"Lars Janqqvist" wrote in message
...
Frightening the yaks, Dale Parfitt just had to say:

re uneven plates: The linearity will suffer but unless you're
calibrating it to something (tuning dial, antenna load) that
shouldn't matter.

I don't see anything linear in F vs C for a resonant circuit. In fact, in
order to make it anywhere close to linear, one has to go to extremes with
the shape of the plates.

My bad. Not linear in the sense of d-pF/d-theta = K. But more
the idea of no sharp changes for slight adjustment of the rotors.
Continuously differentiable, I should have said.

But if a linear variable cap *was* desired, the plates shouldn't
be all that hard to fabricate.

You are correct Lars. I seem to recall this in some early tube sets. I was
able to make the VFO in a hi end homebrew RX I am building quite linear by
making certain the variable cap comprises a relatively small portion of the
total tank C- and this using the typical 1/2 round plates. All sorts of
games can be played by using combinations of series and parallel capacities
along with the variable. An Excel spreadsheet gives quick insight.
Also nice to see a bit of differential calculus on the group!
Dale W4OP