Fancy Tremolo- How to Pan Between 2 LFOs? R.G.?

[Processaurus]

This is really a general question about panning circuits...

I'm working on a EA tremolo, and had decided to make 2 of the LFO sections to get a slow lfo and a fast one, to get a better speed range out what is a pretty limited design.  The light bulb goes on then, why only have one on at a time, why not have a knob that blends between the two sweeps, to get a crazy double lfo in the middle of the knob, a slow one on one end, and a fast one on the other end?  

Its important that at each extreme end of the blend knob there be no bleed from the other lfo, to get the action like a switch on each end, but with everything in between too. so I tried a panning circuit.  I took each lfo's signal off the emitter of the transistor that does the oscillating, to keep its DC bias, and ran it into a pan circuit that I saw on the Geofex panning article (http://geofex.com/Article_Folders/panner.pdf) I made the circuit near the top of the 1st page that uses an opamp (I used 1/2 of an LM358 because its made for single supplies and is low power) to boost the signal after its losses in the panning circuit.  I got this all working great, but noticed on the oscilloscope that in the middle of the knob's travel, when each lfo is equal, the resulting waveform is about twice as large, peak to peak, as each lfo on its own, at the ends of the pot's sweep.

So I was hoping to pick your brains on how to make a panning circuit that would make the composite middle position lfo a similar amplitude P-P as the individual lfos.   Something like the Geo pan circuit but really cuts the signals' amplitude more in the middle of its sweep.  I tried smaller pot values, keeping the surrounding resistors,  the original pot was 100k, I tried 50k, and then 10k, even 1M, but nothing seemed to cut the signal in the middle position more than the two end signals.

This is way too complicated for how simple of a project the EA trem is, but I was hoping to develop something that would work for a more versitile tremolo too (it would be sweet to try one with a modern VCA chip).  This would be relevant to other fancy modulation effects too.  Imagine a phaser where you could blend between an lfo and say, an envelope sweep with one knob, and not have to worry about adjusting things to make them sweep far enough, or sweep too far and clip stuff.

I addressed to subject of this to R.G., because he wrote the article I got the idea from, but I'd definitely appreciate comments from anyone with thoughts on how to go about this.

-Ben

[Paul Perry (Frostwave)]

Are you willing to use a dual linear pot?
Because if you are, then run LFO 1 to the top of pot A, and LFO 2 to the bottom of pot B. And earth the opposite ends of the pots.
And run summing resistors from the two wipers to an op amp stage.
If the pots are say 10K, and the summing resistors 470K, it should work pretty well.

[R.G.]

Have you listened to it?

What's going on is that the panner is designed for half-power, not half voltage. The half-power point is 0.7071 of the full voltage, so yeah, the two .7071 voltages add to 1.414V on un-tremolo-ed peaks. However, that's 3db up, which is a just-noticeable increase in loudness.

There are ways to get really, truly half voltage, but at the cost of some more complexity.

[Paul Perry (Frostwave)]

I think the problem is, that if you want a "full range" of lfo at each end, then the signal has to go over the full LFO range, and in the middle, it is going "too far" so it cuts off for too much of the cycle.
Personally, if it was MY effect, I'd have the two LFOs & send the LFO voltages to a simple mixer, with 2 pots.
Then you can get ANY combo of the two.
And... although you just added a pot, you dont need the depth control any more!! so, same number of pots, really. :wink:

[ethrbunny]

FWIW: there is a stereo trem/pan build on commonsound.org.

[Processaurus]

Thanks for the thoughts, Paul and R.G., and the link, Etherbunny.  I couldn't really figure out how the commonsound dual lfo is implimented, without looking at a schematic or something.  Paul's suggestion of the dual pot is for sure the most direct way to a halved voltage in the middle postion of the knob.  I'll try that.  I can imagine the blank look on the kid behind the counter at radio shack now, if I were actually foolish enough to go in and ask for not just a dual, but a 10k, and LINEAR, pot! :shock: How about a cell phone instead?

The thing about 1/2 power vs 1/2 voltage makes sense, as the pan circuit was made for audio, and keeping what to the human ear is a consistant volume, rather than what is actually a consistant voltage.