"blend" control in standard circuits?

[choklitlove]

is there a way to accomplish this?  like a universal add-on that would provide the capability of "blending" the affected signal with the non-affected one with a pot?  a perfect example is the blue-box.  i know it would have to divide(?) the signal into two parts, one going into the effect.  the middle lug of the pot would go to the output, one of the outer lugs would come out of the effect, and the last lug would be the un-affected signal from the division(?).

please tell me it's possible.  thanks!

 

[petemoore]

  RG has some circtuits that'll do most everything of that nature.
  Also see ROG "Splitter Blend" article

[tcobretti]

Pete, two more posts and your at 10k!

I am no expert, but I would try a high resistance pot and wire it just like you said.  Another option is check out the Ampeg Scrambler, which has a simple buffer that feeds into a blend pot.

[RaceDriver205]

Look up RGs "panning for fun" article at GEOFEX

[John Lyons]

Search Sean M's "Buff and blend" as well.

[R.G.]

It's that impedance thing again.

A single blend pot works well enough with no buffers and such as long as
(a) the source impedances on each end of the blend pot are comparable
(b) the pot value is higher than 10x either source impedance
and
(c ) the input impedance driven by the wiper is high

Under those conditions, a bare pot will form a voltage divider from both sides of the pot.

There is a principle in electronics called the principle of superposition. Under most circumstances when there are two signals you may figure out the output from each one separately and just add them. It applies here. In theory, for a signal on the top of a pot and ground at the other end, the pot acts as a voltage divider and the signal varies linearly with pot position. So by superposition, we assume that one of the two signals is 0, but its impedance is low compared to the pot. Now the blend pot looks like a voltage divider to the opposite signal. Likewise, if we have signal and no signal swapped, the voltage divider now works the other way. By superposition, the same happens when there is a signal at both ends. The middle is half of each signal.

A linear pot in that position is nonlinear for volume versus rotation, most of the effect happening at the lowest settings. In the both-ends-driven blend pot, most of the effect happens at the two ends, with a broad region in the middle where both signals are about half the audible loudness level, letting you have fine control.

There are a bunch of assumptions that go with all of this. The control can also work into a much lower input impedance than either the source or the pot, in which case it's a current-dividing control, and there is no interaction between the two signals at all, just a current equal to the signal voltage divided by its section of the blend pot. Think of the inverting input to an opamp with two resistors in.

In fact, it still works with medium input impedances, but less predictably. As a quick and dirty blend, it is great. But for a universal add-on, you need more circuitry to make sure the results are predictable.