A 10 A 10-element Yagi Antenna for SixRod Mackintosh, ZL3NWDesignHaving had success with one NBSdesign I decided to have a look at a 2.2 wavelength, NBS Yagi on a 13.2 metreboom (ref 2).I had already begun theconstruction of this and by now had the boom made by welding one continuouslength from second hand aluminium tube (50mm dia, 4.7 mm wall) and the elementscut (19mm, 1.4 wall).I discussed the project withBob, ZL3NE and he suggested that whilst the NBS designs were OK, you could getbetter performance for a given boom length using designs created by variouscomputer programs. Bob offered to carry out some design work using BrianBeezley’s ‘Yagi Optimizer A06’ program and thus the final design wasborn.The welded boom arrives home asone length. Note the pillow used as a pivot point on the roof of the car.It is interesting to note thatusing the same program for both antennas, the final design gave 1.5 dBadditional gain above the NBS plus much lower side and back lobe levels (theywere all greater than 20 dB down, reference to the main lobe).Final antenna dimensions. ElementElem length (mm)Spacing mm(centre/centre)Reflector28961270Driven3030.0Director 12743381Director 226421524Director 325912794Director 425914064Director 525655842Director 625657341Director 723889754Director Boom diameter = 50 mmElement diameter = 19 mmT match shorting bars at 970 mmT match to driven element = 80 mm centre to centre. Driven element length - seetextConstructionFor minimum maintenance I usedaluminium, #316 stainless steel and galvanised steel hardware throughout. Forexample in my view there is little point in using plated steel bolts where onehas difficult access to the antenna; one only needs to look at a domestic TVantenna after a year’s operation to confirm this. I was fortunate to have acollection of used, ex-commercial, cast aluminium element clamps and these wereused on all elements, apart from the driven element where I used stainless steelM10 U-bolts and aluminium plate.
The 19mm elements were sleeved with 16mm tubefor about 2/3 of their length to increase their strength. I also threaded alength of rope within the full length of all elements to reduce their mechanicalresonance. This was in response to a problem I had with my five-element NBSyagi.
Although it used 25mm diameter elements, after several years of use onehalf of an element broke at the boom clamp and it was also found the other halfof the element had a hairline crack around 2/3 of the circumference! I had notedwhen I had climbed to the top of the tower that this element, presumably becauseof its length, would vibrate in light wind conditions. Clearly, after time metalfatigue can be a problem.The feed connection box.Half-wave coaxial balun connected to the two outer N connectors and 50 ohmfeeder to the centre N. Note the two 50pf custom-made capacitors feeding the Tmatch arms.TestingI am fortunate to own a HewlettPackard T141 Spectrum Analyser with the appropriate plug-ins and trackinggenerator.
This, combined with a directional bridge, made return-lossmeasurements at any frequency very easy. The advantage of using return-loss as ameasurement is that one can add any test set-up losses including testcable-loss, and thus give an accurate return-loss measurement right at theantenna. This can then be converted to VSWR if you wish, see the conversionchart. The fact you have the ability to look at return-loss over a widefrequency range is also helpful, e.g. Looking for stray resonances.Return loss – VSWR conversion table.
Return loss, dBVSWR101.9:11.22:12.06:1The antenna was assembled on atest pole at a height of about two metres above the ground - which was fine forassembly but not for testing, since the close proximity of the ground changedthe resonance of the antenna and also the feed impedance. In my case theresonant frequency was lowered by 600 kHz, so all serious testing was carriedout with the antenna pointing skywards.
While the antenna was pointing skywardsI carried out two simple tests to confirm surrounding objects were having littleeffect.Themechanical assembly arrangement.Firstly I raised the antenna toincrease the reflector height above ground from about 300mm to about 600mm whileobserving the return loss and resonance frequency; no change was observed. Thesecond test was to confirm that the nearby shed with a high metal content washaving no effect was achieved by rotating the antenna (i.e. The antennaelements) from being parallel to the shed to end-on. What was interesting herewas that as I rotated the boom a second resonance appeared above the mainresonance. Folder guard 8.2 serial keys. After several more tests I proved that this was caused by thevertical support pipe, which in the normal test position was directly behind theboom, and thus had no effect on the antenna.
