The Ladder Line Antenna

Matt Roberts - matt-at-kk5jy-dot-net

Published: 2021-05-05

Updated: 2024-02-08

If you build enough antenna projects, you will eventually collect plenty of spare parts and components. I have no shortage of such project debris at my house, including generous piles of wire and feedline.

For some time now, I have been looking at segments of hand-made ladder line hanging idle on my fence, wondering if it could be used for anything interesting. The feedlines are of high-quality, and have resisted many years of weather, but I simply have no use for them anymore.

Just for fun, I entered the dimensions into EZNEC+, to see what kind of antenna patterns could be produced by using a length of ladder line as an antenna, feeding and/or terminating it in various ways. Ladder line encloses enough area to make an effective receiving loop antenna for HF, but it can also do more interesting things.

Below is a discussion of using lengths of ladder line as a receiving antenna for the HF and MF bands. For convenience, the resulting antenna will be called a "ladder line antenna," or LLA. Depending on how the antenna is wired, it can be unidirecitonal, bidirectional, or omnidirectional.

The Model

To make a directional antenna out of a piece of ladder line, the circuit is simple: the ladder line is fed directly from one end, using a proper beverage antenna transformer to adapt the coaxial cable to the much higher Z₀ of the ladder line. The other end is terminated using a non-inductive resistor. The entire antenna is elevated a few feet off the ground, and the plane of the ladder line is held vertical.

The resulting antenna has an end-fire main lobe, pointing off the feedline end of the ladder. This means the direction of an LLA is the reverse of a typical long beverage antenna, which has its main lobe off the terminated end.

To make the same antenna bidirectional, similar to the SRL and LoG antennas, the terminating resistor is replaced by a short circuit, forming a continuous loop of wire. To make an omnidirectional LLA, similar to a vertical antenna, the terminating resistor is removed, leaving an open circuit at the end opposite the feedpoint. Since all of these changes can be made at the termination end of the LLA, converting between uni-, bi-, or omnidirectional configurations on a single antenna could be done with switches or relays.

As with many receiving loop antenna designs, the LLA is broadband, and if the length is selected carefully, it can be used unmodified on multiple bands. Unlike many of my other loop antennas, however, the terminated (unidirectional) LLA is impedance-matched to the coax. Like its beverage cousins, the terminated LLA uses the termination resistor to maintain a near-uniform impedance along the length of the antenna and into the matching transformer.

The length of the LLA is a trade-off between signal levels captured, and pattern beamwidth on shorter wavelengths — a trade-off that is not uncommon with loop antennas. Through some experimentation, I found that a 33' terminated LLA has a useful pattern for wavelengths of 40m and longer, while a 50' LLA is useful for wavelengths of 80m and longer. Such an antenna is not limited to long wavelengths, however. Shorter LLAs can produce directional patterns on 20m and shorter wavelengths.

According to the models, a length of λ / 4 is a good target for a terminated LLA on the shortest wavelength to be used on that antenna. A longer LLA starts to "crush" the peak of the main lobe in the azimuth plane. For a bidirectional LLA with a shorted termination end, that figure is closer to λ / 8.

Also according to the models, the best terminating resistor value is roughly that which will match the Z₀ of the ladder. For my 3.5"-spaced ladder, a 470Ω resistor produced the best overall patterns. This is a typical spacing for commercially-produced ladder line products, but other spacings can be easily made if desired. Wider or narrower ladders will likely require a different termination value. A wider ladder should generate more power output to the receiver per linear foot of line, since it encloses more are, since it encloses more area.

Figure 1: 33' LLA on 80m (Elevation)

Figure 2: 33' LLA on 80m (Azimuth)

Example model plots from a 33' terminated LLA on 80m are shown in Figure 1 and Figure 2 at right. Click to enlarge each plot image.

The Physical Antenna

The physical installation isn't much more complicated than the model. I used four-foot fiberglass electric fence poles to support the ladder, and hold the spreaders in a vertical orientation, which is important to the pattern cleanliness. I used two small guy lines, one on each end, to pull the end poles vertical, and the wires taut. The installation height isn't critical, but since I pushed the poles about 6" into the ground, the top wire of the ladder was at about 3'6", or just over one meter (1m) above the soil.

Using the computer models as a guide, I selected a length for the prototype that would give a nice directional pattern on 160m, 80m, and 40m: just over 33' long between the terminating resistor and the feedpoint transformer. This length was easy to test-stand in my front yard, where I have plenty of room to work without getting close to the permanent antennas behind the house.

Although the LLA might appear similar to beverage antennas, the LLA does not use the ground as one of the antenna conductors, and does not require any ground connection at or near the antenna. Neither of the antenna wires should not be connected to the ground. The coaxial shield can be grounded near the antenna if desired, but this is not necessary. Grounding of the coaxial shield at the building entry is always best practice for any antenna, for lightning safety and to drain static charges. Since the transformer provides a DC-shorted feed point for the coaxial cable, grounding the shield at the building entry point is sufficient to completely drain static charges from both wires.

The Electrical Antenna

The transformer is a hand-made isolated beverage transformer, wound for 75-ohm cable, following the same W8JI design used by DXE for their commercial transformers. The resistor chosen was the same 470-ohm, 2-watt, non-inductive resistor used by DXE in their beverage transformer kit.

