On Sat, Sep 5, 2009 at 2:02 PM, Ed Stallman<firstname.lastname@example.org> wrote:
> Carl, A year ago I bought the DXE-ATK65 and like you felt the 65ft
> was to short, so I added 2 ft to the top hoping that would get me to
> 3520 of about. with 8 radials on the ground , the length and match
> was good. Each time I added 8 more radials the resonant freq went up
> the band . After 32 radials were in place the match was up around
> 3620 and the 2:1 bandwidth was not that good. Using 6 turns, 1/4 in
> copper tubing , wound on 2.5 inch PVC got me to 3510 with a 1: 3 SWR
> / 38 Ohms . The bandwidth was still not that good , all I can think
> of is the narrow width tubing
In situations like yours the antenna system losses *decrease* as
radials are added. Without the losses which are equal opportunity
absorbers of all frequencies, the SWR "curve" narrows and gets steeper
sides. This is a good thing for performance.
For a single vertical with an extensive commercial grade radial
system, a tuned spot-on-to-50-ohms system does not have an SWR
"curve". Measured right at the base with no coax, it has an SWR "V",
with the sides of the V coming right down to 1.000:1 at the desired
frequency. I was quite surprised the first time I saw this on a
graphing analyzer with no "U" shape at the bottom.
Other than getting cranky linears with cost-saving restricted matching
networks to put power into these antennas without faulting off, SWR is
vastly overrated as an indicator of performance. A dummy load has
absolutely the best SWR going, and very broad.
Part of preparation for the kind of curve to expect is to model an
antenna with perfect conductors and perfect ground. This will give a
theoretical best bandwidth when low losses are achieved. Any broader
than that and one is converting some portion of the RF to plain old
heat and only transmitting the left-over power, most likely with a
"nicer looking" SWR.
Topband mailing list