RE: NEC-LIST:helical ant. modeling

Hi,
My hands-on with a 22 turn helical at 435 MHz highlighted two aspects:
Helicals suffer severe gain saturation if constant radius/constant spacing
is applied. This will presumably emphasise the side lobe radiation and might
also degrade the back-to-front ratio? Maximum length for such a helical is
around 10 turns.

The workaround I applied was to group turns of this helical into 10 at 1.1
circumference for maximum gain, 6 at 1.0 circum, and the rest at .95
circumference while maintaining the same spacing. This increased the gain by
~6 dB and the resulting performance exceeded that of a yagi of the same
length based on Gunther Hoch's "Extremely Long Yagi Antennas" published in
VHF Communications in the 1980's.
Long helicals (22 turns) in gain terms are at least equal to, or better than
"standard" yagis of the same boom length, around 4m in this case.

If circular polarisation is required, antennas such as crossed yagis do not
compare, even in narrow band operation.

Impedance hardly changes with different helical element diameters - for
wideband operation avoid a lumped constant type of match and concentrate on
something such as a sloping line impedance transformation of sufficient
length. Impedance is more related to circumference of the turns.

In the case you mention of constant circumference, gain saturation will be a
problem. On such a physical structure the sole parameter available to avoid
this is to use a changing pitch angle (spacing between turns): either
grouped turns, or based on a parameter such as a log spacing function.

You might do a search for the Web site of Mike Cook (AF9Y) who published a
helical design based on a NEC variant for use in receiving a NASA probe at
435 MHz (Mars Surveyor?)

Please let us know what you discover.

Cheers,
Ian zs6bte

-----Original Message-----
From: Trevor Marshall [mailto:tm_at_well.com]
Sent: 04 October 2001 05:30
To: nec-list_at_gweep.ca
Subject: Re: NEC-LIST:helical ant. modeling

Cornel,
The approximation I have seen most often used is that the modelled
equivalent radius of the strip equals one quarter of the width of the
strip. This is not a theoretically rigorous answer, but it should get you
reasonable results.

I have modelled helicals on NEC2, and the results are an indictment of the
poor theoretical basis for reported helical gains. NEC found an optimum 6
wavelength helix had only about 14 dB circular gain, much less than is
often claimed. The reason seems to be excessive intensity in the sidelobes.
The MINIMUM off-axis intensity was -12db in the centre of the 'nulls', only
26 dB down on the main lobe! Major sidelobes are at -8dB and -14 dB at 30
degrees and 45 degrees off axis. A plethora of other sidelobes reach -20dB
on the main lobe.

Kraus calculated gain by looking at beamwidth, and I think the wasted power
going into the sidelobes gives the lower computed results.

Sincerely,
Trevor Marshall

>Hello ,
>
>I try to model a helical antenna, made of a printed line on a flexible
>substrate that then I roll onto a cylinder.
>Does anybody know of an equivalency "printed line width" - "wire
>diameter", so I can accurately obtain some impedance information?
>
>Thank you,
>Cornel Gazdaru

-- 
The NEC-List mailing list <nec-list_at_gweep.ca>
http://www.gweep.ca/mailman/listinfo.cgi/nec-list