresponding take-off angle for inverted-L
antennas, as a function of radial length and antenna height. In each case, maxi-
mum gain occurs in the plane containing the radiating element, and is oriented
opposite to the direction of the horizontal portion of the L.

Efficiency and other data

EZNEC has the ability to estimate the average gain of an antenna and compare this
with a theoretical lossless antenna operating in a lossless environment. Its average
gain (over all angles of elevation and azimuth) is exactly 1, or 0dB, and its efficiency
is therefore 100%, while an antenna whose average gain is -3dB must have an
efficiency of 50%. Table 5.9 provides a compilation of the efficiencies of our inverted-
L antennas, based on the computer-predicted values for their average gain.
If the antenna height is held constant, we can see that making the radials longer
always increases the efficiency. This intuitively makes sense because we expect a
larger ground screen to reduce losses in the system. If the length of the radials is
fixed at either 30 or 50ft, making the antenna taller always improves the efficiency.
The story changes, though, if longer radials are installed. For 70ft radials, maximum
efficiency is achieved when the height is 70ft (a height of 90ft works almost as well,
followed by a height of 50ft). If the length of the radials is 90ft, a 70ft vertical again
performs best, but now the 50ft tall element takes second place, with the 90ft vertical
in third position. Notice that the inverted-L with the highest efficiency of all those
tested (39.4%) uses 90ft radials in combination with a 70ft vertical section. In con-
trast, the worst antenna (which uses 30ft long radials and a 30ft tall vertical element)
is roughly half as efficient (19.8%).
Fig 5.19 showed that the inverted-L did not have a significant amount of
directionality in the azimuthal plane and this was true for all the antennas analysed.

Height (ft) 30 50 70 90 Height (ft) 30 50 70 90

Radial Efficiency (%)

Radial 2:1 SWR Bandwidth (kHz) Length (ft)
Length (ft) 30 19.8 28.2 30.8 31.6
30 57 74 97 119 50 25.2 33.3 34.9 35.0
50 44 62 86 106 70 29.5 36.7 37.6 37.2
70 36 55 78 99 90 32.7 39.1 39.4 38.8
90 32 52 74 94 Table 5.9: Efficiency of the various inver ted-L
antennas, as a function of radial length and an-
Table 5.8: 2:1 SWR bandwidth for inverted-L tenna height. In each case, the efficiency is cal-
antennas, as a function of radial length and an- culated from the average gain of the antenna,
tenna height. In each case, the reference im- as given by EZNEC. (Note that the efficiency of a
pedance for the SWR is the input resistance full-size /4 vertical, calculated by this method,
value given in Table 5.6. is only 40.6% - see text.)

141
 CHAPTER 5 VERTICALS

For purposes of comparison, a full-size /4 vertical was also modelled, using 60

/4 (134.368ft) radials buried in average soil. A height of 130.826ft was required to
obtain resonance at 1830kHz. The resulting input resistance was 38.25. The
gain was 1.15dBi at 21.8 take-off angle, the 2:1 SWR bandwidth 109kHz and the
effi- ciency 40.6%. It is interesting to see that the performance of this antenna is
not significantly better than that of the best inverted-L designs in our study, although
it is much taller and has a much larger ground system.
Computer models are imperfect representations of the real world, and cannot
possibly include all of the features that are actually present, such as buildings,
vegetation, other conductive objects, irregularities in the terrain, non-uniformity of the
ground constants and other local parameters. However, it is hoped that the informa-
tion in this study will be helpful to those who are considering the use of an inverted-
L on topband.

G3PJT 160M T-VERTICAL

The simple vertical T antenna shown in
Fig 5.20 is often recommended for 160m
use because of its low angle of radiation.
One of its principal advantages is that its
dimensions are more realistic than those
of a full-size 160m vertical. However, most
of the recommendations on making a ver-
tical T are somewhat casual, usually
couched in terms like for 160m use your
doublet with the feeders strapped to- Feed
gether, fed against ground via your ATU.
Such a nonchalant approach is unlikely to
lead to the best results. An article by Bob
Whelan, G3PJT, in the July 2009 RadCom
suggests a better way. Fig 5.20: Basic T antenna.

Fig 5.21: Modelling the T antenna in EZNEC.

142
SUCCESSFUL WIRE ANTENNAS

A T antenna erected at a typical height of between

10 and 20m and having a span of 30 to 40m will have
a feed impedance R j X of between 4 - j
190and 23 + j 150. These impedances
not only mean that some sort of antenna tuning
unit will be required to match to 50coax but
also that losses due to ground quality and the
system as a whole need to be mini- mised.
To put this in a practical perspective, G3PJTs
doublet was 17m high with a span of about 30m.
Modelling this using EZNEC+ (Fig 5.21) showed that
at 1820kHz an impedance of around 11+j 16
Fig 5.22: L matching section.
could be expected, to which an estimate of other
losses needed to be added. The TLA program
showed that
such an impedance would need a L section shunt C of 2286pF and series L of 1.8H
to the antenna, as shown in Fig 5.22. (TLA is the ARRL Transmission Line Advanced
program, bundled with the ARRL Antenna Handbook, available from the RSGB
Bookshop and ARRL.) However, when G3PJT matched the antenna with this net-
work, he found that a shunt C of 2850pF alone gave an excellent match at 1820kHz.
Such a simple arrangement is very low loss.
So much for the theoretical calculated impedance, but what was the actual im-
pedance of the antenna? G3PJT used the program l_network.exe (by R J Edwards,
G4FGQ (SK), available at http://zerobeat.net/G4FGQ/index.htm) to perform the in-
verse calculation from TLA. This showed that his T antenna impedance was more
like 14+ j 22, which of course includes all the losses. This indicated that his
ground loss was only around 3, which seemed very low, but it did raise two
questions: (a) what is the range of antenna impedances that only require a
shunt capacitor to match to 50?, and (b) what would be the dimensions of
such a T antenna?
TLA calculated that 5+ j 14, 10+ j 19, 20+ j 24and 30+ j 24would fill the

