Avanti PDLII – Horizontal and Vertical Polarisation Antenna for the 2m Band – October 2023

My 10m PDLII

Since moving back to the UK, I can no longer use my American-made M2 long boom yagis for the 2m band. Instead, my main VHF antenna is a re-scaled PDLII antenna which is installed in my attic roof-space.  In the following articles I describe, model and build an antenna for 144 MHz – 146 MHz. This antenna is dual polarity with a high front to back ratio in excess of 48 dB and gain of 12 dB. The inherent problem of any antenna that looks anything like a dual polarity quad-like design is high impedance at the antenna feedpoint.

The eminent radio ham W8JI says of the design of the Avanti PDLII, “The only dual-polarization systems I've seen that are workable and have minimal interaction are a pair of crossed (X) dipoles for the feed, or the system used in a PDL quad that was marketed for CB in the late 60's and early 70's. The PDL quad had excellent feed-point engineering, although a crossed dipole feed would work nearly the same”.

The patent for this antenna was granted to L. J. Martino October 28, 1969, Patent Number 3475756 of the Avanti RF Engineering Company. The same company also crafted the Sigma IV and Avanti Starduster designs of CB antenna – all having legendary status amongst CB’ers for many years. During the years 1981 – 1987, I used these antennas for the 11m CB, (Citizens Band) and when licensed as a radio ham in 1985, I thought why not re-scale these designs for the 2m band? In those early days I knew little of the hobby and computers were in their infancy. My first homebrew antennas for the 2m band, were the ZL Special, 14 element G2BCX yagi for 2m SSB and his variant of the J-Pole, aka a Slim Jim for FM, (frequency modulation).

How does it work?

Based on what I have gleaned from the patent, the PDLII employs two different antenna-type designs i.e. fullwave loop and a folded dipole. Of course, the horizontal and vertical polarisations have separate feeds yet connect to the fullwave loop. The PDLII employs a full wavelength loop at the desired operating frequency which is excited via a folded dipole elements. The folded dipole elements comprise the driven element spacers (which connect to the driven element loop-wire). The other part of the driven element spacers I call a ‘basket’ which physically connects to the boom. The gamma match connects to the horizontal or vertical ‘basket’ element which enables antenna tuning.  

There is minimal interaction of current at the antenna feed point and I illustrate this later with use of polar plots. I am not by training an RF engineer and do not fully understand what is going on with this type of antenna. The easiest way to tease-out how it works is to model it using the excellent and free EZNEC modelling software.

My approach to modelling this antenna is based on optimisation and balance between three antenna parameters, VSWR, (voltage standing wave ratio), F/B, (front to back ratio) and antenna gain. My preferred antenna bias is high F/B and high antenna gain. What subsequently became clear through modelling this antenna is that low VSWR, high F/B and high antenna gain do not occur simultaneously. The other interesting facet of this antenna design is the reflector wire loop is the same length as the driven element full-wave loop. The spacing of elements i.e. driven to reflector is also very different to that of a traditional yagi antenna. A case in point is illustrated below. Figure 1 and 1A illustrates a low VSWR 3 element design with corresponding poor F/B of 11.07 dB and a desirable high antenna gain of 14.86 dBi.

Figure 1 - Low VSWR plot of 3 element PDLIII

 
Figure 1A- Optimised for low VSWR design, horizontally polarised

The main point of note is that a low VSWR does not equate to an efficiently radiating antenna. Antenna resonance, boom length, spacing between driven and reflector elements and other factors are at play here! Before looking at polar plots it is interesting to look at the current distribution from this design using horizontal polarisation

Figure 2 illustrates current distribution throughout the antenna using horizontal polarisation

Please note for clarity, wires 1 and 2 comprise the horizonal element spacers which connect to the driven element loop wire. The basket for horizontal polarisation comprises wires 5,6,7 and 8. The gamma match, (red ball) is installed on wire 7.

For vertical polarisation, wires 3 and 4 comprise the element spacers which connect to the full wave element loop wire. Wires 9,10, 11 and 12 comprise the ‘basket’. The full wavelength loop comprises wires, 17, 18, 19 and 20. The reflector wires comprise 13, 14, 15 and 16. The polar plots below detail antenna parameters for an original 2 element design. Later, I model a 3 element with slightly higher F/B design.

Figure 3 – VSWR plot for PDLII using Horizontal Polarisation, optimised for F/B and gain

Figure 3A - F/B and antenna gain

Thus, at an approximate value of VSWR at 4.4:1 at a frequency of 144.3 MHz, the F/B is 44.46 with an antenna gain of 12.78 dBi. An antenna gain of multiples of 3dB is very desirable because it equates to a doubling of transmit input power. For example, I use 5 watts input power to this antenna with an ERP, (effective radiated power) of 80 watts TX power. Nothing including me is perfect in any of this and I arbitrarily assume a 50% loss in TX efficiency to roughly equate a 5 watt TX signal multiplied to 40 watts or so.  It is also interesting to note the current isolation between the green/black plot and vertical red plot. This graph illustrates the excellent current isolation engineering W8JI noted for this type of antenna feed point.

What is next?

