VCS-X “expanded synth modularization” Project

Dedicated to the memory of Juergen Haible

… as I move forward, I’m imagining something quite large, with potential for vast forms of expansion … a fully modular version of the EMS VCS-3 synthesizer

Something I started in ’95 with Juegen, and regrettably had to give up a year later due to success with my Triode modeling work. The format cloning of the EMS VCS3 I think was a bigger goal for Juergen, myself I was never that sure about it. Of course, I could understand the challenge – the patch matrix, compact size etc. all had their charm.

But, having played a few VCS3’s during my university days I also remembered what it was like to play with the two ceiling-to-floor MOOG and Roland modulars they had at the school studio. I always wondered what it would be like to have a full-wall EMS modular synth to play with!

I randomly decided to revive the project in the summer 2017, mainly scanning and organizing notes at first and then slowly working out (reversing) how every bit of VCS3 circuitry did its thing using SPICE circuit analysis software.

The work was quite beneficial as it helped spawn perhaps half a dozen new and un-related effector circuits in the subsequent years …


So, here is my list of preferred goals and self-imposed constraints for the VCS-X project :

  • To retain +12v / -9v supply conditions (minimizing re-design)
  • To make the circuitry (PCB cards) individual/modular
  • To re-designing circuitry for use with more commonly available modern devices (eg., 2n3904, 2n3906, 2n5089, 2n5457, …)
  • To add functionality / retain originality … stay close to original
  • To make the oscillators All-In-One units … with improvements
  • Consider a “non-slick/poor-man’s” version of everything


Done …

all circuit sub-systems have been modeled and simulated successfully and thoroughly – this took about a year of educated guesswork. I now have fundamental understanding of how all VCS3 circuitry works and how behavior can be custom tailored.


In progress …

the RAMP generator (engine) is working as it should, with emphasis on studying linearity and noise specs – one of the characteristic aspects of the VCO’s. I am currently moving into the wave-shaping sections …

The “hard-sync output” add-on is working properly as well …

Developing the simulation material alone has been a fun exercise – encouraging me to make stupid use of SPICE software; ie., working backwards at the circuit block level (identifying them was part of the game here) and moving up in levels, while altering a myriad of parameters until something moved or started cycling on its own; and then moving back and forth in levels between iterations to fine tune things. Obviously this is NOT the way things are done in standard engineering practices, but this is what I did to reverse the whole system, as I wasn’t sure how things were meant to do their thing to begin with. Now, I could write a little book about it.

This mucking about has given me a fond appreciation for the old-school single-ended electronic techniques of David Cockerell and friends – very cool stuff !!

more to come as things develop …

Leave a Reply

Your email address will not be published. Required fields are marked *


This site uses Akismet to reduce spam. Learn how your comment data is processed.