Tube Guitar Amplifier Overdrive Webpage

[robrob]

Hey, could you guys do me a favor and review my first draft of a web page about tube guitar amplifier overdrive? Please let me know what I got wrong. It's an early draft and I plan to flesh it out but I'd like to catch any basic mistakes early. I know the guys in this forum know more about tube amp theory than any other group of humans so I know they won't pull any punches. Thanks in advance.

https://robrobinette.com/Tube_Guitar_Amp_Overdrive.htm

[jazbo8]

Have not read the whole page in detail, but the transfer characteristic chart jumped out at me, the chart you show, is commonly referred to as the Plate Characteristic. Instead, the Transfer Characteristic is better at illustrating the Eg1-Ia curvature of the tube vs. a straight line, i.e., a perfectly linear device.

[robrob]

Thanks jazbo8, I agree, I swapped out the chart.

[MarkusBass]

It seems to me that the web page was inspired by this book: http://music-electronics-forum.com/t40953/
I checked few other pages on your web site and I found out a LTSpice simulation of the VHT Ultra tone stack. Since I use LTSpice on a daily basis I have few remarks to it. The simulation does not run since there are many errors in it, starting from missing simulation command, incorrect INCLUDE command and missing AC voltage source (in the download there is only 300mV DC source, which cannot be used for tone stack simulation). And the list of mistakes in such a simple simulation is much longer than that. What was the purpose of posting it?

Mark

[jazbo8]

You have done a good overview of the subject, but perhaps over-emphasised tube amps' soft-clipping properties vs. the solid-state amplifiers. Follow member Teemuk has written quite a bit on this topic, which could be found by searching the forum and/or from his book. Basically, soft/asymmetrical clipping could be done with either tube or solid-state amplifiers when they are properly designed.

A typical over-driven guitar amplifier contains not just "soft and warm" even-order harmonics but tons of odd-order ones as well, especially when you look at the output of typical pentode output stages. The clipped top/bottom and sharp edges of the waveforms resemble that of a square wave, so you can imagine the distortion spectra that they produce...

[Malcolm Irving]

It’s a nice introductory overview. Not sure if you were influenced by the book Ulrich Neumann and I brought out recently, but I have a couple of comments, which I hope you find useful:

The definition of ‘saturation’ is always a problem. I think the way you are using it, for example in:

“Saturation occurs when the input signal's positive lobe on the tube's control grid goes so positive that no more electron flow from cathode to plate is possible -- the tube is giving you all it can.”

is fine for an introduction, but invariably it’s the inability of the driving circuit to push the grid more positive which is the real limitation causing the clipping in this direction.

In the following section:

“When a tube is overdriven to the point where the control grid voltage goes positive, negatively charged electrons flowing through the grid will become attracted to it. As the electrons accumulate on the grid they push the grid voltage more negative (excess electrons = negative voltage) which shifts the bias voltage toward cutoff. Another way to say this is as the electrons flow from the grid through the grid leak resistor to ground a voltage drop forms across the grid leak resistor which shifts the bias voltage of the tube toward cutoff. This bias shift is very nonlinear so it generates nonlinear distortion.”

I think you are mixing up ‘grid leak bias’ and ‘grid current due to overdrive’. When a tube is overdriven to the point where the control grid voltage goes positive (relative to the cathode) electrons don’t “accumulate on the grid” they flow immediately in the grid wires and out of the tube (some via the grid leak resistor but most via the driving source impedance). (Electrons leaving the tube via the grid correspond to conventional current entering the tube via the grid, of course.)

[pdf64]

Yes, the soft - hard clipping of tubes - solid state seems to me to be a myth that doesn't stand up to investigation.
It may be more correct to explain the difference as being due to the different device characteristics lending themselves to different implementations.
It's a coincidence that the thread was raised, as I had been thinking of emailing Rob to review that page.

It seems to me that the web page was inspired by this book: http://music-electronics-forum.com/t40953/

I seem to remember Rob's web page on the topic pre-dating the announcement of the book.

[teemuk]

Some good stuff there. Unfortunately the whole content would benefit from less biased and more objective overview.

Anyone who has actually scoped tube amplifier circuits knows not to expect marvellously soft clipping from them. Thus they will entirely dismiss those DRAWN images (they do not agree with reality one views from scope screen) and will therefore naturally also seriously evaluate what other content on that site could be misrepresented in similar unrealistic manner.

Anyone who has actually researched design of solid-state guitar amps will also dismiss statements like...

A tube guitar amplifier's soft, asymmetric clipping, time-varying grid current bias shift and screen grid gain shift all combine to make tube overdrive tone more musical than solid state distortion.

