18 February 2014

Triband Delta Loop antenna - some theory


Looking for an improvement in my station's configuration, I evaluated a number of antenna designs; the most important thing, wether some like to admit it or not, is what possibilities you have for installation. What type of ground, how high can you put it, is the surrounding space opened enough towards the horizon, are there elevated points you can use for support, what kind of weather it will have to endure, how can you properly make the safety grounding etc.

All these factors are very important in the choice of an antenna, and i'll come up with a quick example: a ground mounted L/4 vertical above salt water (let's say, on the beach of a small oceanic island) for the 40m band will put out a DX signal similar to an horizontal dipole mounted at about 30m high (horizontal antennas need height to obtain better gain at lower takeoff angles), and the dipole will need to be rotated towards the direction you want to work while the vertical is omnidirectional. If you look at the costs of building a 30m tall tower, adding a rotator to it and the mechanical challenge of making it stable in saltwater-impregnated sand versus the costs of a few wires and a 20m tall lightweight fiberglass pole, the choice is a no-brainer. However, move the same setup above a poor, dry ground (let's say, Sahara desert) and the horizontal dipole will be a much better antenna - and while you have a tower and a rotator you might as well make the dipole a 3 or 4 element Yagi and get an ever better signal.

What i tried to get is a versatile 20m/15m/10m antenna that offers good gain and is easy to build, install and work with. I have used full wave loop antennas before and liked them alot, especially for the rewarding performance/complexity ratio. You just need a wire, cut it at about 105% of desired wavelength, spread it as you can and off you go. When properly installed (vertically, as spread as possible), a closed loop antenna has some advantages over a dipole:

- slightly higher gain (about 2dB)
- less noisy
- less sensitive to height
- less influenced by the surrounding objects and enviroment




The maximum gain for a loop is when it gets the most surface area for its perimeter (make it as close to a circle as you can) and you feed it at the lowest point (this forces one of the current maximums at the highest point, if the antenna is symmetrical). A loop for the 20m band has a perimeter of about 22m, and it would be a mechanical challenge to force such a thing easlily in a circle form. However, if we make the same perimeter a triangle, the gain decrease is not very big (about 1dB) but it makes it much easier to build since now it requires just 3 support points. Holding on to the fact that largest surface area = highest gain, we should try to make this triangle as close to an equilateral shape as possible. To put more of the antenna as high as we can, we should install this triangle pointing down, with one of the sides (1/3 of the perimeter) at the highest point.



So, we have this triangle pointing down, wich should be our antenna.




Now, a full wave antenna installed vertically is a rather directional antenna, with most of the radiation ocurring on the direction perpendicular on the loop's plane. This can be viewed as a disadvantage since it should need to be rotated in order to get the most out of it (hey, YO9IRF, you said it was a simple antenna, ain't nobody got time for those rotator things), but the radiation lobe is pretty broad and there's two of them since the F/B ratio is 0dB, so it actually covers pretty effectively about half (180 degrees) of the horizon. If you install it in a fixed position towards a well chosen azimuth, it could cover most of the interesting DX-lands.

A full wave loop antenna doesn't resonate exactly when the loop's physical length is one full wave; it needs to be larger with about 5% or so, depending on the type of wire, shape, surroundings etc. The advantage I want to use in this antenna design is the fact that it also resonates at about 2x the loop length; more precisely, a loop designed for the 20m band should present a reasonable feed impedance in the 10m band. The bad is in the 10 meter band we will see some upwards shooting radiation lobes; more than that, the resonances are not exactly harmonic (L+5% and 2xL) and we can't have optimal matching in both the 20m and the 10m band. The good is that in the 10m band there is still alot of radiation at low elevations (there are radiation plots later on), the antenna is much less directional and therefore the sides are much less attenuated, the band is not opened all the time and when it is you can work the world even if you're a water pistol of a station like me. The ugly is that there isn't really another simple way to cover both 20m and 10m with loop antennas since the 20m loop will resonate close to the 10m band anyway and will mess up the resonance of another paralelled element.

