Steve Telenius-Lowe PJ4DX Has the lowdown on how to achieve an effective earth system for a quarter-wave vertical antenna, dispelling some myths along the way.

 

Most HF operators make at least some of their own antennas, even if they are ‘only’ dipoles or quarter-wave verticals. There is little that can go wrong with half-wave dipoles and usually they just need a little ‘pruning’ to size using an antenna analyser or an SWR meter, after which they work well.

Quarter-wave verticals are slightly more difficult to get right though, one reason being their need for radials. Unfortunately, you can't just put up a quarter-wave long vertical length of wire and expect it to work properly. The necessity for radials underneath the vertical to act as a ground plane, in effect a sort of ‘mirror’ producing an ‘image’ antenna, Fig. 1, – if you like, the missing half of a half-wave dipole – is well known. However, their necessity raises questions like “how many radials are required?”, “how long should they be?” and “should they be buried, laid on the ground, or elevated above ground?” These three questions are in fact all inter-related. Another question might be: “if the radials are elevated, how high above ground should they be?”

 

How Many Radials?

The usual answer to the question of how many radials is “the more the merrier”. Work done in 1937 by a member of the USA’s Institute of Radio Engineers, George H Brown, and his colleagues set the standard as 120 buried radials each 0.5λ (λ, the Greek letter Lambda, the being the accepted symbol for wavelength) long [1]. This was determined to be the closest possible to a perfect ground plane and it’s still the standard used by the Federal Communications Commission for medium-wave stations’ antennas in the USA.

This number is not practical for the average amateur and, indeed, 120 half-wave radials on 160m amounts to almost 10km of wire. Bought new, here in Bonaire that would cost over £2000, more than I want to pay for ground wires for just one band, even if it would improve my 160m signal by several decibels! If cost and physical effort are no object, then undoubtedly such a system will pay dividends. The Czech club station OK7K uses a 39m vertical on 160m and has put down an extensive ground radial system, Fig. 2. It certainly works well: on February 20th 2018 OK7K was perfectly readable here in Bonaire (8500km away) on 160m SSB before 2230UTC, 15 minutes before my sunset, while 15 minutes after sunset their signal peaked at S9+10dB, quite extraordinary for the distance involved.

www.ok7k.estranky.cz

 

How Long?

There's an old maxim that says given a particular amount of wire it's better to put down many short radials than fewer long ones. The 1937 paper states that radials should be 0.5l long but what if you only have enough wire available to put down, say, four 0.5l-long radials? Wouldn't eight 0.25l-long radials therefore be better? What about sixteen 0.125l-long radials? The answer in both cases is “yes”: radials need only be as long as 0.5l if you are putting down as many as 120 of them.

Over the years for some reason it became normal for amateurs to use quarter-wave radials, rather than half-wave ones. However, when buried or laid directly on the ground, radials couple into the earth so they don't actually need to be resonant at all and, within reason, their length is not critical. If you have put up a quarter-wave vertical with a few 0.25l radials buried or lying on the ground, you could probably improve its performance by shortening the radials and adding more of them – read on!

This raises another question and that is “how short can radials be and yet still be effective?” John Stanley K4ERO did some work over 40 years ago in order to answer that question. Summarising data contained in a professional publication, Radio Broadcast Ground Systems, Stanley wrote an article in the December 1976 ARRL magazine QST. His work is quoted in The ARRL Antenna Book [2] and Table 1 is an extract of data presented there which in turn was derived from information in Radio Broadcast Ground Systems.

This table actually provides us with a great deal of information, not least the all-important relationship between “how many?” and “how long?” and the findings are somewhat counterintuitive. For example, the fewer radials you put down, the shorter they can be. That’s not to say a few short radials will work as well as many longer ones – they won't! – but it does mean that if you only have a limited amount of wire to use for making radials, you can maximise your results by carefully choosing the correct length and number of wires.

Table 1 shows that if you only put down 16 radials, they need be no longer than 0.1l. You could make them longer than that but – with only 16 – performance wouldn't be improved by increasing their length: you would also need to increase their number.

