Magic two-element antennas for KV - 6, Double Delta beam according to G3LDO

Honza Bocek OK2BNG provided us with incredible help - he provided us with his articles and took it upon himself to manage a new section on two-element antennas. He suggested contacting Peter Dodd G3LDO and Less Moxon G6XN who could answer your questions via the forum. Thank you! Contact the authors of the article: Jan Bocek OK2BNG, janbocek@mail.tele2.cz, Jiří Škácha OK1DMU, skachaj@volny.cz

Two-element KV directional antenna systems aroused the interest of amateurs throughout the last century, and this interest will probably not disappear in the future either. This is evidenced by the many lectures held every year at the symposium in Dayton [47]. As a reminder, we present in fig. 1 shapes and frame dimensions of the most used two-element KV antennas, which were already described in this series [48-49].

HB9CV, Rudolf Baumgartner designed a super gain phase system with element spacing of 0.125 LMBD. Both elements are full-size with a length close to 0.5 LMBD. The advantage is the all-metal design and the possibility to connect any power supply [48].

VK2ABQ, Fred Caton tried to minimize dimensions while maintaining electrical properties; finally arrived at a square antenna layout of 0.25 x 0.25 LMBD. The wire radiators were suspended on bamboo supports, the solution for bending the ends of the elements is important here. From a mechanical point of view, the system was difficult to build, because the "rubbery" nature of the system manifested itself with less rigidity. The input impedance was high [38].

G6XN, Less Moxon collaborated for several years with Fred VK2ABQ and the result was an antenna with an input impedance of 50 OOHH in the shape of a rectangle. It is very popular in the world under the name Rectangle beam or Moxon's beam. The tubular design was described in the 3rd part of this series [48].

W4RNL, L. B. Cebik is one of the biggest publicists in the field of antennas. For antenna experimenters, his pages are inherently among the favorites. He optimized the Moxon antennas using various antenna programs, reduced the distance between the elements to 0.14 lambda and optimized the critical connection between the ends of the elements [39].

G3LDO, Peter Dodd is another well-known columnist in the field of antennas [51-54]. He devotes a considerable part of his work to the issue of small rotating directional antennas for HF. In an attempt to preserve the full length of the elements, he used a geometric shape of the elements similar to a triangle, therefore he named the antenna Double Delta - DD antenna for short. He achieved the reduction of the overall floor plan dimensions by bending the wire ends of the elements towards the mast. The antenna is shown in fig. 1. Its older wire version had a mutual distance between the elements of 0.3 LMBD, the version created by further development in the tubular design with the extension of the elements with wire conductors has already substantially reduced dimensions - the mutual distance between the elements has decreased to 0.16 LMBD. As a result of the bending of the ends of the elements, the total length of the element is somewhat larger than that of a classic dipole. Peter abandoned the earlier effort to maintain an input impedance of 50 OOHH; at a distance between the elements of 0.16 LMBD, it reached a classic average input impedance of 28 OOHH, similar to most yagi antennas.

Description of the Double Delta - DD-beam antenna

The antenna has undergone its development over the past twenty years; in the following, we will focus only on the implementation of the basic part of the antenna from metal tubes according to fig. 1b, see also fig. 3. The boom and both elements form a "hook", similar to the constructions of other directional antennas. One element – ​​the radiator – will be divided as a dipole. Peter, G3LDO, uses both elements undivided and supplies the radiator with a shunt, but we do not recommend this solution, because measuring and tuning the element into resonance can then be more complicated. Copper insulated conductors are connected to the ends of the tubular elements, which are stretched by a non-conductive cord towards the mast, roughly following the course of the edges of the pyramid (see Fig. 1a and 1b).

From an electrical point of view, the antenna radiator is tuned to resonance in the middle of the band, and the antenna is then fine-tuned to the maximum front-to-back radiation ratio F/B by adjusting the length of the reflector. The gain in the forward direction is relatively constant, decreasing somewhat with increasing frequency. The front-to-back F/B ratio is quite dependent on the frequency, but in practice it has no significant meaning. The situation is shown in fig. 2, where the horizontal radiation diagram obtained by modeling [53] is plotted, see [53] for a similar result. The forward gain value is less frequency dependent than the reverse gain value. The ends of the elements are close to each other with mutual bonding, similar to the Moxon beam. Although the coupling is looser, it still affects the resonant frequency of the antenna.

