When you need more distortion...

[Bill Mountain]

When you need to increase the amount of distortion in a stage do you prefer to increase the gain in that stage or to increase the size of the input signal?

I'm playing around with these ideas for a few projects and I was wondering how y'all aproach this.

[Mark Hammer]

It wold depend on how easily the gain is modified in the existing circuit, and where in the circuit the clipping occurs.  In some instances, altering the gain of the existing circuit would change the tone-shaping in ways thatmight have to be compensated for.  In other instances, it might be the tone-shaping of the external boost that brings out the best in the circuit you're trying to clip.

[Bill Mountain]

It wold depend on how easily the gain is modified in the existing circuit, and where in the circuit the clipping occurs.  In some instances, altering the gain of the existing circuit would change the tone-shaping in ways thatmight have to be compensated for.  In other instances, it might be the tone-shaping of the external boost that brings out the best in the circuit you're trying to clip.

All great points.  Let try something more specific then.  Say you have an opamp clipper (soft or hard - your choice) and you want to increase gain.  So the easy thing would be to change the feedback resistors.  But like you said that can change the filtering in that section as well.  Or you could put a boost to increase the size of the input signal.  Are there any tonal differences if the tone shaping isn't factored?

Personally, I don't like getting all of my gain from one stage but I don't have any hard science to back up my preferences.  But I also don't like having too many components when a simple part swap would get my 90% of the way there.

I guess it also has to do with your tonal goal.  Do you need to level match different guitars so they sound the same in the circuit?  Or do you just want to be able to go from light overdrive to balls out distortion with the same circuit?

[FiveseveN]

Are there any tonal differences if the tone shaping isn't factored?

In "soft" TS-type yes, in diodes-to-ground types there shouldn't be, provided the booster itself doesn't add saturation. Which brings me to another point: you could also increase overall saturation by using a compressor, another dirty boost/overdrive/distortion being an extreme example of this. That would certainly cause a tonal change, also might help achieving a gradual transition between distortion levels (when one stage clips before the other).

[Mark Hammer]

I would think that, in a perfect world, cascading two 10x gain stages, with no audible restrictions on bandwidth, should not result in anything audibly different than a single op-amp stage set for a gain of 100x.

As has probably been mentioned many times in the past 820,000 posts, frequently what varies between distortions is how much gain is applied in the circuit.  The same amount of gain, withtwo different input signal levels will appear to produce different sound qualities, and conversely, the same input signal into two largely similar circuits with slightly different maximum gain will yield different tone, just as the same circuit, gain, and input will yield different tone if the clipping threshold is different.

Finally, keep in mind there are limits to gain, dictated by the chip properties and the power supply.  You can't get a gain of 1000x from a chip powered by 9v, and seeing 100mv at its input.  Let me correct that.  You MAY be able to get a gain of 1000x for the teeny miniscule parts of the signal that can still be amplified 1000x and still come in under +/-3.5 to 4v.  But you won't be able to amplify the firmly picked low E all that much, on a 9v supply, before you run into headroom issues.  At that point, you are hearing the chip clipping in addition to whatever the diodes or intended clipping element is doing.

Does that make sense?

[Earthscum]

You may try simply boosting the signal however much you need, and use inter-stage LP filtering.

I've been playing around with shelving lately and found that (for bass, applicable to you) if I pass everything above, say 400Hz (rough mark) at 4x gain, a .1V (.4V now) signal won't clip the schottkys I use, but they will come close to the threshold. The rest of the lower fundamentals get amplified and passed right through the clippers (this is diodes to ground) and end up being just a bit louder than the lower gain signal, so it gives the appearance of blending clean and distortion.

The easiest way I found to play with this was to make a buffer to feed a BMP tone stack. Then I just used a non-inv opamp  clipper. 1M pot feedback (220p parallel cap) and a 1k to 4.7uF to ground (fairly generic values for getting low to high gains that works great for experimenting). I found that I can filter down the highs a bit after all is done and still be left with a fairly clean, and reactive, near ultra-gain. If you want to toast the signal a bit more, use a pre-gain stage in front of the tone stack.

Anyways, that's what I've played with that may be of help. The setup helped me kind of get a grasp on how I want to start handling my signals.

[StephenGiles]

I would think that, in a perfect world, cascading two 10x gain stages, with no audible restrictions on bandwidth, should not result in anything audibly different than a single op-amp stage set for a gain of 100x.

As has probably been mentioned many times in the past 820,000 posts, frequently what varies between distortions is how much gain is applied in the circuit.  The same amount of gain, withtwo different input signal levels will appear to produce different sound qualities, and conversely, the same input signal into two largely similar circuits with slightly different maximum gain will yield different tone, just as the same circuit, gain, and input will yield different tone if the clipping threshold is different.

