[Top] [All Lists]

Re: [TowerTalk] Problems on a 2el vertical Array

To: "Mark Beckwith" <>, <>
Subject: Re: [TowerTalk] Problems on a 2el vertical Array
From: Jim Lux <>
Date: Fri, 09 Jan 2004 11:54:12 -0800
List-post: <>
At 12:11 PM 1/9/2004 -0600, Mark Beckwith wrote:
> Hi guys,
> over Christmas I set up a 2el vertical Array for 40m.


>The pattern shows, that I should have at least 10dB
>on EACH station and 15-25dB f/b on DX stations.
>I heard only a few DX stations where I really had
>25dB f/b... but this was the big minority (maybe 5
>out of 100).  Often I didn't notice any f/b at all.
>(especially on stations closer than 2500km).


Where you may not be realizing your model is: The model probably allocates
power equally (50/50) between each vertical, right?

In real life when you feed it, the SWRs on the two verticals are different
because of the fact that there's a second energized vertical nearby having
some kind of effect on it, and each is being fed in parallel with something
else with a different length of coax, blah blah blah.  Gets too complicated
for me, anyway.

The model actually assumes equal element currents, and the Christman phasing scheme deals with the oddball mutual impedances, etc. I will say, having tried, that calculating the necessary lengths of coax analytically (by solving simultaneous equations, which involve hyperbolic functions) is extremely tedious. Much better to let a simple optimizer (like that in Excel) find the lengths by iteration. Or, you can use excel to calculate out the currents and voltages along the feedlines and look for a place where they are equal. (this is basically the technique described by ON4UN in his book).

If you are a glutton for punishment, I have spreadsheets for 2,3, and 4 elements to do the Christman technique. You'll need calculated or measured mutual Zs.

Another factor which doesn't help is when DX signals arrive from
non-straight-line directions or multiple directions.

This is probably the dominant effect.

The real-life way to get the power to split 50/50 is to put a tuner at the
base of each vertical and a tuner right before the TEE.  Once you get all
tuners to tune out all reflected power at all three points, then your power
is being split in half, and you will experience more like the F/B in the

You don't want a 50/50 power split, but, a fairly effective way to do the job is to put a tuner at each element and a tuner at the power divider.

I wish that LDG or MFJ would provide a version of their tuners that allows you to command the L and C over a serial line. Essentially a computer controlled tuner (but not an auto tuner). I modified a stack of LDG QRP tuners to do this (by feeding new software in at boot time), but it has some serious complexity problems (too many wires, to much fiddling around)

I also wish someone would make an inexpensive current probe that measures phase and amplitude (relative to some reference) and returns it over a serial line.

The people reading who think tuners are blasphemous will probably propose
other answers with a lot of strange symbols and words I don't understand.
Anyway, while they're going through all those histrionics, I'll be on the
air working guys.

Considering that the broadcast industry uses the tuner approach, it's probably a winner. The real problem is the user interface and the "variable angle of arrival" problem.

Congratulations Toby for getting your MFJ to actually make your parts
imitate your computer model.  You must be one of those Better Men.  I'll
take a parasitic array any time, mostly because phased ones drive me to
drink.  Hic.

You think that will drive you to drink, you should see the scheme I'm working on.. individual receivers and transmitters at each element with a wireless LAN for interconnect. No more coax for me, thank you! My feedline will be 110V extension cords.


See: for "Self Supporting Towers", "Wireless Weather Stations", and lot's more. Call Toll Free, 1-800-333-9041 with any questions and ask for Sherman, W2FLA.

TowerTalk mailing list

<Prev in Thread] Current Thread [Next in Thread>