To verify that the antenna was wired properly, and connections good, I did an impedance sweep of the antenna using an antenna analyzer. It showed a steady reading of 75Ω +/- ~10%, from around 1.2MHz up to above 20MHz. So the transformer, feedlines, and termination were all working as expected.

The overall gain of the antenna is quite low, even for a receiving antenna. To bring the signal levels up a bit, I added a DXE RPA-1 preamplifier, which has a gain of about 16dB. This is a bit more than offered by the preamplifiers in most transceivers, and the resulting signal levels were just about right. The S-meter was zero on band noise, but once the band became busy, the S-meter readings were S9+. I placed the preamplifier next to the radio, and this was more than sufficient; if one wanted to place an amplifier out at the antenna, this would be fine too, and might provide a slightly better SNR at the radio.

How Does It Play?

The prototype antenna was installed so that the center of the main lobe was pointed due east from my location in central Oklahoma.

I started testing on 40m, so that I could make the most of available FT8 signal traffic to get a picture of the directivity of the antenna. This is near the shortest wavelength for which the 33' antenna will have a good main-lobe pattern.

Initial results did not disappoint.

After several hundred reception spots on 40m, the resulting distribution of received stations showed a visible preference for stations to the east, and a weaker response for stations to the west.

As summertime is approaching, the 80m and 160m bands are not nearly as popular now as they are during the winter months. Even so, I was able to collect enough 80m reception spots to observe that the 33' antenna response on 80m was similar to that seen on 40m. This further supports the patterns described by the model, and shows that the antenna can be used unmodified on multiple bands.

Very few amateur operators have the ability to measure the front-to-back (F/B) ratio of their antennas, and I am no exception to this. Truly measuring F/B for an HF antenna requires equipment, and plenty of flat, unobstructed space, due to the very long wavelengths involved. However, since this antenna is experimental, I wanted to find a way to systematically use the data I could collect to get an estimate for the effective F/B I was getting on real-world signals. Models are great, but if the real installation shows no evidence of F/B, it may not justify the trouble of building and installing that design for most people.

I collected a few evenings' worth of FT8 spot data on 40m, and then used some scripts and math to extract an estimate of the F/B ratio of the antenna on that band. I used different termination resistor values on different days, to see how the model compared to the real antenna as the termination value was changed. The calculations used were a little involved, but the basics are roughly as follows:

Gather data from a couple of hours before sunset to a couple of hours after sunrise; this makes sure to cover the typical propagation cycle for both east and west directions.
Pick the best SNR estimate for each unique callsign from the data during that time, and exclude all other spots from that station.
Exclude spots from transmitters within an exclusion radius, to limit NVIS influence on the results.
Normalize each spot to a single metric that considers both the SNR estimate and station distance. This way, close transmitters with good SNR don't dominate the average over distant transmitters with lower SNR estimates.
Average the metrics from the front and back of the antenna separately; the "front" and "back" were defined by two equally-sized arcs, centered on the axis passing through the centerline of the azimuth repsonse of the antenna. Spots outside these arcs were excluded.
Use the ratio between the front and back metrics to estimate the F/B ratio based on the observed stations, their SNR estimates, and their reported distance.

What I found was that the termination value did influence the F/B, but the value wasn't as critical as I expected. Both the 470Ω and 680Ω terminations did well in real-signal tests. Straying from this range causes the F/B to suffer noticibly.

The best F/B estimate was with the 470Ω resistor, as predicted by the model. As expected, the F/B estimate varies with propagation and stations available to receive, but after a few nights of measurements, the value seems to be in the range of 6dB to 10dB. When you consider that the overall F/B is estimated over many different arrival angles, and that it is still competitive with the theoretical F/B for an ideal two-element yagi antenna, the experiment appears to be rather successful.

This antenna is still very early in its testing, but the initial results are good enough to warrant more testing and experimenting on other bands. Because of the variability in the F/B estimates caused by station participation, propagation, and all the exclusion angles used in the calculations, a better F/B estimate could be done if many weeks or months of data were considered, instead of just a few days. Likewise, the calculation of the per-station metric itself can be done in several different ways, since the received SNR and distance combination can be combined in different ways, and even together are an imperfect metric of true antenna performance for each station.

How Does It Work?

Honestly, I don't know. ;-)

However, there are a few things I can say about the design. The antenna is not dissimilar to other large or elongated loop antennas. Terminating such loops opposite the feedpoint is also not a new idea. VE7CA has experimented with antennas that are a more regular shape, but essentially the same idea, with a loop of wire, terminated on one side, and fed on the other. If you calculate the physical area enclosed by a 3' diamond antenna, and then compare it to the area enclosed by 33' of 3.5"-wide ladder line, they are very similar, so the difference between the 33' LLA and VE7CA's antenna is mostly one of geometry.

That may be one advantage of the LLA, which is that increasing the area inside the loop is very easy, and doesn't require making the antenna any taller. The LLA can be completely and easily hidden inside a yard with a privacy fence, with the top wire well below the top of the fence. Regardless of how long the antenna is made, the height of the antenna can remain below shoulder height, making it easy to modify and maintain.

Using relays to remotely lengthen or reverse such an antenna should also be rather simple, and the relay wiring could be run along the ground, to limit interaction with the antenna element itself.

Copyright (C) 2021 by Matt Roberts, KK5JY. All Rights Reserved.