...as blatantly false accusations. All such characteristics are trivial to add to solid-state circuitry (and countless examples of this exist). Such dismissal of solid-state solutions only reveals author's lack of knowledge and therefore casts a shadow of doubt over any real informational value of the article. That's the danger of people with limited knowledge posing as "teachers".

Not to mention, clipping of some "classic" tube amps with highish power push-pull output stages turns out to be fairly symmetric and hard under tighter scrutiny. Also, not many people fuzz over "musicality" of distortion of solid-state Crate amplifiers, although they are an excellent example of solid-state circuitry that happens to have fairly soft, asymmetric and dynamically time-variant clipping characteristics that should have been non-existent in SS according to author. By that sense SS Crate amp's distortion should have a greater degree of "musicality" than, say, an overdriven Trainwreck tube amp that happens to have very symmetric and very hard clipping characteristics, which is evidenced by simply scoping their output.

So what to believe? Author or references to real amp designs out there? The two seem to be in major controversy with each other.

If I want to learn math and a book states 1+1=3 I stop reading. Basically applies to this website too. Elementary points of theory should be presented correctly, otherwise the educational value of the content is practically zero. What's a point of an article when the author obviously can't be trusted and it's reader responsibility to filter what claims actually have base on reality and what were simply self-perpeatuating myths.

There's also not much addressed about intermodulation distortion and what effects exaggerated amounts of inherent device distortion, asymmetric clipping or soft clipping have on it. Yet, IMD is key element in characteristics like "note separation" of higher gain tones or in overall clarity vs. muddyness of the signal. It is also a much more distinctive distortion mechanism than plain harmonic distortion (though related), not to mention more relevant to real life signal amplification. It also forces to evaluate the sense of evaluating harmonic frequencies as even or odd (because a practical signal is combination of multiple frequencies and IMD will by nature generate distortion where harmonic frequencies bear no "musical" ratio to each other) Higher amounts of harmonic distortion (introduced by characteristics like soft clipping or asymmetric clipping of practical, real-life circuits) also mean higher amounts of intermodulation distortion and resulting degraded signal clarity. That can work as a musical effect (e.g. "warmth") or it can not, and it's largely defined by context and application.

If pristinely clear signal amplification is goal then harmonic distortion effectively needs to be reduced as much as possible because any harmonic distortion will create intermodulation distortion. Of course pristinely clear signal amplification is seldom goal in guitar amplification whereas it explains why many keyboard, bass, PA, acoustic, HiFi, etc amps are solid-state instead of tube. It is easier to achieve higher power (headroom) with less harmonic distortion. Hard clipping characteristics may be preferred in thse because increased harmonic distortion from "soft clipping zone" of device operation boosts IMD and similarly to degrading note separation in an individual instrument IMD would turn complex musical arrangements with several instruments to muddy, undetailed and distinctively distorted mess. No, a fair bit of sterility is actually damn good in certain applications.

What's considered "sterile" in guitar amps is often be mandatory for many other amplifiers that do not have a thinly veiled purpose of acting as effect processors that amplify with ridiculously poor linearity.

So yes, asymmetric soft clipping can sound great in certain musical contexts. Push too much and it turns to muddy signal with no note separation at all, and still quite never reaches that "high gain" -tone territory. So its sort of retained for those medium gain "warm" tones where distortion is evident, yet not too obtrusive. Good high gain tone with decent note separation, on the other hand, may actually require fairly hard, symmetric clipping. When distortion becomes an intentional musical effect and needs to serve certain individual purposes there are no longer "rights" or "wrongs" in that type of design. What ever fulfills design goals is the "right" way. If preferred tone requires high order harmonics introduced by harder clipping then harder clipping is the goal. If higher overall harmonic distortion of asymmetric clipping is bad for note separation then goal is likely more symmetric clipping.

These are all design choices, dictated by what goals the circuit designer has. On that note, often the goal is not to make a "one-trick-pony" -type device so probably a practical circuit also needs a fair bit of versatility so its characteristics can be altered to serve current preferences of the user. An interesting point is also that introducing shifting DC offsets, shifting clipping (a)symmetry and shifting harmonic distortion pattern - controlled by dynamic envelope of the input signal - will in right combinations of "stacked up" clipping offer solutions that work decently in "both worlds" (medium and high gain). It is no coincidence that many practical guitar amps display such characteristics. They also naturally make the circuit react in more "touch sensitive" and "organic" manner than circuits devoid of such dynamic alteration of clipping characteristics, where harmonic pattern of distortion does not change but harmonic frequencies merely either rise or decay in amplitude depending on magnitude of overdrive.