Antenna production

Tab. 2. Approximate experimentally verified dimensions of the DD-beam antenna, the designation corresponds to fig. 3. The total length of the emitter is LZ = 0.576 LMBD = 173/f, the total length of the reflector LR = 0.5875 LMBD = 176.25/f. The dimensions of the gamma adaptation elements are given for R 28 OOHH. Formula Dimension [m] 21 MHz 7 MHz A 71.2/f 0.23 LMBD 3.36 10,10 B 71.2/f 0.23 LMBD 3.36 10,10 C 51.4/f 0.17 LMBD 2.42 7.29 D 50.9/f 0.17 LMBD 2.40 7.20 E 52.17/f 0.173 LMBD 2.46 7.40 F 29.9/f 0.10 LMBD 1.41 4.24 G 56/f 0.186 LMBD 2.64 7.94 gamma 15.65/f 0.052 LMBD 0.74 2.22

The design and actual manufacture of the antenna includes a number of elements, subjects and possibilities, detailed in the previous parts of the series. We will therefore not repeat them in the following text and will only draw attention to some specific moments.

Tab. 4. Cable lengths of the transformation l/4 line 50/28 ohm (in m). Cable 75 ohm, k = 0.66. Band Length l [m] 7 MHz 7.05 14 MHz 3.84 21 MHz 2.35 28 MHz 1.68

The antenna can be powered typically in two ways: Antenna with a divided powered element according to fig. 3 has an impedance at the edges of the band of around 28 ohms - the measured frequency course of the impedance for the antenna for the 40 m band is shown in tab. 3. We will use for power supply transformer impedance – vf line with an electrical length of 1/4 LMBD – see fig. 4. The lengths of the transformation section for cables with a solid dielectric with a shortening coefficient of 0.66 are shown in tab. 4. We will create the transformation line by connecting two sections of the 75 OOHH cable in parallel, so that its resulting impedance will be 37.5 OOHH. One end of the line will be connected to the terminals of the split radiator, the other end of this line can be directly connected to a standard coaxial cable 50 OOHH.

Tab. 5. Dimensions of the gamma adjustment section according to fig. 5 Band A [mm] B [mm] C [pF] D [mm] L(C) 14 MHz 1110 1310 150 85 640 21 MHz 740 940 100 100 510 28 MHz 550 600 75 100 220

Another possibility is, according to the original source, to power the emitter using the gamma section - see fig. 5. Dimensions for this arrangement listed in tab. 5 are calculated for the conversion of 28/50 OOHH. It is worth paying attention to the solution of the tuning capacitor, which is formed by the inner conductor with the PE insulation of the RG-213 cable left. This conductor is inserted into a pipe with an internal diameter of 8 mm. Between the conductor and the pipe we measure a capacity of about 200 pF/m. By moving the conductor in the tube, we change the capacity of this capacitor and thus we can compensate the reactive component of the impedance to a minimum value. By setting the dimensions A and L, we are looking for an optimal SWR. We must seal the pipe against moisture.

Antenna settings

First, we tune the radiator - a divided element made of tubes with connected wires. Compared to the dimensions from the table, we first choose the actual total length about 10% larger, so that we have something to shorten. We fix the element on the boom and raise it to a height so that the wires hang freely vertically down and their ends are at least 3 m above the ground. On the conductors, for example we mark the length marks with insulating tape so that we do not overdo the shortening of the wires during adjustment. We will look for resonance so that the length of the entire element is close to the value of 0.576 LMBD (LMBD corresponds to the center of the band); resonance is usually found in the upper part of the band or slightly above the band. An SWR meter is sufficient for this measurement. For now, we may not be interested in the absolute value of the SWR, which can be (compared to the normalized value of 50 OOHH) even around 1.2, but the position of its minimum and the frequency course.

After fine-tuning the radiator by changing its length, we check by calculation, the total length seems to correspond to the relation 0.576 LMBD. If this is not the case (it may be caused by different pipe diameters and a different extension cord), we must adjust the formula for calculating the reflector so that the length of the reflector is in the ratio of 0.5875/0.576 to the length of the radiator, i.e. that the reflector is about 2% longer. In practice, the length markers on the ends of the wires will serve us well - we mark whole meters and tens of centimeters for the last meter.

The second element, the undivided reflector, will therefore be approximately 2% longer than the radiator. We fasten the reflector in the correct position on the boom, let the extension wires hang down freely for the time being, raise the antenna to the height again and measure the SWR again. The minimum will be slightly lower in frequency than in the case of the radiator itself, the inherent SWR value (compared to the normalized 50 OOHH) will deteriorate to about 2.3 to 2.5; but that's okay, because the bond between the beam elements is already showing, causing a drop in impedance. If, on the other hand, the SWR was still relatively good, that is about 1.2, it would mean that the reflector is long or that the distance between the elements is not 0.17 LMBD. Those who have measuring instruments RF1, VA1 or MJF-259B can measure Z and X - an example of DD-beam measurement for 40 m is in tab. 2. Let's note that the Ra value is small and the Xa value is very low during antenna resonance. By adjusting the dimensions of the antenna, a zero reactance component can be achieved, but it is laborious and zeroing Xa can only be achieved at a single frequency. At this stage, when we have not yet definitively attached the extension wires, we will therefore be content with Xa values ​​not exceeding 20 OOHH at the edges of the band. Although the data are only indicative, this approach is completely sufficient for practical purposes.