As the antenna was rotated, andbecause of the temporary clamp arrangement, the support pipe appeared out to oneside of the boom. Thus the effective diameter was increased and, moreimportantly, some elements then had a shorter electrical length, hence thesecond resonance. I confirmed this by placing a length of pipe next to the boomin the normal test position.
The results were the same.Extensive tests were carriedout with the antenna over a period of time but the only adjustments were made tothe driven element length and matching (T match). It is very important that noother adjustments are made. To quote the ARRL Antenna Handbook (ref 3):'Fundamental rule of yagi matching: Never alter the parasitic element lengths orspacing of an optimised pattern. Driven element length has virtually no effecton gain or pattern, so you’re free to adjust this dimension when matching'.I ended up with the drivenelement being longer than I expected, in fact longer than the reflector.
This issomewhat unusual and I think the main reason for this is that the effective boomdiameter at the driven element is considerably greater due to the metal platesetc used to support the feed box. The extra length of the driven element iscompensating for this.Adjusting the T-match for bestreturn loss resulted in the results still not being exceptional: 20dB at 50.110MHz and 14dB at 49 and 51MHz (nominal). I then added two variable capacitors inseries with the two T-match arms and found that with the pair adjusted to 50pF Icould improve the match to give a return loss better than 30dB at 50.86 MHz and20dB at 49 and 51 MHz (nominal). Two fixed capacitors around the end of the Tmatching rods were fabricated from Teflon sheet and copper strip.
To confirmtheir reliability I tested them with 6 kV DC with no sign of breaking down.Theten-element, 13.2 m boom yagi on the 25 m tower at the completion of theproject.The final feed arrangement was50ohm coaxial feed to the antenna with a 1:4 half-wave coaxial balun, which alsoserves to go from unbalanced to balance feed. This balanced feed point wascoupled into the T-match via fixed capacitors and the T-bar shorting stripsadjusted for best match.MechanicalAfter the testing, the antennawas lowered and mounted about 2m above the ground so that the support strutscould be finalised. To minimise any de-tuning or upsetting the radiation patternsynthetic materials were used. Parifil® for the four outer supports and doublebraided Dacron rope for the four inner struts.
Both these commercial productsuse Kevlar as the inner protected fibre for maximum strength and minimumstretch. They have been developed for supporting radio masts and aerials andhave the advantage they are non-metallic.A supporting T arrangement wasused to give vertical and some lateral support for the boom. The T support pipeused was galvanised thick wall API line pipe for maximum strength.Because this yagi has a longboom, and because this would present most of the wind loading, I decided tomount it at the boom centre point rather than the centre of weight point. I thenloaded the inside of the front of the boom with steel bar to give weightbalance. This steel bar was secured within the boom and end-caps added. Theantenna was then lifted into position on the top of the 25m towerRotator systemThe rotator I use is home-made,with a substantial reduction box in series with a second smaller reduction boxand a DC motor.
Yagi Optimizer Program Free
Selsyns are used for indicating direction. The motor powersupply is designed so that the voltage slowly ramps up and at the end of travelramps down and the rotation speed changes accordingly. With a large heavyantenna this is important.The tower head arrangementsconsist of a thrust bearing, sleeve bearing and alignment coupling. There is noweight on the rotator. If I were rebuilding the rotator system I would includeHarry ZL2SQ’s excellent idea - he has added a surplus car suspension coil springin series with the rotation pipe.
This helps with rotation starts, stops andwind loading on his 40 metre yagi.PerformancePrior to taking my five-elementNBS yagi down I carried out polar plots and signal level measurements over a25.5 km path to John ZL3AAU. The new antenna gave 3dB more forward gain and afar better polar plot. All lobes (side and rear) were measured and found to bein excess of 20dB down relative to the main lobe.
It was interesting to note thepolar plot including the main lobe beam width was found to be very similar tothe computer plot. As far as absolute performance goes, only time will tell butpresent indications are encouraging.AcknowledgementsI would like to acknowledgehelp with this project from the following people: Bob ZL3NE for his help withthe design, John ZL3AAU for his help with testing and Gavin ZL3LS for taking avideo and photographs on the day of the crane lift.