Too simple a formulation of issues.
The loss is ground induction, by *whatever* means. You appear to be
thinking that the radials are the only source of induced current loss in the
ground. The vertical radiator has that kind of loss as well, if you are
not talking about *dense* uniform radials. The placement of the current
center and the actual effective amperage in conductors near the ground is
what matters, not what Z you manage to get with the various feed Z
controlling devices. EZNEC is adding them all up. The individual "nuggets"
of loss that get added up are roughly proportional to amps in conductor
segment times 1/d squared where d is distance from wire to the dirt.
Effectively it adds up those from each single wire segment to each single
square foot of dirt.
Don't use an inverted L over four quarter wave radials. The induction loss
from those is substantial. Only *dense* uniform radials are without ground
induction loss.
73, Guy
On Wed, Oct 19, 2011 at 1:20 PM, wyc <wycpublic@gmail.com> wrote:
> I did some EZNEC modeling of a planned inverted-L, and came up with some
> results that surprised me, and caused me to re-think my understanding of
> Jerry Sevick's work.
>
> I modeled an inverted L over perfect ground and got a feedpoint impedance
> of
> about 13 ohms. I then switched to real ground model, and added a few
> radials. The feedpoint impedance went to about 50 ohms.
>
> This is expected. I've heard it explained as, effectively, a series
> circuit
> of the 13 ohm antenna impedance, and 37 ohms of return path resistance
> through the ground. (Three-fourths of the TX power is 'warming the
> worms').
>
> So far so good.
>
> I then repeated the exercise with a folded-inverted-L (i.e., with
> twinlead).
>
> As is the case with a folded dipole, the feedpoint impedance of the folded
> antenna over perfect ground is about 4 times the impedance of the antenna
> with a single wire, in this case 50 ohms. Again no surprise.
>
> What then surprised me was the feedpoint impedance when I switched to real
> ground, and the same few radials. Instead of going to about 87 ohms (50
> ohms ideal antenna impedance plus 37 ohms of return path resistance), it
> went to about 200 ohms.
>
> Thus, with the higher impedance antenna, the return path resistance (ground
> loss) now looks like 150 ohms, instead of 37 ohms.
>
> In one way, this makes sense. The radials don't do their job better simply
> because they are used with an antenna having a higher nominal feedpoint
> impedance; there must still be a return path for all the antenna current.
> The same three-fourths of the TX power still warms the worms with the same
> radial system.
>
> But it contradicts an earlier impression I'd had - that ground loss becomes
> more of an issue with lower impedance antennas. I'd had this impression
> since Jerry Sevick's 1970s articles about really short verticals, where he
> stressed the importance of really good radial systems because of the
> antennas' low antenna radiation resistances.
>
> His articles give a familiar calculation of antenna efficiency as the
> quotient of radiation resistance divided by (radiation resistance + ground
> loss + ohmic loss in antenna). This formula assumes ground loss is a fixed
> ohmic value. But the EZNEC data suggest this view is simplistic. While
> ground loss with one antenna may look like 37 ohms; with another antenna it
> may look like 150 ohms - all with the same radial system.
>
> So now I'm thinking that my earlier understanding of Jerry's work, that
> antennas with lower radiation resistances require better ground systems, is
> wrong. At least if the EZNEC models are to be believed.
>
> Can anyone give the technical reason that ground return path resistance
> seems to vary in proportion with antenna radiation resistance?
>
> Tnx,
>
> /Bill, K2PO
> _______________________________________________
> UR RST IS ... ... ..9 QSB QSB - hw? BK
>
_______________________________________________
UR RST IS ... ... ..9 QSB QSB - hw? BK
|