[sdiy] Pot linearity towards extremes

[Steve Lenham]

On 14/12/2016 16:03, Rutger Vlek wrote:

> I'm working on what's basically a continuous pot turned into a
> 16-position switch. It's a circuit that uses a simple free-running
> ADC to convert a 0 to 5V output from a potentiometer digital outputs
> that control switching of a 16-way CMOS switch and a 16-way LED
> driver. The led driver lights a LED in a ring of 16 around the pot to
> indicate the selected position. However, in the prototype the LED
> being lit doesn't line up with the position of the pot. The pot is a
> 10K linear one, but from my measurements it seems it is non-linear
> towards the extremes. The angle of physical rotation doesn't line up
> so well with the angle over which the resistance changes linearly. Is
> that a common property of pots? I hadn't seen it so clearly before.
> Any ideas on how to easily overcome it without needing both a gain
> and offset trimmer in the circuit?

Any pot with less than 360 degrees of mechanical rotation will have at 
least a small deadzone at the two extremes of travel where the wiper is 
fixed at 0% or 100%. Manufacturers cannot risk the scenario where the 
mechanical rotation is slightly _less_ than the electrical rotation - in 
your application, that would mean that you could not quite reach 0V or 
5V - so they err on the side of caution and make the mechanical rotation 
slightly larger. The deadzones are where the wiper is running over the 
minimally-resistive metallised end contacts rather than the resistive track.

How big the deadzones are will vary from manufacturer to manufacturer, 
introducing yet another way in which crap pots can be crapper than good 
ones.

Cheers,

Steve L.
Benden Sound Technology