Tout compte fait, une end-fed c'est peut-être une solution simple pas mal compliquée ;-)
Salut Jacques,
Je recopie ici une partie des commentaires d'un maître, Roy W Lewallen W7EL, l'auteur des diverses versions de EZNEC. Dans la dernière mouture de EZNEC + v 5.0.59, il inclut ces commentaires sur le fonctionnement des End-Fed et les directives à suivre pour les modéliser correctement et de manière fiable. Ça revient à dire que dans un montage En-fed, ""l'antenne"" est en fait le fil alimenté par le bout et aussi d'autres choses:
The current flowing out of one terminal must equal, in both magnitude and phase and at every instant, the current flowing into the other.
So why does a real end fed antenna work? The answer is that there's a path for current from the source terminal not connected to the "antenna". (I'm going to put "antenna" in quotes from here on to emphasize that the end fed wire isn't the whole antenna.) Let's look first at a single wire connected directly to the "hot" terminal of a transmitter coaxial antenna connector. The other, "ground", terminal of the connector is its inside surface -- the current flowing on that inner surface has to equal the current on the center, "hot" conductor. Current from the "hot" terminal can only flow to the single wire "antenna". But the current from the "ground" terminal can flow from the inside of the connector over the edge to the outside, and from there over the chassis. If the chassis is completely isolated from the Earth and any large or long conductors, the other "half" of the antenna will be the chassis itself. Like the EZNEC model with a source a half segment from the end of a wire, the transmitter will see a high impedance and there will be very little current flowing to the "antenna". In practice, there will also be some capacitive coupling to the Earth and probably other nearby wiring, which will increase the current somewhat. If the chassis is coupling to other wiring, that wiring is carrying antenna current and is a part of the actual antenna or radiating system. If you connect a "ground wire"-- a wire to the Earth, the mains ground system, or just a long wire to nowhere -- to the transmitter chassis, you've provided another path for current from the transmitter "ground" terminal. It now flows from the inside of the connector to the outside, over the chassis, and down the wire. The wire current consists of current induced by mutual coupling with the "antenna" and, if there's a path to the Earth, current conducted from the Earth which was induced into it by the "antenna". The real antenna is now the "antenna" wire and the "ground" wire. Both are carrying current -- equal amounts at the transmitter -- and both are radiating. Calling one "ground" doesn't change its properties -- it's just as much a part of the antenna as the intended "antenna". This is why adding a "ground" wire to a receiver often makes a marked improvement in reception when its "antenna" consists of a single wire. If you want a model to tell you what the real antenna system will do, you have to include the whole antenna system -- which includes the outside of the coaxial feed line and any other conductors connected to the transmitter chassis -- in the model.