A homemade 5-band Hexbeam antenna for 10, 12, 15, 17 and 20 m

                                                        J-F. Zürcher, HB9MCZ

                                      updated April 7, 2009: new broadband hexbeam

 

After reviewing dozens of antennas, I was convinced that the hexbeam represents the best compromize in terms of performance, dimensions, weight and wind load. So I decided to make my own one ... if possible with some original contributions! I took a long time to study and compare the various descriptions found on the web, among them:

[1]   http://www.hexbeam.com/     the excellent commercial version of Mike Traffie

[2]   http://www.leoshoemaker.com/hexbeambyk4kio/general.html     a very detailed description of the construction of a hexbeam

[3]   http://www.dl7io.de/reflectedw/     a lot of informations by a "hexbeam fan"

[4]   http://www.karinya.net/g3txq/hexbeam/index.html     understanding the hexbeam, with a lot of informations and simulations

My main source of inspiration has been reference [2]. Many thanks to his author, K4KIO !

 

As I have access to various machines (lathe, milling machine, ...), I decided to make a "semi-professional" realization: a rigid central baseplate made of two aluminium plates (6 mm thick!) holding aluminium tubes on which the supporting rods will be fixed. The mounting on the mast is done using a pair of brackets originally used on a Yaesu rotator (bottom of Picture 1). All screws and hose clamps are exclusively stainless steel. For the rods, I used 4 meters long 4-section fishing rods (the thinnest section has been removed, the remaining 3 sections measuring approx. 3 meters long). Picture 1 shows a detail of the antenna base:

Picture 1: the antenna base with the fishing poles mounted, the central post and its epoxy-fiberglass isolator.
A pair of connections to the 10m radiating element can be seen at the top of the picture.

 

The whole hexbeam concept is extremy elegant, excepted the feeding section in most designs I found on Internet: I don't like the assembly of short coaxial cable section, which may become a problem in case of heavy rain or snow. This is the reason of the particular concept used in my realization: the central post itself is the coaxial feeding line for all 5 bands of the hexbeam. It is realized as a square coaxial line made of aluminium profiles, as shown in the next pictures:

 

Picture 2: a detail of the central post forming the square coaxial feeding line. Inner dimensions of the tube: 26x26 mm,
central square part: 10x10 mm, impedance about 54 Ohm (I was not able to find aluminium profiles with the dimensions
required to obtain exactly 50 Ohm)

 

Picture 3: detail of the top of the central post, showing the SO-239 connector, the top PVC cap, a pair of connections to the wires (20m band)
and the transparent (polycarbonate) profile for holding the reflector wires. The center square conductor is ready for being introduced into
the external square tube; the small brass screw will then be soldered to the SO-239 pin. The screw connected to the center conductor (top of the
picture) is insulated from the external tube with glass-epoxy parts specially manufactured. The other screw (bottom of the picture) is connected
directly to the external square tube.
 
 
Once the square coaxial feeding line assembled, its characteristic impedance has been measured with an Agilent 86100B/54754A TDR. Next picture shows the result: inside the line, the impedance is about 53.9 Ohms, and the discontinuity due to the transition with the SO-239 is clearly visible at the left of the picture. This discontinuity looks important, but one has to remember that the TDR has a very short rising time (less than 40 ps!), extending in fact the measurement well inside the GHz range! With the frequencies used here (max. 30 MHz), this discontinuity is neglectible.
 
TDR measurement of the square feeding line (central post)
 

 

Picture 4: the top of the central post during construction. The two connections to the wires (20m band) are clearly visible,
as well as the isolating (transparent) piece for the reflector wire. The green Nylon lines at the top are only used during
construction to bend the fishing rods to the proper shape; they will be removed (excepted two of them) once the antenna
is equipped with all its conductive wires.

 

As it was very difficult to obtain the special copper wire without insulation, standard isolated wire with a section of 1 mm2 was used. All wire length (as given in reference 2) had to be corrected to take into account the new wire diameter and the presence of the isolation. The correction factor has been determined experimentally with Automatic Network Analyzer (HP8720D) measurements made at the LEMA laboratory.

The antenna has then been equipped with the wires for the various bands, starting with 20, 17 and 15m bands. Preliminary measurements have been made with the antenna mounted about 6 m above ground on a small fiberglass telescopic mast. The last two bands (12 and 10m) have then been added, and all antenna parameters re-measured in the same conditions. These measurements have shown that the influence of the two added bands was very small on the first three bands implemented.

