Soft clipping with predominantly 3rd order harmonics – possible?

[lion]

A question I've pondered for a while:
Can a transistor or opamp circuit be designed to soft/medium clip a guitar signal and add predominanty 3rd order harmonics? And what are the rules/theory and the parameters to tweak in order to "bias" different types of clipping circuits for predominantly 3rd order harmonics?

I've done some research. Lots of info on even vs odd order harmonics - symmetric/asymmetric distortion. Also RG's Distortion Primer and other good articles. However, in relation to SS circuits the essence I can find seems to be that "any circuit produces various levels of even and odd harmonic distortion", and very little/nothing about what determines which harmonics comes out and why - and how to design around it.

Erik

[antonis]

I'm not familiar with advanced maths (Fourier analysis e.t.c.) so I'll focus on <"any circuit produces various levels of even and odd harmonic distortion">...

Maybe it should be easiest to concentrate in dumping all unwanted harmonics (by filtering them) and leave "free" harmonics of concern..

[amz-fx]

The signal coming from the guitar isn't a pure tone; it will have even and odd harmonics naturally.

So, to even come close, you would have to low pass the guitar audio and then amplify it (a lot) and clip in a symmetrical manner. Even then you will get a series of all odd harmonics, so you have to low pass again to get mostly third.

The amplitude of the harmonics added are equal to 1/n * n1, where n is the harmonic and n1 is the fundamental frequency. For a square wave, the third harmonic will be 1/3 of the fundamental. The fifth will be 1/5 of the fundamental, and so on.

Lowpass, amplify, clip, lowpass.  Even then it will be only a rough approximation, and it's not soft clipping.    You can get this type of response from a modified BMP.

regards, Jack

[TejfolvonDanone]

Can a transistor or opamp circuit be designed to soft/medium clip a guitar signal and add predominanty 3rd order harmonics?

Isn't every symmetric clipping circuit qualify for this? I mean symmetrical clipping adds odd harmonics predominantly 3rd harmonics.

[karbomusic]

The signal coming from the guitar isn't a pure tone; it will have even and odd harmonics naturally.

Wouldn't we just make the determination with a scope/fft using a pure sine wave, then let the chips fall where they may when using a normal guitar signal? Obviously all those interactions based on a complex guitar signal, it's overtones having their own odd harmonics added, IM (not to mention chords) and so on would be involved but my quest of 'steering' harmonic structure is always performed on a simpler sine wave.

All that being said, an OD I designed is creates only odd harmonics, I also have a switch that brings in some even harmonics. The only point is I did it based on the above.

[merlinb]

Isn't every symmetric clipping circuit qualify for this? I mean symmetrical clipping adds odd harmonics predominantly 3rd harmonics.

Correct. Symmetrical clipping introduces only odd harmonics, mostly third. Asymmetrical clipping introduces at least some even harmonics.

[lion]

So, SYMMETRICAL is the keyword, and how the clipping is achieved makes no/little difference. Meaning, there's no one topology (diode/LED, silicon/germanium, MOSFET/JFET/bipolar, opamps etc.) which has better properties for 3rd order harmonics than others (as long as they are "biased" for symmetrical clipping), it's more about filtering. Amplitude of the harmonics are drive/level dependant (of course) and for higher amplitudes outside soft clipping range - and again filtering is essential. All this leaving the complexity of the guitar signal aside. Have I got this right so far?

Erik

[karbomusic]

it's more about filtering. Amplitude of the harmonics are drive/level dependant (of course) and for higher amplitudes outside soft clipping range - and again filtering is essential. All this leaving the complexity of the guitar signal aside. Have I got this right so far?

Erik

I need some schooling on how filtering does anything other than roll off *all* harmonics, even odd or otherwise. Any 'modification' I've done of harmonics when dealing with even/odd were due to how I changed the symmetry.

[anotherjim]

Bear in mind that if you already have a square wave - such as when a hard clipper drive amp is itself clipping - you can make the clippers as asymmetrical as you like, it will still be a square wave out of them & no even harmonics will be created. But if you introduce an RC high pass filter between amp & clippers (just insert cap), the differentiated square wave out of the high pass will be asymmetrical because the different clippers makes the time constant different between +ve and -ve cycles in the filter. You get plenty of even harmonics then, even when the drive amp is clipping. Also you can return the clippers via ground, instead of Vref, thanks to the added cap.
Down side is a thin bottom end and/or it stops making evens when the fundamental frequency is above the filter cut-off and passes the whole square wave.

[robthequiet]

Here is a link with some ideas about phasing and harmonics http://www.allaboutcircuits.com/textbook/alternating-current/chpt-10/harmonics-polyphase-power-systems/ It's dealing with transformer-generated phasing, which is mechanically trivial, but also thre is an iteresting point about harmonics and phasing.

The article does mention how you can shift AC current into 120-degree phase delays and then the third harmonic is miraculously in phase across all three lines. I'm not sure if this means that you could take a delay tank from a phaser and tap selected shift stages to form a composite 3rd harmonic signal to be amplified and mixed back (without killing the fundamental) -- my experience with pitch shifting is that you get some weird sideband effects if your frequencies are not perfectly lined up, but maybe that's a cool effect unto itself. hth 

[mac]

Most simple soft symmetrical clipping circuit I can think of is a darlington Bazz Fuss with another diode in the opposite direction.

mac

[Rixen]

So, SYMMETRICAL is the keyword, and how the clipping is achieved makes no/little difference. Meaning, there's no one topology (diode/LED, silicon/germanium, MOSFET/JFET/bipolar, opamps etc.) which has better properties for 3rd order harmonics than others (as long as they are "biased" for symmetrical clipping), it's more about filtering. Amplitude of the harmonics are drive/level dependant (of course) and for higher amplitudes outside soft clipping range - and again filtering is essential. All this leaving the complexity of the guitar signal aside. Have I got this right so far?

