From this you should be able to guess my final conclusion about the antenna. It is a fine antenna, just not good enough. It's a shame to waste that effort, but then that is what antenna experimentation is all about.
Propagation has been excellent the past few days. This kept the polar paths open and also favoured low-angle radiation paths where those were present. I was able to compare the delta loop to the dipole over many paths, including to Central and South Asia, the Indian Ocean, East Asia, and Australia. These are paths that ought to favour the vertically-polarized delta loop.
However, ought to is not will. The dipole kept winning or at least being equal except in rare cases. Ground losses either in the near field or far field are either greater than expected or the radiation angles on most 20 meters DX paths are not low enough to disadvantage a dipole up only 10.5 meters.
Keep in mind that if you have excellent ground quality in your locale you might very well have more success than I had. Of course that may require moving to the middle of a marsh or a small offshore island, but still.
Above is a recap of the broadside elevation pattern (right) and azimuth pattern at 15° elevation (left). The feed point is 25% up from the bottom corner (λ/4 down from the apex). The model places the antenna bottom 7 meters above typical (poor) suburban ground, but no effort was made to include the effects of metal in the vicinity of the antenna. The picture of the antenna installation is in the previous article.
For comparison, the 20 meters broadside pattern of the TH1vn up 10.5 meter models with a maximum gain that is 7.2 dbi but at an angle of almost 30°. At 15° it is 5.1 dbi, which is 4 db greater than the delta loop. In practice it may be even higher. Off the ends it is a vertical radiator with a gain of -3 dbi at 30° elevation.
If you compare the gain at the full range of elevation angles up to 30° the TH1vn is superior, but only in the broadside directions. Over better ground the delta loop models as the winner. It is also more omnidirectional, which matters for fixed antennas. In other words, the antenna models are arguably ambiguous with regard to declaring a favourite. This is what makes building and testing experimental antennas so interesting to me.
Ok, that's a lot of preamble. Now let me supplement my first look with two more days of observation:
- To Central Asia and Japan the antennas were roughly equal. While trying to work a UN (Kazakhstan) station I tried switching between antennas but without any tangible effect on either transmit or receive -- he had QRM and I had QRP. I did finally work him. I had the very same results with an EX station. At another time I listened for about 10 minutes to an EA6 running JAs. I toggled between antennas while listening. It was close, though more often slightly favouring the dipole.
- VU and VK were consistently stronger on the dipole. Both should be low angle paths and, paradoxically, the path to VU is not much different than that to UN and EX.
- The dipole was the clear favourite to 3B9 and three 3B8 stations.
- All the findings in the previous article were confirmed.
Rather than make this article any longer I will follow up with another on the delta loop's performance on other bands. It did better than expected on 17 and 30 meters, especially on the former. This is the only real loss I am suffering after dismantling the delta loop -- for now I have nothing else for these bands.
I have another experiment planned that should prove interesting. I'll say more about it after I finish with the delta loop in the next article. I also have some planning and construction work to do in preparation for that experiment. Of course I must also find the time. The weekend beckons.