Op amp Clipping

[PacoPicoPiedra]

Found this: http://experimentalistsanonymous.com/board/viewtopic.php?t=1504 on experimentalists anonymous,
I've been looking for something to replicate this sort of distortion for ages and have some questions:

As I understand it, this just produces your normal hard clipping, so why does it sound so different to say silicon diodes or anything else?
Is there any other (safer? ) way to get this sound?  Everything I read seems to say op amps hate being clipped, so obviously I'm misusing them by doing this, but how likely am I to break anything? What's the benefit in this application of using the non inverting configuration, just wondering why the guy used it over inverting. Also, assuming this is the only way to go, any recomendations on op amps to use?, I keep hearing things about tl071's latching up, so what's more suitable? low noise seems to be the most important factor atm

[bioroids]

Strange, because the image posted at the link is for an Inverting configuration...

[R.G.]

As I understand it, this just produces your normal hard clipping, so why does it sound so different to say silicon diodes or anything else?

It is far more abrupt and far more solidly clipped. This is because even when clipping there is some slope to the voltage versus current characteristic of the diodes, so the clipped waveform is never truly flat on top, and there is always some roundness to the leading and trailing edges of the clipped waveform. In an opamp driven past its power supply limits, the opamp goes from acting normally to dead clipped instantly; this produces no detectable rounding on leading or trailing edges and a really flat topped clipped waveform. In the frequency domain, which is where we hear, the sharp corners and flat tops produce a different balance of harmonics - higher harmonic number and hence less "musical". They sound harsher.

Is there any other (safer? ) way to get this sound?  Everything I read seems to say op amps hate being clipped, so obviously I'm misusing them by doing this, but how likely am I to break anything?

Well, it's not dangerous per se. Opamps are used as comparators, and in that use they spend 100% of their time saturated one way or another. Some opamps tolerate this better than others. Some get downright ugly. As my doctor says when I say "Doc it hurts when I do this...",  don't do that. Use an opamp which states "no latchup when common mode range exceeded" and you'll be fine. At least, the opamp won't die. By the way, there is no civil or criminal statute about not torturing opamps. If it sounds good to you to twist their legs off, do it. There are no opamps' rights organizations that I know of.

What's the benefit in this application of using the non inverting configuration, just wondering why the guy used it over inverting.

He didn't. That's an inverting circuit.

Also, assuming this is the only way to go, any recomendations on op amps to use?, I keep hearing things about tl071's latching up, so what's more suitable? low noise seems to be the most important factor atm

Your mistake is listening to the internet. The internet will tell you anything at all, most of it not true. You know the old saying "Lies, damned lies, and statistics"? A good translation today would be "Lies, damned lies, and the internet". Don't "keep hearing things" about your parts. Go look up the data sheets and educate yourself to figure out what's good and what's not. Anecdotal evidence is famous for being misleading.

[petemoore]

If it sounds good to you to twist their legs off, do it. There are no opamps' rights organizations that I know of.
  I know it's serious business here, but I'm getting a big chuckle just the same...

[PacoPicoPiedra]

cheers guys, sorry for thinking it was non inverting, just repeating the description but having read up a bit yeah, i can see it's inverting. i've read the datasheet, tlo72 seems to be fine for the job. so clipping occurs when the gain of the op amp tries to push the signal through a larger range than the supply voltage, right? so if i were to run it straght of a 9v supply it's going to clip at +/-9v.
basically, would it be useful/necessary to put in another gain stage to recover the volume lost from the clipping?
because +/-9v of swing is pretty large isn't it, so i'm thinking I'd need to attentuate the output rather than amplify it again to have everything at sensible guitar levels unless i've misunderstood how the clipping threshold works

[bioroids]

With an opamp like the TL072 you cannot swing up or down that close to the power rails. With a 9v supply it will clip about 1.5v away from each rail, so you'll have a 6v max swing.

Luck!

