[Antennas] Unused resonant antennas - shorted to ground or floating?

[Terry Conboy]

On 2009-12-24 9:03 PM, Jim Miller KG0KP wrote:
> I can do that (use inductors across the feed point to the radial field(and 
> ground them at the station)) and what I had in mind was a pair of phased 
> verticals (actually Butternuts) and looking at 90 degree separation on 80 
> but also realizing I will probably be better off with an omni pattern on 
> bands above 40 (20 maybe) so wanting a way to inimize the effect of the 
> unused vertical.  Still have to try modeling it but I am in kindergarten on 
> modeling.
>   
Using an inductor to ground and disconnecting the coax center conductor 
at the feedpoint is a good approach.  Since your Butternut verticals are 
resonant on all the bands, you definitely want to present a fairly high 
impedance to ground for RF to prevent high coupled currents and omni 
pattern distortion.  It's amazing how much RF current flows in a 
grounded resonant antenna due to mutual coupling, even when the unused 
antenna is over 1/2 wavelength away.

However, in other situations, a high Z to ground can be a problem.  For 
example, a simple 1/4 wavelength 40m vertical left floating will 
resonate as a 1/2 wavelength dipole on 20m and couple strongly when 
operating on a nearby 20m vertical.

As for where the high impedance is placed, it either needs to be at the 
feedpoint or a multiple of an electrical 1/2 wavelength down the 
feedline.  If the high Z is 1/4 wavelength (or odd multiple) away, it 
will effectively short the feedpoint to ground for RF.  For a single 
band antenna, you could short the coax at 1/4 wavelength from the 
feedpoint, but then most of the higher bands would then appear shorted 
for RF since the coax would be a multiple of 1/2 wavelength.  This 
probably isn't a good idea with a multiband antenna such as the Butternut.

Modeling trapped antennas like the Butternut can be quite challenging.  
You need to wait until at least the second grade to try this!  Of 
course, once you're in the first grade, you can get a pretty good idea 
what happens to the pattern by modeling them as 1/4 wavelength verticals 
on each band to see the effect of varying the feedpoint termination 
impedance.

73, Terry N6RY