Negative feedback bad for audio?

[maty]

Is Too Much Amplifier Negative Feedback Bad? by Bruno Putzeys, May 05, 2020
https://www.audioholics.com/audio-amplifier/amplifier-negative-feedback

[ This was helped by the fact that getting the equivalent amount of feedback with a global loop alone is not at all obvious. Reliable methods are math heavy and were only developed in the 90’s by designers of sigma-delta ADC and DAC chips. Naturally unburdened by feedback phobia, some of them turned their attention to class D amplifiers. Even then it took a surprising amount of work to mature the technology. The current high degree of acceptance of class D amplifiers in hi-fi is almost entirely due to work on feedback loops, so feedback avoidance has not much traction in class D circles.

Of the many myths against feedback, the idea that feedback causes TIM is probably most noteworthy for being not just wrong, but also the exact opposite of the truth. TIM happens in the input stage. An increase in global feedback makes the input stage work less hard. That causes a disproportional reduction in TIM... ]

 

[David Silva]

Strangely enough, all the NAD DAC models reviewed by Amirm are not recommended by him.

 

[egellings]

The sun revolves around the earth, dag nab it!

 

[maty]

Audiophile Myths About Amplifier Negative Feedback Debunked, by Bruno Putzeys, July 18 2020
https://www.audioholics.com/audio-amplifier/amp-myths-negative-feedback

[ Summary of Key Talking Points about Amplifier Feedback

• Feedback is a natural process that happens all around us. Engineers have learned to use it to solve many difficult problems. Building precise amplifiers and filters are some audio related examples.
• The central myth about feedback is that it's easy to pile on more at will. This is only true at subsonic frequencies. Higher up the audio band every dB of feedback is hard fought for.
• The proverbial 70's amp that had lots of feedback only did so at low frequencies, but had very little left at high frequencies. Unfortunately they used a type of input stage that only works properly if you really do have a lot of feedback. That caused TIM.
• Increasing global feedback quickly produces a massive reduction in TIM. This is the exact opposite of a widely circulated myth.
• Nesting local and global loops is the most common way of increasing feedback. Almost all modern class AB amplifiers work like that. They are not low-feedback amplifiers, even if they're often marketed as such.
• Full global feedback is not easy but is technically superior to nested or partial feedback.
• Feedback theory is highly advanced in the field of Sigma-Delta AD/DA converters. Methods developed there are also often used in class D amplifiers.
• Feedback flips the shape of the distortion over. This can turn a distortion with few higher harmonics into one with more higher harmonics or vice versa. In practice the effect is only significant in lab curios specifically built to demonstrate the effect.
• Feedback is one of the most effective tools to improve sound quality. Although the R&D outlay can be substantial, using feedback barely affects the cost of the circuit. ]

 

[KSTR]

I only skimmed through it but Mr.Putzeys is known to deliver real world knowledge backed up by a huge theoretical/math background so he'll be sure spot on, detailing on the core of feedback theory and practise.

Interesting tid-bit is the schematic of a composite amplifier shown in the summary section, which is a nice ultra-high gain and high bandwith building block with two (similar) opamps. It runs closed-loop nicely at 40dB..50dB gain and still have way more loop gain available for feedback at 20kHz than the single opamp in unity gain. The single intergrator cap does it all to get that thing stable when component values are properly selected. There are many more examples of composite amplifiers that live off of huge monstrous amounts of gain available for global negative feedback at audio frequencies and beyond.

 

[Feanor]

A few years ago Nelson Pass wrote this article which can be found on the Pass Labs website ...
https://www.passlabs.com/technical_article/audio-distortion-and-feedback/

Simplifying Pass' conclusions a bit, the basic argument is that feedback, while reducing overall distortion, creates more and stronger high-order harmonics which cause a harsh, discordant sound.

At one point in his treatise, Pass says, "I performed my own version of the experiment, using a power Mosfet in a single-ended Class A gain stage driving 1 watt into an 8 ohm load:", and displays this graph:

I'm not entirely sure whether this apparent increase in size of 4th and higher order HD is entirely typical of increasing feedback or whether they may different under other parameters. One thing is clear, though, that total distortion decreases with feedback -- also 15 dB of feedback, the maximum that Pass shows, in not all that much.

My own thesis is that what tube lovers and other low-feedback advocates really like is that their equipment is that it produces more 2nd and/or 3rd order HD which sounds pleasant, NOT that it produces less higher order.

One thing I do know, anyway, is that my Pass Labs X150.5 class A/B amp sounds great despite have unremarkable distortion measures.

 

[AnalogSteph]

Simplifying Pass' conclusions a bit, the basic argument is that feedback, while reducing overall distortion, creates more and stronger high-order harmonics which cause a harsh, discordant sound.

That's actually true - as long as the amount of NFB applied remains moderate. Bruno Putzeys, however, showed that at some point all the harmonics start duly going down further as you apply more feedback. The turning point is somewhere around 20 dB. Hence he concluded that you shouldn't be timid when applying NFB - the more the merrier.

 

[NTK]

...
Simplifying Pass' conclusions a bit, the basic argument is that feedback, while reducing overall distortion, creates more and stronger high-order harmonics which cause a harsh, discordant sound.
...

Here is Amir's Purifi 1ET400A 1kHz THD measurement. With highest HD at -127 dB, the 1ET400A curve won't even show up in Mr Pass' graph (lower y-limit of -120 dB). So whose view of NFB is better?

