I need to correct a couple of statement I made about HFTA. I said HFTA
always gives 3 dB gain for any antenna stacking distance. That is not
correct. It gives 3 dB gain for very closely stacked antennas, less than
about 0.5 wavelengths. For other wider spacings it seems to function
correctly. Gains for closely spaced stacks, can be manually corrected
with data from NEC, but you will discover that anything less than about
0.4 wavelengths yields so little gain as to not be worth the effort of
stacking. N6BV recommends stacking no less than 0.5 wavelength for HFTA
because of the gain errors produced.
Also the gain error is because of the math models used for the stacking,
not the ray tracing algorithm.
Jerry, K4SAV
K4SAV wrote:
>Be careful when using HFTA to model stacks. HFTA does not model antenna
>interaction. All rays are assumed to go outward from the antenna
>between 0 and 34 degrees elevation. No rays directed upward at the
>other antenna are modeled. So HFTA will always give 3 dB gain for ANY
>stacking distance. This is not the way it works in the real world. NEC
>will show the correct amount of gain from the stack.
>
>The big advantage of HFTA is that it can handle irregular terrains where
>NEC cannot. Your ground can also affect the optimum stacking distance.
>In other words, an optimum stack over flat ground is not necessarily an
>optimum stack over sloping ground. You can use HFTA results up until you
>get to the point where you start to loose gain because of a stacking
>distance which is too close. You can determine this distance using NEC.
>If you do not have flat ground, using HFTA is to evaluate the stack is a
>good plan, because in some cases the terrain can have a huge affect. If
>you go closer than that stacking distance, calculated by NEC which
>starts to decrease the gain, you can manually correct HFTA gain numbers.
>
>Jerry, K4SAV
>
>
>
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|