Information

I've done quite a lot of research whilst working on the design of my synthesizer, and also just out of interest on other topics, like the Hammond organ.

I find that the discipline of writing down the results of these researches helps me to organise my thoughts and to state what I've learned clearly. At least, that's the idea. The pages in this section are the results of my investigations.

They cover my research into different oscillator designs that it might be possible to implement on the PIC chips I'm using. I looked at the DCO design from the Roland Juno synths and at general Direct Digital Synthesis and wavetable oscillators from the Korg DW series and PPG Wave. I decided to try and design a digital oscillator mainly because I've always felt that the three or (maybe) four standard waveforms from an analogue oscillator don't provide anything like enough variety of source material. Also all analogue synths have the same waveforms, so they all sound much the same - purist's points about the differences between one and another aside. However, digital oscillators are one of the hardest parts of a synth to get right, since you usually have to cope with aliasing.

Once I learned how to use the dsPIC 30F4013 to generate audio, I started experimenting with various oscillators designs and configurations. I wanted something that was both simple to implement and which also provided depth and movement in the sound. Phase Distortion and the related area of FM synthesis both fit the ‘simple to implement’ criteria, so I tried those. Eric Brombaugh has already programmed a commercial FM oscillator for Modcan Electronics using the 30F4013, so I avoided repeating his work and played mainly with phase distortion.

I also did some research into the SSM2044 filter, since there are still a reasonable number of these chips available. The 2044 is a minor classic in the synth filter chip world, and appears in a number of famous instruments, as well as some not-so-famous ones.