Improved Ionospheric Propagation With Polarization Diversity, Using A Dual Feedpoint Cubical Quad Loop

by George Pritchard - AB2KC



This Quad antenna project covers a practical Polar Diversity QUAD antenna design and related experiments with Polarization Diversity to enhance reception of fading signals in the Amateur Radio Service, specifically on the TEN meter amateur band. The selective fading when using the AM mode can be so severe as to cause total intelligibility loss of the received signal. Deep selective fading of AM signals can be reduced and possibly eliminated during a QSO using one form or another of diversity reception. For other modes such as SSB and CW selective fading is also a problem, but to a lesser degree due to the narrower occupied bandwidth of those signals. Many Amateur radio operators have noticed, DX signals appear to “shift polarity” during signal fading. Operators have used two antennas and switched between them to see this effect. The problem is that one cannot predict the best polarity to use at the transmitter to compensate for the polarity shift at the receiving station. For this reason, I have decided to investigate and design an antenna that would be effective in providing the standard Vertical and, Horizontal polarities in addition to a Circular polarization mode using the standard full-wave quad loop, to reduce severe distortion of AM signals caused by deep selective fading. It is widely known that two separate antennas of the Yagi type can be easily set up to affect a dual polarity antenna, and use of the phase relationship of 90 Degrees so that a circular “cork-screw” wave front could be generated. However, no data or practical antennas have been published that make use of the Single Quad Loop as a functioning multi-polarity radiator that can be Circularly Polarized. The design details for such an antenna is described herein. For additional reading, an interesting link to attenuation values as a function of “cross polarization” can be found at:


Without polarization diversity, a station has one receiving antenna, typically either horizontal or vertical. Over the long term, ionospheric fading mechanisms impact both polarizations equally so the choice of polarization of this single antenna can be made for other considerations (mobiles / DX). However, over the short term measured in fractions of a second to seconds, ionospheric fading can be substantially different between the two polarizations. An interesting link to a paper concerning “Faraday rotations through the ionesphere”, can be found at:


In addition to the theoretical information obtained at the above stated links, Amateur radio station operator W0RPK collected polar diversity data using two separate antennas at right angles to each other, and two separate receivers. To see the study in detail, go to his website via and use the associated links he provides. The W0RPK investigation provided propagation data as plotted on a graph. Dual Receiver AGC data points were logged and plotted on a computer at his station. The brief but important differences in polarized fading can be seen in the graph of a constant transmitter amplitude 10m signal (each horizontal division represents 1 second). The radical differences in horizontal vs. vertical signal strength as ionospheric fading mechanisms impacted one polarization and then the other is evident.

In as much as a the receiving station benefits from diversity receiving antennas, the transmitting station reciprocally benefits in providing a more diverse signal to the receiving station. It should be noted that the algebraic sum of both graphs could be the mutual amplitude benefit for both receiving and transmitting stations. A station receiving such a diverse signal would be rewarded the benefit of a "polarization match" regardless of the polarity of the receiving antenna. The typical 10 to 30 dB fades in a single polarity antenna would be "held up" in amplitude to the best case fading of the other arriving polarities. The -3dB power split and resultant 1/2 power per polarity of transmitted power for each of the separate feeds, more than outweighs the benefit of reduced overall fading into the noise level at the receiver thousands of miles away. The following pages describe the practical application of a multi-diversity antenna system using a dual feed point cubical quad. The driven element only is described herein as the basis for the antenna. The typical parasitic element design can be any, and is not considered here. Note: Closed loops are recommended for maximum Circular diversity.


1.0 OVERALL DESCRIPTION Triple Polarity AB2KC Quad

The purpose of Triple Polarity AB2KC Quad is as follows:

1. Provide Horizontal polarization for DX and ground wave contacts while both stations are mutually horizontally polarized.

2. Provide Vertical Polarization for ground wave contacts for stations mutually vertically polarized, and specifically for local mobile contacts that use vertically mounted mobile antennas.

3. Provide Circular Polarization for DX contacts to reduce AM / SSB selective fading due to Faraday shifts in the arrival of the signal. Vertical to horizontal rotations is a result of the reflections off the ionosphere, resulting in an undefined received polarity. The affect is also reciprocal for transmission, in providing a polarization consistent at the receiving station. The power is split 3dB to either polarity during this mode.

1.1 Two Gamma matching sections in separate weather proofed enclosures.

A weather proofed containing the wide spaced 100 pF variable capacitor for tuning, and attachment for the Gamma wire (on cap shaft). An SO-239 female connector mounted for feedline connection and center of quad loop attachment permits easy mechanical connections.


