Here's what I did to investigate John's scenario.
First step was to create a model of the C-3S using variables to
control the wire coordinates. The AutoEZ "Create Stepped Diameters"
dialog makes short work of building Yagi models that have elements
constructed with telescoping tubing. For example, here is what the
dialog window looks like for the 20m Reflector. Variable "M" controls
the half-length for this particular element. Other variables control
the boom spacing and height. There is no need to make *any* entries
directly in the wires table.
https://s18.postimg.org/p6f1g15rt/N4_EEBa.gif
The antenna was placed at 33 ft over average ground and fed via 97 ft
of LMR-400-UF. Line loss is included in the calculations. Using
John's initial mod of 10m Dvr tip section from 36.5" to 37.25"
(element half-length from 99.5" to 100.25") I ran a sweep from 28 to
30 MHz (green trace below). Then I lengthened the 10m Dvr from
half-length 100.25" to 104.50" but left the 10m Ref at the factory
half-length of 106.5" (blue trace). Finally, Dvr half-length same at
104.50" and Ref half-length increased to maintain the original Ref/Dvr
ratio, half-length 112.0" (red trace).
https://s18.postimg.org/c29h3byah/N4_EEBb.gif
Note that for John's current settings (green trace) the model doesn't
quite match his reported SWR points but it's close.
Here is a comparison of Free Space azimuth patterns at 28.3 MHz for
all three half-length combinations.
https://s18.postimg.org/xbx3e66vd/N4_EEBc.gif
There's really not much practical difference in the patterns between
leaving the reflector at its original length (blue) vs changing its
length to maintain the original Ref/Dvr ratio (red). However, if SWR
bandwidth is the deciding factor it looks like adjusting both elements
is the way to go.
Anyone who would like the AutoEZ format (.weq) C-3S model can just
email me directly off-list.
Dan, AC6LA
http://ac6la.com/
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|
