One pot, two op-amps?

[hanumanlangur]

I would like to have a single potentiometer control the feedback/gain of two op-amp gain stages. Is doing something like this possible?

[xFallen]

One mechanical approach is to use dual ganged pots (on the same shaft, not concentric).


Barry

[Mark Hammer]

A dual-ganged pot may or may not be necessary, depending on how you want to control the gain. You will note that in a noninverting configuration, the gain is set by the feedback plus ground-leg resistance, divided by the ground-leg resistance ("ground leg" refers to the resistance between the inverting pin of the op-amp and ground). If a pot is wired up such that wiper (middle lug) of the pot is tied to ground, and the wiper-to-side-lug resistance either forms or supplements the ground leg, then moving the pot in each direction increases the gain of one amp while decreasing the gain of the other. Although there are some bugs in the circuit, you can see and example of it in practice here: http://img.photobucket.com/albums/v4...driveschem.png

Alternatively, you can use a combination of fixed-gain with variable attenuation plus variable gain. Remember that a voltage divider (volume pot) sets the "fraction" of the output signal, relative to the incoming signal, as a function of the wiper-to-ground resistance, compared to the input-to-wiper resistance. You can produce more attenuation simply by reducing the wiper-to-ground resistance.

Okay, hold that thought.....

If you had one op-amp set up for a fixed gain and followed it with a voltage-divider type attenuation, you could in theory wire up a single pot as described above (wiper to ground), using one side of the pot's resistance (wiper to outside lug) to be the gain-determining ground leg of one op-amp, and the other side of the pot to set the amount of post-opamp attenuation following another gain stage.

Now that I think of it, that is also illustrated in the linked-to schematic. You can see a Mix/Blend pot over on the right of the drawing. That pot functions as a voltage-divider type attenuator. You can see that the attenuation/level of the upper channel depends on the ratio of R18 to R20 plus whatever fraction of VR2 there is between R20 and ground. The attenuation of the lower channel is increased as you move the wiper of VR2 closer to R12, and the upper channel's output gets quieter as you move the VR2 wiper towards R20.

Okay, now imagine you had a single 100k pot with the wiper tied to ground, and let's say it's VR1. Now lift the end of VR1 that goes to R8 in the lower channel, and instead connect it to R12. As you move that pot towards R12, you are decreasing the resistance to ground from R12, therefore attenuating the output of the lower channel. At the same time, you are increasing the resistance between R16 and ground, therefore reducing the gain of the upper channel. Move the pot the other way and the R-to-ground for the attenuator side gets larger (conserving more output signal) while the R-to-ground for the variable-gain side gets smaller, turning up the gain. So, in one direction, both channels get louder, and in the other they both get quieter.

Admittedly, complicated, but it demonstrates that you CAN do it with a single pot. Just note that you have a bit of math ahead of you to do it right. Also note that you may need to place an additional fixed resistor in parallel with one or both legs of the control pot to achieve the "right" amount of attenuation and or gain change.

[Artie]

Another approach is to use a Vactrol, as available from PerkinElmer Optoelectronics. Its simply a photocell who's resistance is controlled by an internal LED. Use your pot to control the brightness of two LED's, (series or parallel), and then place the photocells in the respective circuits.

The nice thing about photocells is their dynamic noise is even less than expensive plastic-element pots. The only real down-side is their reaction time. But they should be plenty fast enough for a manual control.