We have the antenna with the wires hanging freely downwards roughly tuned, those who are impatient can do the first one as well QSO. The next step is to adjust the antenna to its final state, i.e. with the wires firmly attached to the mast as shown in fig. 3. The shape of the antenna allows for two arrangements: The wire extension can be attached upwards above the boom to the extended part of the mast, or downwards to the mast. The first variant simplifies the issue of rotating the antenna, because we can already use a weaker tube above the plane of the boom; in the second case, the pipe must be stronger.

First of all, we tension the extension wires of the radiator with the help of cords, and for the time being, we twist the ends of the reflector wires into a ball so that the reflector does not affect the measurement. Again, we measure the resonance of the radiator and adjust the length of its conductors so that the resonance frequency is slightly above the band. Then we stretch the conductors of the reflector to the mast and check the resonant frequency of the entire system, which should already be in the required band. For the antenna for 40 m, e.g. the length of the radiator was 24.5 m and the resonance frequency was 7150 kHz. The length of the reflector was then 1.02 times larger, i.e. 24.5×1.02 = 25.0 m.

Example of experiment for 40 m band

From the dismantled Delta Loop for 15 m, 4 pieces of trapped elements with a diameter decreasing from 35 to 16 mm, each with a length of 5.1 m, remained. The ends of the tubes with a smaller diameter had an M6 threaded hole for connecting the wire chord of the original Delta Loop. For the new beam DD at 40 m, these pipes were used without modifications (later, during a storm, it became clear that with these dimensions, the 16 mm pipes are at the limit of usability, the ends of the elements bent slightly at the point of weakness). Two of these tubes were attached in isolation to a plate made of insulating material (divided radiator), the other two to an aluminum plate (undivided reflector). With four additional clamps, both elements created in this way were attached in their center to the boom of the OWA antenna for the 15 m band. For the first experiments, the length of the extension wires was 6.4 m. The antenna was raised 3 m above the ground; it resonated slightly below the band and had an impedance of 24 OOHH. For the original OWA antenna, the mast was extended by 8 meters above the rotator to anchor the long 15 m antenna boom. Therefore, it was chosen to fasten the ends of the extension wires of the DD-beam upwards. After fixing the wires at the top to the mast, the resonance compared to the wires hanging freely down changed by approximately 300 kHz towards higher frequencies, to 7.35 MHz. Therefore, it was necessary to extend the length of each radiator conductor to 7.17 m and the length of each reflector conductor to 7.4 meters. The total length of the radiator was 24.54 m and the length of the reflector 25.0 m. A transformation line LMBD/4 (length 7 m) was connected to the radiator, made of two parallel cables 75 OOHH with a solid dielectric and a shortening coefficient of 0.66, as indicated in fig. 5. The cable was twisted into the shape of a coil and thus forms a vf choke. Cable joints must be made carefully and treated with vulcanizing tape against moisture. The final height of the antenna is 20 m above the ground, the tops of the extension wires are at a height of approx. 28 m, and the average height of the antenna is LMBD/2.

Operational experience

For comparison, a dipole 10 m above the ground, an inclined dipole from 18 m oriented to the west, a Moxon beam fixed in the E-W direction 10 m above the ground and a vertical beam 30 m high were used. Opinions on the situation in the 40 m band and experiences are summarized in the part of the series on Rectangle beam for 40 m [48]. It really is a magical area. With DD-beam, you will experience completely different experiences than you are used to. A turn signal for 40 m is not such a matter of course, as, for example, for the 21 MHz band. You will certainly experience pile-ups very soon, not only from the peripheral parts of the EU, which is relatively common, but also such an hourly pile-up from JA and that is a great experience. You will think you are on the 15 m band.

For example, during a very long QSO in the 7 MHz band OK2BNG with JA2DPC, Setsuko asked in great detail about the dimensions of the antenna, because she was using a Moxon beam on 40 m under the brand N8YL and A35PC. She only knew the older, wired version of DD beam. She sent a long letter asking for photos. She herself currently uses a rotating dipole on 40 m.

However, it is also necessary to admit that the extra dBs were often not heard during the comparison, and the differences between the antennas can often be understood as subjective. The results of the comparison depend on several factors, e.g. on the propagation conditions producing a signal under different angles of incidence, antennas of opposing stations and on their IQ. In all cases, however, the DD-beam is a turn signal, showing distinct directional radiation. If we read, for example station to an oblique dipole with a power of S5, i.e. at the noise level of the 40 m band, then after re-directing the DD-beam, the signal is about 1 S better and therefore more readable, and communication is possible. Even if the gain of this antenna compared to a dipole is only 3-4 dBd, it should not be underestimated. The main advantage is the presence of a lower radiating lobe, which is simply not present in other wire devices. The difference in the signals can be around 20 dB and that already means that we hear or call stations that we simply do not know about when using the wire. And therein lies the advantage of even a non-optimally executed two-element turn signal compared to a straight wire, which does not exactly show significant directional radiation.

