Polarity Diverse Antenna System

Introduction

Recent experimentation by a number of amateurs has clearly demonstrated the advantage of combining orthogonal yagis to overcome undesirable losses due to mismatched polarities over the EME path. This is especially true on 144MHz and was clearly demonstrated in the last ARRL EME contest as Europe and NA were out of phase for extended periods.

Credit for information that I based most of this page on must be given to:

• Mike Staal K6MYC of M^2 for the 2MXP28 design / data.
• Leif Asbrink for his extraordinary presentation of information on this subject presented on the SM5BSZ website.
• A comprehensive article entitled " Spatial Polarization and Faraday Rotation by Tim Pettis KL7WE can be found in the Proceedings of the 22nd Conference of the Central States VHF Society (Lincoln, Nebraska, 1988) on page 95.

I will try to use this page to share information specific to an array of 2MXP28 antennas. Hopefully, you won't fall into the traps and make mistakes that myself and others have made! I intend to do much experimentation with the goal of determining an ultimately practical implementation. If you have experience with this type of antenna and wish to contribute to this page I would be happy to include your data / documentation.

4 x 2MXP28 Testing

I tested each of the 4 x XP28 antennas and found that their SWR / Return Loss Plots were nearly perfect for 144.1 MHz. Pictured is my portable test setup consisting of a White Lawn Tractor and Trailer with a HP 8714C Network Analyzer ... I like to rough it!

Tip from K6MYC: Mike says tuning a power divider is easy ... just put it over your knee and give it a little bend until you get what you want!

4 x 2MXP28 E&H Plane Spacing

That's me with my array of 4 x XP28 antennas during the construction phase. I hang the whole thing on my winch / aircraft cable and then simply crank it up into place.

I have mounted these yagis in the "X" as opposed to "+" orientation" stacked at a distance of 14'6" high and wide. This spacing was chosen as it was the maximum that I could easily attain using existing H-frame components. The manufacturer says that 14' is the correct spacing, while VE7BQH recommends a much larger spacing (15'3" or 4.66m minimum, with 15'7" or 4.75m optimum). I ended up in the middle!

The H-Frame construction:

• Horizontal - Aluminum Tube 14' Long, 2.5" OD, 2" ID, qty 1.
• Vertical Centers - Aluminum Tube 46" Long, 2.5" OD, 2" ID, qty 2.
• Vertical Ends - Morrison Molded Fiber Glass Company (MMFG) Extren 500 (All purpose utilizing isophthalic polyester resin) 90" Long, 2" OD, 1.5" ID, qty 4, each sleeved inside the Vertical Centers 8".
• Vertical Bracing - Aluminum Angle 1"x1", x3/16" thick, 42" Long, qty 4.

4 x 2MXP28 AZ EL

My TIC1032 Ring Rotor, 24" Saginaw Linear Actuator, and Boom Cradle / Elevation Mechanism now sits at 64' on my 72' Heavy Duty Freestanding Trylon Tower. I moved it up 16' to cut my EME path loss (just kidding ... I needed to clear some ever growing maple trees to improve my shot at moon set).

4 x 2MXP28 Phasing

I decided to make my own phasing lines out of Andrew FSJ4-50B (1/2" Superflex). I like this cable because it bends easily, doesn't get water inside, and it handles lots of WATTS! They are electrically 1.75 wavelengths (qty 2), 2.0 wavelengths (qty 4), and 2.25 wavelengths (qty 2) at 144.1 MHz.

I do not recommend using LDF4-50A or equivalent cable for this application, the bending required would be almost impossible!

I thought that if I used 1/2" Superflex for phasing lines, it may be stiff enough to push the antennas out as opposed to pulling them in with the weight. WRONG ... I was forced to install a X-Brace.

The weight of the X-Brace, Superflex feedlines, box, mounting poles, etc combined with the distance from the horizontal pivot point forced me to add an unwanted counterweight system to the front of the array. I had to put 30 lbs of steel weights 11' out in front of the horizontal pivot on each side to get the array balanced. If I had to do it all over again ... I wouldn't! I recommend using the lightest phaselines you can, and avoid all unneccessary weight in the rear of the array!

The Polaris Xplorer 400L 4x4 makes getting around fun and easy! ATV'ing is now my second favourite hobby!