Miguel

[petemoore]

  If you input your guitar to a 'railed out' opamp @9v, I would recommend starting it off with the attenuation set to low volume, expect much greater output voltage from the opamp than what is input.

[WGTP]

SPANK THOSE OP AMPS ONE TIME FOR ME.   

If you using a Distortion +, Ratt, DS-1, etc. your getting op amp clipping, in addition to the diodes to ground clipping.

[PacoPicoPiedra]

ok, did some reading up and I've done this basic layout


input impedance is fairly low, might cause some rolloff but wouldn't have thought it would make any difference once its all been clipped to death. erm, don't know how to calculate the output impedance so i'm a bit lost as to the value of the output cap, help appreciated
i'm taking the gain to be the resistance in the feedback loop/input resistor so the 500k pot should give a decent range methinks.
also have seen people putting caps in the feedback loop (to stop oscillation?) but i've no idea about this and it's not mentioned in my op amp handbook so just left it out til I understand why and what value to use

if it's any help, i'll be using this on bass so i just want a flat-ish response and i'll stick a 7 band eq in front

any flaws/improvements/stupid misunderstandings people can point out would be much appreciated, i'm very much still learning

[Joe Hart]

Don't you have to filter the power supply?  Look at a Dist+ schematic and just take out the diodes to ground after the opamp. That should work for a starting point. Then play with caps and resistors to get what you like.
-Joe Hart

[bioroids]

The impedance of that circuit is too low, you will loose a lot more than some highs! And they will be lost before the clipping may take place.

Also the gain is the feedback resistor divided by the input resistor. Don't forget the guitar pickup has an impedance, that counts as a resistor too (and it varies with frequency), so the gain will be less than you'd expect, and even less for the higher frequencies.

The output cap is not calculated with the output impedance of your circuit, but with the input impedance of the next one. As you never know what will be connected after your effect, just put a big cap like 1uf and you are done with it.

Also, the input cap has the reverse polarity, the + side should point to the higher voltage. In this case you have 0v at the guitar output, so the + side should be at the inside of the circuit.

Greetings!

Miguel

[WGTP]

Check out the new Mockman at ROG.   

[earthtonesaudio]

If you want to use a single battery for your 9V, look up how to bias an op-amp for single supply operation.

[PacoPicoPiedra]

that mockman seems to be pretty much what I'm looking for, the sound clips werent really my thing but i think thats more to do with the bass cut than the basic design, how would i go about flattening out the response? also on the layout i see the gain is really just a volume control on the input, is this a normal/good way to implement this? wouldn't this make it even more prone to noise at high gain as you're increasing the input impedance?

[PacoPicoPiedra]

Edit: *Low gain-it was late when i posted that   

[PacoPicoPiedra]

update:
breadboarded the mockman with a few cap changes and loving it. just one thing, didn't have too much sustain, so figured I'd add a couple more gain stages, just repeated the previous two with another TLO72. problem is now though that it's started gating like it's not getting enough voltage, but all the pins are at the right voltages still. the gating always starts when I connect the positive rail to the second op amp, even if im just listening to the output from the first. can anyone explain?

[Sir H C]

Another fun thing to do is load the output of the op-amp to get it into current limiting.  Should be safe, but might kill it over time (metal migration).  Latch-up is not usually the op-amp issue I knew of, more the inversion of the inputs.  To me latch up is an SCR turning on on the chip.

[PacoPicoPiedra]

current limiting? whats that sound like then?

[Sir H C]

It is just a different way for the op-amp to clip.  Depending on the current limiting circuit used in the op amp (or even if it has one), it will usually be less sharp than limiting from the rails.  Many circuits that have diodes to ground right off the op-amp output will get current limiting instead of rail clipping.

[WGTP]

I don't guess I had thought of that before.  That may account for why some diode to ground circuits have 1K resistors before the diodes and others have 10K.  Hmmmm.