Is a single knob "additive" active blend possible with more than two inputs?

[Andon]

Howdy. I'm wondering if it's possible to have an "additive" sort of active blend on a single knob, where as you turn the knob up or down it gradually adds or takes away inputs in the blend? Say I've got four differently voiced drives or modulations (and just to make things easier let's say that all four inputs to be blended have volume parity so we don't have to worry about that), when the knob is fully to the left (counterclockwise) it's just an overdrive or such, but as you turn the knob to the right (clockwise) it gradually adds in the other inputs until you have all four effects present when fully clockwise. Does that make sense? Are there any working examples of this to be gleaned from?

I'm assuming this would be something akin to a middle ground between a summing amplifier and an active blend circuit, a la R.G. Keen's example in Panning For Fun.

Using the boxed diagram from Panning For Fun, I've currently got a project breadboarded where I have two of the effects tied to one side of the blend circuit, and then all four tied to the other side, and it works pretty well, but I just didn't know if there was a way to gradually add each effect as you turn the pot, even though there are only two lugs on the pot unless you're using something like a dual gang pot (which I don't currently have at my disposal). Any thoughts? And if not, thanks for reading all the same!

[FiveseveN]

as you turn the knob up or down it gradually adds or takes away inputs

Can you graph that so we can have a better picture of what you're looking for?

there are only two lugs on the pot

Two pins means you can only blend between two signals. You would need a pot with extra taps in order to involve more signals.
You could use the pot to control LDRs, VCAs and the like. The most efficient solution depends a lot on what exactly you're trying to achieve (see my first question).

[ElectricDruid]

It's not going to be possible using the pot to control the signals directly. The only way would be to use the pot to create a control voltage that then sets the levels of the signals using OTAs/VCAs/LDRs or something as FiveseveN suggested.

The AS2164/V2164 quad VCA chip would be good for such a thing. There's a quad mixer circuit in the datasheet. If one signal is permanently wired in (So you start with one signal, then add more) you could have up to five signals mixed with the one chip.

Only remaining problem is how to create the four separate CVs such that one comes in, then another, then another, then another. You'd have to use op-amps to offset the CVs to the different VCAs. That doesn't sound too hard.

[amptramp]

This sounds like a job for a joystick.  Not just that it gives you two pots (which should be enough) but you have an indication of what the proportions are that could not be presented with one pot.

[Andon]

Looks like VCA may be the route, then. Per FiveseveN's post, I was basically looking for something where as you turned the knob up or down it would add or remove inputs proportionally, so if you had four inputs, fully CCW would be 1, 33% up would be 1+2, 66% up would be 1+2+3, and fully CW would be 1+2+3+4.

[PRR]

Howdy. I'm wondering if it's possible to have an "additive" sort of active blend on a single knob, where as you turn the knob up or down it gradually adds or takes away inputs in the blend? Say I've got four differently voiced drives or modulations (and just to make things easier let's say that all four inputs to be blended have volume parity so we don't have to worry about that)...

Decompose. A 4-in mixer to mix the 4 inputs, and a cross-fade blend to mix that with the just-overdrive.

[Processaurus]

Interesting subject. I don't think it would be very ergonomic to have one knob blend more than two things, because you quickly lose options of what can be blended with what, and you'd have to replug in the different things to get the last combinations.  There are some esoteric, center tapped pots out there that could do three things, A<>B<>C, but you already lost the ability to mix A and C, as well as the ability to mix all three.

You could always use a rotary switch masquerading as a knob, and select between different summing resistor groups, if you didn't mind chunky graduations in the mixes. Like A, B, C, A+B, A+C, B+C, A+B+C

You could make some kind of single pot DCA constant volume mixer digitally, that could "fade" between those positions, with digi-pots, if you had heavy skillz.

[amptramp]

If a single control is what you need, maybe a concentric control with concentric knobs and two pots would work if the single control is for ergonomic reasons.  Finding knobs for concentric pots is getting to be difficult.  Maybe you could rob a junked oscilloscope or European radio.

[ElectricDruid]

Seems to me that you guys are telling them the problem is wrong rather than solving it. OP didn't ask for a joystick, or a 4-input mixer. The request was pretty specific:

I was basically looking for something where as you turned the knob up or down it would add or remove inputs proportionally, so if you had four inputs, fully CCW would be 1, 33% up would be 1+2, 66% up would be 1+2+3, and fully CW would be 1+2+3+4.

I can see uses for such thing (mixing in further delay line outputs in a multi-band chorus, for example - label the control "thickness"). But it's not a circuit I've ever seen implemented, and it's not a simple thing to do, since the four input case requires three VCAs and three different control signals.

Ben's idea of using a rotary switch is definitely the simplest way to get the required effect. A typical 12-position rotary switch is available as 4x3, so you'd have 3 poles and 4 positions (A, A+B, A+B+C, A+B+C+D?). Saves a lot of circuitry, but you don't get to blend in each input - you just get to choose how many.

