Peter Blasser’s Sidrassi circuit was the first incarnation of an evolution of his design for a seven-oscillator organ. Others have told the story well.
Fans of the Sidrassi/Sidrazzi/Sidrax have never to my knowledge had access to a schematic diagram, either by Peter B. in his inimical style (as we have of the Tetrazzi) or by anyone else. I could be mistaken. But we do have a detailed graphic of the hand-assembled PC board for the Sidrassi, pictured above.
Working from a printout of this graphic, I reverse engineered a schematic diagram so that we can get more insight into the design. We already know how it works, as a black box. But I’ve always wondered what the ‘glitch’ inputs actually did. The answer is not surprising, but is revealing.
I hand drew the following three schematics. The first contains the oscillator proper with all of its modulation inputs. (You should be seeing v1.1, which adds a truth table and corrects the circuit around the glitch inputs.) The second contains the dual VCA, operated by the bars, that goes with each channel. The third schematic collects up some of the common areas of the circuit. All op amps are from TL084 quad DIP.
Nodes are the external patch points, six per section: Frequency Modulation in, Pulse out, two Glitch inputs, two Bar outputs. My drawing shapes for the nodes correspond to the graphics around the node lands on the PC board.
I drew the Oscillator #4 perspective to show that the chaos modulation comes from Oscillator #5, and how the bounds modulation oscillators are related to #4. This continues in a ring around all seven oscillators. The upper, middle section of the drawing is the VCO proper. Current from pin 12 of an LM13700 OTA charges the hairy capacitor (value unspecified), which is buffered by an op amp and sent to other parts of the Sidrassi as Internal CH 4 Out. This is the triangle wave (brought out to the red banana jack on the Sidrax). On Sidrassi the output node is a pulse.
Frequency (oscillator pitch) is determined by a mix of four sources. First, the master frequency pot which drives all oscillators, second, by the Scroll sample & hold button, third by an FM input node, and finally from the next Oscillator through a VCA made from the other half of the OTA and operated by the Chaos pot to control the modulation depth.
More Peter Blasser genius. An 8-to-1 selector (4051) has the logic bit A switched by the rate of the oscillator itself. By default it chooses a lower bound of 0.25 Vcc and upper bound of 0.75 Vcc. With Vcc of 9 volts, these points are 2.25V and 6.75V, making for a 4.5V P-P triangle wave. This is the default, until a glitch input is gated on.
The glitch inputs create two binary signals to control the 8-to-1 selector (4051) for choosing a pair of the other internal channel outs (out of six) to impact the upper and lower bounds of the oscillator. Inputs C, B, A of the 4051 constitute a three bit control for the selector. See the truth table in the diagram. Without active glitch inputs C is low, B low, and A toggles at the rate of oscillation to create the default bounds. With glitch inputs active, the multiplexer will select two of the other six oscillators for bounds modulation. Since bit A is driven by the pulse of the same oscillator, the they will be switched between at that rate. Thus, with the B glitch input active, the bounds will be determined by inputs 2 and 3 (which for oscillator #4 are oscillators #5 and #6). With only C input active the bounds are set by inputs 4 and 5. And with both glitch inputs active the bounds are set by inputs 6 and 7. It seems that at times the lower bound will be higher than the upper bound, confusing the oscillator. As intended!
The big red scroll button imparts a random offset to the frequency. It’s a clever circuit that oscillates while the button is pressed and then holds a value on release. These were replaced with slide pots on Sidrax, a concession to ‘playability’. A hairy scroll capacitor sets the rate. I don’t know what the value was.
Bars and VCAs
Pressing the bars gates the signal from an oscillator through to the output mixer. I’m not sure which is the left and right channel. The schematic shows the internal CH 4 output being routed to the inputs of two transistor VCAs. The piezo transducer glued to the bar controls each channel inversely. Bar press and release CV outputs are present at a connector for the two nodes (orange banana jacks on the Sidrax). The outputs of the VCAs route to the left and right output mixers through 22K resistors. (These resistors are shown on both the above and below schematics.)
The output mixers are shown to the left. Just a simple FET buffer, much like Lil Sidrassi. The pots for master frequency and chaos are shared by all seven oscillators. Mido is Peter’s way of referring to a middle voltage between Vcc and ground. I show it as a little circle. Mido is buffered by an op amp and distributed across the board.
I did not draw the power circuitry, because I couldn’t fully understand it. There seems to be a spot for a battery on the board. I can see a power inlet with a switch to cut off the battery if external power is connected, much like other of Peter’s designs. However there is a three-pin location with strange symbols around it. Please comment if you have any idea what this was for.