Finally, keep in mind there are limits to gain, dictated by the chip properties and the power supply.  You can't get a gain of 1000x from a chip powered by 9v, and seeing 100mv at its input.  Let me correct that.  You MAY be able to get a gain of 1000x for the teeny miniscule parts of the signal that can still be amplified 1000x and still come in under +/-3.5 to 4v.  But you won't be able to amplify the firmly picked low E all that much, on a 9v supply, before you run into headroom issues.  At that point, you are hearing the chip clipping in addition to whatever the diodes or intended clipping element is doing.

Does that make sense?

Good morning from very sunny Malaga in Spain! There is a lot to be "gained" by using a dual 15v power supply which would eliminate headroom problems - 6 high capacity rechargeable PP3 batteries should do the trick if you don't want to dabble with mains. I use ExPro AA rechargeables in my camera which run for 500 + photos taken in RAW format (not "in the raw!!), which you can buy from Amazon. I should think that they do PP3 as well.

[brett]

Let's do the math..
A 0.1V p-p guitar signal though an op-amp with dual Si diodes start clipping at a gain of (0.7x2)/0.1 = 14. A lowly 14. Such a thing produces almost a square wave for such a signal at a gain of 100 or more. Not *quite* square (there's a capacitor in the feedback path of most of these thing, for a start).
Does extra gain change the sound of such a thing? Not much. That's why people don't rush around putting bigger feedback pots in their tubescreamers (or boosters in front of them). Extra gain doesn't turn them into fire-breathing monsters.
People are usually more interested in changing the diodes (Ge = soft LED=hard) and the filter cap (22pF=hard 220 pF=soft) for different tones. One of my favourite fuzzes - the MFZ-1 - is high gain (max ~1000), but has a big feedback cap, which gives a lovely, fat tone. Not a "high gain" sound at all.
cheers

[Mark Hammer]

Brett, you'd be absolutely correct...if the unaltered guitar signal were stable and had constant harmonic content.  The reality is that it ISN'T stable, nor is the harmonic content during the first 100msec representative of what will be present 300msec or 2 sec later, in a freely vibrating string.

What the additional gain often does is apply the same sorts of tonal/spectral/amplitude changes to the guitar signal well after the initial pick attack, that, at much lower gains, only apply to the initial pick attack.  I've maintained for a while that often the difference between what people call an "overdrive", or call a "fuzz", pertains to the duration of the tonal changes produced.  If a picked string sizzles after a second and a half the same way it did when you first picked it, you're more likely to think of it as a "fuzz".  If the additional harmoic content generated quickly subsides, you're more likely to think of it as an overdrive.

One of the things people like about overdrives is that it makes emotional expressiveness more a function of one's picking because it can quicklly subside.  Whereas with a "fuzz" you'll get fairly constant harmonic content almost regardless of how you pick - it will almost depend more on the knob setting than your pick attack.  Being alble to spontaneously moderate harmonic content is a huge part of inserting emotion into guitar tone, and keeping the gain just right allows one to do that.

[Bill Mountain]

I really need to get a scope of some sort.

I have this mental image of a signal reacting differently depending on where the gain is generated.

I have an ultra high output passive bass that when I build a dirt box for it a normal bass can barely clip it.  If I build it for a normal bass then my bass distorts much earlier than I planned for it.

When trying to even out the gain I often prefer to boost the signal versus raising the gain in the distortion pedal.

Why?  Because it never sounds right if I just turn up the distortion.  Either the frequency response changes or the clipping sounds different.  My only explanation has been that distortion stages will distort differently when higher input gain is present versus asking the distortion engine to produce more gain itself.

I know this is dependent on so many different variables but it is a design element that I always consider.

[Earthscum]

In a lot of circuits, when you adjust the gain, you are also adjusting a filter. Take the Tube screamer... the gain pot and the small cap make a HP filter in the feedback (which results in LP filtering at the output). That is most likely why it sounds different boosting pre vs. turning up the gain.

[Bill Mountain]

In a lot of circuits, when you adjust the gain, you are also adjusting a filter. Take the Tube screamer... the gain pot and the small cap make a HP filter in the feedback (which results in LP filtering at the output). That is most likely why it sounds different boosting pre vs. turning up the gain.

Oh I know.  I've been studying filters for a while to make sure my gain stages work with all of my guitars and basses and sometimes it just doesn't.  So I got an MXR Micro Amp and It has helped me a lot but I'm trying to avoid having to include it in all of my builds!

[Bill Mountain]

One thing I will be working on this weekend is an op amp distortion.

I'm going to have 2 pots in the feedback path.  One will set the gain ratio and the other will be the actual gain control.  I plan to set the "gain ratio" to just before the signal clips and then the regular "gain" pot will have a better range of clipping.  My high output bass sounds great with a voltage gain of 100 where my back up (quieter) bass needs somewhere close to 600 to get the same amount of distortion.

I'm hoping to set up the "gain ratio" pot to not have any effect of the low frequency roll off and let the "gain" pot roll off the lows as I turn it up.  This way I hope to be able to get similar filtering with different input voltages.