When the antenna is modelled and built, a question remains, will the gamma match the antenna to a useable VSWR value between 50 ohms – 75 ohms? I have modelled these antennas for 10m, 6m and 2m and in all cases, it is possible. I have tried various gamma match programmes as a starting point to solve this issue and to be honest, finding that ‘sweet spot’ on a curved element is difficult. The easiest way to match this antenna at the correct location on the basket element is ‘trial and error’. Have an educated guess on what dimensions of gamma match to use. A theoretical consideration of the gamma match suggests a capacitance of between 8 pf (picofarad) – 12 pf should adequately match a workable VSWR value. My old-skool 2m transceivers, an Icom 290E and Trio 9130 quite happily transmit on a VSWR of between 1.5:1 – 2.2:1 without hassle or fold-back in TX power. By habit, I have always used a field strength meter when building or using my radio equipment because it is the electromagnetic field on TX that tells me if my signal is getting out OK.

In closing Part I of this article, I detail the performance characteristics of a 3-element variant of the PDLII. I call this antenna a PDLIII. Incidentally, the letters PDL stand for Phase Diversity Loop and the II signifies the number of elements. At an operating frequency of 144.9 Mhz, the VSWR is 4.28:1, the F/B ratio is 48.08 dB and antenna gain at 12.57 dBi

Figure 4 – VSWR 3 element variant of PDL

Figure 4A F/B and antenna gain plot

I briefly illustrate performance parameters for the vertically polarised side of the PDLIII. Figure 5 illustate the fairly low angle of radiation relative to the horizon of 11 degrees at 8.4 dBi antenna gain.

Figure 5 - Vertical polarisation

Figure 5A Azimuth plot with 27.62 F/B and antenna gain of 8.4 dBi

Fig5d VSWR plot for Vertical Polarisation

As mentioned previously, tuning these antennas can be a bit tricky. This is caused by the curve of the basket element where it makes contact with the gamma match. The EZNEC modelling software is really good in this respect. If I follow the model dimensions precisely, 90% of the antennas work accoring to specification. In order to build your antenna to the specs detailed in the EZNEC model, I suggest you do the following but as always, follow your gut instincts as well. If you have tuned a yagi with a hairpin match, you should have no problems tuning a PDLII.

1. Place the gamma match on the basket element exactly as modelled. If the gamma match model is 50% along the length of the basket element, use that as a point of connection. With a bit of jiggery-pokery, pushing the slide bar of the gamma match in or out, you should be able to get a decent VSWR match. If you are unable to achieve your VSWR value, alter the point of connection of the gamma match on the basket element. Make very small adjustments here because the antenna is optimised for maximum F/B or gain. I find it useful using a field strength meter for this part of the antenna tune. When you achieve your 'sweet-spot' you achieve a workable VSWR where the antenna is resonant and be pleasantly surprised by the amount of swing on your FSM - it is very evident. In fact, the FSM swing from a well tuned PDLII is simialr in magnitude to that found when tuning a magnetic loop antenna.The width and length of the gamma match is critical to adjustment because the desired capacitance to alter VSWR is fairly small for antennas at VHF.

1. My gamma matches, both horizontal and vertical have the following dimensions, (a) the fixed tube of aluminium is 90 mm in length and 10 mm wide. The sliding rod which slides in and out for matching purposes is 180mm long and 5mm wide.

1. There are two operating frequencies of note for the 2m band. One is at 144.3 MHz for SSB work and the other is at 145 MHz for the FM section of the band. In other words, the antenna can work at different frequencies dependent on its polarisation. The easiest way around this is to keep the radiating element spacers for both vertical and horizontal radials the same length and simply cut the SSB or 144.3 MHz ‘basket’ slightly longer than the vertically polarised FM 'basket' used for 145 MHz.

1. These antenna types are very compact with a boom length of about 45 cm for a 2 ele design and about 55 cm boom length for a 3 elel variant. They are perfect for hilltop portable work and give the operator the choice of working either SSB or FM.

In the second and final part of this article I show how to build an antenna of this type using materials commonly available. These antennas were manufactured in the early 1970s and very difficult to find or buy.
In closing, the EZNEC modelling software is now free and I have been using this software for about 15 years – it is an excellent piece of kit and very useful for modelling all sorts of antenna types. The developer of this software is W7EL, his website is here in which you can download EZNEC Pro, Version 7.

The only caveat I can add is the modelling software allows you to measure antenna characteristics using increments of 1 mm antenna spacing. In other words, you can optimise the desired antenna characteristic very precisely whether it is VSWR, F/B or antenna gain. When it comes to building an antenna based on what you  model, what we produce is dependent on skill and what equipment we have available. I have a workbench, drill and hacksaw to work with and of course, my aged, tired eyes. Despite all of that, the antennas I model do work well. My 2 ele PDLII opens the world to the north and west of my QTH and for an indoor antenna, I am pleased with its performance. The 3 ele PDL is for portable work but I have not had the opportunity to test it properly – that will come next summer!

If you are interested in obtaining EZNEC files for this antenna, drop me a line at: ellis@sandrelli.net and I will send you a copy of a two and 3 element design. I also have a feeling that the patent for this antenna design is now expired and perhaps an antenna manufacturer could re-introduce these types of antenna back into the ham radio world?  

73, Ellis G1PDA