Is there a way to make something like this Harmonic generator?

[Esppse]

Hello,

This device has dedicated controls for 2nd and 3rd order harmonic generation. I'm curious what clipping circuits control both.

https://vintageking.com/vertigo-sound-vsm-2-mix-satellite-mk2-full-version

Is there a diy way to create such a device?

Thanks

[Rob Strand]

From what I can see the page claims it is an entirely analog circuit.

Without trying to re-invent the wheel.  One possible way is like this:
So from a mathematical perspective you would have say a squaring circuit and a cubing circuit to create the 2nd and 3rd harmonics.     The squaring and cubic circuits will also produce other harmonics representing distortion.   So perhaps there's a multiband splitter feeding a number of squaring and cubing ckts then perhaps a some filtering and mixing (a multiband-combiner) putting the signal back together again.

That's a fairly "brain dead" approach.  Your best bet is probably to find more clever ways to do this by looking through some patents.    Keep in mind you want methods suitable for audio (not things like triplers used for RF circuits, which are much easier).

[amptramp]

This may be generalized even order and odd order harmonic generation than explicit control of just the second and third harmonics.  If you do symmetrical clipping, you get odd order harmonics.  If you do asymmetrical clipping, you get even order harmonics.  If you use squaring or cubing circuits, you have to use frequency splitting as you would get from an equalizer and do these operations on narrow segments of the audio spectrum since they would also introduce intermodulation as well as harmonic generation.

[ElectricDruid]

If you do asymmetrical clipping, you get *both* odd and even harmonics. Generating even harmonics alone requires a wave shaping function that is a mirror image left-to-right. That's a pretty unusual shape and certainly not one that comes up when you change the bias on a typical clipping circuit. Rob's squaring circuit is one example though, so it's possible.

I suspect that the OP's harmonic generator just isn't as exact as its front panel controls seem to imply. The Second Harmonic part could be something like R.G.'s FET octave circuit, and the Third Harmonic part is easy with some basic symmetrical clipping. Both of these circuits will generate more than a single harmonic if pushed or if fed with complex waveforms.

http://www.geofex.com/Article_Folders/jfetdoub/jfetdoub.htm

HTH,
Tom

PS: There's some good theory on wave shaping (which is what clipping is a subset of) here:

http://www.rs-met.com/documents/tutorials/Waveshaping.pdf

[anotherjim]

I think it's probably an aural exciter thingy. Basically, it will be adding small amounts of low order distortion, but dynamically according to level and content. Dull or quiet gets more than loud or bright.
I have some records that brag about being recorded with an Aphex Aural Exciter, but as I've never heard the music without, I've no idea what it added!
I owned a Korg digital synth with an exciter effect, couldn't hear that it did that much, to be honest.

[Rob Strand]

If you do asymmetrical clipping, you get *both* odd and even harmonics. Generating even harmonics alone requires a wave shaping function that is a mirror image left-to-right. That's a pretty unusual shape and certainly not one that comes up when you change the bias on a typical clipping circuit. Rob's squaring circuit is one example though, so it's possible.

One very poor aspect about using squaring and cubing circuits directly is the level of the harmonic does not go up and down in proportion to input signal.   The level goes up as the square and cube respectively which makes it fairly useless.   So you really need to limit the input signal (either soft or hard),   then second order = original signal * limited signal.   third order = original signal * limited signal ^ 2.    Maybe adding a bias term will help?

Part of the magic of that box is to be able to individually dial-up the individual 2nd and 3rd harmonics.   Exciters tend to add a bit of everything.

Trying to split it into odds and evens then filtering could be better and easier.

I'm sure there's a patent on it with some tricky maths technique that lets you cancel out some of the lower harmonics and just filter the very high stuff.

[Esppse]

Ah I see. I've always found those studio "magic" boxes fascinating. I've got Analog Heat and  Culture Vulture, very cool toys!

[Mark Hammer]

Is your desire to get 2nd and 3rd-order harmonics specifically, or simply more lower-order harmonics than anything else?

[teemuk]

Well one could speculate what lies behind "2nd Harmonic Fet Crusher" and "3rd Harmonic Zener Blender". An asymetrically clipping FET stage and a pair of symmetrically clipping Zener diodes perhaps?


Well, I researched into Vertigo Sound website (https://www.vertigosound.com/products/vsm-2) and they actually show harmonic spectrum plots of these effects. "Fet Crusher" produces the typical diminshing pattern of -all-harmonics (both even and odd) inherent to generic asymmetric clipping schemes. So it's not a "pure" 2nd harmonic generator circuit at all. "Zener Blender" again produces a diminishing pattern of -all- harmonics, with the exception of distinct de-emphasize of even order harmonics. Yet again somewhat typical results for a generic symmetric clipping circuit. The latter effect overall seems to produce a tad more of higher order harmonics.

IMO, the additional features seem to be more interesting package overall: Typical "Drive" (gain) and output signal "Level" controls aside there seem to be some pretty helpful features for shaping the timbre of the distortion included, such as "Input Filter" (selects what frequency range goes to harmonic generator, this is IMO a very clever feature), "Shape" (hi-cut filter for reducing amplitude of too obnoxious highest order harmonic distortion products), "THD Mix" (blending mixer for clean and distortion signal paths) and "Style" (soft vs. hard clipping).

Yep, it overall seems to be rather similar effect with the generic "Aphex system"; there's a side band for harmonic generation with tunable input filter and a blend feature for mixing wet and dry signals. However, this one imcorporates that basic idea with vastly more advanced feature set for tone shaping. Nice.

How this works? Well, I unfortunately don't know any specific details of this particular effect unit but seriously... Isn't this "DIY"...? One doesn't have to use too much imagination for figuring out how to put together something similar to this. Following the generic Aphex system topology gets quite far already in basics and input filter, soft vs. hard clipping, blend and hi-cut are actually rather trivial features if you think about it. Employ asymmetric distortion for producing even order harmonics and symmetric for producing odd ones. Voila.   

[Esppse]

Ah, I guess it's approachable once you break it down like that. I'm probably not looking for strict 2nd and strict 3rd order harmonic generation, since that may not sound too interesting.

Something that generates more of each, and then a blendable control. Just would like analog domain to stay in instead of digital recreations. I like what this does to a mix.

Are there symmetrical and asymmetrical clipping circuit recommendations similar to this I should look for?

[Rob Strand]

Are there symmetrical and asymmetrical clipping circuit recommendations similar to this I should look for?

You should look-up some Aphex schematics because they have already done it.   One had a knob for odd/even harmonics.

For even harmonics some sort of multiplier, say using an OTA, is usually involved.  One of the Aphex circuits has this.  There's also something in the old Lab series amps (L5,L7,L9).

Odd harmonics could be done with a differential amplifier as this has fairly symmetrical clipping but any symmetrical clipper would work.   Any clipping introduces harmonics but it can get angry.

For simpler circuits there's too many options.  That would be a matter of playing around experimenting.  You could try using JFETs or MOSFETs for even harmonics.   In theory rectification will give you a second harmonic but again the result can sound fairly angry.