That just shows their lack of sophistication.
Understand a basic point- reviewers are hired NOT because of any special abilities or insights regarding sound or electronics, but their ability to generate entertaining articles month after month which allow magazines to make money.
Understand a basic point- reviewers are hired NOT because of any special abilities or insights regarding sound or electronics, but their ability to generate entertaining articles month after month which allow magazines to make money.
Grey wrote
So, it's not that I'm pathologically opposed to NFB under any circumstances, ever, period, no way...it's that I'm unwilling to pay that imaging penalty or the upper midrange glare. And if I can flog a circuit into decent distortion specs without it, what benefit do I get from NFB? Lower distortion? It's already below the threshold. Wider bandwidth? My goal on anything I build on my own is 200kHz minimum, preferably 250. At the moment I'm running twice that (I have measured the bandwidth--it's actually over .5Mhz, but I'm rounding down). Do I really want to go to 2 or 3Mhz? Higher still? Nah. Maybe next time. I just don't see why I would want to claw my way up to where I'm able to put something together that performs that well (albeit by standing on John, Nelson, and Charles's shoulders--thanks guys) then gunk it up by adding bells and whistles. It just ain't elegant, you know?
Grey, this is absolute baloney . . . but its entertaining, so keep on writing this stuff. It just a pity that some people believe it.
So, it's not that I'm pathologically opposed to NFB under any circumstances, ever, period, no way...it's that I'm unwilling to pay that imaging penalty or the upper midrange glare. And if I can flog a circuit into decent distortion specs without it, what benefit do I get from NFB? