preamp distortion vs power output distortion

[Jazzarian]

I hear many an old wive's tale regarding the subject. The old thinking was power amp distortion sounded better, "less compressed" etc. Great example being the late Duane Allman and his 50 watt Marshall Plexi. To get that distortion, he literally cranked the volume all the way to 11. No Hotplates and PowerBrakes back then.

Along came master volume amps, Mesa Boogie being the most notable of the early 70s. Now distortion could be dialed in at the preamp, and done so at relatively low listening volume. Larry Carlton and Santana used the venerable Boogie Mark I.

Preamp distortion is touted by the "now" crowd. 12AX7 tubes in the Class A configuration of a normal preamp is though of being the best place for such overdrive to occur, especially if the power section is Class A/B.

[randyc]

OK, I'll jump in since this doesn't appear to be a popular topic. Sort of esoteric but interesting to some, dealing with non-linear characteristics of semiconductors, vacuum tubes, ferrite devices (transformers, loudspeaker coils) and so forth.

Music Man, for one, took the position that the "desirable" distortion came from the power amplifier, so they made solid-state preamplifiers followed by vacuum tube power amplifiers. I liked the sound of those amplifiers.

The newer approach of sticking a 12AX7 in front of your favorite solid-state amplifier is, IMHO, driven largely by economics and marketing. Those little dual triodes can be had for just a couple of dollars, distributors of course charge a LOT more for them. These can sound pretty good, too.

(I have to state, though, that I don't like distortion, despite the fact that I have four tube amplifiers. I do like that certain sound frequently described as "warm" that comes from these amplifiers.)

The problem with the commonly employed master volume approach is the increase in signal/noise ratio (and hum) due to excess gain followed by excess loss, noise contributions from frequency mixing due to non-linearities and several other phenomena that bore just about everyone except communications engineers.

This approach is mostly used in amplifiers intended for lots of distortion, however, so a little excess noise is .... well, just more noise. No offense intended for those who like lots of distortion. Given this, there's little disadvantage in putting a vacuum tube "up front", it's cheap and the description sounds so cool in the advertising.

cheers

[randyc]

Just occurred to me that I have an illustrated example of the tube preamp setup. In the photo below, if you look carefully, you'll see that the second "input" of the little Fender Champ (on the left side of the photo) has a cable connecting it to the input of the little Peavey on the far right.

The Champ actually has only one input, the second jack was disconnected from the preamp input and re-wired as a line output (from the second preamp stage of the Champ). This gives the nice, "warm" tube sound that drives the solid state Peavey. The Peavey was free and all but useless until I worked out this setup ...

My Champ also has a 10 inch speaker in it, instead of the tiny little one that came with it. The only other change was to replace the output transformer from the standard 3.2 ohms to an 8 ohm unit, to drive the new speaker.

What's the point of all this ? Warm tube sound and 40 watts instead of 6 watts. Neither of the amplifiers weighs more than ten or fifteen pounds, so it's easy to carry both. The Champ is used as a monitor and the Peavey is the main transducer.

cheers

[bluemood]

I think that I like both ...The way that I'm setup at the moment is my modified Fender HRD volume at breakup around 5 or 11 o'clock(clean channel and then my Fulltone fulldrive II and Fulltone Fat boost...Not high gain ...Strat with high output single coils.

[riovine]

Larry Carlton and Santana used the venerable Boogie Mark I.

I always assumed that all that great Larry Carlton stuff that he did with Steely Dan was with a Mark I. But from his website, he says he played Kid Charlamagne, Don't Take Me Alive etc with a small Fender Tweed Delux, with his back pickup & the tone down to 3. Came as a total surprise to me, what a great sound.

[Jazzarian]

The problem with the commonly employed master volume approach is the increase in signal/noise ratio (and hum) due to excess gain followed by excess loss, noise contributions from frequency mixing due to non-linearities and several other phenomena that bore just about everyone except communications engineers.
cheers

I'm an engineer, try me.

I've never had a problem with high gain amps being noisy. At least not with humbucking pickups. Single coils require noise reduction. Stratocasters are not the ideal high gain guitar, obviously.

