What HF antenna should I erect?

The choice of HF antenna is vast so which shall I go for?

It's easy if you are wealthy with acres of real estate, multi-element horizontal beams mounted a quarter wavelength+ in the air over rotators. For the rest of us, the 99%, it's not so easy. I live in a typical semi with an area of around 35m x 10m with close neighbours so antenna options are a challenge.

Over the last eight months, since returning to the hobby after several decades absence, I must have explored many of the available HF antenna options Here's a simple summary of my initial conclusions.

1. Magnetic Loops

For some strange reason, maybe because I saw this as a difficult challenge, I jumped in at the deep end and began my return to amateur radio by building a couple of magnetic loops.

A magnetic loop simply comprises a conducting loop 0.10-0.25 wavelengths in circumference with a 25-50mm gap across which a tuning capacitor is connected. The loop can be fed by several methods, the simplest of which is a coupling loop 20% of the main loop diameter, mounted close to, but not in electrical contact with the primary loop, and opposite the tuning capacitor.

Fig.1 Magnetic Loop

The key requirement for an efficient magnetic loop antenna is that its resistance must be extremely low to display the high Q necessary to efficiently radiate RF energy. This consideration leads to the high cost and complexity often associated with this antenna. To keep resistance losses low the loop ideally needs to be constructed from relatively large diameter (5-20mm) copper tube, the tuning capacitor is usually motor driven and needs to be of the vacuum type, or at the very least the split stator variety, and all electrical connections need to be of very high quality.

As it turned out the route I chose, to evaluate the concept of a magnetic loop as simply and as cheaply as possible, worked rather well whilst still guaranteeing the very low resistance required. OK, this initial project didn't incorporate a motor driven capacitor for tuning, but I'm not sure one was needed as the following explains.

The first hurdle, to construct a 1m diameter circular loop using 15mm copper plumbing pipe, was much easier than I expected. A pipe bending spring was purchased from eBay for £7.50, and a 3m copper tube (cost £15) was bent into a near perfect circle using a wooden former mounted in a workmate.

Instead of connecting a variable capacitor across the loop gap a simple fixed capacitor was produced by using a gas blow torch to solder a 70cm length of RG214 coax across the loop gap. The target capacitance needed for this loop to resonate on 14.200MHz , as specified by an online loop design calculator, was 64pF. RG214 coax has a capacitance of around 100pF/m, so it was a fairly straightforward process to tune the loop by trimming the coax to the optimum length.

Fig.2 Simple fixed frequency magnetic Loop

Having a fixed rather than a variable capacitor wasn't a big problem. The finished loop tuned from around 14.170MHz to 14.230MHz. If one decided to use this design for a permanent installation it could be tuned to cover most of the cw or data portions of 20m, and two such loops mounted side by side, with a suitable gap between them to prevent interaction, could cover a sizable portion of the ssb segment of this band whilst still taking up less space than a dipole. Also, the top of a 1m loop only needs to be mounted 2m above ground level, a significant saving on the 5m minimum height needed for a dipole.

So how did this loop perform. I now regret that my trials with this loop were somewhat limited because I was very keen to erect a multi-band antenna to get fully operational while designing and then building a multi-band loop incorporating a variable capacitor as a longer term project. On receive the loop was very promising, not least because it's sharp tuning characteristic rejected a lot of the QRN ever present in the urban environment. The wire antennas that I went on to build often had noise floors with s-meter readings of 5-7 on some bands. This loop had a noise floor of 1-2 which I'm sure would have given it great potential with weak dx signals. Comparing the loop to a temporary 20m ¼ wave vertical with 8 buried radials, most signal strengths were the same or similar.

On transmit, only a couple of QSO's were completed (the antenna wasn't up long) with 59's recorded both ways into Europe.

Overall, I rated this experimental loop quite highly. It was cheap (<£30), much easier to build with the necessary low resistance than I expected, and very low visibility for the neighbours, but on the downside my limited evaluation didn't promise much better performance than that of a ¼ wave vertical. Then again, mounting it higher up and on a rotator might have been a different story.

