[Lf] Re: LF antenna gain]

[Andre Kesteloot]

James Moritz wrote:

> At 10:04 25/01/2002 +0000, you wrote:
> >...So nearly all the energy that goes into the ground is dissipated and
> >does not return to the feedpoint.  Therefore it cannot reinforce the
> >radiation pattern.  In that case, does the theoretical gain still hold?
> >
> >Gain is only obtained from directivity.  Directivity can be calculated
> >from physical considerations but the equation to obtain gain from
> >directivity is  G = e*D , where G = power gain, D = directivity, and e =
> >radiated power/total power...
> Dear Walter, LF Group,
>
> Not sure if I'm an expert, but...
>
> I think the losses that the signal is subject to between transmitter output
> and receiver antenna can basically be divided into 2 types:
>
> Firstly, the losses that are incurred in producing a current in the antenna
> in the first place. In the case of a vertical it is necessary to generate a
> huge voltage gradient between the antenna element and earth in order to get
> useful current to flow, and this field gives rises to large losses in
> materials in that field; the earth itself, buildings, trees; the antenna
> wire and it's immediate environment can be modelled as a lossy capacitor.
> The theorists say that it is the RF current in the wire that is actually
> causing the propagating E and H fields, so the losses due to the electric
> field are not really "radiation" losses and make up the "e" part of the
> gain formula. These are the losses which make the feed point resistance of
> the antenna greater than the radiation resistance - the power dissipated by
> the radiation resistance is that part which is being radiated away, and
> wether it is absorbed or carries on propagating, it does not come back and
> affect the impedance of the antenna
>
> Given that a particular current is produced in a wire of a particular
> geometry, the radiation pattern (and so the "D" part of the formula) can be
> calculated either analytically or a numerical result obtained using NEC or
> something similar. "Image" currents in a ground plane are taken into
> account. The radiation pattern for a monopole over lossy ground uses a
> model for the ground which, as I understand it, is a plane surface with a
> reflection coefficient of magnitude < 1 to account for the energy "radiated
> into the ground". This type of modelling is one way of producing ground
> loss curves. As PA0SE pointed out some time ago, The lossy ground model
> produces different radiation patterns at different distances from the
> antenna. The default radiation pattern displayed by EZNEC and shown in the
> text books is the "far field" pattern, mathematically the radiation pattern
> at a distance tending to infinity. For lossy grounds this inevitably has a
> null at ground level, because over very large distance, the ground wave
> will be attenuated very much more than the signal propagating through
> space, while a monopole over perfect ground has maximum signal at zero
> elevation. However, as the distance decreases, the attenuation of the
> ground wave is reduced, until at short distances, where the excess loss due
> to ground losses is negligible, it turns out that the monopole over lossy
> ground has much the same radiation pattern and D as the perfect ground case.
>
> This actually seems to be close to the truth - calculating the field
> strength produced by my piece of wet string with a certain current in it by
> using a NEC model consisting of perfectly conducting wire over a perfect
> ground gives results that are within a few dB (usually a few dB higher) of
> reality, over distances between 1 - 10km. It seems reasonable that the few
> dB's additional "site loss" that occurs could be explained by the obstacles
> close to the antenna which the radiated signal runs into - since higher
> antennas that reach above this local clutter have reduced site loss, at
> least on the basis of the very limited data that is available.
>
> So in summary: There are "antenna" losses, the power dissipated by the
> antenna and its near environment whilst a certain current is flowing, which
> are responsible for the efficiency "e" part of the gain formula, and which
> determine the loss resistance of the antenna.. The directivity D is
> affected by the distance from the antenna - close in, D is much the same as
> an ideal monopole, far away the ground losses modify the radiation pattern.
> Due to the very low values of e for all amateur antennas, G will always be
> a large negative number of dB, however. Calculating the radiation
> resistance and measuring the antenna current is the best easily done way of
> determining what ERP is being radiated, since it does not require knowledge
> of the antenna losses, only D, which is much the same for any electrically
> small vertical antenna, ie. 1.8 relative to a dipole in free space. So ERP
> = D.Iant^2.Rrad. In practice, the actual ERP may be lower by a few dBs
> "site loss"
>
> There are lots of loose ends to this - it means that the radiation pattern,
> gain, and so by definition the value of ERP, depends on the distance
> at  which it is measured where ground losses are significant. Also,
> modelling the ground as a 2 dimensional plane is not very satisfactory -
> currents and fields exist within the ground, as the cave radio guys well
> know! Significant LF radiation from "earth" antennas has been demonstrated.
> NEC just does not seem to model real grounds well where one part of the
> antenna actually has a ground connection. More work needs doing on the
> "site loss". The lobed radiation pattern of verticals shown in the
> textbooks is fine at HF, but at LF the "upwards" components of the radiated
> signal will be bumping into the ionosphere and coming back down long before
> the ground wave has decayed. One could spend a whole lifetime on this sort
> of thing...
>
> Cheers, Jim Moritz
> 73 de M0BMU