On 6/5/2010 8:36 PM, Art Unwin wrote:
On Jun 5, 6:59 pm, Art wrote:
On May 27, 6:22 am, wrote:



On May 23, 2:17 am, Art wrote:

I just modeled two coils counter wound and one inside the other as a
complete closed circuit , end fed. Counter winding is quite the norm
with special purpose machinery to prevent inter tangling of coils when
compressed.
With the coils wound opposite to each other
it made quite a nice antenna when placed on a flat plate. Beam was
pencil shaped without side lobes and sensitive to all polarities.Thus
any such dishes could be used for 'point to point' using any type
circular polarization. Smaller in size compared to other designs is
rather obvious. Where Kraus determined such an antenna single wound
form demanded a certain angular pitch such a requirement is not
apparent with this design. Kraus assumed a certain pitch was required
when wave composition clashed, but the opposite happens when particles
vectors add. Another example of particle versus waves disputes.It also
confirms Maxwells equation by removing the effect of lumped loads
which was not included in his equation as was not the stipulation that
a radiator must be straight.

Caution:
Above details not yet "printed" in a book so it must be considered
questionable.
Regards
Art

you have a nice little web site, please publish the design file and
picture of the antenna so we can all have a good laugh at it.

I expect to put some more things on my page
especially the model that I am honing now.
Now I have a 8 inch dia coil about 4 inches high relatively close
wound. Inside is the counter wound coil connected to the other coil to
make a closed circuit. It is placed on a perfect ground in leau of a
reflector. What I have done is balanced the inductance with the
capacitance, as you move around in frequency it will always be close
to resistive because of the balancing I have done.
It takes a while for the program to work from near
near resistance (.01 ohms) and I have now reached 11 ohms for 16.5 dbi
gain where the lower resistance gave me about 30 dbi
The resistance values are wave like as is the variation in gain per
cycle. Field shape so far varies from elliptical to something like the
double. zepp. My intention is to plot up to 20 ohms and that will take
several times longer than that taken already. The model has about a
dozen sections per
winding and I am pretty much limited to 10 segments per section. All
very interesting because I am following the idea of canceling all
lumped loads in line with Maxwells laws that deals only with
distributed loads. Frequency range covered up to now is 685 to 1200Mhz
With this amount of information you can start laughing now well
before I put it up
Art

just did a check at 3 Ghz. 20 dbi at 5 ohm resistive
so I may not get to 20- ohm resistive So after I get a reasonable
graph I will have to modify the model
to see if I can increase the resistance spam.
Note that the metallic resistance calculated is to gauge losses and is
not the source impeadance

I assume you are using AOPro, since that is what you have previously
stated when asked what you use for analysis. AOPro is an old version of
Mininec wrapped in an optimizer with 2D and 3D result presentation. It
has none of the current corrections to fix the mathematical mistakes
that were made in the original Mininec. Sometimes those mistakes are
relevant, sometimes not. In the case of an antenna like you describe,
they probably mostly are.

If AOPro say 5 ohms + or - j something, the 5 is the real part of the
feedpoint impedance. It is not the loss as you state.

The maximum wire diameters of your model seem to be impractical for use
with Mininec where it is not advisable to go beyond .001 wavelengths
from what I'm able to determine. Based on that assumption I get a
maximum usable wire gauge for 685MHz of 25, and for 1200MHz 30. At 3000
it's a number not even on the AWG table I have -38. So to model this
antenna at 3000MHz you should use 38 AWG or smaller wire for any chance
of accurate results.

You also stated that it is close spaced but gave no numbers, so I can't
comment on that. But I have a feeling that too may violate the limits
of Mininec's calculation engine.

Be careful of what you model and how you model it. Distrust miraculous
results not confirmed by other methods.

tom
K0TAR