OK, so with the 20m and 10m bands out of the way, there is still the 15m band to be covered. This is done by alot of amateurs in a number of ways: either some trap system, shortening the loop via a switch, matching feedlines or feeding the loop via a 4:1 balun and using a tuner to further tune the matching. While these solutions work and there's nothing very wrong with them, I chose a more basic way to get the 15m band: since there's already a huge triangle-shaped frame up in the air, why not insert a smaller one inside it, resonant in the 15m band ? The 20m loop has a high impedance at 21Mhz and most of the energy will go to the smaller resonant loop so no need to switch or match anything, just connect the two loops in paralel at the feedpoint. This way there's no compromise in the 15m band performance and you actually only need to add a few more wires. No lossy matching circuits or traps, no switches that need to be powered, no tuner required, just wires, make them thick enough and well isolated and you could throw unlimited power into it.

OK, so now that we have a concept of an antenna, let's throw it in an antenna modelling software to see if it's any good. I'm used to EZNEC, there probably are better ones out there but i'm more familiar with this one so EZNEC it will be.



To reduce the mutual influence between the two loops but keep the optimal equilateral shape, there needs to be some distance between their feedpoints. For this, I decided to use a piece of 450ohm balanced (ladder) line, 68cm long. This will shorten the total perimeter of the 15m band loop and reduce the gain slightly, less than 1dB, but it's necessary.

I've simulated the antenna at a height of 30m above a city-type ground (it's one of the worst), wich is close to my conditions. The wires are 1.5mm (1.75mmp / AWG 15), copper, PVC coated. There is however further on a batch of radiation plots for a height of 6 meters above normal ground.

Okay, everything seems pretty, let's see how the SWR looks.



The SWR troughout the 15m band is very good, 1.4:1 or less, but as I said earlier, you can't have optimal resonance in both the 20m and the 10m bands. I chose the middle path, and the resonances are at 14.5 and 28MHz; not exactly inside the ham bands, but close enough to cover a good part of the bands (14.1 to 14.35 and 28 to 28.5) with a SWR under 3:1, wich even a transceiver's internal tuner could match down to optimal. This solution provides much lower losses than using a 4:1 balun and a tuner, or a trapped antenna.

Radiation plots for 20m band:



The 10.6 dBi gain figure is very impressing, considering a dipole is at about 8 and a 3 element Yagi a bit over 11. The takeoff angle of 8 degrees is caused by the good height over ground, but in reality it should be a bit spoiled by the building's rooftop at about 8m below the antenna (hey, I don't have a 30m tall tower). The side rejection is better than -22dB, wich may or may not be what you want.

Radiation plots for 15m band:



The small piece of 450ohm balanced line takes it's toll, and for the 15m band the maximum gain drops about 0.8dB, to 9.8dBi. Still a very good value, especially since the takeoff angle is only 6 degrees. Side rejction is about -20dB.

Radiation plots for the 10m band:





As expected, here the radiation doesn't go just towards the horizon, but everywhere. Luckily, there are enough strong lobes at low angles and the one with 4 degree takeoff angle is at 8.66dBi, just 1.28dB below the strongest one. Side rejection is almost non existant, making it a pretty omnidirectional.

All in all, this antenna looks like a pretty good compromise and a reasonable choice for my station. However, maybe some of you guys would like to see how it does in other conditions. Here are the radiation plots for the same antenna mounted at a height of only 6m, above a medium quality ground - let's say it's a typical installation if you live in a house and you have some back yard, or if you decide to use it in portable / SOTA / DXpedition type conditions.

20m band:


Clean looking plot, the low elevation radiation is pretty hard to achieve with a vertical and at medium elevation is still a considerably better choice than a dipole or an Inverted V.

15m band:


The upwards shooting lobe is caused by the choice of height, but you will need a 3 element Yagi to get better results overall.

10m band:



At lower heights there is some sidelobe rejection in the 10m band, not very strong but still observable.

The second part of this article will discuss construction and results of such an antenna, so please check in once in a while.

No comments:

Post a Comment