This PC board appeared in my mailbox, a gift from Josh Rodriquez. It’s hard to believe that it has been nine years since my Modular Tetrazzi project to put Peter Blasser’s SMT Tetrazzi board into a MOTM panel. This time the task is to analyze this large PCB. It helped to have an original Tetrazzi on hand to open up and probe around.

Fixing a problem with the etch at the piezo inputs

To get the piezo bars working a trace needs cutting and three jumpers added.

Next to a TL082 is found the oblong symbol with two connections. Make the following changes:

• Cut the trace between the connector and diode at the blue X.
• Jumper the end of the diode to ground.
• Jumper the connector hole to Mido (indicated by a square symbol at various points on the board). This hole also gets the wire from the outer part of the piezo.
• Jumper the other connector hole to pins 5 and 3. The center of the piezo connects here.

Here’s the correct schematic for the piezo processor. Note the connections to the piezo. Also note all the mido connection points. Mido is Peter’s word for the middle voltage, 4.5V, distributed around the board. Mido acts as a virtual common for Peter’s symmetrical designs. The graphic shows mido as little squares (as do my schematics). However the graphic does NOT show the traces that distribute mido.

This circuit is a pair of inversely controlled VCAs, with the input coming from one of the oscillators and with four open collectors going off to the output mixer (see below). Note the colors green, blue, yellow and red, which correspond to the similarly marked points on the output mixer.

Output mixer connections

The output mixers for left and right outputs are in the center of the board. There are four inputs, which I’ve marked with four colors. Peter’s graphic shows slightly different symbols (two of them have an extra circle around them) with arrows. Connect all five points of the same color together.

I’ve labelled the four mixer input connections with the color codes. The way I knew which to connect was by looking at the original Tetrazzi board. Here I’m guessing at capacitor values. I think that those marked with two little stripes are 102, i.e. 1nf or 1000pf. I measured the series caps on the outputs. Probably any value from 100nf to 10uf would work here. I picked 4u7.

Scroll circuits

The scrolls are operated by the red buttons on the original Tetrazzi. The button connection is indicated by a crown symbol. When the button is pressed, an LFO is created at the output. When released, the value is held. It’s used to generate a random pitch. The capacitors in the scrolls on the original are 1uf mylars. In addition, the feedback resistor on the comparator is 22K on the original, but marked 100K on this large PCB. I suggest experimenting with the value.

Scroll connections to the oscillators

This kind of stumps me. The large PCB has traces connecting each scroll to two oscillators, but on the original each button impacts only one oscillator. For additional mystification two of the oscillators are connected in two spots, one with a 470K and the other with a 100K resistor. So this large PCB has three connections to two of the oscillators from scrolls, and two connections on the other oscillators. I wasn’t able to trace further on the original than to one 470K hooking between each scroll and its oscillator. Here are the relevant traces.

Oscillator schematic

This schematic is a more readable version, plus some additional details, from the one Peter published. I’ve shown how the cross modulation inputs from the other oscillators are hard wired to the UPMOD and DNMOD signal inputs. The 4051 is an 8-to-1 switch controlled by a three bit code. The LSB (A) comes from the pulse out of the oscillator and is used to set the upper and lower boundaries of the output. The two other bits (B, C) come from nodes. By default, the upper and lower bounds are fixed at 6V and 3V, making for a 3V p-p signal. When the other bits come into play, the bounds are limited dynamically by the value of the outputs of the other three oscillators: Peter’s famous bounds modulation, which he used also in the Sidrassi/Sidrax and in the Denum Eurorack module.

There are three node inputs. One for B, one for C, and one that switches both B and C. This input is wire-OR with the separate B and C inputs. I noticed that this node input circuitry was changed for the SMD Tetrazzi. In this large PCB, the inputs are DC-coupled, while in the SMD version they are AC-coupled (probably better for finger operation). The DC coupling is an advantage in a way, in that holding one of these inputs high (or low) for a long duration can keep the selection of the bounds partner oscillators fixed for a while. With the AC coupling the switch will fall back after the series capacitor (on the SMD version, not here) discharges.

Note that this is the upper right oscillator on the board. The other are similar, except for the 4051 wiring and the way the scrolls connect. The node input circuitry is inverse polarity on two of them.

Internal modulation routing

Star symbols on the graphic indicate the modulation connection points and also the routing of each oscillator output to its corresponding VCA input. Notice two stars of the same color in each of the quadrants. One is the ramp wave output and the other is the VCA input. These get connected. In addition, there is a star of the same type/color in each of the other three quadrants. All five of the stars of each color are to be connected together. (You could, if you wanted to, bring out these four signals as individual oscillator outputs in the 3V to 6V range. I would add a series 10K to 22K resistor between the op amp and the extra output jacks.)

Node external connections

The ‘nodes’ are indicated by cross-shaped symbols. There is one output node and three input nodes per oscillator. I discussed the operation of the node inputs above. On the original Tetrazzi these were brought out to brass screws mounted on the bars for touch connection. Do not connect these together internally! They are mean to be brought out to be touched or patched together.

My next steps

I’m cogitating on how I might like to build this board. I’ve some ideas about bringing out some inputs and outputs of the various sections to banana jacks for patching. The scrolls outputs could be patchable to CV inputs, for example. More thinking to do!