Finally, the antenna has been mounted in its final destination, and the SWR on all bands re-measured. Minor corrections (a few centimers of shortening) had to be done to optimally center the resonance for the various bands (especially for the 10m band). Picture 5 shows the antenna in place:

Picture 5: the hexbeam in place on a rotating telescopic mast, about 15 meters above ground. Its weight is approximatively 5 kg.
The mountain behind ("Rochers de Naye") is 2045 meters high, about 5 km East of the station!

 

The hexbeam is connected to the station (Yaesu FT-897D) with a low loss Aircell-7 cable having a length of about 12 meters (attenuation: approx. 0.4 dB @ 28 MHz).

After the telescopic mast has been raised to its final position, the SWR for all 5 bands has been measured accurately (September 2, 2007) with the Palstar AT1KM instrument, and gives following results:

 

   

 

    all five bands are well matched ...

 

There is a HUGE difference between this new hexbeam antenna and the previous antenna (a 50 meters long wire, about 12 meters above ground): depending on the band and position of the counterstation, the hexbeam brings 1 to 6 S-points more, and in average about 4 S-points... The front-to-back ratio has been estimated to be between 2 and 3 S-points (12 to 18 dB), which is consistent with values found in the litterature.

 

************************************************************************************************************************************

The new "broadband" version

End of February 2009, after surviving many storms with wind >100 km/h and a lot of snow and ice, two of the fishing poles broke suddenly (their wall thickness was approx. 0.8 mm!). This was the end of my marvellous antenna ...

But it was also time for improvement, since a new concept (the "broadband hexbeam") appeared end of 2007 [http://www.leoshoemaker.com/hexbeambyk4kio/broadhexcomparison.html] and [http://www.karinya.net/g3txq/hexbeam/broadband/]. So, I decided to rebuild my hexbeam with mechanically much better spreaders which I ordered in the USA from MaxGain. Once they arrived (only one week after ordering them!), they were assembled and painted. The feeding system remained of course the same, I just changed the epoxy-fiberglass isolators against better UV-resistant polycarbonate ones (Picture 6). I also added a small piece of metal mesh at the bottom of the square coaxial line to prevent the installation of insects or spiders inside the line ...

The spreaders will be clamped between the two aluminium baseplates, each spreader drilled at its base with two holes (diameter 6 mm) for the M6 screws (see Pictures 7 and 8).

Picture 6: the broadband hexbeam base with its new Makrolon (polycarbonate) isolating parts

Instead of constructing the complete antenna in the garden and later installing it on the mast, I pre-fabricated all parts (cords, wires, wire spacers) and made the final construction directly on the roof. It was finished in a couple of hours. I again used the same isolated wire (1 mm2), correcting the wire dimensions given by K4KIO [http://www.leoshoemaker.com/hexbeambyk4kio/broadhexspecs.html] by the factor (0.975) determined for the first hexbeam. The 20m and 17m band wires were installed first (Picture 7), and after a few minor length corrections, the SWR was good across these two bands. A few qso's with USA, South America and South Africa with excellent reports confirmed that the antenna was working well. A few more hours were necessary to install the wires for 15, 12 and 10m bands (see Picture 8), and to make the final tune of all the five bands.

After final tuning, the correction factor for the (insulated) wire length was re-calculated, and found to be close to 0.96 for this "broadband" version of the hexbeam.

Picture 7: the broadband hexbeam during construction with 20m wires already installed, and 17m wires ready for installation

 

With the antenna mounted in its final position (about 15 meters above ground), the SWR on all 5 bands was measured accurately (April 5, 2009) with the Palstar AT1KM instrument, and gives following results:

 

 

 

 

 

all bands have been tuned "a little high" for following reasons:

- I do not operate CW

- with moisture (rain, snow), the frequency goes down a few tens of kHz

 

Picture 8: detail of the center post of the broadband hexbeam

 

The first evening after the antenna construction and tuning were completed, nice qso's with Brazil, Chili and Argentina were made on 14 and 21 MHz with only 100W and reports between 55 and 59, despite the very poor propagation conditions ...

 

J-F. Zürcher, HB9MCZ      9.8.2007 / updated 21.9.2007 / updated for new broadband construction: 7.4.2009