Ever try two LFOs in a Phase 90?

[Mark Hammer]

I was preparing the layout for a board for a Phase 90 with envelope control.  The envelope follower is only going to require one op-amp, so I was wondering if I could just use a dual and find a purpose for the 2nd op-amp.  I was wondering if it would be contra-indicated in any way.  I'm assuming that the 2nd LFO would cover a range of very slow and shallow sweeps, and be mixed in with the primary LFO, essentially moving it around.

I figure if the resistor mixing in the primary LFO with the bias voltage is 3M9 (or 3M3 like some issues), then the secondary LFO would be something like 4M7, just to keep it in the background.  Does that make sense?

[Electron Tornado]

Mark,

Are you talking about two different LFOs you switch between, or running at the same time? If the latter, would there be intereference (addition and subtraction) in the combined waveform that would create some potentially interesting effects?

If you're looking for something for a second LFO to do, how about using one to sweep a resistance in place of the internal trim pot.

[Mark Hammer]

Are you talking about two different LFOs you switch between, or running at the same time? If the latter, would there be intereference (addition and subtraction) in the combined waveform that would create some potentially interesting effects?

If you're looking for something for a second LFO to do, how about using one to sweep a resistance in place of the internal trim pot.

I was thinking of two summed LFOs, not alternating ones.  Part of the intent is to aim for aperiodic modulation, which can be a nice feature sometimes.

The bias trimpot sets the overall range of the sweep, although the "sweet spot" can be fairly wide.  As per the drawing on the GEOFEX site, I replace the stock 1M fixed resistor with a smaller resistor and panel-mount pot to dial in different "offsets" for the sweep so that it can be higher and swirlier, or lower and gurglier. 

Normally, the LFO simply feeds a summing voltage/current on top of whatever arrives via the bias source.  You can see that the 1M and 3M9 - 3M3 in some issues - fixed resistors from the bias and LFO, respectively, arrive at the same point: the gate of all the JFETs.  I figured if you can sum two sources there, you can sum 3, so long as they don't sum too high; hence the thought of using 4M7 as the summing resistor for the 2nd LFO.  And subbing a larger cap (47uf?) for the stock 15uf unit in the normal LFO would allow for very slow sweeps that gradually nudge the main faster sweep higher or lower.  Coupled with the envelope control of speed that was planned, a person could get some interesting textures.

In effect, this would accomplish more or less what you have suggested - tampering with the bias - except that I was figuring on being able to tamper with the bias as well.

[Electron Tornado]

The bias trimpot sets the overall range of the sweep, although the "sweet spot" can be fairly wide.  As per the drawing on the GEOFEX site, I replace the stock 1M fixed resistor with a smaller resistor and panel-mount pot to dial in different "offsets" for the sweep so that it can be higher and swirlier, or lower and gurglier. 

I've done a mod to a couple of Phase 90s removing the trim pot and simply installing an external pot in it's place. I used the schematic here: http://aronnelson.com/gallery/main.php?g2_view=core.DownloadItem&g2_itemId=2122&g2_serialNumber=2 since that reflected the circuit in the actual pedal. I used a 25k pot since that was the value of the trim pot. The sweet spot, however, was useable but narrow. If I do another, I'll use a 10k pot.

Are the differences in R.G.'s circuit part of a mod or is that found in some stock Phase 90s?

[Mark Hammer]

That particular schematic is from the second issue ("non-script", I guess).  The first issue uses a 200k trimpot, and a 1M fixed resistor, rather than the 25k trimpot and 750k resistor.  I find that using the 1M and setting the trimpot, THEN replacing the 1M with a 330k fixed resistor and 1M pot, works well for me.

[duck_arse]

mark, just yester I was buggerising w/ a p45 with 2 unswept stages (would that make a p90?) and stuck a phase shift osc on the breadboard, mixed w/ a triangle osc. I thought of using a switch between a higher range osc one side, a lower range the other, and the centre position would switch in equal-to-the-mixer-resistor resistors, so the total amplitude of the mixed was about the same as a single.

my triangle osc was jerking me about, I didn't know what I was doing, but it does work.

[duck_arse]

above is the circuit I worked up last night. you get a high range osc, a low rangeosc, and a mix of the 2 in the centre position.

the dpdt centre-off kills the other oscillator, and the 2 100k resistors mix the outputs 50/50 when both are running in the centre position. they start up slow like all good pso's. the kill sends that osc output to about 3V9, and the output level doesn't jump too much as a result. an external bias pot makes helps iron out the differences.

I'm not sure about the mixing resistor setup, there is probably many more correct or better ways to do it, but it does work. the values in one or both oscs could use some tweek, as they don't quite overlap. you can sub your favourite triangle osc as well if you like, I started with that setup.

[Mark Hammer]

First, thanks!  The diligence of people in response to curiosity-provoking questions here never ceases to amze me.

Second, I suppose there are two ways to approach the use of dual LFOs.  One is to switch back and forth or use both.  Another way is to have a variable blend of them.  That would suggest a blend/mix pot instead of SW1b, and no SW1a.  They both have their respective merits, depending on what it is one wants to do.

[duck_arse]

diligence? it was just lying there already scribbled.

it occured to me that the ground connection of SW1A could instead be connected to one of RG's "spiff up your" circuits from the ean tremolo. just for some colour, mind.