Is there a better way to do this?

[phaeton]

What i'm trying to do, is basically 'pan' between two elements of a circuit:



"Tone Mangling Circuits" could be filters, distortion units, modulation types, whatever...  it doesn't matter- just understand that I want a ratio like 100:0 at one extreme, 50:50 at the middle, and 0:100 at the other extreme.  I realize pots don't *really* work that way, but something "close enough" is "close enough".
If i use a pot like that, it doesn't really seem to work.  Mainly, it sounds like I'm panning between 40:60/60:40.  value of the pot (1K, 2K, 5K, 10K, 20K, 100K, 1M) doesn't really change a whole lot.  I could do this instead:



Either using 2 separate pots, or using a dual-ganged pot (with one half reversed).  Makes sense that this would do the true 100:0/0:100, but is this at all possible to do with an unganged, single pot (or any other means?)

As you all know, pots, switches, jacks, enclosures are the most expensive part

thanks for any and all.

Next I'll ask lame questions about pots acting like voltage divider circuits

[LyleCaldwell]

You want a blend pot for that. 

[charbot]

DJ mixers use a single  slide pot to cross fade  between channels. Sounds like what you are looking for.     Im not sure how this works but it might be cool in a dual pedal... using your foot to blend effects.  Im sure there is a diy mixer project out there some where.  Maybe you could start there.

[TELEFUNKON]

http://www.diystompboxes.com/smfforum/index.php?topic=40095
http://geofex.com/Article_Folders/VCA%20Applications.pdf

[George Giblet]

The first circuit you gave is by far the simples at and is usually find for doing this sort of thing.  Somewhere in the details of your A and B circuit is why that methods doesn't work.  Two possibilities:

- the gain of A is much different to the gain of B; in this case equalize the levels first.
- the output impedance of A or B isn't low impedance and the blend pot is intrefering with one or both of the circuits; in this case buffer the output of A and/or B.

What actually went wrong with what you tried?

[Processaurus]

Hi, if your first idea didn't have enough seperation between the two tone mangling elements, a real panning circuit does that nicely.  Have you seen the GEO panning article?  There's several schematics to choose from.  I've gotten some milage out of the one that uses an inverting opamp to make back the losses in the panning circuit.  Also if you use a dual opamp you can use the remaining one to either change the phase of one of the circuits if you need to, buffer one, or change the phase of the panning output.  It looks like a lot of parts, but it can easily fit on a 1 1/4x 1 1/4 piece of perfboard.

If you have plenty of output on your two circuits, and it could still be loud enough at a 5th of the level, you can skip the active gain recovery off and just do put the passive part of the panning circuit, with a pot and 4 resistors.  Look at the blend pot on the MXR bluebox project at tonepad.com if that interests you.  If you want a bit more gain, you could tack on the last gain recovery stage used in the big muff, if you just wanted to use one transistor. Or if the gain was fine but there were impedance issues, you could buffer the ouput with a FET buffer, like the nice sounding one at GGG.

PS The second one won't work because at each extreme of the pots rotation, the output of the one you're trying to get loudest gets shorted out, because the wipers of the dual pot are connected together.

[phaeton]

What actually went wrong with what you tried?

With the knob all they way over to one side (100<------>0), I would get 60% of Circuit A and about 40% of Circuit B.  As soon as I move off the one end with the knob (80<----->20) I would get about 25% of both (the volume would drop about in half).  It would be like this all the way across the knob's travel, until I got to the other side (20<------>80), at which point i'd have about 40% of Circuit A and 60% of Circuit B.

I checked over both and A/B'd between the two outputs with a SPDT and they were even in volume, but panning them didn't do as I expected.  Increasing the value of the pot (up to 1M, for instance) further emphasized the dropout in the middle, and decreasing the pot (down to 1K or so) gradually flattened it out until the difference between extremes was subtle or non-existent.  In my test example, Circuit A and Circuit B were both single-NPN-transistor boosts with some filtering applied, so I'm assuming that there arent any phase issues involved.

[TELEFUNKON]

What Processaurus meant with the second circuit:
don`t connect the pot wipers directly, but via 2 resistors (try at least 4.7k or higher),
one from each wiper to the summing point.