Project synth

From Spivey's Corner
Jump to: navigation, search

The aim of this project is to build a hybrid music synthesizer with approxiamtely the capabilities of the Minimoog, mostly using Moog designs with updated components for the analog part, and an AVR with DACs for the digital part. Oscillators, Moog ladder filter and VCAs will be analog, while the LFO and envelope generators will be digital. The plan is to control the synth from an external MIDI keyboard, and maybe also from a wind controller.

One simplifying factor is that op-amps have improved greatly since the Moog days, and it's possible to settle on a single FET-input chip for most purposes. I've decided to standardise on the LF412 dual op-amp.

A stock of LF412s

What follows is a blog-style log of progress, in reverse chronological order.

16th July 2016

This afternoon I made a sawtooth-to-triangle converter following Bob Moog's brilliant design that uses a single transistor.

triangle and sawtooth waves

13th July 2016

The design for the voltage controlled oscillators is a hybrid of the Minimoog (sawtooth core, CV summers and waveshapers) and the Moog Prodigy (exponential current source based on a CA3046 transistor array, with thermostatic temperature control). Yesterday I built the current source and core, and (after a few false starts) got it to oscillate. The VCO is capable of a range from about 0.1Hz up to 19.6kHz; at the top end, the current source saturates and the frequency will rise no higher, and at the bottom end, the slope of the sawtooth starts to become nonlinear, and I got bored with waiting 10s or more for a cycle to complete. I haven't tried to measure the linearity of response to control voltage, but I did get the oscillator to modulate its frequency in response to a sinusoidal control voltage from a signal generator.

The design frequency of the oscillator at zero control voltage is 1kHz, with negative CV inputs able to produce lower frequencies. The frequency is set by a combination of the timing capacitor (1% polypropylene) and a resistor calibrating the current source (1% metal film), but it was still unexpected that as soon as I connected up the scope it showed an oscillating frequency of 1.02kHz.

  • Here's a scope shot of the oscillator being modulated with a sine wave.
VCO frequency modulation
  • A circuit diagram for most of the VCO, omitting the CV summers.
  • A photo of the board as it is now. There's space to add a second VCO on the same board, and also waveshapers and CV arithmetic.
VCO Stage 1
  • A layout for the board. (I'm using hand-drawn layouts done on my own templates, and constructing on 160 x 100mm cards intended for DIP packages. I can't be bothered with CAD software.)

Next steps: build the waveshapers for deriving a triangle and a square wave with variable duty cycle from the sawtooth; experiment with thermostatic control of the current source, using one of the transistors in the array as a temperature sensor and another as a heater.

Project documents

Personal tools