rickman wrote:

On 10/30/2015 12:01 PM, John S wrote:
On 10/30/2015 10:46 AM, rickman wrote:
On 10/30/2015 9:07 AM, Jeff wrote:

I looked up some materials for fixed capacitors and found dielectrics
with ?r change with temperature as low as 10 ppm/°C. These materials
also have a loss tangent less than 0.001, some much less. I'm
wondering
if they would be practical to use for the dielectric in a variable
capacitor.

Me thinks you are overlooking the very high voltages involved.


I would have thought that glass was a good candidate and in plentiful
supply in various thicknesses, and would withstand very high voltages.
The Er is in the range 5 to 10 depending on the actual type.

It is the high voltages that makes the dielectric useful.


No, it is the increase in capacitance that makes the dielectric useful.

That sounds rather argumentative. I explain this in the next paragraph
which you seem to be agreeing with. So which is it?

It's clearly both, as an insulator with a higher breakdown voltage than
air would enable the plates to be closer and thus smaller for a given
capacitance, as well as more fitting into a given length, even if the
dielectric constant was the same as air.



The plates in
these capacitors have to be widely separated and the use of dielectric
allows this spacing to be reduced, that's one dimension. The Er
increases the capacitance which allows the capacitor to be reduced in
the other two dimensions.

Yes.

The problem is the change in Er with temperature which will cause the
resonance of the antenna to change, potentially outside the bandwidth if
the Q is high enough.


Are you planning to operate this antenna over a wide range of temperatures?


I'm not sure how low the loss tangent would need to be to minimize self
heating to a point that higher Er changes with temperature won't matter.
Even if self heating is not a problem, larger Er changes will
temperature would mean you could not retune the capacitor to the same
value with environmental temperature changes and so the tuning would not
be repeatable. Possibly this could be compensated for by measuring the
temperature and calibrating for temperature.

Sure. Have you mathematically analyzed any of your proposed scenarios?
That might help.



--
Roger Hayter