After I had sussed out the Cocoquantus schematic and understood the internal relationships among the five oscillators that Peter B. calls the Quantussy section, I began to wonder what it would take to build a modular version of the Quantussy. It is comprised of five triangle wave VCOs, each clocking a pair of sample and holds (S&H). At any given time the inputs to the sample and holds come from one of the other four VCOs. (The selection of which other VCO is made by a solid state switch controlled by a logic circuit whose status is given by three LEDs in the central lozenge.) One S&H receives the triangle output of the selected VCO, while the other receives the castle output of the selected VCO. Peter calls the output of the S&H to the right in my diagram the castle output. “It is a result of sample and holding another oscillator at the speed of this oscillator.” The output of the S&H looking at the castle output of the other VCO is fed into the FM section of this VCO through a VCA to become the chaos modulation. This topology, combined with the switching matrix that selects inputs to the sample and holds, gives the Quantussy its amazing properties.
I looked at a number of simple VCO and S&H circuits, but the task of breadboarding these or designing a circuit board was too daunting. Then the CGS Serge SSG came along! It has two circuits that can work either as a triangle wave VCO or as a S&H. Replicating the Quantussy would require eight SSG boards, with one unused circuit. (Five for the 10 S&H and three for 5 VCOs.) The number four makes for a better board count, using two SSGs for 4 VCOs and four SSGs for 8 S&H. Hence: Quartussy. The block diagram above details one of the four identical sections. Let’s examine it.
The VCO block is a Smooth Generator, internally patched to self-cycle. Its output is the Triangle Out. Its cycle output, besides being patched to its own input, is hooked to the dual-edge one-shot circuit in the SSG. The one-shot output, a short negative pulse, is routed to the Hold inputs of the two sample and holds, which are like the Stepped Generator of the SSG. The left S&H accepts an input, labelled CV in, and its output is hard-wired to the CV attenuator pot of the VCO. This corresponds to the S&H in the Quantussy that accepts castle inputs. The slew rate for this S&H is internally set to the maximum. A panel switch allows disconnecting the Hold input, with the result that the CV will be tracked at the output, effectively bypassing the S&H. This allows the VCO to be controlled by a CV in the normal manner or to sample the incoming CV at a rate of 2X the frequency of the VCO, just as in the Quantussy. (This is due to the fact that the one-shot triggers on both the rising and falling edge of the cycle signal.)
The right S&H input and output (labelled ‘Castle in’ and ‘Castle out’) are available for patching, but its hold input is always being clocked at 2X the rate of the VCO. I added a panel pot to control the slew rate of the ‘Castle’ S&H.
Patching replaces the solid state switching used in the Quantussy. To replicate the Quantussy signal flow, the four sections can be patched in a ring, such that the Triangle out of each patches to Castle in of the next, and Castle out of each patches to the CV in of the next. Rings of three and two are also patchable. Patching allows more interconnection topologies than in the Quantussy! You have four independent VCOs, each clocking a S&H. They can be used separately or in combination with each other or any other synthesizer module.
Another switch could be added between the one-shot and the Castle Hold input. Or both Holds could be switched at once. The rate pot on the Castle S&H could be omitted. Bringing more signals to panel jacks would overly complicate the design and I will avoid it.
I’ve sketched out panel features. Each section has three pots, one switch, and four jacks. Four sections could easily fit into a 2U-width MOTM panel. Banana jacks would facilitate the cross patching and, by stacking, allow the outputs to be taken out as controls to other synthesizer modules. I haven’t by any means finalized a panel design. How to fit six CGS PC boards is a problem, too. But I’ll be working on it.