Topband: Coax Shield vs. Ground & elevated radials

[Telegrapher9 at aol.com]

When elevated radials are mentioned they are often full 90-degree radials. 
The low impedance presented by the radials makes decoupling the coaxial cable 
shield easy. The counterpoise current will divide between the radial impedance 
and the shield impedance. 

But what about short elevated radials? Very short radials can present a 
substantial reactive impedance and make decoupling the shield a non-trivial task. 
Four 50' elevated radials placed 10' above the ground present a reactance of 
approximately 200 ohms capacitive. To decouple the shield so that the shield 
current is 12 dB below the radial current (my arbitrary number) the balun (unun) 
needs a CM impedance of about 800 ohms. Alternatively one can place a 
reactance of 200 ohms inductive in series with the radials. I use the the latter 
technique. 

Tuning the radials can be performed with a clamp on RF current meter such as 
MFJ manufactures. Since my radials originate in the shack I use an MFJ tuner 
with the ground tuning feature. I a series inductor to get the radials close to 
resonance and use the ground tuner to tune for maximum radial current. 

For roof mounting a top band antenna with 'floating' radials the series 
inductor can be a great help in keeping the counterpoise current on the roof. For 
this I like to simulate the antenna like so:

1) ground mount the antenna on perfect ground and resonate it. 
2) move the antenna to the roof and change the ground to average.
3) Place the roof counterpoise in the simulation and series tune it so that 
the antenna input impedance has zero reactance. 

Build the actual antenna with the simulated counterpoise inductor value. This 
will get things close enough so that tuning things up is not such a chore. 
Even if the counterpoise inductor is not adjusted - is left as per the 
simulation - the counterpoise will present a much lower impedance than short untuned 
radials. Then the balun CM impedance need not be as high to keep the 
counterpoise current on the roof. To determine the approximate balun CM impedance needed 
I add a 90-degree radial in simulation. I then add balun impedance in series 
with the radial (R + X) until the 90-degree radial current is 12 dB below the 
counterpoise current. This procedure allows us to design the balun and 
counterpoise on paper, assemble things, and avoid a some of the cut-and-try emperical 
design work. 

The tuned short radials are used on several commericially built amateur 
antennas. I think that Moxon advocates resonated short radials over 90-degree 
radials. 

Has anyone slese had good success with tuned radials? 

    Dave WX7G