[Elecraft] Modelling an EH Antenna

[Charles Greene]

Stuart and John,

I have done some modelling of my 20 M EH antenna in EZNEC as a short, fat 
dipole, with the physical elements of the EH antenna including the network 
and coax, and got results.  The network was at the antenna just like the 
real EH antenna and the coax a .5" wire to the antenna mast base. I plugged 
this network into the KM5KG Network program and assuming a Q for the coils 
and caps got the losses in the network, which were low.   So for kicks, I 
modelled the antenna without the matching network and got the R J 
components of the antenna, which showed a very low R and very high -J, as 
to be expected.  Then I used the KM5KG program to model a L network to tune 
the antenna with these  r J components and plugged this network in my 
short, fat dipole model in EZNEC instead of the EH network at the same 
point; i.e., at the antenna.  I had to tweek the L and C values quite a bit 
in EZNEC (which may be a modelling artifact) to get the short, fat dipole 
to load, but got a 50 +0 impedance.  The losses in the L network modelled 
in the KM5KG program were low too.  Not what I expected, but if valid, 
indicates a way to load a short, fat dipole without the loss with a L only 
loading network.  Prior communications with Roy Lewallen indicated that a 
reasonably short dipole can have a decent gain, but the problem is in the R 
losses in the loading network and the antenna itself.  I don't know of any 
way to measure or model the losses in the antenna elements themselves in a 
short, fat dipole and they may be high, but if there is a way to get the 
losses in the matching network down, that is a step in the right 
direction.  This shouldn't be a surprise for anyone, as L networks have 
been around for a long time.  I just wonder why they haven't been used more 
in the antenna itself.

I would be interested in exchanging model files for EZNEC and continuing 
this discussion for the sake of advancing out understanding of what is 
going on.

You wrote:

"John, a clarification please,  certainly the efficiency of the large
cylinders is large, taken as themselves alone, right?  That is the 98 per
cent efficiency you quote?   Because of the losses in the matching network,
(practical network which surely would have a Q of far less than 1000, )  you
then see the overall antenna efficiency as only 1 per cent, correct?  IE the
antenna has to be taken as a system and the inefficiency of the matching
detracts from possible efficiency of the whole.

The standard method of measuring small antennas is a Wheeler cap, a closed
cylinder that can be placed over a test antenna above a ground plane.
Measurements of Z of the antenna with and without the cap can be entered
into calculations to produce efficiency based on the low radiation
resistance component of Z, and the high loss resistances of the antenna
system.  We have such a Wheeler cap at work, and are measuring locally
produced FLEX antennas.   I hope to use the cap after hours someday to
measure an EH for which I collected parts.

Thanks for doing the modeling, it is very informative.   Very low radiation
resistance is a big hurdle, but even dummy loads radiate some.
73,
Stuart K5KVH"




73, Chas, W1CG
K2 #462