View Full Version : Determining the function of logarithmic potentiometer

02-20-2013, 09:44 PM
I am using a logarithmic rotary potentiometer. The voltage output should be related to the rotation of the shaft as a logarithmic function. So the relation should be of the kind y= log_b(x). But the potentiometer I bought was very cheap and in wikipedia I found the following article.

"Most (cheaper) "log" potentiometers are not accurately logarithmic, but use two regions of different resistance (but constant resistivity) to approximate a logarithmic law. The two resistive tracks overlap at approximately 50% of the potentiometer rotation, this gives a stepwise logarithmic taper. A logarithmic potentiometer can also be simulated (not very accurately) with a linear one and an external resistor. True logarithmic potentiometers are significantly more expensive."

If this is true then the function should be more complicated. Can anyone please tell me how can I determine this function accurately?

P.S:- I also bought a 1k linear potentiometer but its shaft was too small for any knob to fit in. So interfacing it with the device I am making will be much more expensive and tough. The shaft of the logarithmic potentiometer was the perfect size for my device. Also note that the maximum rotation required of the potentiometer is 180 degrees.

Thanks in advance.

02-21-2013, 02:03 AM
If you have some method to accurately measure angle, you have to get V vs theta (upto the accuracy you want it to be). And I say plots because you need to get atleast 5-10 plots to be very sure that no human measurement error is involved..

Get that data in excel, plot with V as y theta as x, and the use the curve fitting tool to come up with a close approximation function.

More preferably, use matlab's cftool to do a way better job at the approximate equation.