Tube Amp with the "Good Kind of Distortion"

[Ken Tajalli]

Reading the whole article it's apparent that Pass fairly explicitly advocates that a desirable distortion profile minimizes higher order harmonic distortion while tolerating substantial 2nd and/or 3rd order.

Thanx for the link. I read it and hopefully understood it.
At no point did I get the feeling that he was advocating for "good distortion". At worst he was saying that if we accept some distortion is unavoidable, he rather have simple low-order ones as opposed to complex ones (as a last resort).
The bulk of the article was about irresponsible use of feedback.
What I understood was that, making a multi-stage high-distorting gain stages and then hoping to use vast amounts of global feedback to bring distortion down is wrong, as it changes the nature of distortion pattern.
He rather advocates, designing an amp with inherently low distorting stages to begin with, would be the ticket.
Then one could use just enough global feedback to iron out the rest.
I find nothing wrong with that argument, but then again I am not as knowledgeable as some on ASR.
Did I miss something?

 

[Jim Matthews]

It's important to note that the design discussed in the advertising video (and it is advertising) is dependent on readily available power pentodes (6550 to Ktxxx) that are run "single ended" so that there's no crossover distortion.

These are chosen for cost, stability and power handling. It is important to note that this design emulates direct heated triodes which are much more fragile, expensive and desirable.

In the 1980's these were considered starter circuits for newbies.

Tubes

 

[MattHooper]

Distortion is distortion, how can it be good?

I see you haven't spent any time on guitar forums or playing electric guitar

In other words, at least speaking to the principle of "how can distortion be good?": different types of distortion, including tube distortion, can be found pleasant and preferred by some. It suggests, in principle at least, an amp could add distortion that some find pleasant. (It does seem to be the case for me with my tube amps, though I haven't objective data to validate this). And after all, anything that deviates from the original signal is a "distortion." If a tube amp impedance interacting with the right speaker was causing a bit of base bloat or rolled off highs or both, some may prefer that "warmer" sound.

Also: I am among those who get some enjoyment from some of the distortion in vinyl playback.

Whether you find it "good" is a subjective call.

 

[Gorgonzola]

Thanx for the link. I read it and hopefully understood it.
At no point did I get the feeling that he was advocating for "good distortion". At worst he was saying that if we accept some distortion is unavoidable, he rather have simple low-order ones as opposed to complex ones (as a last resort).
The bulk of the article was about irresponsible use of feedback.
What I understood was that, making a multi-stage high-distorting gain stages and then hoping to use vast amounts of global feedback to bring distortion down is wrong, as it changes the nature of distortion pattern.
He rather advocates, designing an amp with inherently low distorting stages to begin with, would be the ticket.
Then one could use just enough global feedback to iron out the rest.
I find nothing wrong with that argument, but then again I am not as knowledgeable as some on ASR.
Did I miss something?

I think you understood Pass pretty well. However avoidance of feedback is not something all designers agree with.

It is probably true that higher order harmonic distortion is inherently disagreeable and to be avoided, however negative feedback reduces all orders of distortion, including higher order as well as (relatively tolerable) 2nd and 3rd order distortion. Many current designs reduce all forms of distortion to miniscule levels using feedback. Class D designs, e.g. Hypex, Purifi, are use lots of feed; as I recall Purifi (e.g.) used over 70 dB of feedback.

 

[Ken Tajalli]

I see you haven't spent any time on guitar forums or playing electric guitar

In other words, at least speaking to the principle of "how can distortion be good?": different types of distortion, including tube distortion, can be found pleasant and preferred by some. It suggests, in principle at least, an amp could add distortion that some find pleasant. (It does seem to be the case for me with my tube amps, though I haven't objective data to validate this). And after all, anything that deviates from the original signal is a "distortion." If a tube amp impedance interacting with the right speaker was causing a bit of base bloat or rolled off highs or both, some may prefer that "warmer" sound.

Also: I am among those who get some enjoyment from some of the distortion in vinyl playback.

Whether you find it "good" is a subjective call.

Really!

 

[MakeMineVinyl]

I listened to The Occasional Podcast yesterday while mountain biking and heard something which almost made me crash. The designer essentially said that negative feedback makes the signal 'go round and round again' over and over and thus was bad. That's the definition of an oscillator! Perhaps he was just not great at explaining it, but it was startling to hear that from someone who was supposedly a designer. On the other hand, in every episode of that podcast they throw out information which is clearly wrong. Awhile back they had an episode about reel to reel tape machines and the bullshit got almost too much to bear since I specialize in analog reel to reels.

