Topband: Tuning elevated radials

[Ford Peterson]

I'm working on building an elevated radial vertical.  After installing the radials, I have a question about what I'm finding.

Using a small toroid that I split in two, I wound 9 turns on each half, glued the halves to the inside of a plastic clothes pin, and installed a full wave bridge using 1N914s and a 0.01uF cap.  The small clip attached to an elevated radial provides a relative voltage indication, which provides a current sense.  With 100W into the antenna, a reading of 4vdc to 25vdc is obtained from an individual radial.  The readings a quite repeatable using a digital VOM.

There are three elevated radials, each about 1/4wL and each about 120 degrees apart.  At resonance, I find that the current was not well balanced between the radials.  So I started trimming.  It turns out that by trimming the highest current radial, the system came into relative balance.

By trimming the excess length off a couple of radials, I found that I could balance the relative current reading pretty good.  At resonance, this is what I obtained.

                      trimmed         pre-trim
North radial = 10.0 vdc       21.8 vdc
SW radial =    11.0 vdc       10.5 vdc
SE radial =      10.5 vdc        5.6 vdc

Here is what surprised me.  

A 1.4% change (up) in frequency moved the trimmed values to:

North radial = 12.7 vdc
SW radial =    11.3 vdc
SE radial =       8.0 vdc

A 4.2% change (up) in frequency moved the trimmed values to:

North radial = 17.2 vdc
SW radial =      6.2 vdc
SE radial =       4.5 vdc

Now admittedly, this is somewhat kludge in that the 706 I was using to drive the antenna was shutting down due to a bad load, but it is the relative reading that I believe to be important.  I was expecting to see the values remain somewhat constant over frequency after trimming.  Instead, I find a nearly 4:1 difference between them.

The goal of the process was to minimize pattern skew.  I'm thinking it might be quieter if I achieve a good null at zenith.  Perhaps this is a twisted notion to begin with.  I find that the position of the radials is very important.  The SE radial needs to snake its way through some bushes and repositioning that radial has a significant impact (negative) on the current carried by that radial.

Is this normal for an elevated vertical?  I know I am splitting hairs on this, but I'm trying to understand better what is happening.  Modeling using NEC2 does not seem to predict these behaviors.

Any empirical evidence from others who have experimented with this would be appreciated.

73

Ford-N0FP
ford at cmgate.com