[sdiy] Good precision pot choice for VCO CV panel tuning ??

[David G. Dixon]

> David G. Dixon wrote:
> > No pot will be linear (no matter how expensive) if it's output isn't
> > buffered before hitting a finite impedance, such as the input resistor
> to a
> > summer.
> 
> General rule of thumb is that the wiper should see a resistance x10
> that of the track resistance.  So if you have a 100k pot then you
> should have 1Meg connection to the wiper.  Obviously higher is better,
> offset by the problems inherent in high-impedance circuits.

Ah yes, the ol' "10x for a stiff source" rule!

However, as one who often follows 100k pots with 100k input resistors
without a second thought, it actually matters very little for most inputs,
and it spares me from having to keep a bunch of different pot values on
hand.  There are very few instances where I really need the pot response to
be linear, and in those cases I'd rather use a buffer than have to buy
special pots or load up my signal chain with huge resistors.  Especially
when it comes to pots between the rails, I'd rather have a larger pot and
live with the non-linearity than load up my power supply with a bunch of 10k
pots with 30V potential differences (each one of those would eat up 0.2% of
my power supply).  Plus, the non-linearity actually gives you more voltage
sensitivity near the centre of the pot, where you usually want it most.