Psuedo-Log-Taper Pots

[List, Christopher]

On Wednesday, April 15, 1998 7:23 PM, John Speth [SMTP:johns at oei.com]
wrote:
> I looked into this about a year ago.  Here are a couple of
ASCII-matics for your use:
> 
> simulation of a lot pot:
> 
>         o vi
>         |
>         >
>         <
>         > rh
>         <
>         |
>   wiper +-----------+----o vo
>         |           |
>         >           >
>         < rl        < rx
>         >           >
>         <           <
>         |           |
>         +-----------+
>         |
>        gnd


Note that you'll get the same results if the bottom of the parallel
resistor (Rx) is the going to the summing node of a summing op-amp.
Virtual ground rather than a real ground. In this case, your result
voltage (Vo) is the voltage "seen" going into the summer. Good for
providing lod-taper volume control on a mixer. - Important to keep in
mind this interplay of pot with summing resistor anytime you put an
attenuator on a summing input! 

> 
> I even went so far as to write a little C program that generates the
raw data vo/vi curve for various values of resistors and pots.  
> 

This was a pet project of mine as well - I actually did this all in
Excel with formulas - not too tough. Just series-fill one column with
values from .00 to 1.00 in .01 increments and use that as your
percentage of pot rotation and work from there. The resistance of the
bottom half of the divider is just the parallel resistance of the two
resistors going to ground. Use that result in the formula for a voltage
divider and you're there. Unfortunately, I don't think I have the spread
sheets around anymore. I found that a resistor with a value of 1/3 to
1/10 the value of the pot gives good results, with 1/3 values being more
linear and 1/10 values being very "log-ish". As John mentioned, it's not
a true log curve, but it does a pretty good imitation. The only problem
is that you always have a little linear "bump" close to zero before it
starts acting like a log - if you graph it you'll see what I'm talking
about.

Another fun way to experiment with this (if you're REALLY bored) is to
use a dual pot - one with a resistor in parallel, one without. Hook the
wiper of the linear one to the X input on your scope and the other wiper
to the Y - the proof is in the trace!

- Chris