VCO-1 – The Oscillator One

Since Coolaudio started reissuing old IC’s used in various synths I always eyed the V3340, which is a reissue of the CEM3340 chip as used in many classic synths such as the Roland SH101, Prophet 5, Memory Moog and plenty of others. It’s, in essence, a fully featured 1V/oct oscillator on a single IC and thus takes away a lot of the headaches inherent to VCO design. Things like linearity and temperature compensating are taken care of for you. This takes away a lot of the pain of building VCO’s, in fact, since these IC’s are available again a DIY polysynth starts to become a possibility.

First prototype, looks a lot better now.

So, I ordered one – just one, they’re quite expensive – and started testing. It really doesn’t take that much effort to get the oscillator running and I based the circuit on the CEM3340 datasheet, various implementation examples on the electric druid website, the nonlinearcircuits VCO-1 and various other sources. It’s no surprise, but there are very little modern resources available for this IC.

Since I was implementing an oscillator anyways it looked like a perfect opportunity to add some interesting wave-shapers to the circuit. First and foremost a triangle to sine wave shaper, because I really wanted a sine output and the CEM doesn’t offer one. I tried a few ways and settled on a design I found on Thoma’s Henry’s excellent site which seemed to produce a suitable sine. It’s certainly not perfect, but it does the job.

In addition to the sin wave shaper I’ve also added a linear ramp waveform. It’s sort of a mix between a square wave and a saw wave and offers a different kind of feel to it then either of them. I got the idea from the Intellijel Rubicon oscillator which features this waveform and it’s not that hard to think up a suitable circuit to produce that waveform from the combination of a saw and a square wave.

While looking into various wave-shaper designs I found a schematic to phase shift a saw wave. That gave me the idea of trying to implement a ‘super-saw’ by shifting several saw waves and mix them back together. Turns out it’s not really a ‘super-saw’ in the Roland JP way, but it does sound nice. I didn’t went all the way with this idea as I quickly ran into some problems when you have several waveforms added on top of one another. Shifting a single saw wave around mixed together with the original one already sounded very nice. I fixed the 2nd shifted saw at 90° and it can be toggled on or off for a different tone.

Overall I’m very pleased with the sound quality of the oscillator. It produces really nice and rich sounding waveforms.


  • 1/V oct tuning
  • Coars tuning (+/- 1 octave)
  • Fine tuning (+/- a semitone)
  • Octave switch
  • Exponential & Linear FM
  • Individual outputs for sine, triangle, saw, super saw, Linear Ramp, Pulse and Square wave
  • CV input for Pulse Width Modulation
  • CV input for saw wave shifting
  • Hard & Soft Sync



Board A: Download the pdf
Board B: Download the pdf

BOM and Build notes

In order to make trimming easier, it’s recommended to install the IDC power connector and all trim-pots on the back of then BOARD B PCB.

There are plenty of trimmers on this one. Here’s a list of what they are all supposed to do.

  • RV1: sine symmetry trim
  • RV2: High-freq trim
  • RV3: Freq scale trim
  • RV4: Sine roundness trimmer
  • RV5: Pulse-width modulation range
  • RV6: linear ramp
  • RV7: Freq trim shape trimmer
  • RV8: saw wave-shifter trim 1
  • RV9: saw wave-shifter trim 2
  • RV10: max saw shift level


Update History

16-07-2018: Some updates on the schematics.

21-05-2019: Thanks to Tyler for pointing out that R21 should go to -12V and not to ground, the sine-shaper won’t function otherwise. The schematic has been updated to reflect the correction.

  • Changed Linear Ramp output volume so it’s in line with the other ones (R59 to 20k instead of 10k)
  • Changed the sensitivity of the coarse and fine knobs (R10 and R11)
  • Used an LM4040 to create a 5V reference instead of L7805 and L7905 voltage regulator. I hope this will improve stability when switching octaves.
  • Added a potentiometer for the saw animation in

13-09-2019: Updated the post to reflect the latest version of the build. Also added build guides and BOM.

12-11-2019: Updated BOM, build guides and added a calibration document.


  1. Hi there, thanks for sharing your schematic.
    Do you get pitch interaction with your PWM input? When I apply LFO or whatever to PWM input, I get a pretty significant pitch wobble. Not sure if this is a bug in my design or a “feature” of the chip.

  2. Author

    I haven’t noticed anything like that on the build I’ve done so far. The IC shouldn’t do that. The CV input for the pitch is very sensitive to fluctuations though. Let me know if you find out what causes the issue.

  3. Hi thanks for detailed build notes and circuits of both boards. However I’ve noticed some issues which may have been overlooked in the build guide of board A and the BOM.
    The inclusion of R21 5.6M and R36 30K are both in build A guide and circuit diagram but not in the BOM. Is this an oversight or have they been excluded for some other reason?
    The de-coupling capacitors C21 and C12 both 0.1uF ceramic dont appear in the BOM or the build A guide but are in the circuit diagram and on the circuit board. Are these actually required or is this an oversight since the inclusion of de-coupling caps on the power lines are common in all the circuits and PCBs I have designed and built?
    Many thanks for an excellent couple of boards and the front plate and hope to enjoy incluing the finished VCO in my main frame Synthesizer… Cheers

    1. Author

      Redoing and expanding the document of the VCO-1 is on my to-do list. It has gone through many iterations and as a result I messed up the documentation as I released the schematics well before a production version was build. Lesson learned I guess. But to answer your questions:

      R21 5.6M will work fine, but has been replaced in the latest version (of which changes are indeed not reflected in the build guide and schematics) with a 10M. Which gives the fine-tune control a range of about 1 semitone up/down.

      R36 should be indeed 30K and that’s an oversight that it’s not included in the BOM.However, the value isn’t that crucial, 33K would work as well.

      The decoupling caps aren’t on the BOM while they are on the schematic. It’s intended they’re there. However, since it are decoupling caps it won’t break functionality if they happen to be absent.

      I’ll update the documentation as soon as possible.

  4. Hi
    Built module and working well. It does tend to pull my +12v line down to 11.5v whenever I have it powered up. The current draw is about 200mA which seems a bit excessive I am using fixed 1 amp 12v fixed Regulators which get quite hot but still power the unit ok.
    What sort of current drain should I expect with the Oscillator one??

    1. Author

      That’s indeed excessive. I’ve measured 58mA on the positive rail and 56mA on the negative one, so it should generally be somewhere around those lines.

Comments are closed.