I built a Jupiter 8 VCO

I’ve spent about 3 weeks studying the Jupiter 8 schematics. It’s immensely rewarding as I realise something new each time I look at it. So I built up a VCO from the schematics and got it scaling to 1v per octave running from +/-9v. It is possible to build it from currently manufactured components.
Follow the development here.

Listen here.

Cool stuff! So what is the secret sauce/ingredient that makes it different from any other saw-core VCO? I gave a quick glance at the schematics and all I saw was pretty standard. I’m curious to hear your findings…

Yes, good demo, but I only hear a square with pulse width and saw waveform. Where is the sine and tri? Good job on the modern component clone.

Hi Olivier, well a curious fact is the use of the 3046 as a differential pair for the expo converter. They are only matched to ~5mV. Although I used the LS318 which is matched to <1mV, so maybe there’s a little magic gone. I think it’s use of the PNP to give it a very wide range is quite interesting. As far as I can tell the tuning, range, etc are all controlled in software, so it needs to be as linear as possible over a wide range without adjustment via digipots etc. Also it uses a 7000ppm tempco, but really it’s affect after considering the compensating resistors are the same as a standard 3300ppm.

I’ve been inspired by your comment elsewhere on the forum that the IR1309 is just a four OTAs and an expo converter on a single die. I think it would be quite possible to recreate all of the voice module using LM13700s and modern opamps.

@audiohoarder, I just built up the saw core and the pulse section of the waveshaping, Really I’m focusing on fully understanding the saw core and making it as accurate as possible.

Neat. I hope that you are able to make it as accurate as possible. I just wanted to make sure I didn’t miss a part of the demo. :slight_smile: