Op amp distortion via limiting supply voltage

[Acryliceater]

Hello

I am still a relative newbie when it comes to electronics, but I was wondering about using op-amps for a distortion type circuit

My idea is basically to either amplify a guitar signal beyond the voltage levels supplied to the op amp to achieve distortion and then attenuating the signal with a voltage divider.

I also thought about limiting the voltage supplied to the op amp to do the same.

Does anybody know if this idea could be viable or has been done before?

[teemuk]

Perfectly solid idea and yes, has been done before.

[MrStab]

Generally, that signal decapitation is done by clipping diodes at a lower threshold than the supply rails, as op-amps tend not to clip in a pleasing way.
The Sansamp GT-2 is an exception, though - the TLC2262 it uses clips nicely. Or so I'm told.

[antonis]

Hi & Welcome..

In case you want to implement just overdrive (no clipping diodes or any other item) I'd suggest a single BJT CE amp configuration, in the form of high gain booster..
(FETs are not recommended due to lack of transconductance, hence not enough gain - among other flaws..)

But, as already Grant said above, for an effective (in the mean of audio pleasing) op-amp distortion circuit, it should be a good idea to implement some form of clipping..
https://generalguitargadgets.com/how-to-build-it/technical-help/articles/design-distortion/

[Steben]

Generally, that signal decapitation is done by clipping diodes at a lower threshold than the supply rails, as op-amps tend not to clip in a pleasing way.
The Sansamp GT-2 is an exception, though - the TLC2262 it uses clips nicely. Or so I'm told.

Yes. Not very very soft - it still is hard clipping - but nice with some small roundness and not that much "latch-up" (not needed to be explained here).
TL07x opamps are omnipresent and good allround opamps in diode clipping circuits yet are bad at full clipping since they act very odd beyond the rails....

[marcelomd]

Lots of pedals use opamps for clipping.

Several Sansamps do just that. Later models use zener diodes.

Pedals in the Guv'nor family, Shredmaster, Suhr Riot, JHS Morning Glory, King of Tone,  etc. clip the opamps first, before some kind of diode clipping. Same for the Klon. They all use TL07x. RATs do that too.

But I've never seen changing the actual voltage supply to do that. Might be interesting.

And I have to disagree with Antonis, unless I'm missing something. . FETs do clip in a pleasing way. Bunch of good designs overdrive JFETs, MOSFETs and even MOSFET logic gates.

[garcho]

Set up an op amp with variable gain (~100), and then use a LM317 to adjust PS voltage. Breadboard city!

even MOSFET logic gates.

Tube sound fuzz et al

[garcho]

This makes me think, maybe there's a place for 1.8V op amps in diy pedals after all! 

[MrStab]

Are those pedals meant to clip the op-amp, or were they designed for an era when pickups were more sensible, before EMGs and Invaders started taking things a bit too far?

Invaders are stupid! I hate mine! #regrets

[antonis]

And I have to disagree with Antonis, unless I'm missing something. . FETs do clip in a pleasing way. Bunch of good designs overdrive JFETs, MOSFETs and even MOSFET logic gates.

I DO agree but not in case of a single (J/Mos)Fet CS amp, as I proposed above..
For a +9V single supply and symmetrical clipping there isn't enough gain for a notable distortion - not for ordinary guitar sinal amplitudes, at least..

[MrStab]

Using this lazily-searched chart as a reference:

https://blueguitar.net/new/misc/pickup_mv_chart.pdf

Assuming max gain in the Marshall Guv'nor, a bog-standard strat pickup at 40mV (I assume Vpp) will reach 1.86V after the first op-amp, then attempt to reach 126V after the 2nd, but you're not hearing the curve of the TL072 hitting the rails because that signal will be clipped by the LEDs far below that threshold. Characteristics of op-amp clipping are therefore discarded. No?

So it likely does clip the op-amp, but doesn't end up in the output signal.

[marcelomd]

Using this lazily-searched chart as a reference:

https://blueguitar.net/new/misc/pickup_mv_chart.pdf

Assuming max gain in the Marshall Guv'nor, a bog-standard strat pickup at 40mV (I assume Vpp) will reach 1.86V after the first op-amp, then attempt to reach 126V after the 2nd, but you're not hearing the curve of the TL072 hitting the rails because that signal will be clipped by the LEDs far below that threshold. Characteristics of op-amp clipping are therefore discarded. No?