Power distortion is impractical. Without some kind of "shunt", they're far too loud. Power Brake ("shunts") tend to rob amps of their characteristic tone, which makes them rather moot.

[Jazzarian]

I always assumed that all that great Larry Carlton stuff that he did with Steely Dan was with a Mark I. But from his website, he says he played Kid Charlamagne, Don't Take Me Alive etc with a small Fender Tweed Delux, with his back pickup & the tone down to 3. Came as a total surprise to me, what a great sound.

I thought is was a Fender Princeton Reverb. Oh well.

On his very first record, he used his Boogie Mark I, in 1978. 10 years later he switched to Dumble, with occasional Boogie forays now and then.

There's a Live in Japan CD from 1982 where his live sound with the Mark I is just unreal.

[randyc]

I'm an engineer, try me.

Sure, no problem.

I've never had a problem with high gain amps being noisy. At least not with humbucking pickups. Single coils require noise reduction. Stratocasters are not the ideal high gain guitar, obviously.

It's good that you've never encountered a problem and my statement certainly wasn't intended to suggest that this amplifier architecture has problems that prevent or limit it's use. Obviously, with the many thousands of this type of circuit in use, others are as content as you seem to be.

You didn't pose a question in your original post, you simply suggested a topic, as far as I can see. I saw that no one was responding, so I jumped in to keep the ball rolling, OK ?

Overall noise has to be increased by the addition of excess gain, which is required in order to drive a subsequent stage into distortion. The excess noise will be manifested when there is no input drive level, i.e. one stops playing. The pocket calculator suggests that noise degradation will be in the order of 4 dB and hum degradation will be about 5 dB. Whether this is a problem depends on overall amplifier gain, personal preference and the musical environment.

Power distortion is impractical. Without some kind of "shunt", they're far too loud. Power Brake ("shunts") tend to rob amps of their characteristic tone, which makes them rather moot.

If Music Man were still in business, I'm sure that they would be eager to entertain your suggestions I just said that I liked the sound of those amplifiers. I also said that I liked the other (preamp distortion) configuration and even showed a practical use for that technique, with photo.

My impression that this was to be a discussion, not an argument. I didn't join this forum to argue with anyone, just to contribute my experience as a musician and as a design engineer and enjoy reading about the experiences of others.

cheers

[Jazzarian]

Sure, no problem.



It's good that you've never encountered a problem and my statement certainly wasn't intended to suggest that this amplifier architecture has problems that prevent or limit it's use. Obviously, with the many thousands of this type of circuit in use, others are as content as you seem to be.

You didn't pose a question in your original post, you simply suggested a topic, as far as I can see. I saw that no one was responding, so I jumped in to keep the ball rolling, OK ?

Overall noise has to be increased by the addition of excess gain, which is required in order to drive a subsequent stage into distortion. The excess noise will be manifested when there is no input drive level, i.e. one stops playing. The pocket calculator suggests that noise degradation will be in the order of 4 dB and hum degradation will be about 5 dB. Whether this is a problem depends on overall amplifier gain, personal preference and the musical environment.



If Music Man were still in business, I'm sure that they would be eager to entertain your suggestions I just said that I liked the sound of those amplifiers. I also said that I liked the other (preamp distortion) configuration and even showed a practical use for that technique, with photo.

My impression that this was to be a discussion, not an argument. I didn't join this forum to argue with anyone, just to contribute my experience as a musician and as a design engineer and enjoy reading about the experiences of others.

cheers

Actually my Boogie preamp is about as quiet as a tube amp gets. Near virtual silence with no input.

Gain stages add noise because they amplify everything, desired signal and noise. Noise can be mitigated by simply twisting the wires to the input jack, thus improving its Common Mode Rejection Ratio. Most manufacturers do this, along with encasing the V1 tube in a metal jacket.

I've found I can turn the gain knob all the way up the noise is still very tolerable.

[randyc]

Actually my Boogie preamp is about as quiet as a tube amp gets. Near virtual silence with no input.

Gain stages add noise because they amplify everything, desired signal and noise. Noise can be mitigated by simply twisting the wires to the input jack, thus improving its Common Mode Rejection Ratio. Most manufacturers do this, along with encasing the V1 tube in a metal jacket.