The described antenna is structurally simpler than hexbeam, described in the last part of the series (its advantages are, however, the attractive input impedance value of 50 OOHH and a slightly better gain). A DD-beam can be made in one to two weekends with only basic workshop equipment, without additional help and with a little luck.

So what is the strength of this antenna? Let's look again at fig. 1 and we will see that in the dimensions and then in the radiation resistance. DD-beam with dimensions corresponding to a classic antenna on 15 m is functional in the 40 m band. It's unbelievable, but when the finished and well-functioning OWA antenna for 21 MHz was used as the construction basis for the DD-beam for the 7 MHz band, it was black and white. The finished DD antenna for 15 m has dimensions corresponding to the classic antenna for the 6 m band.

Other types of two-element antennas have approximately similar electrical and radiation parameters. From a theoretical point of view, there is of course a lot of room for various discussions, but practice tends to be more merciful. A two-element system will remain a two-element system, even if phased systems (HB9CV) or "coupling" systems (G6XN) have their advantages. In the end, the common prerequisites are mainly a solid mechanical design and the possibility of somehow getting the antenna into the space.

The DD antenna is taught to all experimenters who, for various reasons, do not buy commercial antennas, but it also enables the construction of those for whom, for various reasons, "normal" dimensions simply do not work. For example, the rudder for the 20 m band is as big as a shack; The DD antenna is four times smaller. And that's already worth the experiment.

What to say at the end of the series?

A total of six articles in our series were devoted to simple directional antennas for KV bands. The main reasons and our internal arguments for the fact that we devoted time, energy and work, connected with concentration, verification of essential information and finally with writing for interested parties and readers, are based on two basic statements:

1. Even the simplest directional antenna, especially if it is rotatable and placed at a suitable height, opens up new horizons for amateur operation and moves the station into a completely different technical and operational category. It makes available an area of ​​the world of communication, often inaccessible when using improvised or otherwise rigged antennas, especially "somehow cut long wires", brings previously unknown experiences and connections with stations that would not have happened otherwise.

2. Complicated, heavy, large, optimized and professionally manufactured directional antenna systems requiring massive masts, massive rotators, a vast plot of land, intensive inspection and maintenance, insurance (remember this year's storms), etc. they can also have very good electrical and communication parameters when placed appropriately; but they usually also have an appropriate price. However, those who do not have the mentioned possibilities or who derive satisfaction from equipping themselves with a self-made directional antenna have the prerequisites to move up the station ranking. Modern two-element directional systems, especially in designs with suitably reduced dimensions, offer rich and interesting possibilities for this, especially today, when it is no longer a big problem to play with computer models or to measure relatively delicate electrical parameters with generally available antenna analyzers. It was not intended to reject wire antennas. For LBDXing, they are hard to beat when placed at a sufficient height. They exhibit radiation lobes that can illuminate the Earth from BY to KW6 better than the DD-beam. But that would be about something completely different.

In the end, maybe just a modest wish: It is clear that there are operators equipped with rich knowledge, rich experience and rich possibilities of realization or use of high performance antenna systems. For you, the information from the series was not very interesting. But there are also many others for whom the individual parts could be a source of inspiration, provide topics for searching for the necessary information and encourage the determination to do something to improve their equipment so that they can be better used on amateur bands. Believe that the peace of mind over mastered problems is really worth all the effort. We wish everyone good luck, success and satisfaction.

Literature
[47] www.cebik.com/FDIM, Dayton 2002
]48] Jan Bocek, Jiří Škácha, Magic two-element directional antennas for KV – 3, RA 3/2002
[49] Jan Bocek, Jiří Škácha, Magic two-element directional antennas for KV – 4, RA 3/2002
[50] http://www.cebik.com
[51] Peter Dodd, G3LDO, Wire Beam antennas and the evolution of the G3LDO Double-D, RadCom, 6/7 – 1980
[52] Peter Dodd, G3LDO, Further Evolution of the G3LDO Double-D antenna, RadCom, 4/1990
[53] Peter Dodd, G3LDO, The Antenna Experimenters Guide, RSGB 1991,1996
[54] Peter Dodd, G3LDO, Backyard Antennas, RSGB 2000, 2002
[55] Peter Dodd, G3LDO, Review of the MQ2 Mini-Beam Antenna, Practical Wireless, 8/1999