Monday, January 20, 2014

Narrow Diamond Loop for 40 Meters

Full-wave loops come in all shapes and sizes. At one extreme is the circular loop (maximum interior area) and at the other is the folded dipole (minimum interior area). All can be fed for vertical or horizontal polarization, or any selected mix, by suitable placement of the source (feed point).

I want to finish up my modelling and analysis of full-wave loops with one more: the diamond loop. Earlier I looked at the delta loop (my current antenna for 40 meters) and, what I call, the chevron loop, and compared their DX (low-angle) performance to other wire antennas at apex heights ranging from 15 to 25 meters over medium ground. I skipped the most common type of loop, the square loop most often seen in cubical quad beams, since it requires more than one high support.

The diamond loop is most often deployed as a square loop turned 45° (left current plot), with all interior vertex angles 90°. If the interior angle of the top and bottom vertices is increased you get a narrow diamond loop (right current plot). In this narrow diamond the interior angle is 120°. Like other loops they can be fed for horizontal polarization (top or bottom) or vertical polarization (side), or a selected mix of the two. For these 40 meters loops I will stick with vertical polarization since I am focussed on antennas that are low to the ground in terms of wavelength. An apex height of 15 meters is less than λ/2. Horizontal antennas, especially full-wave loops, are poor DX performers at these heights, as I'll come back to later in this article.

If you keep the apex height constant and further increase the bottom and top interior angles the antenna get narrower and, importantly, the average height increases. However beyond 120° the vertically-polarized diamond loop's Q rises sharply and the impedance gets low. This is similar to what happens in the chevron loop, but without the same benefit of exceptional low-angle gain at low heights. The antenna also becomes difficult to build since the tie-down points would have to be very far from the support. Therefore my choice of 120° is a reasonable optimization between greater average height and matching performance.

The SWR bandwidth of both the square and narrow diamond are sufficient to cover the entire 40 meters band. The above plot shows the narrow diamond antenna SWR for an apex height of 15 meters and cut to favour the CW band segment. Resonance can be shifted higher to keep the SWR below 2 across the band. A 75Ω quarter-wave transformer (RG-59 or RG-11) is used to match the high loop impedance to 50Ω coax.

All four legs are of equal length -- 10.98 meters -- and constructed of 12 AWG insulated copper wire. The antenna is 12 meters high so the bottom is 3 meters up when the apex is at 15 meters. The feed line can be run along one of the tie-down ropes to either side vertex.

With all the cautions and caveats of earlier articles on all the presented loop antennas I present the updated low-angle gain (DX performance) numbers to include both the above kinds of diamond loops.

Notice that the diamond square gain is indistinguishable from that for the delta loop. It also cannot be installed with an apex height less than 17 meters due to its greater height. This is not a good choice for a 40 meters antenna.

The narrow diamond loop's gain is not quite 1 db better, and so is midway between that for the delta loop and the chevron loop. It can be built lower to the ground (13 meters apex) than the delta loop and is less structurally complex than the chevron loop.

As the apex height is increased all the vertically-polarized loops top out at less than 2 dbi gain at 10° elevation. Gain is limited by ground losses and the appearance of second, high-angle lobe when above 20 meters height. Above 20 meters height an inverted vee outperforms all these loops. A dipole does better yet but would have to made from a full λ/2 of aluminum tubing (~20 meters long) to require just a single centre support.

The narrow diamond loop is moderately omnidirectional, as can be seen in the adjacent azimuth plot for this antenna at an apex height of 15 meters. As with the broadside gain, this antenna's omnidirectional performance is midway between a delta loop and a chevron loop.

When fed for horizontal polarization all the loops perform poorly at these heights. I did not bother to plot them for that reason. To give you some idea of what to expect the narrow diamond loop fed at the bottom vertex has a gain of -3.8 dbi or 1.9 dbi at 10° elevation for an apex height of 15 or 25 meters, respectively.

This is one reason behind the decades-long argument about whether a quad or a yagi is a better antenna. A loop has better gain but it typically also has a lower effective height. This makes its low-angle DX performance often no better than a yagi, and typically worse on bands below 20 meters. Depending on the basis of comparison either antenna can be shown to be deficient.

I will pursue this topic further when I model some 2-element loop arrays for 40 meters and compare them to the wire yagis I described in earlier articles.

Concluding Remarks on Loops for 40 Meters DXing

There is a lot of information in this set of articles on full-wave loops for 40 meters so I'd like to distill it to a few points and recommendations.
  • For utter simplicity the delta loop remains a good choice. It is omnidirectional and gives good DX performance at apex heights from 15 to 20 meters. The gain isn't great, which is in large part why it is omnidirectional: power fills what would otherwise be a side null.
  • For maximum gain the chevron loop does best, even at apex heights below 15 meters. The price paid is some loss of omnidirectionality, more complex structure and narrow SWR bandwidth. A second antenna to fill the deeper side null should be considered. Also beware environmental interactions at the lowest heights that can lower its actual performance.
  • The narrow diamond loop, as discussed in this article, is a compromise in gain and complexity between the delta loop, one that works well below 15 meters height.
  • Above an apex height of 20 meters an inverted vee or dipole is a better choice in regards to gain and complexity. This applies when the interior apex angle is at least 120°, otherwise the break-even apex height rises. The side null of these antennas is much deeper so a second inverted vee is needed for global coverage.
Choice of antenna is also influenced by various interaction, which deserve thoughtful consideration:
  • The greater the interior angle of the loop's apex the greater the desired separation between the loop and high-bands yagi above it. The delta loop therefore is a better choice if the separation is less than ~2 meters.
  • Vertically-polarized loops are susceptible to interactions with the tower/mast that supports the apex. It is a good idea to include the tower in the model for vertically-polarized loops. The tower plus mast I use for my delta loop does not resonate on 40 meters, so I excluded it in the models of antennas that interest me. It does however resonate on 30 meters.
  • Try to maintain at least 1 to 2 meters separation between the loops high-impedance points (where current is lowest) and the metal tower/mast.
  • The length of the antenna is dependent on how it's fed. When fed for horizontal polarization you should run a model of the antenna first, and do so at the intended height, since the length will require some adjustment. Usually this means shortening the antenna a small amount, often no more than 1% from the lengths modelled for these various vertically-polarized loops.


  1. Thanks for the article.
    Very interesting.

    When you compares several antenna performance at 10 degrees elevation angle, I'm just wondering what do you mean for a dipole. Is it a vertical dipole hung from top of mast/tower or a horizontal dipole?

  2. Ricardo, I do mean horizontal.

    73 Ron VE3VN

  3. how do you feed the vertically polarized narrow diamond loop? I can do balanced. Thank you 73 see you on 40 Dan K5PSO

    1. Feed it in the side corner, as shown in the article. It will be difficult to use a balanced line and avoid common mode current since it would have to extend horizontally straight away from the feed point. At least with coax you can use a choke at the feed point.

      73 Ron VE3VN


All comments are moderated, and should appear within one day of submission.