Stubs have their place, but not in an application like this.
Stubs have very low Q (typically 20-60 for less than 1/4 wl
of open wire line) so they increase loss. The stub itself
decreases bandwidth (particularly when the stub is long in
terms of factions of a wavelength), and bandwidth was the
goal in this case. Besides radiation issues, they also
transform voltages to new levels, and peak voltages can be
much much higher than excitation voltages. They create a
world of problems.
If I were doing this, I'd be VERY careful what I was
switching. I didn't follow the thread, but if these
capacitors are on a shunt feed wire you'd likely have many
thousands of volts across open contacts. You'd have to have
armature to frame and coil and contact-to-contact breakdowns
both over the maximum peak voltage (plus safety) of the
system, both in series and to ground considered.
I might use a "BCD" arrangement of parallel caps switched
with relays. A 25,50, 100, and 200 in parallel will give you
25 pF steps up to 375pF. Two relays allow four steps.
Three relays eight steps. Four relays provide 16 steps. You
get the idea.
Most large open-frame relays breakdown in a dry test below
5kV, and many break down below 1kV peak. Current in the
relay from capacitance (say from the movable contact to the
coil) can also be a problem. Vacuum relays are about the
only choice for shunt fed systems.
A series-fed structure has much less voltage and can use
standard relays, like those in antenna switches. Other than
that, it's motor driven capacitors.
I generally find it easier to just make the antenna a
broader bandwidth design.
73 Tom
_______________________________________________
Topband mailing list
Topband@contesting.com
http://lists.contesting.com/mailman/listinfo/topband
|