Hero and Sidekick VCOs

This is the Hero VCO, my new voltage controlled oscillator.

It started off as kind of a mashup of the Kassutronics VCO 3340 and the LMNC 1222 VCO. You can still find traces of both in the schematic, as well as bits of the Thomas Henry VCO Maximus, although little has evaded modification. Like those, it’s built around a 3340 VCO chip. There are separate output jacks for each waveform: The standard triangle, ramp, and pulse, and also sine. Two V/oct pitch CV inputs, one exponential pitch CV input, PWM CV input, and linear FM CV input — the latter three with attenuators. Separate from the PWM CV attenuator is a knob for setting the initial pulse width, which can go from near zero to near 100%. There is a sync input jack and a switch to select hard, soft, or no sync. There’s a fine tuning knob and an octave switch with 11 positions. Why? Not just Spinal Tap fan service. It’s pretty nearly as easy to make an 11 position octave switch as a five position one, and the 3340 can track that whole range pretty well. So, y’know, why not? At the low end it’s into LFO territory, and at the high end the central pitch is up around C8, the top of the piano keyboard.

Also on the front panel there are holes for access to four trimmers, for adjusting tracking, high frequency tracking compensation, center frequency, and pulse width range.

Under the hood there are fixes for some of the 3340’s less welcome attributes. There’s regulated -5 V power which eliminates the infamous pulse width frequency shift, and an added resistor that gets rid of the ringing on the falling edge of the pulse wave at low frequencies. All the waveforms are scaled and shifted to make them ±5 V.

Kosmo format, of course; 10 cm wide.

Altogether I think it’s a pretty nice VCO, but wait, there’s more. This is the Sidekick VCO, my other new voltage controlled oscillator.

Kosmo format too, but only 5 cm wide. So it lacks some of the Hero’s inputs — there’s just one each V/Oct, PWM, and linear FM — and controls: There’s a fine tune knob and an octave switch with a mere 5 positions. There’s a sync switch, which may seem odd, since there’s no sync input jack.

That’s a hint. While the Sidekick can be used on its own, it’s really not intended for that. It’s inspired by the old Moog 921b oscillator, which also was half the width of its full-featured sibling, the 921, and was meant to be used mainly for building multi-oscillator voices — maybe two or three oscillators slightly detuned for a fatter sound, or tuned at octaves or fifths or twelths or something, and all tracking the same input control voltages. Rather than using several multiples and a dozen cables to patch those CVs to all the oscillators, you’d patch them to a 921a oscillator driver module. It would sum up the inputs and panel knob settings, and send the resulting sum CVs out on a card edge connector, behind the panel, to all the 921b VCOs connected to it. It made for easier setup and use, and took up less room than a bunch of wide 921 oscillators.

The Sidekick works kind of similarly. But instead of an oscillator driver module, the Sidekick is driven by — you guessed it — the Hero, which has an output link header for a ribbon cable running to the input headers on one or more Sidekicks. It sends the same summed pitch, PWM, and linear FM CVs it uses to the Sidekicks for them to track. It also sends its ramp wave output, which the Sidekicks can sync to.

image

And no, the link headers (8 pins) are not the same as the power headers (10 pins), so you can’t plug the wrong cable in.

Aside from the reduced front panel contingent and the input instead of output link header, the Sidekick is the same circuit as the Hero. The layout is very different, though, because to fit in 5 cm width it had to go on a PCB perpendicular to the front panel instead of parallel ones.

I’ve built one prototype each of Hero and Sidekick and they seem to be working well. I did identify some faults with the PCBs, nothing that can’t be worked around, but I’ve corrected them in the design files and will be building the new version in the not too distant future. But for now, be warned, technically the latest version is untested. It should be fine…

Panel design files, Gerbers, and documentation in GitHub repository at GitHub - holmesrichards/herovco: Full featured 3340 based VCO in Kosmo format and GitHub - holmesrichards/sidekickvco: Subsidiary 3340 based VCO in Kosmo format.

Quick demo:

26 Likes

analogoutput thanks again for sharing your work with us ! . I can see ordering 5 each [ the min. you can order at JLCPCB ] of these.

5 Likes

which 3340’s are you using ?

1 Like

Bravissimo! I really appreciate the work you’ve done here to merge some of the best features of some classic 3340-based VCO boards.

3 Likes

I’ve got these two in my JLC cart already :grinning:

4 Likes

I doubt it matters a lot but I’m using the AS3340A.

3 Likes

LOL yeah I have 5 of the AS3340’s and was thinking the same …

2 Likes

I have the AS3340 . was looking into the AS3340A’s but dam they will probably cost more than the pcb’s .

Be sure to let me know what I did to screw up the new Gerbers this time. :laughing:

2 Likes

Sometime last year I ordered three AS3340 chips from AI Synthesis because taking postage into account they were the cheapest I could find at the time… and they sent me three AS3340A. I don’t know if that was a once only thing or if they’re routinely supplying the A chip, I suppose you could ask. They’re charging $7 each.

2 Likes

yeah their listening is for the AS3340 doesn’t show the AS3340A . I sent an Email so we’ll see.

1 Like

CabinTech charges only $7.40 each for three or more, but then they charge $4.72 each for four or more AS3340.

2 Likes

Thonk has an AS3340HYB . the spec. sheet says it has better internal PWM noise separation but for $12.00

1 Like

yeah that [ cabintech ] looks like the best deal out there for the AS3340A’s

2 Likes

Brilliant work. Very impressive. Many thanks

5 Likes

Just when I was thinking I needed no more oscillators, @analogoutput comes along with this beauty.

How many Sidekicks can one Hero control? I mean could I have 5 sidekicks making a 6 oscillator voice?

6 Likes

ermmmmm this would be immense if so!!

5 Likes

Well, well. I’ve been thinking of making a dual oscillator module, as I like especially when the second oscillator is tuned one fifth up of the other. Seems that just by changing resistors I could make the octave switch in Sidekick to -2, -1, 0, +5th, +1.
And multiple of those, mmmm… analog super saw.

4 Likes

Good question. The Hero’s output link has voltages straight from op amp outputs and the Sidekick’s input link is high impedance. I would think five Sidekicks would be no problem. But someone would have to build five Sidekicks first to find out, I’ve built one so far…

6 Likes

Nothing’s stopping you from using more switch positions; in principle you could do e.g. -2 oct, -12th, -1 oct, -5th, 0, +5th, +1 oct, +12th, +2 oct or something…

(You’d need resistors in ratios 7:5. The closest you get with E24 values is 5.1k:3.6k which is 1.2% off making your fifths… hmm… about 3 cents sharp (relative to equal temperament). In fact darn close to pure Pythagorean fifths, about 1.5 cents sharp. Hmm.)

3 Likes