At 04:24 PM 7/29/2006, Mark Beckwith wrote:
> > Modeling it is
> > simple enough
>
>I have modeled a lot of antennas for a lot of years, and I have concluded
>that varying the yagi height will change the take-off angle much more than
>any amount of varying element lengths.
I agree. That height parameter so dominates all the others, I think that
all you can really do with trying to do beam steering (in the vertical
plane) is possibly suppress something that is coming from the "wrong"
elevation angle. This is for relatively low (<60ft) installations. Get
high enough, and I think it starts to have more value.
If you think about it in the "pattern multiplication" scheme for doing
array patterns, you'd multiply the elevation pattern of an isotrope at the
relevant height by the elevation pattern of your array in free space. The
problem is that the "height derived pattern" has such deep nulls, that no
amount of directive gain from the array can overcome it.
>Varying element lengths may return nickles and dimes; varying the height
>will return $$$. If you want to control the take-off angle in a useful way,
>then stacked yagi arrays where you can control which antenna gets how much
>power at what phase angle will give the greatest control of this variable.
And even there, I suspect the big return comes from running the power to
the antenna that's at the "right" height. The steering you from phasing
them differently is going to be pretty small (a few degrees?). Placing a
null is a different matter. You could place a very sharp null at a
particular elevation angle, which might be useful.
The other case is if your array of SteppIRs is high enough off the ground
that you're in the delightful situation of the "ground reflection pattern"
being a big broad lobe. Get your array center a lambda or two off the
ground, and now you can do some serious fooling around.
Jim
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|