Pan circuit without the hump

I need to pan between two control signals with as little of a hump in the middle as possible using the circuit pictured below. Ideally if there were for example 5V DC at each input, then the output would stay at a constant 5V as I moved the pot from left to right with no hump in the middle. As I understand it though there will always be a hump with this type of circuit so I was wondering if there is a way to minimise this with the resistor selection. I think the pot should be low value compared to the input and output resistors, but I'm not sure how to optimise the input and output resistors.

So in summary. Can I minimise/change the hump in the middle of this pan control with resistor selection? If so where should be large and where should be small?

Thanks!

For those wondering why I want to use this circuit when there are others better suited: I need a way to pan between two control signals with a mono pot but have two outputs. One output should have both control signals inverted, one output should have only one of the two signals inverted. I have found a way to do it with a more complicated version of the circuit above involving multi taps and inverting op-amps. But if anyone else knows of another circuit to fit this need please chime in. I also thought of VCAs but would rather not if possible.

Thanks!

Not sure if this helps your situation but I highly regard the circuits in D. Self Small Signal Audio Design which suggests this circuit from page 438 for panning with near ideal sin/sin^2 behavior:

Hump?

What are you doing with the two panned results? Unless you are summing them to mono there is no hump....

There are many deep discussions about tradeoffs between constant power pans, constant voltage pans, and variants between the two.

JR

JohnRoberts said:

Hump?

What are you doing with the two panned results? Unless you are summing them to mono there is no hump....

Click to expand...

I am using the two output signals to drive individual VCAs . If for example I put a 5V DC signal into each of the two input nodes, and then set the resistors so that the output of the opamp reads 5V when the pot is either hard left of hard right, I then get a hump when the pot is in the middle (eg 6-7V).

This is for control signals, not audio, so I want as close as possible to constant voltage within my limitations

squarewave said:

Not sure if this helps your situation but I highly regard the circuits in D. Self Small Signal Audio Design which suggests this circuit from page 438 for panning with near ideal sin/sin^2 behavior:

Click to expand...

Thanks, I tried something similar to this in the beginning, but unfortunately at the moment I need to do this with a single mono pot AND be able to create the two outputs as described above.

If you want the sum of two DC voltages to be hump free then you need the same total gain irrespective of the pan pot position. To do this make the pan  pot R equal to the input series R. Unloaded tis will have 6dB loss at extremes and 12dB loss each side in the middle. To make it unloaded the right hand Rs need to be much higher than the pan pot R. Perhaps buffer each pan pot output with a fet non-inverting buffer and sum the outputs.

Cheers

Ian

I guess I wasn't clear, the philosophy between pan laws is concerned about how the two results are eventually used.  If a mono stem is separated into two stereo feeds for say playback in a listening room, the constant power (both legs down -3dB at center) will result in similar perceived sound level as a signal is panned left to right. That said if/when this stereo mix gets summed to mono there will be more than unity signal present when panned center. 

The most accurate way to make a constant voltage pan that always results in unity when summed is to use a subtractive circuit. One control generates the left signal, and the right signal is simply the input with left output subtracted from it. This approach will always mono sum to unity. 

JR

electric_diaries said:

Thanks, I tried something similar to this in the beginning, but unfortunately at the moment I need to do this with a single mono pot AND be able to create the two outputs as described above.

Click to expand...

You can re-arrange the circuit (inputs across single pot with wiper grounded and so on. But I don't know if the negative impedance converter will work as expected which is the important part that gets to the center of your issue. I would use LTSpice in this case to see and nail down the specific resistor values.

JohnRoberts said:

The most accurate way to make a constant voltage pan that always results in unity when summed is to use a subtractive circuit. One control generates the left signal, and the right signal is simply the input with left output subtracted from it. This approach will always mono sum to unity. 

JR

Click to expand...

I'm not sure I fully understand this, but do you mean a circuit like THIS?

electric_diaries said:

I'm not sure I fully understand this, but do you mean a circuit like THIS?

Click to expand...

More like that backwards, feed the pot first with the original signal and use it as left, then substract the output of the pot to the original signal and use it as right. Then trim the center point so both have the same level.

JS

electric_diaries said:

I need to pan between two control signals

Click to expand...

What about this?

joaquins said:

More like that backwards, feed the pot first with the original signal and use it as left, then substract the output of the pot to the original signal and use it as right. Then trim the center point so both have the same level.

JS

Click to expand...

Do you have a link to a schematic or article on this? I still can't quite understand how to get your subtractive method to work, unless it's in the configuration of the link I posted.

Thanks for your help!

abbey road d enfer said:

What about this?

Click to expand...

It works great for simply panning between two signals, but my needs are a little more complex:

I need a way to pan between two control signals with a mono pot but have two outputs. One output should have both (or none) of the signals inverted, and the other output should have only one signal inverted.

Here is the circuit I currently have. It works well for the outputs I need, but I still have the hump in the middle. Hopefully I can reduce the hump with Ian's suggestions, but maybe there is a way to simplify this circuit?