Volume Pot "Crosstalk"/Interaction?

[AndrewCE]

I'm looking at this simple mixer schematic (the 3rd one pictured):

http://www.all-electric.com/schematic/simp_mix.htm

It seems that, even with the volume pots all the way up, you would still get half of your voltage thrown away to ground by the input pot and 10K resistor. Am I wrong or is this a design error?

[JKowalski]

Your wrong. The input pot will control the voltage, the 10k resistor to the inverting input will not load down the voltage divider in any noticable way.

[AndrewCE]

Your wrong. The input pot will control the voltage, the 10k resistor to the inverting input will not load down the voltage divider in any noticable way.

I've drawn up the current situation (top) and the equivalent circuit, demonstrating that this does  reduce volume. Am I still wrong? Am I missing something? I know that the bottom demonstration is "backwards" of the way volume pots are normally wired, but Gibson Les Pauls seem to use it consistently.

[JKowalski]

That's not an equivalent circuit - notce that in yours, at full pot rotation you would have the equivalent of a direct wire connection to the inv input with a 20k resistor to ground, giving you infinite gain....


You are thinking of this circuit in the wrong way. Think of the circuit as in two seperate parts. The voltage divider, and the op amp input resistor. The 10k resistor plus the op amp impedance is far to high to load down the 10k voltage divider. When you have the pot at full rotation in the original, it is the equivalent of a 10k resistor to ground, and then a 10k resistor to the op amp input, and that is not a voltage divider, so you get the full signal still.

[Processaurus]

It won't exactly reduce the volume by half, but the original poster is right that there will be undesired loading/interaction from the virtual ground on the inverting input.  The virtual ground is made by the opamp output making an inverted signal and running that back through the resistor in the feedback loop at such an amplitude that it cancels the signal comming in from the other resistor (from the 10K pot in this case), and leaves what it sees on the non inverting input.  This is because opamps, when used as amplifiers,  work by using their output to try and keep their inputs at exactly the same voltage, and any difference causes it to go to work to try and correct the situation.

The virtual ground is great for mixers because it means that nothing going on on the other channels can interact with another channel.

Most inverting mixers I've seen use summing resistors 10x the value of the pots.  The designer here used the same values, what that will do is behave like a 10K resistor tied across the wiper and ground lugs of the 10K pot.  Namely it will make it have a changing load on the input as the pot gets moved around and have a weird taper.  To get rid of that, you could use 100K summing resistors and a 100K in the opamp feedback loop. Or larger on the feedback resistor (500K or 1Meg) to have the possibility of getting above unity gain.

[AndrewCE]

That's not an equivalent circuit - notce that in yours, at full pot rotation you would have the equivalent of a direct wire connection to the inv input with a 20k resistor to ground, giving you infinite gain....

no no no, you're misunderstanding me. I failed to stress the importance of the physical positions of the pot arrows in the diagrams. What I'm saying is that as the first diagram is at full pot rotation, it behaves the same as the second one at half pot rotation. And what I'm asking is does this reduce volume?

i'm starting to suspect that the ORDER of the resistors makes a difference; by switching the in and out, volume will be reduced, but in this config it isnt. Ill test it tomorrow

[earthtonesaudio]

Alright, that explanation makes sense.  But these two circuits are equivalent ONLY when the 10k/10k is at full rotation and the 20k is at half rotation.  At any other setting they are not equivalent.

But Processaurus is correct, the summing resistors are too small compared to the voltage dividers.

[AndrewCE]

i'll try bigger summing resistors and let yall know what i find