Double Delta Loops Can Double Your Fun

Long wire antennas have always intrigued me over the years. Without exception these long pieces of wires seemed to perform better compared to dipoles or verticals. The problem for most of us is how to put up so much wire and get it squared off for best performance. In thinking about delta loops I really liked the fact that they do not need to be elevated too far above ground. . Every morning when I would drive off from our home in Escondido I would sub consciously drive by two huge pine trees in the backyard. One day my brain said "look at them" - they could be about 34 feet apart. That night I got out the tape and voila they were exactly 34 feet from trunk to trunk - the perfect quarter wavelength on forty meters – my very favorite band.

The Trees with the two separate antennas and the two feed points on each antenna

It did not take me long to buy 500 feet of wire at the local Home Depot. The math works out that each antenna needed to be 140 feet long from end to end. I started reading up on various configurations and feed options and decided that end feeding them would provide the best match and lowest angle of radiation. In thinking about how they might best be lined up it was an easy choice. Broadside to Northeast and Southwest would work for rag chewing and DX into Europe or the South Pacific.

A number of years back I became very sold on phasing antennas to increase gain and performance. At the time we were living in Kansas City. I put up two mono band 40M verticals spaced 90 degrees apart – a quarter wavelength. They were ground mounted with a lot of radials. They performed almost too well with a backside rejection of about 20db and a forward gain of three to four db - enough to basically double my transmitted signal. The problem was when I would get into a conversation on both ends of the country – like California and New York - switching back and forth just would not work for all parties. This would not be the case from my new QTH in Southern California.

The following illustration helps. Two ground mounted verticals are placed about 90 degrees apart – ¼ wavelength (at 40M this is about 34 feet). If we feed them in parallel they are in phase (typically this would be done with two 75 ohm ¼ wavelength (electrical – explained later) – coax to match the 25 ohm center feed point impedance – two 50 ohm antennas in parallel). By inserting a phasing line like a 90 degree xelectricalx piece of 50 ohm coax in front of one or the other 75 ohm feed lines magical things happen.

Figure 1

It is important to understand that the electrical length of coax is related to its velocity factor and is shorter in its physical length. For example 75 ohm coax has a nominal velocity factor of .75 and 50 ohm somewhere around .85. The electrical length can be calculated by multiplying the physical length by the velocity factor. This is usually good enough in practice. Alternatively if one has an antenna analyzer you can find the exact length by leaving one end open and the other end connected to the analyzer. This will allow one to trim the coax to the frequency desired. Just start out a little longer than the calculated length and trim an inch or two at a time until the analyzer resonates at the desired frequency. Back to figure 1 to visualize what happens. Letxs assume we place the phasing line in front of the feed line to antenna "2". A signal coming from the opposite side of antenna "1" shifts 90 degrees (the 34x spread between the two antennas) by the time it arrives at antenna x2x. The receiver hooked to both antennas sees a further 90 degrees in phase shift because of the ¼ wavelength phasing line or a total shift at antenna "2" of 180 degrees compared to antenna x1x. Now the two antennas are 180 degrees out of phase and tend to cancel each other. The signal is attenuated. If the signal is coming from the other side of antenna two the opposite takes place and the signals at both antennas as seen by the receiver are in phase and enhanced. The same dynamic happens on the transmit side. This phenomenon is not black magic but rather moving the signals around to achieve better reception or more of the power concentrated directionally.

Setting up the two delta loops in this same way presented a few challenges. Initially I started out by feeding them with two 90 degree (electrical) pieces of 50 ohm coax. In checking the center feed point impedance it was exactly 25 ohms – somewhat like putting two 50 ohm resistors in parallel. This would present a huge mismatch with the 50 ohm feed. In discussing this with my technically oriented California ham friends they all chuckled and chimed in – "you need to use two 90 degree sections of 75 ohm coax like RG-11". Armed with this new knowledge I dug out the handbook which provided the following solution. The formula was Z= the square root of Z1 and Z2 where Z is the characteristic impedance needed to match the network, Z1 is the antenna impedance, and Z2 is the impedance for which it is to be matched. Plugging in the numbers as the square root of 25 ohms plus 50 ohms the result was 35.4 ohms. Putting two parallel 75 ohm sections of coax (1/4 wavelength or multiple thereof) would be like paralleling two 75 ohm resistors yielding 37.5 ohms). Almost a perfect solution for the 35.4 ohms required. This yielded an almost 1:1 match for the network.

The math for the antenna shown in Figure "2" is simple. We know that one kilohertz has a wavelength of 300 meters. If we divide 300 by seven MHZ we get 42.89 meters or about 137 feet of wire required. For the two 75 ohm feed lines we need ¼ wave length (electrical). Since the velocity factor of RG11 is nominally 75% the physical length ¼ wave length of 34 feet needs to be reduced to about 25 ½ feed or .75 x 34. This same math for the 50 ohm 90 degree phasing line would be 34 x .85 or about 29 feet.

Figure 2

The apex of each antenna was hoisted about 35 feet into the two trees and the lower wire stretched out so that it was about three to six feet above the ground. The variation was due to the uneven slope of the ground.

The next issue was the phase line and how to insert it in such a manner that it could be switched back and forth from the East antenna to the West antenna.

Note the two fittings on the coax relay – a male "T" with two female connections on opposite ends. From the above relay picture you can visualize the action. In the non-engaged position the 50 ohm feed enters the relay and flows through the upper "T" to antenna "1" and concurrently the 50 ohm phase line connected to the bottom of the "T". The phase line is now in front of the 75 ohm feed line to antenna "2" and the feed to this antenna is delayed 90 degrees from antenna "1". The opposite action takes place when the relay is engaged. The feed to the opposite antenna (no. "1") is delayed 90 degrees. The phase shift enhances the signal in the direction of antenna "2" when the relay is not engaged, and antenna "1" when activated.

The phasing line was originally cut to ¼ wave length electrical out of 50 ohm coax and later reduced to about 80 degrees in length. I wanted to experiment with the phasing line so I took the chance of reducing its length to 80 degrees instead of the original length for ¼ wave length. In practice I noticed an improvement in the front to back ratio and left it at this cut. In actual performance the front-to-back ratio shows up at about five to six “S” units or 25-30db. This results in significant rejection on the back side and gain on the front side.

I am not a DX hound but with the phased loops I have worked South Africa, Europe, New Zeeland and Australia. Reports have all been excellent. In my experience it is one of the best performing antennas I have ever used to 40 meters.