Heavily altering a pot's center value.

[stonerbox]

Jump to here.

Got a tricky one for you. I am trying to figure out a way to get a 20k (minium), 50k or 100k pot (rheostat) to dial in around 3-5k when set at 50%. Altering the taper with a resistor is no good as I need the full range of the pot.


Could it be done with some sort of gate?

Rotary is not an option.

[FiveseveN]

You're describing a log pot: a 25K one should give you around 3.75K at center.
Also note that the taper resistor trick doesn't alter the "full range of the pot", it just doesn't work that way when you're using it as a rheostat, which sounds like what you're asking.

[stonerbox]

You're describing a lone should give you around 3.75K at center.
Also note that the taper resistor trick doesn't alter the "full range of the pot", it just doesn't work that way when you're using it as a rheostat, which sounds like what you're asking.

By limiting the full range I mean limiting the maximum resistance. Put a 4.7k across lug and wiper and the 50k will never reach above 5k.

Not a bad idea (log 25K ) only problem is it now works backwards. A reverse log potentiometer could work but It would still limit me to the very minimum of 25k. A 50k or preferably a 100k would be best, if even possible?

[stonerbox]

This is the circuit.

[FiveseveN]

One more time: the "tapering resistor" trick only works when you are using the potentiometer as a 3 pin device, a voltage divider. You are using it as a 2 pin device, a variable resistor aka rheostat.

it now works backwards

Swap pins 1 and 3.

Why is the position of this gain control so important and why aren't you doing it properly with an AC bypass cap?

[stonerbox]

Aha, I'm not very familiar with altering tapers and you are right, I'm using it as a rheostat!

it now works backwards

Swap pins 1 and 3.

Why is the position of this gain control so important and why aren't you doing it properly with an AC bypass cap?

The cap is there, just forgot to add it to the schematic.

This still limits me to 22k log (what I got on me) which turned out to be too small.
With the emitter connected to pin 3 and ground to pin 2 the emitter sees 18k to ground at 50%. With the emitter connected to pin 1 it sees 3.8k to ground but now the saturation control works backwards.


I need to be able to control  Q1 with a 50-100k pot. Would I be able to configure a transistor or something else to limit or open the path to ground? Just thinking out loud..

[stonerbox]

I tried setting up a voltage controlled JFET as adjustable resistance. Looked half-okay on multimeter but had very sporadical performance as a real emitter resistor.


For anybody just dropping in now - I am trying to create an adjustable resistor/pot that has 3-4k at middle, a 100k when dimed and 0-200Ω when fully dialed the other way. Rotary switch+resistors are not possible this time.

[mozz]

Just call it "unsaturation".

[m4268588]

Changing taper by parallel resistors is "suppression of increase".  Can't pull the "route" down.


Vactrol?

[stonerbox]

I have tried every potentiometer I got, JFETs, LDRs and redesigning the input stage. It seems impossible to get a result that yields -6dB at lowest, +18-20dB in middle position and a +39dB when maxed out. Guess I am screwed.

[MrStab]

Use an analogue switch IC (eg. 4066) with fixed attenuation resistances, with each switch triggered perhaps by a ladder of diodes & transistors which conduct as increasing DC voltage is sent to them via. a pot.

Less resolution ofc. If you're hardcore, use a multiplexer IC and resistor arrays.

[MrStab]

Play around with the resistance slider on this conceptual mess i threw together:
https://tinyurl.com/2c5dd8zy

Curveball: maybe you could use a dual pot, with one gang handling analogue signal attenuation, and the other gang triggering something like this.

[stonerbox]

Play around with the resistance slider on this conceptual mess i threw together:
https://tinyurl.com/2c5dd8zy

Curveball: maybe you could use a dual pot, with one gang handling analogue signal attenuation, and the other gang triggering something like this.

Use an analogue switch IC (eg. 4066) with fixed attenuation resistances, with each switch triggered perhaps by a ladder of diodes & transistors which conduct as increasing DC voltage is sent to them via. a pot.

Less resolution ofc. If you're hardcore, use a multiplexer IC and resistor arrays.

This looks interesting. Similar to my initial pondering about building some sort of gate-thing. Sure, it is not exactly high resolution but I could build a massive step ladder to make it a semi-smooth.

Thank you Stab!

[MrStab]

I did think it was possibly overkill, just fun to think of convoluted solutions.
The more you creep towards higher resolution, the louder digital pots might scream at you instead.

(That said, a 4066 with its 4 SPST switches is four-bit, meaning you have 16 or so potential combinations. But the control voltage routing for that gets very silly.)

[amptramp]

Maybe a vactrol would help.  You would have high resistance at zero current and lower resistance as the light gets brighter.  You can always use an incandescent lamp to make it even more responsive to the control.  And bonus: the control is isolated from the signal path.  The pot for the lamp can be set up for greater current clockwise or counterclockwise, so you have the equivalent of log or antilog pots.  And the pot for the lamp can be a log pot if you want things to get interesting.

[MrStab]

Just thinking: maybe you could get a pot with a centre tap (they are out there) and set some fixed resistance to ground or Vref from the tap (where you would traditionally just tie it straight to ground or Vref, to outright mute the signal). Probably not what you're after, as attenuation would resume normal operation past 12 o'clock, but food for thought.