Global Feedback - A huge benefit for audio

[sottomano]

Get a hearing test maybe?

So you can hear a difference between music material containing -60 respectively -80dB H2/H3?
I can't, and know some others, with the same result. Don't think they are all in need of a hearing test,

Cheap indeed, such comments. Didn't expect that from you, to me personally disappointing given how I appreciate your contribution on hum/noise in amps and also the MC pre compendium.

OK, what do you think about Mauro Penasa's work? I posted the links some pages ago.
Question for all btw, especially for those very vocal about the "back EMF" topic.

 

[billshurv]

So you can hear a difference between music material containing -60 respectively -80dB H2/H3?
I can't, and know some others, with the same result. Don't think they are all in need of a hearing test,
.

Some claim they do. Personally I think if the 19-20khz IMD products are below 80dB down then that part of the requirements are met and we can move on to other possibly more important parameters. But I make no claims about this being better or inaudible. Some on here claim to be able to hear changes 130dB down...

 

[scottjoplin]

I was surprised by what I thought I could hear in one of Pavel's listening tests, but I know I don't listen to music like that, I was intensely focusing on one small element to hear a difference, it was not about pleasure.

 

[Bonsai]

So you can hear a difference between music material containing -60 respectively -80dB H2/H3?
I can't, and know some others, with the same result. Don't think they are all in need of a hearing test,

Cheap indeed, such comments. Didn't expect that from you, to me personally disappointing given how I appreciate your contribution on hum/noise in amps and also the MC pre compendium.

OK, what do you think about Mauro Penasa's work? I posted the links some pages ago.
Question for all btw, especially for those very vocal about the "back EMF" topic.

Nothing more than humorous dig 😉

Seriously, this subject comes up every 2 yrs. There’s a huge discussion which really is nothing more than people with entrenched views repeating the same stuff We all cool down after a week or two. Rinse, spin and repeat in 2 yrs.

 

[scottjoplin]

Except this thread was started 5 years ago by someone in the business looking for ideas, I wonder if he found any and if he did whether he'll tell us?

 

[sser2]

Not just opinion. I have that article on my website for 15 years.

Anyway, if you do the math, it becomes very quickly clear that pfb increases distortion, worsens the freq response, and deteriorates the output impedance.

Which you would have known too, had you taken the trouble to read the article.

Jan

Wondering which classic I should listen to, Mr. Miller or Mr. Didden.

 

[sser2]

Nothing more than humorous dig 😉

Seriously, this subject comes up every 2 yrs. There’s a huge discussion which really is nothing more than people with entrenched views repeating the same stuff We all cool down after a week or two. Rinse, spin and repeat in 2 yrs.

What you said applies both ways.

 

[sser2]

Never seen a Schmitt trigger in the signal path of an audio amplifier 🙄.
The tube amplifier designs I was referring to did implement both voltage NFB as well as current PFB, if we can agree that current FB means sending back a signal that is proportional to the amp's output current.
Best regards!

Harman Kardon A300, schematic easily found on the web.

 

[N101N]

sser2,
now I need to make a sincere apology, we may be making the same point. Your opinion about NFB is surprising.

 

[sser2]

I'm not sure what you're talking about. What I've observed, however, is that one side clings to very naive and simple views like "(global) negative feedback is bad" and "the distortion spectrum is more important than low numbers" but also creates the counter-positions like "negative feedback is always good" ... which are not only strawmen but also similarly naive and simple.

Reality is a bit more complex. But the science has been known for a long time. The measurements and tests of audibility of distortion also are nothing new. (This is also easy and cheap to make fun of. See the "hearing test" comments.)

But I can see the appeal to belonging to the former group. It's easy on the intellect. That seems to be the most important point.
It's also fun to make fun of the evil science that cannot be true because that would mean a waste of lots of time and money and admitting that cognitive bias/distortion are real ... by attacking self-made naive strawmen that are easy to make fun of. (So effectively they're making fun of themselves.)

Reminds me of flat earth groups.

I wish if it was that simple: NFB good period. But as this tthread reveals, it is not. NFB cannot be applied in infinite amount, so claims that feedback amplifier output is a dead short for speaker back EMF do not hold water. In real circuits, phase compensation of NFB is difficult. Even NFB proponents agree that it has to be carefully designed, and almost Einsteinian wisdom is needed to do it right. But why going into the trouble, I wonder. As quite reasonably stated in several posts of this thread, and not disputed by NFB proponents, low distortion and low output impedance can be achieved w/o global NFB. So the point that NFB must be used no matter what is as extreme as the point that NFB should never be used under any circumstances.

 

[analog_sa]

Never seen a Schmitt trigger in the signal path of an audio amplifier 🙄.

Some amount of PFB does not a Schmitt trigger make.

 

[Bonsai]

There seems to be a lot of 'infinite feedback' and 'zero output impedance' stuff being bandied about in this discussion.

There is no such thing as infinite feedback or zero output impedance in practical feedback amplifiers. The figures are easily measurable and easily calculatable.

If someone is making these claims, clearly they do not understand feedback.

 

[bimo]

IF NEGATIVE FEEDBACK ACTUALLY WORKED, THEN ALL AMPLIFIERS WOULD BE PERFECT, AND ALL AMPLIFIERS WOULD SOUND EXACTLY THE SAME."

No amplifier or no technology is perfect. There is no perfection in engineering. There is always a compromised.

 

[bimo]

So the point that NFB must be used no matter what is as extreme as the point that NFB should never be used under any circumstances.

If you can not design a good amplifier with global negative feedback, it does not mean everyone can not either.

 

[Bonsai]

As has already been explained, it is not possible to design an amplifier without feedback - be it GNFB or local feedback.

 

[jan.didden]

Or disbelievers, for example Achieve good time domain behaviour to produce single cycle sine

Reminds me of a long-forgotten thread 'first cycle distortion'. Another example of not understanding that a single cycle waveform has an infinite bandwidth.
A speaker cannot reproduce an infinite bandwidth signal, no matter how many servos you throw at it.
And so it goes.

Jan

 

[jan.didden]

So the point that NFB must be used no matter what is as extreme as the point that NFB should never be used under any circumstances.

Indeed. That must be why no sane person is claiming that. But it is a nice strawman if you run out of technical arguments ;-)

Jan

 

[edbarx]

As has already been explained, it is not possible to design an amplifier without feedback - be it GNFB or local feedback.

Mister, you might be out of luck with youtube. Rest assured there are 'great sounding' amplifiers without feedback.

I leave to readers what 'great sounding' means.

 

[scottjoplin]

That's an easy one, nothing

 

[bimo]

As has already been explained, it is not possible to design an amplifier without feedback - be it GNFB or local feedback.

But Dr. Arto Kolinummi can design an amplifier without GNFB with very low distortion. If you can not design an amplifier without GNFB with very low distortion, it does not mean everyone can not, either.