How to make this circuit voltage controlled?

[danjel]

I am trying to replace the 50k pot on here (which is acting like a floating resistor from 0-50k)

and make it voltage controllable.


I tried using Vactrols but they do not go low enough in on-resistance and are very non linear to drive.

Lately I tried this analog fet opto coupler H11F1M https://www.fairchildsemi.com/pf/H1/H11F1M.html but I am having trouble with the control circuit. I am using a 2n3904 with the led of the opto coupler connected to the collector with a resistor for current limiting. When I vary the current to the base of the transitor I seem to only get a switch like response to the opto coupler (it is either on with a resistance of about 200 ohms or it is off with an infinite resistance). I need to get a linear response.... e.g. if I attache a potentiometer to the base of the transitor (with a 10k resistor in series with the wiper) I would like to be able to sweep the pot to get a range of resistance on the fet of approximately 200ohm to 50k. Right away I see I need to sum an offset voltage at the base so that the transistor is always on (e.g. sum about 0.6V)

Here is the circuit:
**broken link removed**

How would you make this CV controllable?

 

[crutschow]

The problem with the H11F1M (and similar opto isolators) is the small range of linear operation. If you look at Figure 2 in the data sheet you will see that the voltage across the device must be less than ±0.05V to stay within the linear resistance range (the near vertical part of the curves). If the voltage is greater than that, the device saturates and limits the signal to a constant current (the horizontal part of the curves).

Thus in order for that device to operate as a variable resistor without significant distortion, you would need to limit the voltage across it to about ±50mV pp.

If you put a resistor in series with the emitter to ground of the transistor you are using to drive the device, you will get a more linear response (500Ω would give about 2mA of collector current per volt of base voltage). Also if you put a Silicon diode in series with the bottom of the pot element to ground you will get the needed 0.6V offset to bias the transistor on at the minimum pot rotation.

 

[danjel]

Thanks Carl!

I tried putting the diode into the negative feedback path of an opamp and then fed the CV to the same negative input via a resistor to limit the max current.
This seems to work.

I also put a 100k resistor in parallel with the FET output and this is giving me a resistance range of about 60ohm to 52k which is pretty close to what I would like.

Since I am using this for audio I am now attenuating the inputs so that they are below 100mV pp and then boost them later. This adds noise but at least it works.

So far this is working ok now.

 

[crutschow]

I tried putting the diode into the negative feedback path of an opamp and then fed the CV to the same negative input via a resistor to limit the max current.
This seems to work.

Since I am using this for audio I am now attenuating the inputs so that they are below 100mV pp and then boost them later. This adds noise but at least it works.

Certainly putting the diode in the feedback loop gives a near perfect conversion from voltage to current as long as you don't need more diode current than the opamp can provide.

You can minimize the noise by careful layout and shielding, and using low-noise op amps.