[robrob]

Thanks for the feedback Dr Irving. My overdrive webpage's purpose is to give a basic understanding of tube overdrive so that books such as yours will be easier to follow. I first heard about your book about a month ago and immediately ordered it. I began writing the overdrive webpage while the book was en route but your book has definitely filled in a lot of holes in my understanding of overdrive. I do plan to add some info on the overdrive input impedance change.

You are right, I was confusing 'grid leak bias' and 'grid current due to overdrive' but I'm still confused on grid current. When the grid goes positive doesn't that mean there is a scarcity of electrons on the grid that would pull electrons from the tube's space charge or driving source impedance?

[J M Fahey]

Ouch!!!!

I HATE to write this, but there is so much wrong (mixed with a little right) that it can't be "corrected" but fully rewritten which is not the point.

In a nutshell: looks like you did a great effort reading a lot, and including a lot of what passes as "common knowledge" and is repeated all over the place ... problem is that most of said "source knowledge" is wrong or inaccurate to begin with

One particular irk is that your explanations are *current* based (running out of electrons or accumulating too many), while clipping is much better explained by *voltage* swing hitting limits (+V or ground) which it can not go beyond.

You even show the top waveform "rounding" and passing beyond the +V rail


You also *draw* waveforms which, to put it mildly, don't match at all what scope screens show ...... big time.


*actual* tube clipping, smooth warm and symmetrical my *ss !! ... thanks God.
From one amp famous for its killer overdrive sound: Marshall 18W which relies on power tube clipping:


more tasty NON "smooth/warm/rounded" tube overdrive **scope** screen waveforms:




Or contradict on the same line:

Like electrical charges repel one another so when the control grid has excess electrons (a negative voltage) they repel and trap the free electrons discharged by the cathode and prevent them from flowing through the grid to the positively charged plate

Well, pick one: repel OR trap?

And when negative biased (the normal way a tube works when amplifying) they do not "prevent" (which in good English means "stop") electron flow, since electrons are always going through, only more or less of them .... unless you biased the tube into cutoff, which is not the normal operating point for an amplifying tube.
And on and on and on.

That said, and on a lighter side, it's not worse than its sources but even more: *nobody* is actually designing these days, what passes for it is a combination of cut and paste plus endless tweaking , nobody will take a blank sheet of paper and design based on this, and the very few who may actually want to *design* in the full meaning of the term, will go straight to the real source (Radiotron Designer's book for example, tons of datasheets, study what commercial succesful designs do and *why* ) plus breadboard the projects, measure, *listen* and tweak, so in fact no real harm is done.

Sorry, hate to sound this way, nothing personal meant, quite the contrary, I appreciate you as a Forum member and see a lot of effort has been put into it.

[robrob]

Thanks teemuk for the 'pull no punches' review that I was looking for. The overdrive page is definitely a work-in-progress and I have only mentioned it online here in this thread and on the closed Facebook "Tube Guitar Amp Builders" group also for review. I posted here for feedback to keep my website as factual as possible. I'm always open to corrections. I do admit to a tube amp bias but the bottom line is I don't want to spread bad info. I will definitely replace the drawn clipping examples with actual DSO captures.

Again, thanks for the feedback.

Rob Robinette

[robrob]

Mark, the VHT tone stack LTSpice file was the first thing I attempted in LTSpice and I've been meaning to either fix it or remove it because you are right it is less than worthless in its present state.

[jazbo8]

I HATE to write this, but there is so much wrong (mixed with a little right) that it can't be "corrected" but fully rewritten which is not the point.

Now that you mentioned it, I'm not sure Rob's site was meant to be read by engineers/designers, Ron can correct me if I am wrong, but it's more for musicians that are curious about how their tube amps work. If that's true, then perhaps he can just keep the basic information on the transfer characteristic, and how clipping fits into the picture, and leave off all the fancy bits.

[Mike Sulzer]

It may be more correct to explain the difference as being due to the different device characteristics lending themselves to different implementations.

Yes; for example, an amplifier with a lot of global negative feedback can have a very abrupt transition from "sine" to "flat top" with enough higher harmonics to stand out from the signal in a distressing, ear piercing way.

And on the other side, FETS in a preamp can be made to have almost any sort of characteristic that you like.

I would certainly hesitate to write anything with definite general conclusions: someone has a counter example. I would prefer to stick with examples of what tube and solid state circuits do when used in their most "natural" environments, and some different things that you can do with a bit more effort.

[Malcolm Irving]

.... When the grid goes positive doesn't that mean there is a scarcity of electrons on the grid that would pull electrons from the tube's space charge ...

Yes, that's roughly right. In metallic conduction (as in the grid wires) there isn't really a scarcity of electrons. It's more that the (plentiful) free electrons have a force imposed on them by the local electric field and that causes them to drift along. Electrons arriving at the surface of the wire, from the space charge, replace those that want to drift along.