So it likely does clip the op-amp, but doesn't end up in the output signal.

The LEDs are after the second stage, so you hear the LEDs clipping an already clipped signal. This sounds a bit different from pure opamp or pure LED clipping in a single stage.

[Steben]

Lots of pedals use opamps for clipping.

Several Sansamps do just that. Later models use zener diodes.

Pedals in the Guv'nor family, Shredmaster, Suhr Riot, JHS Morning Glory, King of Tone,  etc. clip the opamps first, before some kind of diode clipping. Same for the Klon. They all use TL07x. RATs do that too.

But I've never seen changing the actual voltage supply to do that. Might be interesting.

And I have to disagree with Antonis, unless I'm missing something. . FETs do clip in a pleasing way. Bunch of good designs overdrive JFETs, MOSFETs and even MOSFET logic gates.

As antonis explained a single jfet stage is quite different from an jfet input opamp. The latter is extremely useful as input opamp for guitars but can sound awful when clipped. A tlc2262 does just that in a perfect manner. Boss discrete opamp circuits are designed for it as well.

[MrStab]

This sounds a bit different from ... pure LED clipping in a single stage.

Why?
I dispute this, unless slew rate or bandwidth degenerates as the op-amp clips.

[marcelomd]

I DO agree but not in case of a single (J/Mos)Fet CS amp, as I proposed above..
For a +9V single supply and symmetrical clipping there isn't enough gain for a notable distortion - not for ordinary guitar sinal amplitudes, at least..

Yeah. One, lonely, voltage starved buffer probably won't sound good. But, in most applications, you will have some pre- and post eq/filtering, which is the secret sauce. That's where I'm coming from.

Why?
I dispute this, unless slew rate or bandwidth degenerates as the op-amp clips.

What I'm saying, and maybe I intercommunicated, is that two stage clipping sound different from one stage clipping. More because what's surrounding the clippers (transfer functions, corner frequencies, impedances, etc.) than the clippers themselves. Each stage adds it's own information, and doesn't completely overwrite the previous one (excluding the most extreme cases, of course).

But don't trust me. Let's simulate or breadboard something =)

As antonis explained a single jfet stage is quite different from an jfet input opamp. The latter is extremely useful as input opamp for guitars but can sound awful when clipped. A tlc2262 does just that in a perfect manner. Boss discrete opamp circuits are designed for it as well.

My n=2 datapoints is that I breadboarded a Guvnor and a Klon to try different diode sets. They sounded ok with none. Absolutely not saying this is always the case.

[MrStab]

A simulation could be informative and fun. There is a TL071 LTSpice model in existence. I'm not entirely sure where to look on datasheets for overload nuances (latch-up aside), though I'd expect op-amps with internal protection diodes to just clamp then be done with it.

[marcelomd]

Got this snippets from the Guvnor. LED, opamp and LED+opamp.

The opamp model is the TL071 from Tina.

[ElectricDruid]

I'm not entirely sure where to look on datasheets for overload nuances

That's the kind of thing that datasheets describe as "not characterised", if they even mention it at all.

[Mark Hammer]

I take the view that placing all the gain in one place/device doesn't necessarily lead up to the most pleasing clipping.

In which case, if one was going to push an op-amp past its headroom, a more pleasing sound might be obtained by use of two single op-amps, the first one having a lower supply voltage than the second.  One's reflex might be to use a dual op-amp, but then BOTH halves would have the same supply voltage.  I'm thinking in terms of the first one/stage having a lower V+ (e.g., 5V), then the second one applying some gain to the stage 1 output, but having a higher V+.

That approach also allows one to shape the signal, in the first stage, to get the best out of the second one.

And if I'm not mistaken, and haven't misremembered, doesn't one of the Boscorelli distortion circuits use cascaded LM324 stages to deliberately exploit their clipping properties?

[MrStab]

What happens if diode-only sim has its input voltage swing set to 9 or 10Vpp?