Its almost impractical in reality. But if you model it, I am sure it will become clear. It will effect the impedance slightly, but you are also trying to produce a good low angle take off. Most verticals are modeled with the plain at right angles to the main element to produce the best results.
Jonathan
Hi Ed,
Can I quote the following Wikipedia entry from a damn good writer Arthur C Clarke?
Clarkeâs Three Laws are three âlawsâ of prediction formulated by the British science fiction writer Arthur C. Clarke. They are:
1 When a distinguished but elderly scientist states that something is possible, he is almost certainly right. When he states that something is impossible, he is very probably wrong.
2 The only way of discovering the limits of the possible is to venture a little way past them into the impossible.
3 Any sufficiently advanced technology is indistinguishable from magic.
Invoking Law No 1, if you lay down enough radials (a wire that runs along a radius from the antenna base) (eg 40 off 0.2 wave long) the ground type is not relevant. The wires simulate a sheet in which the current flows on the top side only hence the underlying ground is not involved. All the ground out to the rf horizon is however involved in determining the low angle reflections. Look for a huge salt water swamp with no trees on the SOTA summit. DXpeditions often pick low islands or put their aerials on the beach to get the best signals.
Otherwise three wires at 120 degrees or 4 at 90 degrees orientation 0.25 wave long, well above ground are required. As current flows on the whole circumference of the radials they will interact with any nearby conductor or material with a relative dielectric constant greater than 1 even if below the radial. 10 wavelengths is the rule of thumb for the extent of the near field of an antenna but the interaction for radials is much reduced if the spacing exceed one wavelength.
A counterpoise is a wire or several parallel wires to âbalanceâ the antenna. Used for standby or city building based AM band station antennas. The wires are insulated at the ends and usually placed symmetrically under the antenna. The antenna itself was usually a wire suspended between masts in a T arrangement. An ATU is required. In amateur circles it is used mainly in single wire form to stop the microphone biting the operator.
Invoking No 2.
Because the nature of the radial spacing and the ground condition are variable and generally unknown some experimentation is required for field ops to get the best result. Or you could use a current balun and ATU at the feed point and get on with operating.
Invoking No 3.
It might seem like magic but antenna science is sufficiently well known to be able to be modeled accurately on a PC. The ARRL used to distribute a CD with an antenna analysis program with every copy of its Radio Amateurs Handbook. Reading a handbook on antennas - one with some theory, not a collection of build it projects - is strongly recommended before you go pruning your radials or running computer simulations. If you donât under stand the basics it does seem like magic.
Any antenna that uses the word âmagicâ in its name or advertising is probably rubbish.
BTW the simulation programs are a great way of filling in a snow-bound weekend.
73
Ron
VK3AFW
Iâve learnt something, I didnât know this was the distinction between the two! Iâve not tried raising the 8 radials off the ground, except on some summits where they went up over stones, or across pits. These situations often led me to think just how the radiation pattern/antenna performance would be affected.
Go for it Jonathan! I was up at the Aston Hill trigpoint last night and suspect Iâve probably left something else up there, as I packed up in the dark with numb hands and no torch 
To be honest, this may be the case, all of my EA8 were over rocky ground so I canât really comment. Some summits did seem better than others, in terms of tuning and performance, so maybe the geology and physical properties of the rock come into play as well?
One strange summit (Senalo = EA8/LA-011) on Lanzarote, pretty much the top rim of a collapsed volcano surrounded by an otherworldly lava field for a few KMâs, caused my setup some issues. After a period of operating, static seemed to build up and not dissipate on the KX3 and I suspect the antenna system. This manifested by really loud static noise and a nasty shock when touching the KX3 chasis. This was the only summit I have ever experienced this on and my solution was to unplug the battery and leave the equipment for a bit, to bleed away, before packing up. Any ideas why this was caused, I suspect the rock didnât provide much of a ground?
Cheers,
James M0JCQ
The resistivity of basalt is of the order of megohms per centimetre, so no, it wonât provide much of a ground. In fact pretty well any rocky summit should be a very good insulator. You only get the resistivity decreasing at the depth where water from rain soaking down starts to permanently fill the network of cracks or joints in the rock, this water may become quite low in resistance from dissolving minerals out of the rock. In the UK the rock on summits will be frost shattered to a depth of several metres by the cold conditions of the ice age and I would guess that this means that while an antenna might be several metres above the ground surface it will be an additional several metres above the earth plane where water has decreased the resistivity. Furthermore this earth plane where it is determined by frost shattering will slope away from you parallel to the ground surface. OTOH if the summit is mantled in peat then you get low resistivity at ground level. Not much of a help but it can explain why antennas sometimes donât meet your expectations on a summit!
Brian
I think you hit the nail on the head and this was helpful as there can be only one conclusion:
On âgoodâ ground you can put the aerial and the radials directly to the ground. The longer the radials are, the better the ant will work as the radials couple capacitively (is there such a term in English?) to ground.
On âbadâ ground you have to get away from ground to minimise loss. Therefore you need resonant radials (1/4 lambda), the more radials, the better.
Ahoi,
Pom
Edit:
@M0JCQ
On the noise issue: Have you tried a resistor (> 10 kOhm) in the feeding point from âhotâ to ground? This should lead static to ground.
You have just contradicted yourself there ! Keep them to the design length ! Do not change them arbitrarily this will cause you problems.
Jonathan
Itâs two different situations. With an elevated antenna the radials will replace ground and then must be resonant.
Close to the ground they may be 1/4 lambda, but longer radials are better then.
Counterpoise before radio was about is defined as a balancing or equalising force or applying balance to a system. As such in radio terms, counterpoise is ideal. It provides the missing ground and lets the system be equalised or balanced. Without a counterpoise you end up with current on the outside of feeder and RF in the wrong places.
If you look in enough books and especially older ones, youâll see that a counterpoise doesnât have to be elevated or radial to a vertical radiating element. Itâs there to provide the missing earth. You see the term counterpoise and elevated counterpoise but is elevated counterpoise a tautology like DC current? Certainly when radials are mentioned you read about elevated radials as being something distinct to radials. But all of this maybe nothing more than how words are used or misused.
How long should they be? Elevated? Well you can buy/buiild designs that people say work and try them or you can run some modelling software and change things about then build something and see if the model is correct. The problem with antennas is the maths behind them can get quite hard quickly and itâs often difficult to quickly evaluate whether changes have worked. The availability of modelling software means you can try out ideas and see if the theory says they are better. Then you can try in practice but youâll need a lot of real trials to remove propagation differences etc.
There is also lots of antenna lore which may or may not be valid. But enough people repeat it enough then it becomes fact!
A particularly good book with lots of maths and theory is Services Textbook of Radio: Vol5: Transmission and Propagation by Glazier and Lamont. They say that ground planes exist to minimise the current on the feeder and suggest 1/4 radials but 1/8 wave will do if you have space issues. They say deriving the feedpoint impedance is not easily predicted and you should measure it so you can get a good match. Radials longer than 1/4 wave tend to affect the polar response such that the antenna lobes point more to the sky, something we normally donât want for good DX performance.