I've found I can turn the gain knob all the way up the noise is still very tolerable.

Well, "tolerable" is a relative term, doncha' think ?

Yes, gain stages do amplify everything BUT large signal gain, i.e. distortion conditions, is significantly less than small signal gain while noise remains the same so signal/noise is degraded.

And under quiescent conditions, the distorted stage reverts to it's higher small-signal gain so output noise increases by the difference between small-signal and large-signal gain.

Twisting input wires mitigates nothing except maybe a little hum (by reducing the characteristic impedance - and therefore coupling - of the wires). Unhappily, it also increases capacitance and degrades high frequency response.

Noise can't be "mitigated", it's a natural phenomenom, N=KtBR. Given the constraints of bandwidth and input impedance, you can do nothing to reduce it except reduce the temperature (pack the preamp in liquid nitrogen or the like). Since this is not practical, noise can only INCREASE from 4.14(10-21)V^2/Hz/ohm/degree K.

And CMMR only applies to a BALANCED transmission mode, like an operational amplifier or a balanced transmission line - your twisted pair is not balanced since one wire is at ground potential instead of being 180 degrees out of phase.

The only balanced configuration in a typical guitar amplifier is the output stage in a vacuum tube amplifier. Noise contribution in the power amplifier is negligible.

Why are you arguing with me?

[TieDyedDevil]

A couple points to add:

High gain amps tend to reduce "perceived" noise by rolling off high frequencies. Even though an amp designer can't escape thermal noise, he can limit the apparent hiss though EQ changes.

"Power tube" distortion is not clearly understood by most guitarists. They tend to think that any cranked amp must be exhibiting "power tube" distortion. While the power tubes do contribute, there are many other factors that come into play at the same time: preamp, phase inverter, operating point shifts (multiple components), power transformer saturation, speaker distortion (which itself has multiple components) and even psychoacoustic effects.

[Jazzarian]

Well, "tolerable" is a relative term, doncha' think ?

I take it you don't play with distortion often?

Yes, gain stages do amplify everything BUT large signal gain, i.e. distortion conditions, is significantly less than small signal gain while noise remains the same so signal/noise is degraded.

Gain is gain, unless you're saying there's some kind of rolloff.

And under quiescent conditions, the distorted stage reverts to it's higher small-signal gain so output noise increases by the difference between small-signal and large-signal gain.

Operational amplifiers always amplify the difference between the inputs. They also tend to employ negative feedback.

Twisting input wires mitigates nothing except maybe a little hum (by reducing the characteristic impedance - and therefore coupling - of the wires). Unhappily, it also increases capacitance and degrades high frequency response.

And CMMR only applies to a BALANCED transmission mode, like an operational amplifier or a balanced transmission line - your twisted pair is not balanced since one wire is at ground potential instead of being 180 degrees out of phase.

I disagree. The input wires are isolated from ground. The return of the input jack is not referenced to chassis ground. Thus twisting a pair of indentical wires is a transmission line and any "L' or "C" consequences of twisting the wires are moot at audio frequencies.

[Jazzarian]

A couple points to add:

High gain amps tend to reduce "perceived" noise by rolling off high frequencies. Even though an amp designer can't escape thermal noise, he can limit the apparent hiss though EQ changes.

You do realize tubes were and still are used in Hi-Fi applications? That last thing you'd want to do is filter their output. A tube power amp is supposed to have a very flat freq response curve.

You'd be surprised how quiet tubes can be. Listening to my Boogie preamp (5 12AX7s) via phones sure bears that out. Sometimes I find it hard to play a combo amp afterwards, their hiss drive me nuts.

"Power tube" distortion is not clearly understood by most guitarists. They tend to think that any cranked amp must be exhibiting "power tube" distortion. While the power tubes do contribute, there are many other factors that come into play at the same time: preamp, phase inverter, operating point shifts (multiple components), power transformer saturation, speaker distortion (which itself has multiple components) and even psychoacoustic effects.

Indeed, that's why I prefer preamp distortion.

[lpdeluxe]

Music Man, for one, took the position that the "desirable" distortion came from the power amplifier, so they made solid-state preamplifiers followed by vacuum tube power amplifiers. I liked the sound of those amplifiers.