4 x 2MXP28 RX

I mounted an aluminum chassis (10W"x17"Lx4"H) out in the middle of the 8 driven elements. This maximizes my RX capability since each antenna is only a short phasing line (<= 2.25 wavelengths), power divider, Nm to Nm adapter, Transco Y Relay, and a Nm to Nm adapter away from the GaAsFET preamp! It doesn't get much better than that! Insied the box is:

• Transco Y - T/R relays (qty 2)
• FSJ4-50B TX Feedline (qty 2)
• SSB Electronics LNA145 (qty 2)
• RG213 RX Feedline (qty 2)
• 20 Conductor Control Cable complete with ferrite sheilding

Each set of antennas ("/" and "\") has a separate Preamp and RX feedline into the shack to allow flexible experimentation until an optimum solution can be verified and implemented.

I acquired a pair of old HP 8475A S Parameter Test Sets. These boxes feature (among other goodies) a 30cm mechanically adjustable line (good from 0.1 - 2.0 GHz). I took the 2 lines and recombined them back into a single case. I run each RX line through one adjustable line and then into 2 port combiner that feeds my LT 2 S. This will allow precise setting of the cable length and thus the combined RX signal. I can also switch power to each of the preamps, thus allowing me to select "/" or "\" or any combination of "/" and "\".

Currently, I am awaiting arrival of a Down East Microwave Transverter modified to have a pair of downconverters sharing a common LO allowing me to feed one signal into the Yaesu FT1000D main RX and the other signal into the sub RX for full stereo mixing.

4 x 2MXP28 TX

This photo was taken before addition of the X-Brace. Currently, I use only one relay to select transmitting "/" or "\". I intend to add a 2 port power divider and swicth in some coaxial delay lines so that I create "V" and "H" as well.

4 x 2MXP28 Phasing

The proper orientation of all components in an array of 4 2MXP28 antennas is absolutely critical. To understand why, first let's review a few basic assumptions:

• Antennas are supported on fiberglass (or some other suitable non-conducting) poles. Use of metallic poles would of course degrade the pattern and seriously decrease peformance.
• Phasing lines are run away from the antenna boom towards the center of the array from a point between the two orthoganol driven elements. The lines must be perpendicular to the rear driven element to minimize their effects on the radiation pattern. This method still represents a small but practical compromise in performance. Ultimately, both feedlines should be run to the rear of the antenna and be routed at least 4" behind the rear most reflector. Although optimal, this method has undesirable tradeoffs:
  • increased phasing line length and,
  • increased weight offset additional distance from the centroid.
• Mounting of these antennas in the "X" as opposed to the "+" orientation has adavntages:
  • There is no interference between overboom dacron support ropes and the elements.
  • Water runs off slanted elements for better wet weather performance. Water droplets hang on horizontal dipoles which causes electrical changes resulting in less than optimum peformance.
  • Shorting routing / length for phasing lines.
  • Even with antennas mounted in the "X" orientation (+ / - 45 degrees), Vertical and Horizontal polarizations (desirable for tropo work) are easily generated using simple switching of relative cable lengths.

Notice that 3 different lengths of phasing lines that are required to properly phase these antennas. It is very easy to get lost and confused in the math associated with array. However, a few simple rules should help to make this all simple to understand. Please reference the drawing for idnetification.

• This array should be considered as two separate 4 yagi arrays (A & B) that happen to be orthoganol.
• Two antennas (A & B) share the same boom and are offset by 20" (1/4 wavelength). Note that A1, B2, B3, and A4 are in the forward positions and B1, A2, A3, and B4 are in the rear positions. This is done (as previously described) to facilitate phasing lines in the most electrically unobtrusive manner.
• A1 and A2 are related as follows:
  • The electrical orientation of A1 and A2 are identical (rotate A1around its own dipole axis results in A2) even thouth physically they are on opposite sides of the booms.
  • Therefore the only real difference is the 1/4 wave difference created since A1 is forward and A2 is rear.
  • Now, to get A1 and A2 in phase, we simply make the phasing line to A1 a 1/4 wavelength longer than phasing line to A2 (waves travel slower in the phasing line than in space). Think of it this way, the signals are all in phase coming out of the power divider, so when a signal reaches A2 and starts to propogate down the yagi, we need to delay the signal to A1 (by 1/4 wavelength) to align the wavefronts! The signal to A2 travels through 2 wavelengths of phasing line and 1/4 wavelength of space to align with the signal to A1 which which travels through 2 1/4 wavelengths of phasing line. Simple, right!?
• It should now be a simple extension of these principles to confirm the orientations and lengths.
• As an intuitive cross check, compare opposite corners (A1B1 to A4B4 and A2B2 to A3B3) are notice that physically and electrically they are equivalent.