With sixteen 0.1l-long radials you will get an almost perfect match to 50W coax, whereas if you go to the effort of putting down 120 x 0.4l-long radials, the impedance will fall to about 35W. As a result, your SWR will increase from a perfect 1.0:1 to about 1.43:1. This is nothing to worry about and just shows that a perfect 1:1 SWR is not always what you should be aiming for!

As the amateur ‘standard’ is 0.25l-long radials, what if you have already put down sixteen 0.25l-long radials on the ground? Interpolating between the figures in Table 1 you could expect to improve your signal by about 1.25dB if, instead, you put down 30 radials each 0.13l long. Note that in each case you are using the same total length of wire: 4l. If you had fewer than sixteen 0.25l radials to start with, the improvement should be somewhat greater. An improvement of 1.25dB may not sound a lot but, bearing in mind there is theoretically only 3dB difference between 16 x 0.1l-long radials (1.6l of wire) and 120 x 0.4l-long radials (a whopping 48l of wire), it is not to be sneezed at. (However, in practice there does seem to be more than 3dB difference between minimal and really extensive radial systems.)

The PJ4DX 80m vertical originally had 18 completely random length wires as ground radials. Some of the wires were considerably longer than 0.25l but after researching this article I reconfigured the set-up to 32 x 0.15l long wires, Fig. 3. I cannot state categorically that it has improved my 80m signal but at least I feel I am now making optimum use of the wire available.

 

Buried, On Ground, or Elevated?

Should radials be buried under the ground, laid on top of the ground, or elevated above ground? The easiest way is simply to lay the radials on top of the ground, whether this is a lawn, concrete yard or the flat roof of an apartment block. But if your vertical antenna is next to your beautifully mown lawn you may not want to have a lot of wires messing up the garden, so the solution is to bury them.

Here, there are two options. The first is to cut a narrow slot in the lawn with a spade, push the wire down into the slot and cover it up again. Fortunately, radials do not need to be buried deep, about a centimetre is sufficient. Nevertheless, this can still be back-breaking work if you are burying many long radials. The second option is to cut the lawn very short at the end of the growing season, push the radials down on top of the grass, holding them in place with pins, then allow the grass to grow over them. You will need to be careful not to cut the wires the first few times you mow the lawn the following season but, eventually, the worms will do their work and the radials will become buried.

Because radials lying on the ground couple into the earth, there is no electrical difference between putting them on the ground and burying them. The choice is just between the ease of simply putting the radials on the ground against the inconvenience of having wires all over the garden.

 

Elevated Radials

The alternative to both of the above is to elevate the radials above ground. This brings one major advantage but several disadvantages. In order to prevent the radials from coupling into the earth they should, ideally, be a minimum of 0.05l above ground. This means that elevated radials for an 80m vertical should be at least 4m high, or twice that height on 160m, in order for them to become ‘electrically elevated’.

Radials also need to be high enough to prevent people from garotting themselves on the wires. The ends of radials will have a high voltage on them when transmitting, so care should be taken that they cannot be touched by people or animals. This all suggests that elevated radials should be at least 2m above ground, regardless of the frequency band.

Elevating radials implies that the feedpoint of the antenna is also elevated. This is not a problem on the higher bands, although it is an issue on 80m and 160m, where it is difficult enough to put up quarter-wave verticals without also having the feed-point several metres up in the air.

If elevated radials are too close to the ground, their one major advantage is lost – the fact that far fewer radials are required, compared with a ground radial system. Most commercially-made VHF ground plane antennas have four radials, though there is little evidence that four provide any improvement over two, provided that the two are in a straight line (at 180º to each other). I use just two elevated radials, 2m high, on my 40m vertical, Fig. 4.

Unlike ground radials, elevated radials do need to be made resonant (electrically 0.25l long). To achieve this, first make a temporary half-wave dipole from two radial wires and position it where the vertical’s radials will be. Using an antenna analyser, the ‘dipole’ can be trimmed until resonant where required. Because it will be much closer to the ground than a half-wave dipole used for transmitting, capacitive loading from the ground will mean it’s likely to be shorter than expected. Once resonated, remove the coax, join the two wires together and there is your elevated radial system.