Erik

Yes. The tube screamer circuit is a symmetrical clipper, and relatively soft. I have seen a spectrum of the output which shows predominantly odd order harmonics. BTW I just visited the Wikipedia page on the Tube Screamer, and see some BS about how it works has crept in. It is not a crossover distortion.

[TejfolvonDanone]

Yes. The tube screamer circuit is a symmetrical clipper, and relatively soft. I have seen a spectrum of the output which shows predominantly odd order harmonics. BTW I just visited the Wikipedia page on the Tube Screamer, and see some BS about how it works has crept in. It is not a crossover distortion.

Just for the record it is more like a logarithmic amp than crossover distortion.

[lion]

Thanks for the comments and suggestions, appreciated. Time for more study and experiments.
Rob - that article is interesting, but mostly over my head I must admit.

I've searched the site for actual harmonics plots from different/any SS clipping circuits- thinking some must have been made/posted over the years - but couldn't find anything but one related to triode emulation (not hitting the right keyword I'd guess). Can someone remember any such posts or related threads?

Erik

[merlinb]

I've searched the site for actual harmonics plots from different/any SS clipping circuits- thinking some must have been made/posted over the years - but couldn't find anything

It's not something I have seen posted very often. Harmonics of a clipping circuit in isolation tell you very little about how it will sound, so it's not something that is likely to get such detailed examination. When it comes to distortion, the method of clipping is only third on the list of important things. Dynamic shifts and EQ are what count.

What exactly are you trying to do / find out?

[lion]

Thanks Merlin - I see what you mean re harmonic plots. If I would go ahead and try testing different circuits myself, I was just curious to see what/how others might have done.

On what I'm trying to do - well, nothing definite right now. I'm just an old DIY'er enjoying modding and building guitar related stuff for my personal use, and there always seems to be around a dozen electronics related issues in my head I'm trying (struggling mostly) to understand, or things I'd like to learn

Distortion characteristics is one interesting subject. I read somewhere that analog tape saturation/distortion is purely 3rd harmonics, but in the same article it was consequently characterized as SOFT clipping, which confused me a little (the SOFT thing vs noticeable amplitudes of harmonics - especially 3rd's and only 3rd's).

I've just found these:
http://home.netcarrier.com/lxh2/web/tapesat.html
http://www.diyrecordingequipment.com/products/15ips-colour

Almost identical approaches. Clearly symmetrical clipping going on with some filtering - but the amplitude of harmonics must be quite low (but probably what's needed). I guess I'm just trying to get my head around it.

Erik

[teemuk]

I read somewhere that analog tape saturation/distortion is purely 3rd harmonics, but in the same article it was consequently characterized as SOFT clipping, which confused me a little (the SOFT thing vs noticeable amplitudes of harmonics - especially 3rd's and only 3rd's).

Ever scoped tape saturation?

I have. It's pretty much just symmetric and fairly soft clipping. There's a slight frequency dependance in it, and some people fuzz about associated modulation effects like "tape warble" but I have never seen or heard them as such a big deal. In scope screen tape saturation looks like plain symmetric soft clipping.  I don't understand what the fuzz is about. That goal is certainly easy to achieve with multiple different solid-state solutions.

As is, it's not going to be purely 3rd harmonic. There's a diminishing pattern of higher order harmonics too, and depending on amount of saturation (and distortion) amplitude of them is either higher or lower in comparison to 3rd. But overall, there's a slight truth to it: Very soft symmetric clipping distortion produces predominantly 3rd order harmonic distortion and amplitude of other harmonics is somewhat negligible. To a point. The more you "overdrive", the higher the amplitude of other distortion products also rises. Pretty much how all clipping distortion works. The idea that tape saturation is "pure" 3rd harmonic is fallacy.

That of course applying to harmonic distortion only. Intermodulation is a different matter. Whenever you have signal consisting of more than one frequency any harmonic distortion will also create intermodulation frequencies. For this reason symmetric clipping also produces less IM overall since it produces less harmonic frequencies of distortion than asymmetric clipping.

[teemuk]

Harmonic plots? I found this site quite illustrative on concepts of different kinds of waveform distortion:

http://www.innerfidelity.com/content/headphone-measurements-explained-total-harmonic-distortion-plus-noise-part-1

[PBE6]

I created a few diode clipping models in Excel last year based on iterative solutions to the Shockley model. They actually line up very well with scope observations for simple circuits. I can either upload or forward them to you if you want to play around with them. The nice part is you can either use Solver or the Fourier function to get an idea of what harmonics are generated by the diode clipping itself.


Sent from my iPhone using Tapatalk

[anotherjim]

Good link Teemuk...
And with reference to that...
I think the easiest S-curve amp has to be the CMOS inverter biased for "linear". Worth experimenting with. For symmetrical soft clip I would suggest the 4069UB or 4007UB devices. Then there is the 4049UB, which is often use in CMOS distortions, but is not symmetrical as its output is designed to sink current rather than source it. Probably that asymmetry makes the 4049 a compound curve soft clipper & may explain why its chosen for so many distortions  - Red Llama, 49er etc.
I'm liking what I get with a common source N channel pre-amp followed by an inverter stage, all from a 4007 chip at the moment.

Here's a conundrum. If you feed a triangle into a CMOS inverter, the soft clip can fold the wave to look very much like a sine wave. As the triangle is all odd harmonics - does that mean that the soft clip has filtered those odd harmonics out?