I had one of the early MM HD-130 Reverb (with a geenormous 2-12 sealed cab) which has a polarity inverter tube in the pre (I know, everyone calls them "Phase inverters") and it had the ultimate clean LOUD sound but would also break up nicely by dialing up the channel gain knob. I've always thought their approach was a little wrong-headed, which is possibly borne out by the fact that MM amps are known as the kings of clean. I played my pedal steel through mine, and it was incredible sounding. Without the tweak on the preamp gain, though, I could never get it to distort (on the positive side, I still have much of my hearling).

It seems like all my other amps have featured pre distortion, which I don't find unattractive. I'm thinking of my Carvin oak amp, from the early '80s, which had what amounted to a fuzz-pedal circuit (I added a Torres Tone Kit, which included a tube socket and pot for an additional gain stage, and that made that amp into a killer).

I don't think there's anything wrong with pre distortion...one of my current faves is my Blues Jr NOS, with 15 watts and one channel, and that may be the only one that relies on power tube distortion. Since I play mostly clean, I haven't really explored it.

[randyc]

I take it you don't play with distortion often?

Never, as I stated in my original post. I also don't play with power tools but it doesn't mean that I don't understand how to use them and how they work.

Gain is gain, unless you're saying there's some kind of rolloff.

You said that you were an engineer but you're not an EE, right ? I don't mean to be offensive, but that statement is just downright silly. Have you never seen/heard the terms "small-signal gain" and "large-signal gain" defined ?

Instead of arguing with me about this stuff, look it up. There are tons of sites on the internet that would be helpful in providing you with a clear understanding of the topic. If you can't find one, I'd be happy to assist you - I'm retired and I have lots of time

The result might be that, instead of getting irritated with me for correcting you, your knowledge would be enhanced and we'd both be on the same page. Imagine how much more pleasant these conversations could be.

Operational amplifiers always amplify the difference between the inputs. They also tend to employ negative feedback.

What's your point ?

I disagree. The input wires are isolated from ground. The return of the input jack is not referenced to chassis ground. Thus twisting a pair of indentical wires is a transmission line and any "L' or "C" consequences of twisting the wires are moot at audio frequencies.

Only on your planet.

Look, this conversation is going nowhere. You can't be the big dog in every situation - there will always be those with either more education, more knowledge or more experience (perhaps all three).

I don't think that there's a courteous way of saying this ... If you can't keep up, maybe you should keep silent and cut your losses

cheers, I'd still be happy to help with those website referrals

[Jazzarian]

Never, as I stated in my original post. I also don't play with power tools but it doesn't mean that I don't understand how to use them and how they work.



You said that you were an engineer but you're not an EE, right ? I don't mean to be offensive, but that statement is just downright silly. Have you never seen/heard the terms "small-signal gain" and "large-signal gain" defined ?

Instead of arguing with me about this stuff, look it up. There are tons of sites on the internet that would be helpful in providing you with a clear understanding of the topic. If you can't find one, I'd be happy to assist you - I'm retired and I have lots of time

The result might be that, instead of getting irritated with me for correcting you, your knowledge would be enhanced and we'd both be on the same page. Imagine how much more pleasant these conversations could be.



What's your point ?



Only on your planet.

Look, this conversation is going nowhere. You can't be the big dog in every situation - there will always be those with either more education, more knowledge or more experience (perhaps all three).

I don't think that there's a courteous way of saying this ... If you can't keep up, maybe you should keep silent and cut your losses

cheers, I'd still be happy to help with those website referrals

Sorry jack, I'm an EE, as ELECTRONICS ENGINEER. You're mumbo-jumbo isn't fooling anyone.

Of course, when I was in school, nobody was teaching tubes anymore.

Perhaps you'd like to learn a few things about transistors? Thevinin? Norton? Filter theory? Smith Charts? Logic circuits? Bode plots? Load lines?

I know, how about operational amplifiers?


As I recall, the subject was preamp tube vs output tube distortion. You haven't added any value in that regard.

[randyc]

Sorry jack, I'm an EE, as ELECTRONICS ENGINEER. You're mumbo-jumbo isn't fooling anyone.

