Reducing Potentiometer Range Without a Fixed Resistor in Series

[turdadactyl]

For some reason my brain is not coming up with a solution to this.  I built a fuzz circuit that I like the sound of using an LM386.  When I turn the gain knob (between pins 1 and all the way up (0 resistance) I get some really out there sounds and I'd like to have the ability to keep that wide open gain.  But when I roll the knob back to greater than 6.1K resistance the signal cuts out.

I'd like the ability to get more than 5K resistance, as there are some nice tones right before it cuts out, so I don't want to use a 5K pot.  But 10K will obviously give me a section of the gain pot, as described above, that cuts the signal.  Is there a solution to this?  (Am I rambling nonsensically?)

[reddesert]

Wire a resistor parallel to a 10K (or other value) pot, such that the combined resistance is close to 6.1K.  For a 10K pot this would be a 15.6K resistor. Since tolerances on pots are often 20%, you may wish to experiment with resistors in the neighborhood of 15K, rather than trying to get exactly a 15.6K resistor. (You can try making oddball values with two resistors in series.)

[turdadactyl]

See!  I knew there was a simple answer.  That whole multiplying by zero thing should have come to mind.  Thanks.

[duck_arse]

what taper is the pot you are using?

[turdadactyl]

I've been using linear (B) taper.

[duck_arse]

so you want a 10k pot that quickly jumps at the start, and then flattens between 6k and 0R, it sounds like. maybe a C10k would do the trick.

[turdadactyl]

Duck, with that thought, I may try a W.  I have a few laying around.  To be continued...

[pinkjimiphoton]

For some reason my brain is not coming up with a solution to this.  I built a fuzz circuit that I like the sound of using an LM386.  When I turn the gain knob (between pins 1 and all the way up (0 resistance) I get some really out there sounds and I'd like to have the ability to keep that wide open gain.  But when I roll the knob back to greater than 6.1K resistance the signal cuts out.

I'd like the ability to get more than 5K resistance, as there are some nice tones right before it cuts out, so I don't want to use a 5K pot.  But 10K will obviously give me a section of the gain pot, as described above, that cuts the signal.  Is there a solution to this?  (Am I rambling nonsensically?)

you don't have to connect a fuzz pot to ground, you can just use it as a variable resistor too... then its more of an "attack" than a fuzz control, and you can crank it, but turning it down won't necessarily turn it off.
i agree on the c or w taper pot.

i used linear pots for years, but audio lets ya get to where the tones are. hate 'em as volume pots on guitars, but they're great for guitar tone controls and for fuzz controls. particularly reverse taper.

gotta schematic of what you're tinkering with? hard to advise without seeing it.

[reddesert]

This pot isn't like a usual fuzz pot. It's specific to the LM386. That chip has an internal feedback network that sets the gain to 20. If you connect a resistor, or pot used as a variable resistor, between pins 1 and 8 , then you can adjust the gain from 20 up to 200.  You are putting the external resistor in parallel with a 1.35k internal resistor, see the LM386 datasheet at http://www.ti.com/lit/ds/symlink/lm386.pdf.

An example of a pedal that uses this is the Clari(not) which can have a fuzz/clean switch that switches the LM386 between high and low gain mode.

Because it's in parallel with that 1.35k internal resistor, values between about 0 and 5-10K will have the most effect. Once you get above about 5-10K, it's like leaving the connection between pins 1 and 8 open. Of course, how it sounds is dependent on how the rest of the circuit reacts to an LM386 running at a gain from 20 to 200. If the sound dies when the pot is run up to 10K, then the LM386 at gain 20 isn't driving the circuit enough to make sound, which is sort of weird.