Paul Burridge wrote:

On Wed, 14 Apr 2004 00:36:17 GMT, "Harold E. Johnson"
wrote:

An 8 MHz filter doesn't have to be physically large Paul,

Hi Q coils in that frequency range in compact sizes... they don't seem
to go together. :-(

and efficiency
drops pretty fast (Think like a rock) as the multiplication factor goes up.
Have you ever actually defined what it is you're trying to do? Some control
thing in your 70 MHz band? Or real power for some application? Hard to hit a
moving target. Or is that the idea?

Yeah, moving target's good. Keep the discussion generalised and it
might help others as well. I'm not sure where the 70Mhz figure comes
from, but it's a good enough guess by whoever made it. However, the
final desired frequency in my particular case is in the region of
40Mhz which will be achieved by mixing down with the output from
another oscillator and filtering.

A high Q resonant circuit can be rather small.
For example, i made a tunable LC with a Q approaching 1000, and it was
not the size of a garbage can (resonant cavity); it was about 5 inches
tall and about 3 inches in diameter.
On one extreme, one uses standard LC parts and get fair Qs in small
size.
On the other extreme, one makes a ersonant cavity to get very high Qs
at the expense of size.
In between there is something that can be called either a "shielded
inductor" or a "resonant cavity with slow wave structure".
One takes an inductor and places it in the center of a metal cylinder;
one end of the inductor attaches to the inner wall (makes electrical
connection and acts as support).
The capacitance to the walls (and added ends) is the other half. Move
an end for fine tuning.
Rather ingenious; ther was an IEE paper 20 years(??) ago covering the
desigh equations.
The terminology used was "Helical resonator".