A while ago, I needed to "improve" my shunt feed system. I used fixed
ceramic high voltage caps for the series element and used a "L" network to
match the resulting resistive load. I found that I needed a series C value
that was in between the available combinations of fixed caps. The window
glass cap idea seemed like a good thing to try.
Copper tape left over from a microwave project was stuck to either side of
a piece of 1/16 inch thick glass of unknown origin. It was easy to trim the
tape to achieve the correct value of C using a hobby knife. The thing tuned
up great. Power was slowly increased and the temperature of the glass was
checked after each increment. At about 800 watts, the glass punched through
at the edge of one of the tape plates. It never did get hot, except at the
arc. Since the calculated breakdown voltage was much higher than the
expected voltage, I assumed the failure was the result of much higher
voltage gradient at the edge of the cut tape.
Possible ways to fix this would be to use a much thicker glass or metal
plates with beveled edges.
The final solution for me was to use a fixed ceramic cap in parallel with a
small value air variable to get the required C.
Some air variables might not work well for high power because of the loss
in the wiper. This can be avoided by using a split stator cap. Connecting
between the two stators and leaving the rotor floating will give you a very
much higher Q capacitor than the conventional method.
I would also like to hear about any other experiences using odd ball
methods of shunt feeding towers on topband.
73,
Larry - W7IUV
|