50K Reverse audio taper, what else can I use?

[MikeH]

Try this: a 450K ish reistor between 2 and 3 and a 5Kish between 1 and 2 of a 100K linear pot.  Should give you something reasonably close to a 50K rev log.  I might have it backwards... so it might be 450K between 1 and 2 and 5K between 2 and 3, but those would be the values.

[John Lyons]

There is more than one way to get to a rev log taper.
By putting a resistor across the outer lugs (1 and 3) of a linear pot
will give you a slight rev log taper. (as I described above).
If you want a more radical taper you can put resistances between the wiper
and outer lugs as described in the links above (not mentioned in the
secrete life of pots write up).

All the details are here with a graphic or two.
http://www.diystompboxes.com/smfforum/index.php?topic=70732.0

john

[svstee]

Thanks a lot guys!

I think I have enough options to get something that works right.

[mdh]

There is more than one way to get to a rev log taper.
By putting a resistor across the outer lugs (1 and 3) of a linear pot
will give you a slight rev log taper. (as I described above).
If you want a more radical taper you can put resistances between the wiper
and outer lugs as described in the links above (not mentioned in the
secrete life of pots write up).

All the details are here with a graphic or two.
http://www.diystompboxes.com/smfforum/index.php?topic=70732.0

john

I looked at the thread you linked, and I realized that I had always been thinking about a voltage divider configuration, and hadn't realized what can happen when the wiper and the CCW lug are the only parts of the subcircuit "seen" by the rest of the circuit (this is arrangement "D" discussed about halfway down the thread linked above).  It's interesting.  I sat down and did the math, and it turns out that the total resistance seen across the wiper and CCW lug is quadratic in the proportion of rotation, and has a maximum at (R_pot+R_parallel)/(2R_pot).  If R_parallel == R_pot, then the maximum is at 1, which is the case John mentioned.  If R_parallel < R_pot, the maximum is somewhere between CCW and CW, which is kind of strange behavior for a pot.

If anyone is interested in the general solution, if we define 0<p<1 as the proportion of clockwise rotation, then the total resistance is p*R_pot*(R_parallel + (1-p)*R_pot)/(R_pot + R_parallel).  If you graph that (as was done in the cited thread), you see that for R_pot = R_parallel, you do get something approaching reverse log for this configuration (it's really a segment of a parabola).  So, mea culpa.  Still, I think the other way is better, because you can vary the intensity of the taper much more without introducing this weird intermediate maximum.

Sorry to beat a dead horse, and to be pedantic about it, but it's how I've been educated to be!

[MikeH]

No that's great.  Always good to see the math.

[blueduck577]

edit: oops wrong thread