Voltage Controlled Switch

[CutRock]

I'm trying to take up diy audio electronics, and for my first project I'm making a voltage controlled oscillator. As part of my design, I'm trying to incorporate a little switch to control the output so that down the line I can chain multiple oscillators together.

Essentially, this is the functionality I'm shooting for:


All that's really going on is that if the first input is high, the second input (carrying the audio signal) is let through. If the first input is low, the second input is blocked.

It's pretty critical that the signal not be distorted by whatever circuit can achieve this. Also, the +15V is just convenient for the circuit I'm building, it can be easily lowered if that poses a problem.

Thanks!

[osufishman]

This can be done a variety of ways, relays, fets, bjts, multiplexers, bus switch, push buttons, rotary dials, or gates, etc.

Is the switch going to manually activated, voltage activated, or current activated?
What voltage and current level do you have available to drive it? If you know that you should be able to pick the one that makes sense in your circuit at the lowest cost.

You say you dont want the signal distorted which means you need very low on resistance if using semiconductors to do this. There are signal mosfets with Rds(on) of 1 to 2 ohms that would do they trick. Just apply between 5 and 10V to the gate and it would give you a closed path between source and drain.

[markusA]

Building a switch like that is easy, the hard part is figuring out the demands and later filling them.

[CharlieLaub]

You could use an LDR to make a voltage divider. An LDR is a four-terminal device that has a current dependent resistance - two terminals are use to supply current (up to 20mA typically) and the other two are the resistor terminals. Assuming that you are using the 15V supply as indicated in your sketch, use a 750 ohm resistor in series with the LDR's current terminals to supply 20 mA current to the LDR when the 15V supply is on. When 20mA is supplied to the LDR, the resistance drops to a few hundred ohms typically. When no current is supplied, the resistance is a few mega ohms. Combine this with another resistor to make a voltage divider - connect the input to the LDR resistance terminals, connect next to the second resistor, then connect that to ground. The output is taken at the junction between the LDR and the grounded resistor. This won't get you exactly an "off" state with 0v output, but you will get close if the grounded resistor has a resistance of a few hundred ohms.

You could also just use a N.O. SPST relay, but that wouldn't be as interesting...

-Charlie

[CutRock]

Thanks for all the replies, I've got a lot of material and possible devices now to start doing some research.

I should note that the signal will be an oscillating voltage which will output at audio frequencies and can be discrete (square wave) or continuous (sine). I'm not sure if this will affect any of the suggested solutions.

[bob123]

What is the required output impedance? There are a number of ways in which this can be done.

The simplest is probably an analogue switch such as the ADG712, they are essentially a couple of back to back fets with some logic and will act as a low resistance of a few ohms when turned on and a very high resistance when off.

Have you got schematics of the oscillator you want to use is there may be a cheeky trick that could be used.

The ADG712 only works to about 5v but im sure there are others that can go higher.

If you need the output to be zero volts when off then a pull down resistor can be used and the output can be buffered.

[paulb]

A relay is by far the simplest solution.

[benb]

Also, what would be helpful would be knowing what controls the control/on/off input.

For analog switches, there's the venerable and cheap 4016 and 4066, and you can get better performance for more money with parts like these:
Selecting the right CMOS analog switch - Maxim

If you're interested in making music synthesizers, Google diy synth. There's a lot out there, surely including good ways to do this.

[bob123]

Originally Posted by paulb

A relay is by far the simplest solution.

IMO not true. A relay requires significantly more drive current. It also will require the drive circuit to be protected from back EMF. It is far slower and mechanical yielding a less eloquent solution.

As a design exercise for someone who is less experienced, a relay may be easier to understand and is likely to be more resilient to mistakes providing the drive circuit is protected by placing a diode in reverse bias across the coil.

[markusA]

Ask yourself if you want a mechanical or electrical switch?