A "better" Ross/Ropez sweep

[Mark Hammer]

The Tonepad Ropez is a clone of the old Ross Phaser, graciously provided by Jeorge Tripps and adapted by Francisco Pena for Tonepad users/clients.  The black Ross Phaser is fundamentally an adaptation of the first-issue "6-chip" EHX Small Stone, using two OTAs to provide the LFO to modulate the phase shift.  The Ross version uses LM13600 dual OTAs, instead of the (now) harder-to-find CA3094 chips (which EHX used a house number for).

The Color switch on the Small Stone did several things at once.  One of those functions was to increase/decrease feedback.  The other was to change the shape of the sweep, from normal triangular to "hypertriangular".  That LFO change also had the effect of altering both the width and speed range of the sweep.  So, in one switch position, the sweep was somewhat slower and wider, with more feedback, while the other was slightly narrower, triangular, and capable of being faster, with less feedback.  The Color switch essentially provided two presets: one optimized for slow dramatic sweeps, and the other optimized for faster bubbly ones.

The Ross/Ropez provided continuously variable feedback, which was nice, but only provided fo hypertriangular sweep.  That's really nice for long slow sweeps, but downright annoying for fast ones, especially in the absence of any control over sweep width.  Attempts to get a nice "bubbly" quasi-rotary sound just couldn't happen.

Looking at the schematic for the 6-chip Small Stone, I recognized what I could do to improve the situation.  In the diagram, I show the critical portion of the Ropez board, with the key components indicated in yellow and labelled A, B, and C.  Their corresponding components on the Small Stone are also shown, with the same lettering.

On the Small Stone circuit, there are three resistors comprised by A and B.  At all times, they are tied to V+, but at different junctions.  "A" is 270K on the Ropez, and 180K+100K in series on the SS.  By switching where V+ is tied to, "A" effectively becomes 180K, while "B" becomes 270K+100K (I'll get to C in a moment).  Changing the indicated components on the Ropez board transforms the LFO from hypertriangular to straight triangular.  It's not quite as dramatic for slow sweeps, but is more usable across the entire range of sweeps, from very slow to very fast, and also provides a little more speed at the high end.  It IS possible to replicate the mode-switching from the Small Stone on the Ropez, but the board starts to get cluttered with "flying" components.  I found it easier just to revert to only triangular.  Your choice.

I noted that the hypertriangular sweep went somewhat wider, sweeping higher than the other mode.  Component "C" provides the current-limiting that dictates how wide the sweep is.  Stock, the resistor is 10k, and one is advised not to drop too much below that, or else risk frying the OTAs.  I replaced C with a 9K1 resistor, in series with a 20k variable resistor, such that the current-limiting ranges from 9k1 to 29K1.  The 9K1 extends the top of the sweep a little more, while the higher resistances rein that in.  It doesn't change where the sweep starts, but only how high it goes.  Trimming back the sweep width makes fast bubbly speeds more tolerable and useful, while the highest sweeps approximate the drama of the hypertriangular ones.

Is it possible to tinker with the values of A and B to get a sweep that leans a little in the direction of hypertriangular, but maintains the utility of the triangle waveform?  Possibly.  I haven't experimented with it.  As well, note the presence of the 10uf cap to ground after C on the Small Stone.  That would "slur" the sweep a bit, and probably reduce width, as speeds exceeded 2hz.  Adding that works the same way in the Ropez.  But overall I'm very pleased with this mod to the Ross/Ropez circuit, as shown, and think you will be too.