Tonefiend Forum

These are LTSPICE simulations of a sine wave going through a number of increasing levels:

50 mV, 100 mV, 200 mV, 500 mV, 1V, 2V

Only the basic distortion element is included in the simulation. The Klon especially has a complex frequency shaping network that further contributes to the pedal's character.

Here's a waveform set for the Klon.

And here's the Tube Screamer.

Now, would you call the Klon a TS clone? I wouldn't!

I was really surprised when I saw the TS waveform. It's important to take these things with a grain of salt, because I've removed all tone shaping circuitry (that is, frequency-dependent things like capacitors), and the highest input signal was 2 volts, which is probably hotter than you're normally going to get coming right out of a guitar.

What I find most fascinating is the spacing of the waveforms. The Klon generates its clipping by running a gain stage into a back to back diode clipper load, so it's a lot like one of those passive diode distortions waveshape wise. As your input gets higher, the output crowds together and gets more distorted.

The TS on the other hand, puts the diodes into the feedback loop of a NON-inverting op-amp. So when the output level gets above the diode forward voltage, the op-amp stage gain goes to 1, not to zero like it would in an inverting stage. A non-inverting op-amp buffer (unity gain) puts the signal to the + input and the output goes back to the - input.

So what you see from the TS circuit at the highest levels is more like a sine wave riding on top of a square wave. It actually CLEANS UP somewhat with higher input signals.

They don't teach you about this stuff in engineering school!

Hi mwseriff,

I do not take issue with your conclusions. If the end result is music, #1 the judgment is going to be personal, and #2 looking at a waveform doesn't tell you as much as you'd like to know about how it will ultimately sound. So the fingers and ears have to be the final judge.

I'm an engineer first and a musician second, so I do look at musical electronics from an analytical perspective. I also readily acknowledge that to the extent I studied audio electronics, the focus was on creating amplifiers with low distortion and flat frequency response - for "hi-fi". So, almost nothing I learned is relevant to creating guitar-oriented musical audio devices. I also did not study tubes in school although one chapter of my semiconductor electronics textbook (written in 1966) does deign to talk about vacuum tube circuit models.

My comparison of these two circuits was triggered by Joe's statement along the lines of "I'm tired of op-amp and diode distortions". I took a look at some (clone) schematics online, and although the Klon and TS are both op-amp and diode distortions, at their core, the way they generate distortion is quite different. You could, for example, take the heart of the TS and drop it into a Klon circuit (or vice versa), and what would happen? Would it sound more like a Klon, or a TS? I don't know. Who wants to try? I think the TS circuit is pretty clever, but it's also certainly not how any tube amp I am aware of actually operates.

I've noticed that almost every diode distortion circuit I've seen uses identical diodes, so the distortion thus created will favor odd rather than even harmonics. This effect has a pretty strong correlation between what the waveform looks like, what the spectrum looks like, and what it sounds like. Once you know the difference between the SOUND of even vs. odd harmonics, then you can make choices in your own experiments. Maybe you want one, the other, a switch or maybe a continuous blend between them. My suggestion to anyone who wants to build a diode-based distortion is to at least put a switch in there so you can unbalance it. If you don't like that sound, OK - leave it out. But maybe that was just what you were looking for.

For people who want to go beyond paint-by-numbers DIY, I think some understanding of how things really work is important. For example, power amp distortion from a single ended tube amp sounds different from a push-pull amp, and furthermore that may also depend on the type of phase invertor circuit you are using. No amount of analysis will tell you which one you will prefer, if any of them. However, once you identify something you like, knowing why it works can only be helpful as you continue your pursuit to recreate it and make it better.

Quote from Digital Larry on October 19, 2012, 09:53
Hi mwseriff,

I do not take issue with your conclusions. If the end result is music, #1 the judgment is going to be personal, and #2 looking at a waveform doesn't tell you as much as you'd like to know about how it will ultimately sound. So the fingers and ears have to be the final judge.

I'm an engineer first and a musician second, so I do look at musical electronics from an analytical perspective. I also readily acknowledge that to the extent I studied audio electronics, the focus was on creating amplifiers with low distortion and flat frequency response - for "hi-fi". So, almost nothing I learned is relevant to creating guitar-oriented musical audio devices. I also did not study tubes in school although one chapter of my semiconductor electronics textbook (written in 1966) does deign to talk about vacuum tube circuit models.