 

[Ken Tajalli]

I think you understood Pass pretty well. However avoidance of feedback is not something all designers agree with.

It is probably true that higher order harmonic distortion is inherently disagreeable and to be avoided, however negative feedback reduces all orders of distortion, including higher order as well as (relatively tolerable) 2nd and 3rd order distortion. Many current designs reduce all forms of distortion to miniscule levels using feedback. Class D designs, e.g. Hypex, Purifi, are use lots of feed; as I recall Purifi (e.g.) used over 70 dB of feedback.

Avoidance of feedback is one thing, I did not get the impression he was preaching for that (on the article).
On more than one occasion he was advocating for low distortion.
Are his amps any good?

 

[DVDdoug]

However avoidance of feedback is not something all designers agree with.

The designer essentially said that negative feedback makes the signal 'go round and round again' over and over and thus was bad. That's the definition of an oscillator!

Right. It's positive feedback makes an oscillator and it's what happens when you stick a microphone in front of a speaker and you get that feedback squeal. (I once built a preamp that turned-out to be an RF oscillator and I fried one channel in my power amp. There was an "unexpected" phase shift that turned negative feedback into positive feedback at very-high frequencies.)

Negative feedback (CORRECTIVE feedback) is an awesome thing!!! No actual engineer would disagree. It's what allows you to drive straight down the road as you make corrections with the steering wheel. It how airplanes find the airport and how ships find the port. Positive feedback would be turning to the right when you start drifting to the right... That would be a bad thing!

I see you haven't spent any time on guitar forums or playing electric guitar

The amp, cabinet, and speaker are "part of the instrument". They are not supposed to be "high fidelity", they are supposed to alter the sound. Usually when a guitar is recorded direct (without an amp, speaker, or cabinet} a "sim" (a simulator effect) is used to simulate the effect of using an amp because the signal directly out of the guitar doesn't "sound right". Every guitar player has their favorite amp and, they often use all kinds of effect pedals, etc.

A high-fidelity amp isn't supposed to alter the sound. It's supposed to accurately reproduce the recording. You wouldn't want to use a guitar amp (or effect pedal) as your hi-fi/stereo amp.

If you like extra distortion from vinyl, or it you like the "warm crackle" of vinyl, or if you like to boost the bass, etc., that's OK but it's not "high fidelity" reproduction.

 

[Grumpish]

Avoidance of feedback is one thing, I did not get the impression he was preaching for that (on the article).
On more than one occasion he was advocating for low distortion.
Are his amps any good?

I have one of the DIY designs that he makes available through the diyaudio forum - a Pass F6. It is an elegantly simple class A design, makes around 20 watts per channel, and probably has no more than twenty discrete components per channel. And yes, it does sound very good. My take is that, at least when it comes to reproducing recorded music, there is no such thing as good distortion, there are, however, some forms of distortion that are less objectional than others. Something I have noticed on this forum is that there seems to be a hard line "ultra low THD is king" element that is resistant to looking at different ways of measuring amplifiers and seeing what, if anything, correlates to what actually sounds best to people.

 

[Gorgonzola]

Avoidance of feedback is one thing, I did not get the impression he was preaching for that (on the article).
On more than one occasion he was advocating for low distortion.
Are his amps any good?

Yes I'd have to say. Beautiful construction and very heavy. I owned a Pass X150.5 for several years.; to me it was a nice sounding amp, a trifle warm, but not as resolved or dynamic as my current VTV Purifi amp.

See Stereophile's measurements for the Pass INT-150 whose power section is identical to the X150.5. Notice that distortion is not especially low and also the distortion rises at higher frequencies. This is much worse THD that the Purifi amp, e.g. in case of the EVAL-1 version reviewed by Amir.

No good distortion? I'm not there is. Take the Purifi which near SOTA for low distortion: I like the Purifi, but the fact is that audiophiles such as those at the old-time Audio Asylum forum find such low distortion amps "lean" and "sterile".

 

[MattHooper]

Really!

Yup.

 

[Ken Tajalli]

Yup.

As @DVDdoug explained, guitar amps are musical instruments and are therefore outside the scope of our discussion .
I thought you were having me on.

