The design objectives are:
|The requirement can be simply
met by an inverted V dipole antenna oriented NW-SE to
present broadside in the NE direction.
Situation in a built up area indicates selection of a horizontally polarised antenna for lowest receive noise.
A dipole mounted close to ground and with the legs sloping downward from the centre tends more toward an omni-directional pattern rather than the classic pattern for a dipole in free space. The figure at the right shows the modelled pattern from EZNEC.
RG6 was chosen as a transmission line. The table shows the key parameters for alternative coaxial transmission lines considered.
Note that the loss of 24.5m of RG6, although greater than for 21.0m of RG213, is almost identical in loss per meter of length.
An alternative transmission line configuration of a full wave of ladder line with a cored balun, but it would exhibit similar losses overall to the RG6 which is easier to install and relatively independent of precipitation.
|Whilst the centre feed
impedance of a half wave dipole in free space is just over 70 ohms,
installation of the dipole close to lossy earth and sloping the legs
down from the centre have the effect of reducing
the feed impedance to around 65 ohms which will cause standing waves on
a 75 ohm feeder (VSWR=1.15).
The matching strategy is to lengthen the dipole to excite a VSWR(75) of 1.5 in the RG6, and cut the cable at the most convenient length where the impedance is transformed to 50+j0. The first point at which the impedance is 50+j0 is at 1.5 metres from the feed point which doesn't suit my installation. This condition repeats every half wave, so the first convenient point is at 24.5 metres. The Smith Chart illustrates the matching scheme.
The model shows a gain of 7.3dBi at 30 deg elevation on the major lobe. The feed system has quite low overall loss, resulting in 86% of the transmitter power reaching the antenna. The achievement of a 50 load impedance for the transmitter, allows the transmitter to achieve rated output without using a lossy ATU.
These are design model figures, which are the starting point for implementation. Implementation will be influenced by real ground conditions, proximity of other structures etc. It will be necessary to jiggle leg length and transmission line length to achieve low VSWR. Note that there are many types of RG6 and the velocity factor varies with the construction. The actual length of the transmission line will depend on the velocity factor.
An EzNEC model of the antenna / transmission line is available for download.
|A Smith Chart view of the
matching scheme is shown to the right. The values plotted here
are as modelled as above.
The chart is normalised to 75Ω. The desired normalised impedance at the transmitter end of the line is 0.66+j0Ω.
The balun was a W2DU balun using a string of inexpensive ferrite cores threaded over the end of the RG/6 feedline adjacent to the feedpoint.
Models are great for understanding how an antenna system will work, but some parameters are not well known and the estimates subject to some error. Nevertheless, the assist in finding the final solution.
The approach taken to implementation was to cut the dipole a little shorter than the design, and use two adjustable capacity hat stingers near each end. The stingers were adjusted in position until the VSWR(75) (measured with a VSWR meter calibrated for zero reflection with a 75Ω load) at the source of of about 21m of RG6/U at the patch panel was 1.5. Next, a fly lead length was found so that VSWR(50) was approximately 1. The total length of RG6/U was 25.2m, which suggests the feedpoint impedance at 14.15MHz was closer to 53+j19Ω.
Seems like an ideal task for an antenna analyser such as the MFJ-259B. Well, no, the MFJ-259B produced misleading results, its VSWR indication did not reconcile with a quality VSWR meter (Bird 43), and the assumption is made that the MFJ-259B's signal source did not dominate the measurement, that there was enough received signal to result in substantial errors. (The MFJ-259B gave correct indication on a dummy load.)
At the end of the set up task, the antenna system VSWR measured no worse than 1.2 across the entire 20m band, and less than 1.1 at 14.15MHz at the input end. The antenna system can be connected directly to a transmitter PA without compromising power output or stressing system components unduly, and without the need for an ATU.
Was there a simpler way? Well, the suggestion that the real part of feedpoint impedance was lower than modelled by about 5Ω leads to the suggestion that the feedpoint impedance at resonance might be around 53Ω which would be quite suited to direct feeding with 50Ω coax (VSWR=1.06). The RG6/U as used has about half the loss of RG58C/U and costs less, but there is no doubt it is more complex to set up. Larger coax such as RG213 would not accommodate the balun as readily.
Achievement of the best VSWR depends on the time one is prepared to spend trimming the antenna legs and cable length. In my case, I abandoned efforts when the VSWR at 14.15MHZ was 1.1.
Signal reports indicate that the antenna is working as intended. Receive signal strength and reports received from other stations are consistent with similarly configured other local stations, and the paths commonly encountered.
The advantages of this design are:
|ATU||Antenna Tuning Unit - in this context, a device for transforming the impedance presented at the transmitter end of the transmission line to a load impedance to suit the transmitter.|
|Balun||Balanced - Unbalanced - a device for transitioning between balanced and unbalanced transmission line, may take many forms.|
|HDC||Hard Drawn Copper|
|VSWR||Voltage Standing Wave Ratio|
VK1OD on the 'net. I appreciate feedback, click on the ... in webm...@vk1od.net for my email address.
© Copyright: Owen Duffy 1995, 2012. All rights reserved. Disclaimer.