A Simplified Directional 40m Antenna (Part II)
Bob Whelan G3PJT completes his design for a 40m directional antenna system.
For the past two years I have been experimenting with parasitic vertical beams for 40m. I described a threeelement triangular array with some of the thinking behind it in Part 1. As part of the computer modelling work associated with that antenna a number of alternative arrangements of two elements were investigated. The antenna to be described here is a further simplification of the basic concept of a vertical beam with a common earth system and consists of two elements with a single support mast. It is a pretty low profile design.
Modelling
One of the simplest forms of a multi-element directional antenna is one of two elements spaced at between 0.1 and 0.5 of a wavelength. One element, the driven element, is driven directly from the transmitter while the other, excited through mutual coupling, is called the parasitic element. If the relative RF currents flowing in these elements are in a particular amplitude and phase relationship then a directional radiation pattern results. Depending on the relationships then the parasitic element can be tuned to behave as either a reflector or as a director.
The relationship between antenna gain, beamwidth and element driving impedance is complex, which means that computer modelling is essential to shorten the time taken in developing a design and in any subsequent practical performance tuning. The approach taken here was to fix the approximate dimensions of the antenna and to vary the loading in the elements to get an optimum RF current relationship and radiation pattern. Computer modelling with EZNEC seems to be good enough such that local environmental factors do not take any practical design outside the range of reasonable adjustment. Fig. 1 shows the basic dimensions of the chosen arrangement arrived from the experiments in Part 1. The parasitic element was tuned to be a director.
The azimuthal pattern at 20° elevation of this antenna is shown in Fig. 3. This shows a useful gain of about 5.9dBi or about 3.9dBi over a ground mounted vertical. The main forward lobe is 137° with the parasitic element tuned as a director. The parasitic tuning capacitor would be about 170pF and the driven element impedance about 16Ω and inductive, though this will vary with the precise setting of the parasitic capacitor. Although the antenna radiation direction could be reversed by tuning the parasitic element as either a reflector or a director, it’s simpler to use a switch box, which just swops the two elements from driven to parasitic.
Orientation
Many of us do not have much choice as to the orientation of our antennas. However, for those who do it is worth considering what might be the best orientation for this antenna. The −3dB width of the forward lobe is about 135° and the width of the maximum front-to-back (F/B) is about 60° and for better than 10dB F/B about 110°. You may choose to orient the antenna such that the maximum F/B is oriented to Europe so that when beaming to USA etc. you have best rejection of Europe. Or you might, for example, have an area of local noise which you would like to attenuate. In my case I decided to orient the antenna such that its plane was 80 – 260°. This attenuates Europe when beaming west but also places the forward lobe over the Far East and SP Australia when beaming east. I found that I have a local noise source to the west of my QTH so at least it is quieter when beaming East.
Construction
Two identical lengths of 2mm enamelled wire 12.23m long were fitted with a insulator (a) at 5.83m from one end and a second insulator at one end (b). Each of the two elements were hung from the top of an 11m mast with around a 1m length of cord. An 8m length of cord was fastened between the two elements to fix the element spacing at 8m when the elements were spaced apart. The height of the insulators (a) was about 5m above ground. The lower lengths of the elements were brought back to the switch box and the ground radial centre.
Ground Radials
All verticals need a ground radial system in order to work. As described in Part 1, the ground radial system comprised 50 radials of insulated thin wire varying in length from 5 to 8m long (30 of ~5m would probably have been enough). These were pinned into the grass surface using best bent wire and large paper clips. The radials were terminated at the centre with crimp terminals and bolted to an aluminium plate with stainless steel set screws.
Switch Box
The switch box, Fig. 2, allows the elements to be tuned and the direction of radiation reversed. The circuit is shown in Fig. 4. The relays are 26V vacuum relays but open frame 10A contact rating relays would work just as well. At 400W about 5A of current flows and voltages are only about 100V. Capacitors were wide-spaced trans
mitting air variables found at various junk sales.
Tuning
Symmetry is important so each element was measured for self-resonance and impedance at resonance. The elements both measured 5.9MHz at resonance with an impedance of 26 to 30Ω (ELNEC modelled 6.1MHz and 26Ω).
Initial tuning of the parasitic element was undertaken by nulling the signal from a vertically-polarised low power signal in line with the antenna. This was a quick and easy way of getting close to the optimum setting. As my antenna is orientated 80 – 260°, then I could null out some signals from Europe too. Taking advantage of the ARRL CW contest in February, I was also able fine tune the parasitic element tuning to null out some of the stronger East Coast stations (K3LR, K1RX, etc.)
Using my digital phase and amplitude meter, the readings of current magnitude and phase differences for this setting were 1-2dB and −162°, which compared well enough with ELNEC figures of 1dB and −156°.
The matching of the driven element to the 50Ω feedline was done by first nulling out the inductive part of the input impedance making it 22Ω resistive. This was then transformed to 50Ω using a quarter-wave impedance transformer made from a paralleled pair of 75Ω transmission lines [2]. The matching of the driven element did not affect the parasitic element tuning even though the tuning of the parasitic element did affect the driven element impedance.
Performance
The availability of WebSDR receivers means that some qualitative measurements can be made of performance. Although I had used the Twente WebSDR in 2019, I tried to use some of the more distant ones in the USA and Russia.
The results were only qualitative, going from audible to inaudible when I changed the direction of beaming.
With the stronger signals from Europe I have measured the front-to-back (F/B) on a range of signals and, given the vagaries of propagation, the rejection of signals from most of Europe is impressive. And consistent with the azimuthal pattern, central parts of Europe (OK, DL, SP, etc) range from 10-20dB. For EA and F, which are more off the side of the antenna, the F/B falls to 0-6dB as expected.
Bandwidth
Table 1 shows the result of modelling the gain and front-to-back bandwidth of this antenna.
Measurements showed that the current amplitude did not change over 7.0-7.2MHz but the phase difference varied from −158° at 7.0MHz, −162° at 7.020MHz and −178° at 7.1MHz as could be concluded from Table 1.
So, depending on your operating preferences you might want to set the centre frequency to suit.
Summary
The antenna described here could readily be scaled for 30m. It represents a very simple antenna design, which is inexpensive to build and offers good performance, certainly better than any simple quarter-wave vertical on 40m.
I do not have a support tall enough for an 80m version but it should be a good performer on there too. For more background, readers are encouraged to look at Part 1 and the references.
References
1. An HF phase detector. G3PJT, RadCom April 2019.
2. ARRL Antenna Book, 18ed., p.24-12. And: https://tinyurl.com/ya9y5wpl
3. RSGB List of WebSDRs at: https://tinyurl.com/y6c3ymok