I've now purchased a lot of the bits needed for a permanent tuned version, and my multi-band antenna installation is complete, so all that's stopping me proceeding with a mk2 version is the biggest hurdle of all – motivation.

2. ¼ wave verticals

I erected ¼ wave verticals for 10m, 20m, and 40m. All were installed with eight buried ¼ wave radials, and much to my surprise worked well even though they were positioned close together (2-4m spacing).

Many QSO's were recorded, including quite a few with North and South America and the Caribbean on 40m, and I would have been happy to use them on a permanent basis. However, no VK/ZL contacts were made so there was certainly scope for improvement. I also thought the 40m vertical was too high for good neighbour relations, and wanted to be active on 80m/160m so my attention focused on to horizontal antennas. The 10m and 20m verticals have been retained as “back up” antennas and for comparison against newer antennas.

2. Multi-band end fed inverted “L”

The successful version of this antenna that I finished up with was 32.6m long, 15m high at its highest point, and fed remotely at the bottom of the garden using an old LDG automatic atu mounted in a water proof box.

Attempts to feed this antenna with various baluns and ununs using a coax feed and an atu in the shack had proved universally disastrous. The SWR's were reasonable on a couple of bands but high on most. Losses in the 15m coax feed must have been high, and on receive the antenna sounded “dead”, often without the characteristic rise in background noise one expects to hear when an antenna is tuned to resonance. I learned a golden rule at this time.

NEVER FEED A MULTI-BAND ANTENNA WITH COAX UNLESS THE ATU IS AT THE FEEDPOINT.

This may not be a universal rule, for example a short length of very low loss coax may be tolerable, but as a rule of thumb I think it's a rule I will stick to in the future.

With the atu in the correct position, at the feed point, I was much happier with the performance of this antenna. It tuned up on all nine HF bands although several were 1.9 SWR. Comparing it with the 10m/20m/40m ¼ wave verticals that were all still up when it was erected, the inverted “L” was perhaps a little down compared to the 40m ¼ wave, but was generally similar in performance or sometimes better, compared to the 10m/20m verticals.

A minus with this antenna was the need for an extensive buried radial system.

2. Ladder fed multi-band doublet

My time with a remotely fed / remotely tuned inverted “L” was cut short when the outside atu stopped working. Faced with the prospect of spending several hundred pounds on another outdoor atu I decided on the cheaper option of a ladder line fed doublet, which would have the advantage of not needing an extensive radial system to maintain, and could be tuned by a manual atu in the shack without the worry of coax feeder losses.

The commonly employed 65ft x 65ft doublet was erected, fed with 38ft 6in ladder line. To fit this within my boundaries it was necessary to allow the outer 15ft of each top to hang down vertically from the end supports. Each doublet half and ladder line half was made from one 103ft 6in length of insulated wire so waterproofing should never be an issue with this antenna as there are no joints anywhere. The ladder feed, spaced with pvc tubes giving a 40mm spacing and 450 ohm impedance, is fed directly into an MFJ941E manual tuner in the shack.

This antenna has only been up a few days, and I've already received a 58 report from North America on 40m, so performance seems just as good as that of the previous inverted “L” antenna. On receive, it is 2-3 S points up on the 10m ¼ wave vertical, very encouraging. Ten HF bands tune up easily too, so hopefully I can now spend more time in the shack and less time on the roof in the future.

So which antenna is the best?

To be honest, I'm not sure. It goes without saying, none of these antennas can match a beam yet they all give similar performance with which I'm more than happy. After all, it must get boring working VK/ZL every day. For mono-band use I quite liked the 40m ¼ wave vertical, but for all band coverage it has to be one of the horizontal wire options, and from a cost and maintenance point of view, I guess the doublet becomes the clear winner.

What's next to try?

Perhaps a quad or a loop might be worth looking at next, but I think I'll stop there for the time being and concentrate on operating.