  • Use two ground lugs to solder the quad loop wire to the SO-239 connector flange.
  • Use a tin plated copper right angle bent strap from the shaft of the cap to the gamma wire. Seal box against weather. Use Plexiglas spacers to maintain spacing of the gamma wires to main loop.
  • Both boxes must be identical.

1.2 Two 1:1 Baluns for Isolation

A balun at each feed point is mandatory, to provide a high impedance to the outer shield of the feed lines to eliminate loading of the loop. The two Baluns are made with coax sections. Wound with 4 turns at 3-inch diameter. USE TEFLON COAX HERE. Each cable used to make the Baluns must be the same length. Tie up each coil in solenoid form using 6 heavy duty nylon tie wraps evenly spaced around the coil. The coils must be solenoid wound to reduce distributed capacitive coupling. The two feed lines to the branch line coupler must be the same length. The Baluns have PL-259 connectors at both ends for easy attachment to the gamma boxes that have the SO-239 females ready for action (Hi).

1.3 Ninety-Degree Phase Shift

To get Circular polarization… you need a 90 Degree phase shift on one of the feed lines. This can be done by simply adding a _ wave section of line to either feed line, and adding a TEE to split the signal. The problem with that is the TEE connection provides no isolation, and reduces the feed point to 25 ohms (assuming a pair of 1:1 SWR lines). I took a novel approach to this problem. I used a “BRANCH-LINE-COUPLER” design, which is a modified design from common microwave dividing and combining technique. In essence, it is a 90 degree quadrature –3dB divider, that reciprocates as the combiner during receive. Normally at microwave frequencies, the network is printed as stripline on double sided dielectric PC board. At HF frequencies, I simply used coax! You will need to make a network using two _ wave lengths of 50 ohm coax, two _ wave lengths of 35 ohm coax (explained below) and connect it to four SO-239 connectors mounted on a plastic box (no common ground between connectors).

Here is how it is connected up: To visualize it, draw a square with 3 inch sides. On the bottom left corner, mark it “transmitter input”. Mark the top left corner “Vertical out”. Mark the top right corner “Horizontal out”. Mark the bottom right corner “termination”. Now, mark the two vertical sides of the square 35 ohms, and the two horizontal sides of the square 50 ohms. The two 35 ohm lines are each made with two parallel 75 ohm lines (37.5 ohms close enough!). Be careful to take the velocity factor of each line into account when determining the _ length needed. Some 75 ohm cable is polyfoam dielectric with a velocity factor of about .8 compared to Teflon at .66 (MAKE SURE). When complete, each SO-239 connector should have three coax connections, a pair of 75 ohm lines, and a single 50 ohm line. The termination connection is for a dummy load connection. When the entire system is working, there will be no power at this point. If there is a small imbalance in the system with VSWR or a bird in the antenna, the difference in amplitude will end up at the termination and NOT AT YOUR TRANSMITTER. The rig will always “see” 50 ohms!. I use a 200W dummy load, good for 1KW carrier of AM. I never saw more than 100 watts across the entire 10M band. You can add coax relays and a 3-position RF switch where needed to feed only the vertical, or only horizontal feed points for “linear polarization” (not circular, Hi Hi). It’s fun to flip the switch and recover a severely faded AM signal while receiving. It more fun to give your fellow ham a circular signal to receive from you! Now all that work you put into your AM transmitter’s audio will have less selective fading on the other end! Of course… when the band goes out, none of this helps and you’re finished!


2.1 Initial Adjustments/SWR and Isolation

Initial adjustments of the gamma matching networks proved to be straight forward, with almost unnoticeable interaction between the two feed points. The variable capacitor will end up about 1/2 meshed. Make the gamma wire longer to start with, and use “back-to-back” soldered alligator clips to find the best length. I did most of the matching with the MFJ SWR antenna analyzer. After adjusting the dual feed points for minimum VSWR at 28.7 MHz, I fed one polarity (horizontal) with 10 watts and measured the power coupled into the vertical feedline. It was less than 0.15 watts, about 18-dB isolation. Vertical to horizontal isolation is 18-dB.

2.2 Bandwidth

The bandwidth of each polarity was excellent, with a VSWR of less than 1.4 : 1 from 28 to 29.6 MHz, with only a single parasitic reflector.

2.3 Balance

With 10 watts drive into the transmitter input port, the branch line coupler’s termination port received only _ watt power worst case, (28 MHz) across the full bandwidth of 13-dB combined amplitude and phase balance. At 28.7 MHz (design frequency) to 30 MHz, the reflected power to this port was almost zero. This should produce little or no skewing of the circular rotational shape in the AM portion of the band. I use a pair of identical VSWR meters (old Lafayette type) to each of the branch line outputs to monitor each feedline during circular mode. A cheap power meter in the termination line gives a good overall indicator of total balance of the system.





8 June 2003