My comparison of these two circuits was triggered by Joe's statement along the lines of "I'm tired of op-amp and diode distortions". I took a look at some (clone) schematics online, and although the Klon and TS are both op-amp and diode distortions, at their core, the way they generate distortion is quite different. You could, for example, take the heart of the TS and drop it into a Klon circuit (or vice versa), and what would happen? Would it sound more like a Klon, or a TS? I don't know. Who wants to try? I think the TS circuit is pretty clever, but it's also certainly not how any tube amp I am aware of actually operates.

I've noticed that almost every diode distortion circuit I've seen uses identical diodes, so the distortion thus created will favor odd rather than even harmonics. This effect has a pretty strong correlation between what the waveform looks like, what the spectrum looks like, and what it sounds like. Once you know the difference between the SOUND of even vs. odd harmonics, then you can make choices in your own experiments. Maybe you want one, the other, a switch or maybe a continuous blend between them. My suggestion to anyone who wants to build a diode-based distortion is to at least put a switch in there so you can unbalance it. If you don't like that sound, OK - leave it out. But maybe that was just what you were looking for.

For people who want to go beyond paint-by-numbers DIY, I think some understanding of how things really work is important. For example, power amp distortion from a single ended tube amp sounds different from a push-pull amp, and furthermore that may also depend on the type of phase invertor circuit you are using. No amount of analysis will tell you which one you will prefer, if any of them. However, once you identify something you like, knowing why it works can only be helpful as you continue your pursuit to recreate it and make it better.

---------------------------------

I agree with all you say and my point is engineering is a good start but then it needs a lot of hands on guitars with ears doing the hard work of listening. This is also true in hi-fi design (where I have also worked in the past) We used to do tons of blind listening tests which were a large part of both speaker and amp designs. We had a facility in a very small farm town in Illinois. We literally got farmers, middle aged ladies and even retirees off the street to come in and listen to different versions of speaker x-overs and amplifier circuit changes. In properly done blind comparisons they provided us with good data. These average people could hear the differences between electrolytic caps with etched foils vs. non-etched foils for example. There was no way to measure those differences at the time (and probably not even today ). We are talking very subtle differences but the kind of thing that makes and amp or speaker easier to listen to for long periods of time (reduced ear fatigue). But we also found many other problems using test equipment like impedance problems in speaker x-over design done entirely by ear (this is a problem where even simple test equipment would have found the problem). Back in the 80's & 90's there was a lot of ear designed esoteric hi-fi and we actually did instrumentation analysis to fix design issues for the "ear guys" (as we called them). So I have seen both ends of design and IMHO you need both instruments and ears. The ear and brain are capable of very subtle discernment but are decidedly difficult to use in an equation. In my experience electronic instrumentation gets you to the ball park but it takes ears to get you to your own seat.

As a long time musician I would never purchase a pedal based on how a bunch of sine waves look after going thru it but as a designer I would use those waveforms to tell me I was on the right road and it was time for ears and guitars to get to work. I have always used as much instrumentation to do my work as possible however they can be very seductive but still require careful listening tests to assure quality of design or repair. A good example is bias in fixed bias output tubes (matched set) I use a set of meters to start but use a guitar to check it by ear (and my customers could tell the difference). What I advocate is use all the tools available and remember to value your own senses.

Quote from mwseniff on October 20, 2012, 03:41
I agree with all you say and my point is engineering is a good start but then it needs a lot of hands on guitars with ears doing the hard work of listening.... What I advocate is use all the tools available and remember to value your own senses.

I'm with you all the way here. One of my formative experiences as an engineering student unfolds as follows:

Professor draws a simple RC low pass filter on the board and writes some component values for the R and the C.

He then asks, "what is the time constant of this circuit?"

All the eager students quickly reach for their HP calculators to get the answer.

Professor jumps up and SCREAMS "STOP!!!!!!! You are engineers, not scientists!!!! Your responsibility is to QUICKLY get an answer that is 'close enough' to allow you to proceed." He had chosen the values so that a mental calculation would have been easy.

A lot of people are put off by the mathematics involved in engineering. This includes the engineers themselves who are constantly looking for simplifying approximations that make getting a quick answer easier. Ultimately, when your tone circuit is filled with 10% tolerance components, your calculated answer was wrong anyway - but at least it was close.