I'll echo Jim Lux but be more specific. If you provide specific details,
you'll probably get pretty good advice.
How high is bottom of antenna/radial height?
How long are your radials? Insulated or not? Copper? AWG?
How are your verticals constructed: length, diameter, or segment diameters
and exposed lengths?
What was the impedance, X+jY, for an antenna with the other antenna not up
yet (or with coax disconnected at the antenna and open; not short
circuited, at the antenna)
[that means you should cal your analyzer at the end of a short piece of
coax]
What's your spacing between antennas?
How are your radials oriented, i.e. perpendicular to a line going through
both antennas?
What phase delay are you planning? -90, or have you optimized it for some
specific goal?
Did each antenna, with the other open or gone, measure nearly
identically to each other?
Are you clear of "nearby" metallic objects which may interact?
When you had all your coax connected, did you have the same X+jY whether
end-fired one way or the other (as they should)?
What are you doing for matching from the common coax point to your radio
impedance, as end-fire is surely quite different from broadside.
A good modeling program should let you know about how the system should
perform and whether your resonant frequency shifts, and whether it should
be a big deal or not. You want it "good enough".
But actual values, plugged into W7EL's software or AC6LA's software, will
be even better for matching. Both of those, I think I recall, focus on
figuring components or lengths for end-fire. Broadside can be figured by
shortening the long side to the same as the short side and making sure the
phase delay in both is the same; observing the common drive impedance when
in parallel, and then (at least for AC6LA's software) matching that for
50 + j0.
Getting proper antenna impedance is critical. Like I said, nothing beats
actual measurement. But you gotta do something with it.
ON4UN has a complete treatment of Drive Impedance, which is what those 2
programs referenced early use. Getting that is a process. All his editions
probably have it, but I'm looking at 4th Ed, p. 11-4 (Ch. 11 is phased
arrays). It talks about measuring one antenna with the other "open" (or
gone) and then also "shorted". And the Drive impedance is calculated with
that. My newer computers won't run his software any more, and the
calculations are complex algebra... so when I do them I use Python. I'm not
sure what others do.
regards,
Dave N3AC
Forest, Virginia
On Fri, Sep 5, 2025 at 7:58 PM R. David Eagle via TowerTalk <
towertalk@contesting.com> wrote:
> Hello all,
> Let me preface this by saying this - I am NOT a phasing expert by any
> means so I thought I would tackle a new antenna project this summer and
> maybe learn something....or lose more hair! So any ways...I decided to put
> up a pair of phased 1/4 wave verticals and phase them via the Christman
> method. Each antenna has 2 raised radials. I ran across an article by
> W4NFR where he did exactly what I am attempting, BUT I am getting different
> results....naturally. I tuned each vertical to 7.100 exactly trying to get
> the lowest possible reactance on each. I checked and double check the
> RG-213 71 degree delay line and the 84 degree feeds to the verticals. Now,
> here is the puzzling part - when I put everything together in my homebrew
> directional switch box, my resonant frequencies are considerably higher
> when I switch between Northeast, Broadside, and Southwest. The resonant
> frequencies are near 7.450-7.500 where R=39 with X=6.
> So I must be missing something here.....Anyone else had success with this?
> Any help would greatly be appreciated!
> DaveKB8NNU
> _______________________________________________
>
>
>
> _______________________________________________
> TowerTalk mailing list
> TowerTalk@contesting.com
> http://lists.contesting.com/mailman/listinfo/towertalk
>
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|