I said you can use as few as two elevated radials but that isn’t strictly true. You can even get away with just one, although the resulting vertical antenna will no longer be omnidirectional − there will be a slight null behind the single horizontal radial. If you think about it, a vertical with a single horizontal radial is just an inverted-V dipole with a 90º angle between the two legs, rotated on to its side.

 

A Recommended Read

In 2010 Rudy Severns N6LF wrote an article [3] about radials that has become an amateur radio classic. It was the culmination of 18 months of practical work (rather than computer modelling) summarising the results of numerous experiments with ground and elevated radials. It has been uploaded to the internet [4] and I recommend it highly.

Two of N6LF's conclusions are: “Laying down a system with at least 16 radials will give you most of the obtainable improvement [over no radials at all]. As we go to 32 and then 64 radials the improvement gets progressively smaller” and “...four elevated radials at a height of 48 inches are within 0.2dB of 64 radials lying on the ground.” There is a caveat to this, however. In a later article [5] N6LF pointed out that the smaller the number of elevated radials, the more likely that even a small amount of asymmetry in the system would lead to significant changes in the resonant frequency, feedpoint impedance, radiation pattern and radiation efficiency of the antenna. He concluded that “you cannot just throw up any four radials and get the expected results”. N6LF therefore recommends using at least 10 to 12 elevated radials rather than four.

 

Final Thoughts

Although the two 0.25l-long elevated radials on my 40m vertical have provided good results, having read [5] I'm wondering whether its performance would be improved by increasing the number to, say, 12.

I have not mentioned ground conductivity at all. Suffice to say there is a vast difference between operating a vertical over, or adjacent to, saltwater compared with the same antenna on poor, rocky, ground. However, you cannot influence your ground conductivity, whereas you can always improve your radial system. The worse your ground conductivity, the more important your radial system becomes.

 

********SUGGEST THIS SECTION IN A COLOURED / SHADED BOX*****

To Summarise

• There is no difference between buried radials and radials laid directly on the ground.
• Ground radials do not need to be resonant.
• For a given length of wire, use more short radials rather than fewer long ones. There is relationship between the two and an optimum number-to-length ratio.
• Counter-intuitively, the fewer ground radials you use the shorter they can be.
• Theoretically there is only 3dB difference between 16 x 0.1l-long radials and 120 x 0.4l-long radials (though in practice it’s usually more).
• Increasing the number and length of radials will increase the antenna system’s SWR but this isn't necessarily a problem provided your rig can cope.
• Elevated radials should be electrically 0.25l long.
• Elevated radials should be at least 0.05l above ground and, for safety reasons, not less than 2m high.
• Two or four elevated radials can perform as well as an extensive ground radial system, although a greater number is likely to work better.

********END OF COLOURED / SHADED BOX*****

 

 

References

[1] Ground Systems as a Factor in Antenna Efficiency, George H Brown, R F Lewis and J Epstein, Proceedings of the Institute of Radio Engineers, June 1937. This paper is available on the internet at k6mhe.com/files/GroundSystems.pdf

[2] Practical Suggestions for Vertical Ground Systems, The ARRL Antenna Book (Chapter 3, pp9 – 10 in the 2003 20th Edition).

[3] An Experimental Look at Ground Systems for HF Verticals, Rudy Severns N6LF, QST March 2010, pp30 – 33.

[4] www.nonstopsystems.com/radio/pdf-ant/article-antenna-elevated-radials.pdf

[5] A Closer Look at Vertical Antennas With Elevated Ground Systems, Rudy Severns N6LF, at http://rudys.typepad.com/files/elevated-ground-systems-article-final-version.pdf

 

Table 1: Six possible configurations for ground radials (source: The ARRL Antenna Book).

 

Length of radials   0.4l    0.25l       0.2l    0.15l    0.125l    0.1l

Number of radials 120     90            60       36          24            16

Impedance             35W    37W         40W    43W       46W         52W

Low-angle loss      0          0.5dB      1dB     1.5dB    2dB         3dB

 

This article was featured in the July 2018 issue of Practical Wireless