[TowerTalk] vertical antenna advice needed

[Jim Lux]

On 6/14/13 5:06 AM, PY1NB - Felipe Ceglia wrote:
> Hello folks,
>
> I'd like to hear your opinions about this topic.
>
> I am looking for an antenna (or more than one) for 10/15/20m that has the
> following characteristics:
>
> - vertical polarization

Well then, either a vertical element or a horizontal loop as your 
radiator(s).

> - smallest possible horizontal far field lobe

Not sure what you mean by this.  Do you mean you want low gain "at the 
horizon"?  That isn't usually what DX wants, where you want good gain at 
low angles.  Are you looking to suppress local noise sources that are on 
the ground?

> - "good" forward gain (at least comparing to a single element vertical)

gain or directivity.  Later you talk about using this for Rx only. For 
HF, receive noise is dominated by atmospheric noise sources, so gain 
isn't usually the big deal, what you want is good directional behavior?

> - greatest possible front to back ratio
Front to back, or good nulls? or front to sidelobes?

>
> I was thinking about using:
>
> - a tribander tilted 90 degrees

It has to be up fairly high so that the interaction of the elements with 
the ground doesn't bite you.

>
> - 2 element end phased array of:
>     - vertical dipoles
>     - sitting on the ground verticals (with or without radials?)
>     - multiband verticals (using W9AD phasing line to switch between bands)
>     - 5/8th wave verticals (at least for 10m, it would be easy to convert CB
> antennas)
>
> - a 4 square
>

All of these are generically the same: none have particularly narrow 
forward beam. 4 squares and the like are nice because you can quickly 
swing the null to put it on the interfering sources.

Two verticals with adjustable phasing can work very well to null signals 
coming from one direction, which may be all you need. There was an 
article in one of the antenna compendiums where an adjustable L/C 
network was used as a variable phase/amplitude shifter.

The W9AD scheme from array solutions is similar, using coax as delay 
elements (although the original scheme with adjustable L and C also 
allows adjusting amplitude, which will get you a much better null)


One thing to bear in mind on any of these approaches is that phasing 
errors don't change forward gain very much, but a small phase error 
turns a -20dB null into a -10 or -3dB null.  (this is probably the 
source of the 4square complaints about installed performance..)