 

[Soniclife]

Simplifying Pass' conclusions a bit, the basic argument is that feedback, while reducing overall distortion, creates more and stronger high-order harmonics which cause a harsh, discordant sound.

At one point in his treatise, Pass says, "I performed my own version of the experiment, using a power Mosfet in a single-ended Class A gain stage driving 1 watt into an 8 ohm load:", and displays this graph:

When you are designing a harmonic distortion generator that does a bit of amplifying on the side I suppose it's critical you get the right mix of distortion, but it's not relevant to people designing hi-fi amps, as illustrated by that purifi measurement.

 

[Feanor]

That's actually true - as long as the amount of NFB applied remains moderate. Bruno Putzeys, however, showed that at some point all the harmonics start duly going down further as you apply more feedback. The turning point is somewhere around 20 dB. Hence he concluded that you shouldn't be timid when applying NFB - the more the merrier.

So the use of feedback is one of those rare situations where if a little makes things worse, much more actually make things better better.

I'm not arguing. However the fact remains that my Pass Labs amp sounds great. Here's John Atkinson's comment on the similarly designed, (though class A), XA60.5 ... "I don't have much to say about the Pass Labs XA60.5 other than this: It is the best-sounding amplifier I have ever used."

... https://www.stereophile.com/content/pass-labs-xa605-monoblock-power-amplifier

 

[Theriverlethe]

It’s negative vibes, man. You can’t have negative vibes disrupting your positivity, or questioning your engineering competence.

 

[SIY]

I'm not arguing. However the fact remains that my Pass Labs amp sounds great.

The question remains, though, whether it actually sounds different than an engineered low distortion amp. And neither Nelson nor Atkinson will do the basic experiment needed to answer that question.

 

[Feanor]

The question remains, though, whether it actually sounds different than an engineered low distortion amp. And neither Nelson nor Atkinson will do the basic experiment needed to answer that question.

What "basic" experiment do you propose? I suspect that a basic, blind "ABX" test limits the applicability of its results by its test conditions. Audiophiles don't typically use ABX testing to establish their preferences. For that matter, both Pass and Atkinson have done a lot of listening and testing to establish their preferences.

 

[Feanor]

When you are designing a harmonic distortion generator that does a bit of amplifying on the side I suppose it's critical you get the right mix of distortion, but it's not relevant to people designing hi-fi amps, as illustrated by that purifi measurement.

This is a rather flippant remark. Nelson Pass, (for example), has been designing hi-fi amps for more that 50 years more than a little success. But I agree that Pass has aimed for "the right mix of distortion" rather than lowest possible distortion for all his career at least to date.

If there is a problem with Putsy's philosophy, (which I do greatly respect), it is that he doesn't accept that not all audiophiles actually want the least possible distortion -- how do you explain the preference of so many for tube equipment?

 

[SIY]

What "basic" experiment do you propose? I suspect that a basic, blind "ABX" test limits the applicability of its results by its test conditions.

No, it does no such thing. It just isolates the "sound" from other non-auditory influences.

Two amps can be distinguished by their sound or they can't. There's really nothing complicated about that.

 

[Feanor]

The likeliest explanations are commodity fetishism, Veblen goods, face, and perhaps some actual sonic preference, but such a sonic preference bears no relationship to what live music sounds like or what is actually on a recording.

Yes, I agree: only the lowest possible distortion will get you closest to "what is actually on a recording". For better or worse that isn't what everybody wants to hear.

 

[Feanor]

No, it does no such thing. It just isolates the "sound" from other non-auditory influences.

Two amps can be distinguished by their sound or they can't. There's really nothing complicated about that.

... If all you want to do distinguish one amp from another, sure. If you want to decide which you like best, not necessarily.

 

[SIY]

... If all you want to do distinguish one amp from another, sure. If you want to decide which you like best, not necessarily.

Sure, if "like" is the criterion. I "like" the amp I'm using now better than the amp it replaced, but I 100% guarantee that I (and I suspect anyone) couldn't tell them apart by sound.

But if you're claiming you "like" it better because of the sound, or as the seller, claim that is sounds different or better, that needs to be demonstrated ears-only to be an honest claim. Nelson hasn't done that. Atkinson hasn't done that. And until they do, their claims are no different than those of UFO abductees.

 

[Feanor]

Sure, if "like" is the criterion. I "like" the amp I'm using now better than the amp it replaced, but I 100% guarantee that I (and I suspect anyone) couldn't tell them apart by sound.

But if you're claiming you "like" it better because of the sound, or as the seller, claim that is sounds different or better, that needs to be demonstrated ears-only to be an honest claim. Nelson hasn't done that. Atkinson hasn't done that. And until they do, their claims are no different than those of UFO abductees.

I really don't know why you suppose that Pass can't tell the sound differences among the the many, many different designs he has developed.

Take 3-4 minutes and read my recent amp review here ... https://www.audioasylum.com/forums/amp/messages/23/235714.html ... you will, of course, scoff at my evaluation method and conclusions.

 

[KeithPhantom]

I really don't know why you suppose that Pass can't tell the sound differences among the the many, many different designs he has developed.

When claiming something, we usually resort to sticking to the null hypothesis (the one who does not claim any change) and test if there is a change. Until Pass proves himself in a third-party, scrutinized and repeatable test with enough positive evidence that he can hear the difference, we believe he cannot detect the differences.