Craig,
you would need greater resolution when you want the tuning range to cover
the entire 40m band (instead of just a 50kHz segment).
currently:
single turn pot => 300 degrees rotation => 1000 steps =>0.3 degree/step
tuning range 50kHz => 50,000 kHz/1,000 steps = 50Hz/step
If we want a tuning range of 400 kHz (8x larger range), then the step size
would be 400Hz per step, which is too much (not enough precision).
It wouldn't help to just use a multiturn pot in this case because there
are still only 1,000 steps.
In order to keep the 50 Hz/step you would need 8,000 steps instead of
1,000, so eight times more steps. More steps means greater resolution. We
need 3 additional bits in order to have 8 times more steps.
For 3 additional bits we need to take 4^3=64 samples and take the average
of them (oversampling).
With 3 bits greater resolution, so 8,000 steps, in combination with a
10-turn pot, we would have 10 turns x 360 degrees / 8,000 steps = 0.45
degrees/step, which is 50% better than we currently have, while the tuning
range would span the entire 40m band!
But as I said I have never tried this - perhaps the tuning mechanism will
become too slow when it needs to take 64x more samples to determine the
pot position. Then a rotary encoder might indeed be easier and perhaps
cheaper too!
73, Allard PE1NWL
Allard PE1NWL