 

[MattHooper]

As @DVDdoug explained, guitar amps are musical instruments and are therefore outside the scope of our discussion .
I thought you were having me on.

No, I explained why it was relevant.

I presumed you were asking a real question, not a rhetorical one. That is, if you have defined distortion as "bad" because your goal is pure accuracy, then asking how it could be "good" is merely rhetorical. But if you were REALLY asking in what sense could distortion be "good?" - as in how can it make sense ANYONE would find distortion to be "good," I gave an answer.

Distortion of various types can be deemed pleasant and desirable sometimes by some people. Right? I mean, even in music production, especially pop/rock/electronic, distortion is added to all sorts of instruments and vocals, for a desired pleasing effect. Therefore distortion is not always "bad." And some forms of distortion are apparently heard by the ear as more pleasant than others, and even pleasant when added to a signal (e.g. 2nd order especially). Plus there's various forms of distortion.

It's not some insurmountable leap from this to the idea that some may like a general subtle distortion in their music playback system.

Remember, even your various room reflections are doing this to your sound, and you may prefer some level of "room distortion effects" (reflections) applied over the entire playback signal as providing a more airy alive sound or whatever. (I play with acoustics like this for different desirable effects all the time in my system).

And I gave the example that plenty of people, including me, can find the distortion introduced in vinyl playback to be pleasing.

So...it was a serious point. Up to you if you want to seriously engage the point.

 

[0bs3rv3r]

In the 1980's these were considered starter circuits for newbies.

That thought struck me too, when I first started reading about the resurgence of tubes for high end audiophile use. It's true that most cheaper audio equipment, back before transistors, used these same simple circuits. I remember for example, one tube record players.

Nevertheless, I would not allow preconceptions to stop me trying it, and I was suitably impressed that the same simple circuit topology, with some basic values adjusted for wider bandwidth, constructed with modern components, better transformers etc, and played with modern hifi sources and into modern hifi speakers, could surprise me with how nice it can sound.

 

[Pinox67]

I'm conducting a study on how nonlinear distortion affects listening pleasure in this thread. It is shown another point of view about distortion effects, according also to other scientific studies, and the discrete models (in the subsequent posts) able to simulate them in a playback chain. A bit technical, but I think understandable in general aspects even by less experienced people.

 

[q3cpma]

No, I explained why it was relevant.

I presumed you were asking a real question, not a rhetorical one. That is, if you have defined distortion as "bad" because your goal is pure accuracy, then asking how it could be "good" is merely rhetorical. But if you were REALLY asking in what sense could distortion be "good?" - as in how can it make sense ANYONE would find distortion to be "good," I gave an answer.

Distortion of various types can be deemed pleasant and desirable sometimes by some people. Right? I mean, even in music production, especially pop/rock/electronic, distortion is added to all sorts of instruments and vocals, for a desired pleasing effect. Therefore distortion is not always "bad." And some forms of distortion are apparently heard by the ear as more pleasant than others, and even pleasant when added to a signal (e.g. 2nd order especially). Plus there's various forms of distortion.

It's not some insurmountable leap from this to the idea that some may like a general subtle distortion in their music playback system.

Remember, even your various room reflections are doing this to your sound, and you may prefer some level of "room distortion effects" (reflections) applied over the entire playback signal as providing a more airy alive sound or whatever. (I play with acoustics like this for different desirable effects all the time in my system).

And I gave the example that plenty of people, including me, can find the distortion introduced in vinyl playback to be pleasing.

So...it was a serious point. Up to you if you want to seriously engage the point.

It's not unbelievable, indeed, as people do like distortion on certain material. But I have some doubts concerning the sincerity of that argument when that distortion is applied to all material and all instruments indiscriminately, which is different from what artists do.

 

[MattHooper]

It's not unbelievable, indeed, as people do like distortion on certain material. But I have some doubts concerning the sincerity of that argument when that distortion is applied to all material and all instruments indiscriminately, which is different from what artists do.

Why?

It's like preferring one sound system over another.

Few sound systems, especially from a speaker/room stand point, are perfectly neutral. One person may like "speaker/room combo A another combo B. To the extent one sounds different than another, one is picking some level of "overall flavor" to the sound and people do this all the time.