[antonis]

That's more a "stepper" rather than a "proportional" bleder, Tom..

[Processaurus]

A typical 12-position rotary switch is available as 4x3, so you'd have 3 poles and 4 positions (A, A+B, A+B+C, A+B+C+D?). Saves a lot of circuitry, but you don't get to blend in each input - you just get to choose how many.

One way  to get finer control would be if you used a 1 pole, 12 position rotary, and the common goes to the inverting input on the opamp summing stage, and the 12 positions select different combos of summing resistors from the sources, 12 separate groups. Lots of resistors. For A, just a summing resistor from A. For A+B (50:50 blend), equal summing resistors. For B+ a little bit of C, use a larger resistor for C than B.

If the sources are all low impedance, preferably direct  opamp outputs, the idle networks of R's connected to the unselected switch positions  shouldn't cause bleed back through the neighbor's summing resistor, the opamp's low impedance should be able to sink the current from the bleed.

[ElectricDruid]

Aah ok! Sorry Ben, I hadn't understood your idea fully.

That's more a "stepper" rather than a "proportional" bleder, Tom..

Well, yeah, but if it's one rotary switch and some resistors, instead of multiple VCAs, several op-amps and a lot of additional circuitry, we might accept the limitations, right?!? I think Mark Hammer is right and that in reality, there's a lot of controls where "Off, Some, Full" is enough options, or perhaps "Off, Some, More, Full" if you want really fine detail!!
This is especially true in the DIY world where we can choose what those particular levels are to suit our own tastes, instead of having had someone else do it for us.

[Andon]

I appreciate all of the insight every has contributed in this thread - thank you! Tom is actually right on the money - I'm looking to essentially blend four differently voiced delay lines for a chorus (using an LFO that also has an inverted output, so two lines are in phase, two are out of phase,) but I thought it might be better to broaden that idea in the initial post so that maybe the idea would be of more use to others who wanted to apply it to a multi-effect setup or such.

[Eb7+9]

... as you turned the knob up or down it would add or remove inputs proportionally, so if you had four inputs, fully CCW would be 1, 33% up would be 1+2, 66% up would be 1+2+3, and fully CW would be 1+2+3+4

I would ignore the nay-saying "experts" on this forum, this is a trivial circuit problem ...
(quite similar to last week's unrequited PWM-OTA challenge)

go look up linearized/degenerated diff-pairs // ...

you then need to figure out how to set up the structure of the solution:

>> give each pair a 2volt control span and stagger them vertically (eg., say 0.5-2.5v, 2.5-4.5v, 4.5-6.5v, 6.5-8.5v)

the rest is a no-brainer

[ElectricDruid]

Eb7+9's suggestion reminds me of something - J.Haible's solution to the "Interpolating scanner" problem, which is basically what has just been described:

http://www.till.com/articles/scanner/

http://jhaible.com/legacy/tonline_stuff/jh_ipscan.html

I didn't hear anyone saying "nay" and I didn't hear anyone claiming to be an expert either. We're just hunting for the most efficient/effective/neatest way to solve the problem.

[Ben N]

I'm not sure I understood the problem, but it seems to me that you could have a single pot into a single analog microcontroller input, and with the uC driving 4 digipots into a mixer, and then you could program the uC to do whatever you want with those 4 inputs depending on where the pot is in its rotation, i.e. the voltage on the input, could even have a number of preset programs as to what happens with the pot rotation selectable with a switch. The OP's description would be easily doable, but with a lot more flexibility. Would be cool with an expression pedal.
I guess this is functionally similar to the 4xVCA idea.

[Eb7+9]

glad to hear Juergen's name mentioned here ...

the idea in his scanner is to open and close channels
in successional
whereas in this thread the channels go from closed to open
and stay open as the Cv pot travels upwards in voltage

—-


the ckt I'm talking about is pretty classic Current-Control
in this case used to drive 4 OTA's from one pot

consisting of

14 PNP devices
(6 for Wilson-mirrors/sources and 8 for pairs)
14 resistors,
four 50k trimmers
and one 10k Cv pot

total draw lies under 6mA on a 9v source
with sources set to 1mA


—-

simple setup:

set trimpots to 1.5v, 3.5v, 5.5v, 7.5v to get
consecutive linear fade-in zones spanning 0.5v to 8.5v

with source saturation levels set to about 300mv this should work ...

—

the OTA side could simply be four open-loop stages
with outputs summed into a common (variable) load

possibly followed by a buffer ...

done

[Eb7+9]

simulator shows the control span would need to be narrowed a little running for 9vdc operation ... (my bad)
here's the output for the 4-banger running at 18vdc to get an idea of intended profile as per above