Pik-a-swirl: envelope control of P90 speed

[Mark Hammer]

I was struggling with this one for a bit, but finally managed to get something that worked nciely, and I'm very happy with it.

I made a heavily modded Phase 90 (speed, variable feedback/resonance, variable offset, 2 sweep-width settings, phaser/vibrato) and augmented it with a little daughter board containing what is essentially the envelope follower from a Dr. Q.  The Phase 90 has an op-amp unity-gain buffer stage at the input that splits the signal and sends it to the mixer stage (clean) and the phase-shift stages.  I tapped the signal from there and fed it to a 1458-based envelope follower.  Rather than using a fixed-gain and input attenuator (Sensitivity), like the Dr Q does, the buffer stage goes directly to the DQ follower stage, and I used a pot to adjust the input resistance to achieve variable gain.

The Dr Q uses the output of the follower/rectifier to drive the base of a transistor, acting like a variable resistance to ground.  I wanted to use the envelope control to provide a variable resistance in parallel with the Speed pot in the Phase 90 circuit, so the as-is Dr Q circuit was not going to work.  Instead, I substituted an LED for the 2nd diode you see here, and used an LDR in parallel with the Speed pot.  The LDR needs to be a noticeably higher dark/off resistance than the pot, and shift to a reasonably low "on" resistance when the LED is lit.  Paralleling the LDR with the Speed pot will reduce the duration of the slowest sweep cycles, but not be all that much.  I suppose one can always use a 1M speed pot if you feel you're missing out on something.

Using an LDR provides a little bit of additional ripple-rejection, compared to a transistor, which is helpful.  But I also wanted there to be some lag to achieve a kind of fast ramp-up but slow ramp-down.  I service of this, I went with a 100uf averaging cap instead of the 10uf cap shown.

I used a superbright white LED, largely because I had no information on the LDR and what wavelengths it was maximally sensitive to.  White should have what the LDR needs.  And because I wanted as much oomph as I could get, I subbed a 5817 Schottky for the first 1N914 (lower forward voltage)

The variable gain control on the op-amp stage allows you to essentially turn the envelope control off by making the gain low enough to not result in illumination of the LED.  Note that, since the LDR is just a paralleled resistance, the impact of the LDR will vary with the Speed pot setting.  Set the Speed for slowest sweep (highest pot resistance) and the LDR will produce more noticeable increases in speed than if the speed is already set to a medium-speed throb.

How does it sound?  Great!  very responsive, and a lot of fun to play with, both in phaser AND vibrato mode.  The lag time really helps to maintain a more emotional feel than if the change were suddenly fast and suddenly slow.  I suppose if a person wanted to, they could use the same circuit to parallel the feedback resistance around the phase-shift stages, reduce the feedback resistance by doing so, and essentuially increase the resonance in response to picking.  Having both the resonance settle back down, and the speed ramp-down at the same time woud likely introduce even more emotionality to the sound in response to picking.

I had the pleasure to repair a buddy's EHX Polyphase last fall, and it does these particular tricks...and a bit more: http://www.youtube.com/watch?v=f2wQstrdrfo  It appears to adjust the offset/bias of the phase-shift stages as you pick harder, such that the range where the notches are located moves upward as you pick harder.  Combining that with envelope modulation of Speed provides a very expressive phaser.  Most of that can be readily implemented on the Phase 90, with suitable use of LDRs.  The Polyphase was a 6-stage phaser, so the sound quality is a little different, but the ability to modulate the modulation in real time is just as much fun.

Note: I hadn't realized that the Empress Phaser also has envelope control, although it has both envelope control of rate, like I describe above, PLUS direct envelope phase sweep, which sounds like a form of auto-wah.

[drolo]

That's funny i am right in the middle of doing a similar thing to a Lovetone Doppelganger derivate :-)
I had done a similar arrangement with a State-variable (mutron/meatball) circuit that had an LFO whose speed was envelope controlled. As you say, its a lot of fun.

Your post comes right on time as I was looking for a way to spare on opamp, that I was using as a buffer after the envelope follower. Replacing one diode with a LED might do the trick.

I'm still hesitating between using the stock(ish) dual LFO from the Doppelganger and replacing it with the TAPFLO chip, that would add a sample and hold option ... options options ...