[TowerTalk] Stacking Tribanders

[Lux, Jim]

On 6/22/22 12:18 AM, Jim Brown wrote:
> On 6/21/2022 7:37 PM, Billy Cox wrote:
>> So outside of the cautions Dean shared, and used
>> with other methods (EZNEC/etc.) why would one not use
>> HFTA as a useful software tool for stack planning?
>
> Because as KK9A explained, it doesn't model interactions between the 
> two antennas, which can be considerable. HFTA is a very useful tool 
> FOR WHAT IT WAS DESIGNED TO DO. Caps for emphasis. These actions come 
> in the form of currents induced and impedances coupled between 
> elements of the coupled antennas.
>
> Those who are saying this are well-educated engineers. And Dean was 
> the editor of the Handbook and Antenna Book when he developed HFTA. 
> Ward Silver took the reins when Dean retired. I know both men, have 
> worked extensively with Ward, and both are excellent engineers.
>
> Software like NEC can model antenna interactions, producing a full 3D 
> pattern over a flat earth or in free space, but it cannot model 
> terrain, and there's no way to transfer that to HFTA in a meaningful 
> way, because each of the two antennas interact differently with the 
> terrain depending on their 3D location with respect to that terrain. 
> FAR more complex software is required to that, the terrain data must 
> be plugged into it. One program that MIGHT be useful is called Hobbies.
>
>
This raises an interesting point, though - is there a "simple" 
methodology to compute far field patterns including the effects of 
terrain (e.g. from HFTA) for a stacked array, where you have the far 
field patterns for each of the stacked antennas.

It's a bit tricky, because of the mutual impedances between the 
antennas, so you can't just model them individually and combine using 
HFTA data for the height to get a weighting factor for magnitude.  Nor 
can you just put the antennas in the model and excite only one, then 
excite the other, since HFTA sort of assumes a point source radiator.

Complicating this is that most antenna modeling tools return the pattern 
with the origin at ground level.

Could you use a freespace model, centered at each antenna in turn, and 
then combine those with magnitude data from HFTA.  This would assume 
that the loading effects of the soil on the antenna are negligible - for 
a 10/15/20 Yagi multiple wavelengths up, this is probably a reasonable 
approximation.   I can think of ways to de-embed soil effects, but it's 
complex.

Another way would be to get near field data, and transform it somehow.


One other thing to think about is that H-pol antennas are really only 
H-pol on boresight.  Over ground, they have significant V-pol response 
off the sides.  (A dipole close to the ground is a fine V-pol antenna 
off the ends of the dipole).  HFTA doesn't model V-pol at all (it's a 
lot more complex, because you need to know the soil properties, while 
for H-pol, you can sort of assume total reflection)


It's an interesting question.