voltage-controlled panner/mixer

[samhay]

I have two low-impedance signals (from op-amps) that I would like to be able to pan between/mix into a single output (which can go via an op-amp/buffer/whatever).
While it would be easy to do this with a cheap-and-cheerful pot, or with a bit more finesse thanks to R.G.'s 'panning for fun' info, I would like to pan using a voltage or current from an LFO or similar.
I know this has been done before, and I can think of rather elaborate ways to do this with a couple of OTAs or a compander (NE570). However, before I re-invent the wheel, can anyone point me in the direction of something that is tried and tested?

Edit: Come to think of it, and while I'm probably asking a lot anyway, I would settle for any ideas on a voltage-controlled potentiometer where all three 'lugs' are floating. I have managed to do this with one resistor and a FET or LDR, but it has two drawbacks: (i) getting it biased to sit at a 1:1 mix (e.g. in the middle of the LFO sweep) is a bit of a PITA, and more importantly (ii) the value of the 'pot' (i.e. the resistor + FET/LDR in series) changes considerably, with less than desirable knock-on consequences.

Edit: Also fixed some typos - must learn to type as fast as I think...

[R.G.]

Use an emitter-biased PNP differential amplifier with the emitter current equal to the max current for a bias pin input on an OTA. Use a Dual OTA and feed one pin from each collector. Hold one base of the diffamp fixed, and arrange your control voltage to move the other base +/-25mv around the bias on the fixed base.

Done.

[samhay]

Thanks RG - that sounds workable.

[Fender3D]

...I know this has been done before...

I think I'd attached this somewhere already... 



It's not mine, I can't remember where I found it, credits must go to author (whoever he is)

[samhay]

I think I'd attached this somewhere already... 

Thanks Fender3D - the VCA section looks to be very helpful and is, reassuringly, pretty close to what I was thinking of last night.
I take it that is is a proper vibrato? Once it is trimmed properly, it should be a pretty cool effect. I did take a stroll through the archives, but didn't find whichever thread this came from. Do you know what this is called and/or where you posted it?

[fuzzo]

Hi,

I did something close to what Samhay posted. It was on a breadboard project of the Harmonic tremolo using ota as final stages. They were mixed into an op-amp before the ouptut . It worked pretty well.

[Fender3D]

sorry Sam,
I can't remember where I found it...
I built my last vibe unit several years ago...

I posted it here http://"the other site"/viewtopic.php?f=10&t=20730&p=210572&hilit=ota#p210572 suggesting it for a stereo blending unit

[samhay]

sorry Sam,
I can't remember where I found it...
I built my last vibe unit several years ago...

I posted it here http://"the other site"/viewtopic.php?f=10&t=20730&p=210572&hilit=ota#p210572 suggesting it for a stereo blending unit

No worries, and thanks again for a very helpful schematic.

[samhay]

Before this thead goes completely cold...
I haven't quite got round to the OTA solution yet, as I thought I might try a quick and dirty solution first - A partial schem is below (ignore where ground is - everything is DC couled to Vb). The LDR and trimmer form one half each of a voltage divider. This then feeds into a non-inverting op-amp. It has to be non-inverting or else you get a constant level of IN2 with the gain of IN1 modulated on top of it. For what it's worth, the three op-amps are not buffers - they all have gain.
It works quite well - certainly good enough for rock and roll - but you can hear the LFO under the audio when there is no signal going through it and I thought I would see if anybody can suggest a fix.
It is almost certainly not related to the power supply, which decouples the audio and LFO sections and has plenty of filtering. It also makes little difference if this is powered with a battery.
In the schem, the right hand side of the LDR is oscillating at the same freq as the LFO. The objectionable noise though (think whine) is a higher-frequency component, which I can't filter out becasue it is in the audio band.
The op-amps are all DC coupled, but I have tried de-coupling them and it doesn't help. Is the problem the high input impedance of the final op-amp? I have tried lowering it with a voltage divider, but I can't get it that low without a significant drop in volume.

As an alternative, I found an old thread where deadastronaut used a 555 and two LED/LDRs to do stereo panning. It seemed to peter out, and I wonder if anything came of that approach?

Cheers,
Sam



Edit - I guess this will work with the final op-amp inverting if I feed it with a series resistor after the LDR/trimmer junction. Not sure if there any reason why it should work better, but the Tremulus Lune etc seem to use the LDR into an inverting op-amp, so I am wondering if that virtual ground is a good thing in this application.

[deadastronaut]

hi sam, sorry man, i never got round to finishing it,

from what i remember , it worked using the 'fading eyes' circuit, with ldrs... it did work , but was quite subtle..(alternating fading leds)

it needed a bit better designing though...but that was the crux of it..

[samhay]

OK - thanks Rob. I figured that it might have been something like that. I initially tried this approach with 2 LEDs out of phase, but decided it was largely unnecessary. However, if I can't get this thing dead quiet, then I might have to re-visit it.

[samhay]

Tried final op-amp (-) input via a 100k series resistor after the LDR/trimmer junction. Didn't help. However, cleaning up the power further did fix it for the most part (LFO noise and breadboards = one wild turkey chase + much confusion) and the partial schem I posted earlier today does work as a cheap and cheerful panner.

I should stress that this is a minimum parts count option rather than a particularly 'good' solution. That said, it works and might be of use to someone else...