> The gain figures I quoted are dBd reference the dipole
> mounted in place of the DB-224 before the test. The
> Scientific Atlanta turntable was connected to a circular
> strip chart and the amplitude measurements were recorded
> directly to the strip chart which was submitted. The
> turntable was mounted on top of a five story building with
> the signal source about 500 feet away mounted near the
> ground. The tilt angle between the source and antenna to be
> measured is to prevent ground reflections from entering into
> the results.
Reducing, or compensating for, ground reflections is one of the biggest
challenges on an antenna test range.
> I don't see your point on where the energy comes from to make
> the extra 3 dB gain, as it obviously comes from the 3 dB
> reduction in gain on the back side of the tower compared to
> an omni antenna.
That's exactly the error I was trying to point out! If you add 3 dB
somewhere, you don't subtract 3 dB from somewhere else. If you add 3 dB
somewhere, you have to take away EVERYTHING from somewhere else. I tried to
use the power divider as an example, but I guess that fell flat.
In terms of power:
+3 dB is 2x
-3dB is 0.5x
-infinity dB is 0x
Maybe it's easier to understand in linear terms. Say you have a 100 watt
transmitter. You feed it into a two-way power splitter (50%/50%). One
output delivers 50 watts, the other output delivers 50 watts. Now we modify
the power splitter so that one port delivers 100 watts. What's left coming
out the other port? ZERO watts. ZERO watts is negative infinity dB gain.
We added 3 dB to one port by going from 50 watts to 100 watts, but you can't
subtract 3 dB from the other port to yield a valid result. If you had
subtracted 3 dB, that would imply you went from 50 watts to 25 watts out
that port. 100 watts out the hot port + 25 watts out the weak port = 125
watts, which obviously makes no sense because we only had 100 watts to start
with.
--- Jeff WN3A