Once again, even the acoustics in most rooms are adding a coloration over everything, and it's entirely reasonable to prefer a certain "overall coloration" if one wants, e.g. a slightly livelier room than a duller one. (Or speaker positioning that may emphasize or de-emphasize certain frequencies, etc). That's a mass "indiscriminate" type of coloration but perfectly reasonable.

Similarly, I've mentioned many times before that I perceive my tube amps to add a subtle but welcome coloration that I like on everything.
It's enough to enhance my overall enjoyment of the sound, but not remotely enough to make "everything sound the same." Very much like choosing how much room reflections I want involved with the system.

 

[q3cpma]

Why?

It's like preferring one sound system over another.

Few sound systems, especially from a speaker/room stand point, are perfectly neutral. One person may like "speaker/room combo A another combo B. To the extent one sounds different than another, one is picking some level of "overall flavor" to the sound and people do this all the time.

Once again, even the acoustics in most rooms are adding a coloration over everything, and it's entirely reasonable to prefer a certain "overall coloration" if one wants, e.g. a slightly livelier room than a duller one. (Or speaker positioning that may emphasize or de-emphasize certain frequencies, etc). That's a mass "indiscriminate" type of coloration but perfectly reasonable.

Similarly, I've mentioned many times before that I perceive my tube amps to add a subtle but welcome coloration that I like on everything.
It's enough to enhance my overall enjoyment of the sound, but not remotely enough to make "everything sound the same." Very much like choosing how much room reflections I want involved with the system.

Possible, but I'm not even sure "tube distortion" is actually audible, as it has good chances of being masked; do you have measurements that correspond to your type of equipement?
It may work if you don't listen to very varied genres, but from a lot of people, it justs sounds like an excuse to spend money on a cool toy. Maybe I'm cynical, but I don't really believe the aesthetic pretentions.

Objectively, I can't see additional distortion always enhancing the sound, whatever the genre (solo violin, for example?) but I can get wanting a (possibly nostalgia fueled) more 'fuzzy and relaxed' sound.

 

[MakeMineVinyl]

Possible, but I'm not even sure "tube distortion" is actually audible, as it has good chances of being masked; do you have measurements that correspond to your type of equipement?
It may work if you don't listen to very varied genres, but from a lot of people, it justs sounds like an excuse to spend money on a cool toy. Maybe I'm cynical, but I don't really believe the aesthetic pretentions.

Objectively, I can't see additional distortion always enhancing the sound, whatever the genre (solo violin, for example?) but I can get wanting a (possibly nostalgia fueled) more 'fuzzy and relaxed' sound.

High amounts of 2nd order distortion in higher frequencies can act to increase 'detail' due to the doubling of every frequency being reproduced and that being added back to the original signal. So essentially frequency extension is 'doubled'. Where if the original program material didn't have any information past, say, 10kHz, with added 2nd harmonic this extends to 20kHz (but at a much lower level obviously).

There is no other electronic component in a typical system which can do that - a treble boost won't do it for instance. Higher orders of distortion >= 3rd also can 'brighten' up the high end but they tend to be more dissonant. Some people may like that effect.

So its perfectly logical that this effect could be behind the 'better soundstage' reported by many tube amp owners due to increased detail. This detail may be 'false', but it is correlated with the original signal.

 

[q3cpma]

High amounts of 2nd order distortion in higher frequencies can act to increase 'detail' due to the doubling of every frequency being reproduced and that being added back to the original signal. So essentially frequency extension is 'doubled'. Where if the original program material didn't have any information past, say, 10kHz, with added 2nd harmonic this extends to 20kHz (but at a much lower level obviously).

There is no other electronic component in a typical system which can do that - a treble boost won't do it for instance. Higher orders of distortion >= 3rd also can 'brighten' up the high end but they tend to be more dissonant. Some people may like that effect.

So its perfectly logical that this effect could be behind the 'better soundstage' reported by many tube amp owners due to increased detail. This detail may be 'false', but it is correlated with the original signal.

It'd need a massive amount of distortion for such low order harmonics to break masking limits, though. I'm dubious about perceptual detail claims, as high distortion effectively drowns low-level detail within multitone material (i.e. almost all material). I really don't see what soundstage has to do with distortion.
That "emulated restoration of lowpassed material" thing is interesting though even if I think that frequencies above 16~18 kHz (or so, you get it) really aren't that important.

On that subject, is there any blind data of quality on the whole subject?