Of course, when I was in school, nobody was teaching tubes anymore.

Perhaps you'd like to learn a few things about transistors? Thevinin? Norton? Filter theory? Smith Charts? Logic circuits? Bode plots? Load lines?

I know, how about operational amplifiers?


As I recall, the subject was preamp tube vs output tube distortion. You haven't added any value in that regard.

Take it easy, there's no need to get defensive and macho, wasn't your intent to generate discussion? Adding value is perhaps dependant upon one's perception. One man's garbage is another's treasure and vice-versa. (Did you really read what I posted ?)

If you're more comfortable discussing transistors than tubes, sure, let's talk about them .... or other engineering topics, if you like. I'm agreeable, talkative, have the time and comfortable with most analog engineering topics, as long as you keep your temper and your boss is OK with you spending your time on the internet.

Want to talk about large-signal simulation and how one can predict the change in gain from the small signal value ?

cheers

[Jazzarian]

Take it easy, there's no need to get defensive and macho, wasn't your intent to generate discussion? Adding value is perhaps dependant upon one's perception. One man's garbage is another's treasure and vice-versa. (Did you really read what I posted ?)

If you're more comfortable discussing transistors than tubes, sure, let's talk about them .... or other engineering topics, if you like. I'm agreeable, talkative, have the time and comfortable with most analog engineering topics, as long as you keep your temper and your boss is OK with you spending your time on the internet.

Want to talk about large-signal simulation and how one can predict the change in gain from the small signal value ?

cheers

I program and test these days.

[BigDaddyLoveHandles]

This thread has blinded me with Science. Thanks, guys!

[lpdeluxe]

So, did we prove which is better yet?









Didn't think so. Pixie dust is SO hard to count, let alone, weigh.

[randyc]

So, did we prove which is better yet?

I like 'em both but if someone was to offer me a choice between a Boogie and a Music Man 210, I'd go for the MM. To me they sound more like the classic clean Fender sound ... and that's what I like for jazz.

cheers









Didn't think so. Pixie dust is SO hard to count, let alone, weigh.[/quote]

[lpdeluxe]

There ya go. I apologize if I came across as snarky, but in my humble non-engineer opinion, I can only tell what I like, and, at that, only when I hear it, and, often enough, only when I've played through an amp in the context of whatever ensemble I'm playing with.

I'm not even sure, always, whether I'm hearing one sort of distortion or the other -- but then, I go a clean tone. Occasionally the material I play requires a little grit, and, if so, I take along my Floor Pod Plus so I can stomp on one of the pedals and get it without boring the audience with knob turning.

I mentioned the Music Man I had: I ended up setting it up in our practice space and plugging it into a Hartke 410XL cabinet for my bass, and I don't remember how long it had been since I touched the channel gain. As an aside, it was set up there because the head and cab totaled 144 lb...I gigged with a 65 lb Ampeg.

It's the music that matters.

[Jazzarian]

[quote=randyc;49050]

I disagree. The input wires are isolated from ground. The return of the input jack is not referenced to chassis ground. Thus twisting a pair of indentical wires is a transmission line and any "L' or "C" consequences of twisting the wires are moot at audio frequencies.

Only on your planet.

That is interesting, as somehow DSL gets 10 MB per second over a twisted pair. Yet a foot or so of twisted pair wiring is somehow "less hi-fi" in terms of "L" or "C" consequences of twisting bell wire? I mean at frequencies humans can hear, 20 to 20 Khz.

With all your fancy test equipment, you could provide actual numbers, if you were so inclined.

[randyc]

That is interesting, as somehow DSL gets 10 MB per second over a twisted pair. Yet a foot or so of twisted pair wiring is somehow "less hi-fi" in terms of "L" or "C" consequences of twisting bell wire? I mean at frequencies humans can hear, 20 to 20 Khz.

With all your fancy test equipment, you could provide actual numbers, if you were so inclined.

Why not spend your time doing something that is productive instead of arguing with me.

cheers

[Jazzarian]

Why not spend your time doing something that is productive instead of arguing with me.

cheers

You've got the test equipment. Prove me wrong.


Imagine, 10 megabits per second over twisted copper bell wire.