Fig 5.23: G3PJTs T antenna elevation radiation pattern.

CHAPTER 5 VERTICALS

143
SUCCESSFUL WIRE ANTENNAS

bill. To use these values as the basis in EZNEC you

Height (m) Span (m) need to insert a load that is representative of your
earth loss. In G3PJTs case, based on the above, he
11.0 50.0 estimated this at about 5and so inserted a
17.5 34.0 5 load in EZNEC. Modelling his own antenna
23.0 23.4 in EZNEC (Fig 5.23) showed that its
performance was pretty good, with maximum gain
1.49dBi achieved at an elevation angle of 24.
Table 5.10: Suggested dimension for a T
More modelling suggested the T antenna dimen-
antenna requiring only a shunt capacitor to
sions given in Table 5.10 for an efficient and simple
match to 50.
match. These dimensions may need to be adjusted
to trim the antenna to the frequency desired. Of course, ground losses will affect
these values too and should be minimised. If you want to have a very efficient and
simple matching arrangement as well as an effective DX antenna on 160m these
dimensions are the ones to go for.

THE FOLDED VERTICAL ANTENNA

One of the problems with an ordinary vertical antenna is the relatively low radiation
resistance that can lead to low values of efficiency, especially when the earth con-
nection is poor. This can be overcome to a degree by using a folded vertical antenna.
By using a length of 300 ribbon as a half-folded dipole or folded unipole,
the radiation resistance of the antenna is raised to a value in the region of 80 -
150, depending on configuration and height. This means that even when
used with an average earth system the antenna will have an efficiency of around
40%.
The version described here uses a shortened 300 ribbon section to
which is added a length of single wire. To calculate the length of the ribbon
needed, an electrical quarter-wavelength at the desired frequency must be
multiplied by the velocity factor of the ribbon used (slotted ribbon has a velocity
factor of 0.87). This

Fig 5.24: The folded vertical antenna which uses 300ribbon for most of its length. The use
of the folded dipole principle raises the feed impedance of this antenna from around 15 to
four times this figure. A reasonable match can be obtained with 50coaxial feeder.
 CHAPTER 5 VERTICALS

144
SUCCESSFUL WIRE ANTENNAS

length is less than a quarter-wave, and it must have an additional wire connected at
its end to make it up to be an electrical quarter-wavelength. This technique is very
similar to that used when constructing folded half-wave dipoles from 300
ribbon (see Chapter 3).
The folded antenna illustrated in Fig 5.24 is designed for 3.7MHz operation and it
only needs 10.6m (35ft) high supports. The efficiency of the antenna is proportionally
higher than a single wire vertical because the length of its vertical section is in-
creased as a proportion of the total quarter-wavelength. A minimum of six buried
radial wires, each being at least a quarter-wavelength long, are recommended for a
suitable earth system, although with the limitations of many garden plots this may
not be achievable. For a given length of wire, it is better to use many shorter radials
than fewer longer ones. Versions of this antenna may be scaled up or down for use
on other bands.
The step-up of feed impedance brought about by using this folded dipole tech-
nique allows the use of a 50 coaxial feeder. The greater distance
between the vertical part of this antenna and any buildings etc, the more effective
the antenna will be for low-angle long-distance communication.

THE QUARTER-WAVE SLOPER

Sloper antennas have become popular as low-band antennas that can be erected
alongside an HF beam on a tower. Although sloping dipoles as described in Chap-
ter 3 are very similar, these slopers comprise a single section and are usually
classed as a form of vertical antenna. They provide a little directivity and can be
erected very easily, making them an ideal antenna for many stations.
The quarter-wave sloper or half-sloper antenna is really half of an inverted-V
dipole. A quarter-wave antenna is normally arranged to be bottom-fed, which means
that the maximum radiation is at the base of the antenna. By inverting the feedpoint of
a quarter-wave as shown in Fig 5.25(a), the position of the current maximum and
therefore the greatest radiation can be moved to the top of the antenna.
The ground against which slopers are fed is usually the metallic mass of the
support tower, although for this to be as efficient as possible, a good low resistance
ground at the foot of the tower remains important. The slope angle, L, of a half-
sloper is usually 45, and its maximum radiation is in the direction of the wire away

(a) (b)

Fig 5.25: (a) A quarter-wave sloper antenna used with a metal tower; (b) Using a quarter-wave-
length of wire which is almost vertical to replace the metal tower. The feed impedance of the sloper
depends upon several variables, one being the angle L.

145