Maximally linear RF inductor with minimal loss

[thadman]

For an amplitude modulated plasma loudspeaker, we might seek a resonant frequency of 27Mhz with a maximum quality factor (Q). A high Q will result in an inherently stable plasma. I believe a minimum Q of 1000 would be desirable, the higher the better.

A variable vacuum tube capacitor will satisfy these constraints (linear, extremely low losses, extremely high voltage capability).

A superconducting electromagnetic coil will be very linear and will have negligible losses, but the cost associated with such a system would be immense. As a result, we might dismiss this type of inductor as inherently impractical.

I believe a traditional coil using copper wire would be linear, but it will also dissipate a significant amount of heat due to the extremely high voltages. This will not result in a RLC circuit characterized by a maximum Q. Intuitively, we might increase the cross section of the wire to minimize resistance. However, the skin effect will not allow these gains to be realized due to the extremely high frequency resonance. The electrons will travel along the edge of the wire, minimizing the cross sectional area and countering the lower resistance normally associated with larger gauge wire. As a result, we might dismiss this type of inductor on the grounds of not meeting our constraints.

What type of inductor would be capable of 27Mhz, be maximally linear, handle extremely high voltages, and have an absolute minimum of resistance?

We might also desire the inductor to have variable inductance. This will allow the opportunity for tuning and experimentation with different capacitor types, while still maintaining our desired resonant frequency.

In addition, I'm not sure how accurate the manufacturing tolerances are or how accurate my simulations would be. Also, purchasing components on-line is accompanied by an element of variability, the component I receive might not match the distributors description or specifications exactly.

Could anybody offer further insight on this?

Thanks,
Thadman