> Back when we had to submit an antenna pattern in order to get
> a repeater license for the ham bands, I mounted four elements
> of a DB-224 directly on one leg of a Rhon 25 tower and
> mounted the two tower sections on an antenna test pedestal
> and ran the pattern. With the antenna sections directly in
> line and pointed away from the tower, we had 9 dB gain in the
> favored direction, 6 dB gain at plus and minus 90 deg, and 3
> dB gain off the back side of the tower.
>
> The plot was perfectly round with the 3 dB offset for the
> center point of the plot.
I'm not trying to pick a fight here, but this seems to defy the laws of
physics. There would have to be nulls that were quite deep elsewhere in
the pattern aside from the four azimuths you cited if those values were
indeed valid.
Think about this, a grossy simplified example that would be difficult to
achive in the practical world, but valid in theory nonetheless.
Say you have an omnidirectional antenna that has some gain due to elevation
pattern compression (for example, a stacked dipole array like we're talking
about). Or it could even be an inefficient antenna (rubber duck) that has
negative gain relative to a dipole. The omni gain isn't really important,
it just serves as reference, but for sake of argument say it's a 4-bay
dipole array that has 6 dBd gain omni. So, 6 dBd would be 0 dBr
(r=reference).
Now we do something to alter the azimuth pattern which results in there
being a 90 degree arc that has 9 dBd (+3 dBr) gain uniformly over that arc,
two 90 degree arcs that have 6 dBd (+0 dBr) gain uniformly over those arcs,
leaving a fourth 90 degree arc. What's the gain over that remaining 90
degree arc assuming it, too, is unform across its 90 degrees?
If you answered 3 dBd (-3 dBr), you're wrong.
If you answered 0 dB (-6 dBr), you're wrong.
In fact, you'd be infinitely wrong if you answered anything other than
negative infinity dB's (i.e. ZERO radiation).
Using your examples/numbers above, and assuming a "perfectly round" pattern
offset from the center by 3 dB plotted as a typical logarithmic polar plot,
you're saying that over an arc of 180 degrees (the "forward" direction) the
gain was a minimum of 6 dBd (0 dBr), and a maximum of 9 dBd (+3 dBr),
correct? That would be impossible as you couldn't take enough power out of
that remaining 180 degree arc on the backside to ever balance out the
forward gain you claim to have achieved! There would have to be other deep
nulls elsewhere in the pattern for there to ever being a chance of those
gain values being being possible.
If it helps to understand this better, think of a two-way power divider. A
perfect two-way power divider would yield -3dB at each of its output ports
relative to the input signal. Let's say the input power is 0 dBm, so at
each port we have -3 dBm. Now say we modify the power divider so that one
port is 3 dB hotter than what it was originally, resulting in 0 dBm coming
out this "hot" port. What's left coming out the weak port? Hint: it's NOT
-6 dBm!!!
Antennas can't create power, they can only direct it where you want it to
go. An antenna that has 100 watts at its input terminals can't radiate more
than 100 watts total. It can concentrate more power in certain directions
(i.e. gain), but the sum-total of the radiated power can't be more than the
input power.
--- Jeff WN3A
