This is my build of the J3RK Buchla 258J VCO board, as discussed on Muff’s message board:
I put it into my famous 1U 16-position MOTM grid panel. Here’s the Front Panel Express file.
This is a music VCO with 1V/octave tracking over several octaves and a temperature compensated exponential voltage converter. I would say its most notable features are a voltage-controlled wave shape, and a very nice linear FM. The schematic diagram provided by Mark Verbos matches pretty closely to the PC board.
My build and modifications
- The builder decides between mixing a square or sawtooth wave with the sine wave in the wave shaper. The choice can be made by a selection of resistors at build time or by adding a rather complicated switch. For simplicity I chose the sawtooth version.
- The PC board offers three outputs: the main wave shape output, a triangle and a square wave output. These outputs are roughly 5 volts peak-to-peak. I added a quad op amp to the board to buffer these outputs and also the sine wave output. The buffers have gain to bring the levels to a standard 10 volt p-p level. (I’ll add the buffer schematic later.)
- The PC board has frequency control voltage summing for a coarse tune pot, two 1V/octave inputs, and one reversible CV input (which can easily be doubled to two inputs). My exponential CV inputs include a 1V/octave input jack, one attenuated FM input, one reversible CV input, and also a fine tune pot for which I had to add a 4M7 resistor to the board.
- For the linear FM section I selected a 1uf polyester film capacitor with a 39K series resistor and a 50K audio taper attenuator pot. These values work well for me, giving a good depth of FM, and control over the lower levels of modulation. Low frequencies pass quite well, allowing for slow linear tremolo.
- I added some 100nf bypass caps to the underside of the board.
- As to the much debated choice for switching transistor (a rare 2N3638 is called out), I substituted a BC560C, which works fine. I always print out a data sheet when installing transistors to insure correct lead placement. The PC board marks E, B, C, for the transistor leads on the top of the board, which helps.
- I added a 100K linear pot to attenuate the wave shape CV in and left off the 150K resistor from the board.
The initial frequency trim pot lets you set a bias point for the coarse range. At first I set the lowest frequency to 20 Hz. As this produced a high end of 35 KHz, I decided to set the high end to about 25 KHz and let the low end fall where it might, which turned out to be under 1 Hz! A very nice range.
I was able to set the 1V/octave tracking for at least three octaves of accuracy and perhaps more.
Adjusting the wave shape trim pots, I had a surprise. The VC wave shape affects both the sine and the wave outputs! I was very glad that I brought out the sine wave (which is tapped from the drain of the J201 FET and buffered to bring up the level). Nicely, the sine output maintains a constant amplitude as it changes shape and becomes edgy. The wave shape output that mixes in a sawtooth increases in amplitude as it shifts from ‘sine-like’ to ‘saw-like’ in shape. The ‘sine-like’ shape yields low harmonic content, but is not as smooth as the sine output. (I’ll post some demo recordings.)