Addendum and Errata to Backyard Antennas

26th February 2002.

The Moxon Rectangle antenna shown on page 99 created a lot of interest due to the fact that it is simple to build, light weight, compact and appears to perform well. It has been built by several readers who report good results.

The configuration of this antenna originated from W4RNL's website although I devised the multiband arrangement, as shown in Fig 6.18. However, although this antenna works well as a single band antenna, tests done subsequent to publication indicate that the antenna does not appear to function as a multiband antenna as shown. This is probably due to the coupling between parasitic and driven elements being more critical than in the VK2ABQ, or the Double-D antenna shown in Fig 6.21, where a multiband configuration appears to work when tuned for optimum performance.

The original VK2ABQ antenna is a square structure, see Fig 1. 

Fig 1: The original VK2ABQ antenna structure compared with the G6XN and the W4RNL. The G6XN has a centre section spacing of the elements of around 0.17 wavelength spacing, while the W4RNL has element spacing further to 0.14wavelength.

The centre portions of the driven element and the reflector are quarter wavelength apart although the tips the elements support each other using insulators. These insulators are constructed so that the tips of the elements are 6mm (1/4in) apart and, according to the original description [1], capacitive end (voltage) couples the driven element to the reflector, as opposed to the coupling that occurs on a Yagi. The gap between the tips of the elements is described as 'not critical'.

Multiband editions of this antenna were constructed without any known difficulty.

G6XN [2] changed the structure from a square to a rectangle thereby reducing the centre section spacing of the elements from 0.25wavelength to 0.17 wavelength, which resulted in improved gain and directivity.

C.B. Cebik, W4RNL [3], reduced the element spacing further to 0.14wavelength and obtained yet more gain and improved directivity. This antenna he called the Moxon Rectangle and is the one described in on page 101. The downside of this higher performance is the multibanding problem.

Tony Box, G0HAD, built a G6XN antenna back in 1993. The overall size of this structure was 6.1m x 3.8m (20ft x 12ft 6in) as recommended by G6XN; this antenna out performed his previous commercial mini beam. I used a computer model (EZNEC) to check these dimensions, which produced the antenna dimensions 6.92m x 3.8m (22ft 10in x 12ft 6in). The reason for this discrepancy is that G6XN used loops in the elements at the element support points. This makes for a more compact antenna but increases the mechanical complexity slightly.

Below are formula to calculate dimensions for the G6XN antenna (without loops at the element support points).

The size of the antenna, using dimensions A and E from Fig 2 (Fig 6.18 in the book) can be found by:

Fig 2: The Moxon Rectangle antenna. The dimensions A, E and C are described in the text.. Element lengths are calculated using the formula in the text

G0HAD found by experiment that the gaps between the tips of the elements needed to be larger than published [2]; being 560mm (22in) for 20m, 380mm (15in) for 15m and 250mm (10in) for 10m, (values for C in Fig 6.18). This is confirmed by computer modelling, which shows an improvement in the front-to-back ratio when these gaps are increased.

To work out the length of each support (cane or fibreglass rod) structure required the formula is:

56.09/f = length of diagonal in metres. 184/f = diagonal length in ft.

The units of ft are indicated as a decimal number. To convert 12.5ft to feet and inches multiply the part after the decimal by 12, e.g. 0.5 x 12 = 6; 12ft 6in.

If you wish to operate on several of the HF bands and you don't have an ATU then parallel dipoles may be the answer. An example of such an antenna is described on page 20 and 21. However if more than three parallel dipoles are used, using the construction method shown in Fig 2.8, the structure becomes complicated and difficult to manage.

Stewart, GM4UTP, uses a parallel dipole design that tidies up all the wires of the multiple dipole. This arrangement uses the lowest frequency dipole to support the higher frequency dipoles using spacing insulators made from 11mm plastic electrical conduit. The construction is shown in Fig 3.

Fig 3: The GM4UTP multiband antenna. The detail shows the larger spacers to accommodate 6 wires. The outer spacers are progressively shorter with holes drilled for 5, 4, 3 and two wires respectively. The 24MHz dipole is not shown but the lengths are 2.84m (9ft 4in)

The antenna is configured as an inverted V with the weight of the centre insulator and the 1:1 balun mounted on a 10m high aluminium scaffold pole. Low centre band SWRs are possible if some time is spent tuning each dipole. This can be achieved by arranging the ends of the elements so that they are clear of their support insulators by about 200mm. The dipole lengths can be reduced or increased by folding back the end and securing with plastic tape.

The resonance of these dipoles can be interactive - when you adjust one it effects the resonance of the other so be prepared to have to re-resonate elements.

A similar arrangement is described by K0GPD [4]. In this design all the spacer insulators are made the same length and a nylon cord is run from the end of the highest frequency dipole to the end of the multiwire sections of the antenna to improve mechanical stability.

Page 63, Heading to Table 4, G3OLB should be G0LMJ

References, Page 67, G0OUJ should be G0LMJ

Page 115. The dimensions for Table 7.1 are for one side of the quad rather than the whole element length as shown in the UHF "circular" quad in Fig 7.16 on the same page.

Fig 7.22, page 120. This drawing implies that Yagis 2 and 4 are fed 180 degrees out of phase with Yagis 1 and 3. The driven elements of all Yagis must be fed in phase.

Fig 9.7, page 151. The first value of transmission loss should be 0.5db rather than .05 as shown. 10m, in Fig 9.7 title should be 30m.

FURTHER READING

[1] 'VK2ABQ Antenna', Fred Caton VK2ABQ Electronics Australia, October 1973,

[2] HF Antennas for all Locations, L.A. Moxon, G6XN

[3] http//www.cebik.com/. Note that this site replaces that shown in the reference on page 107.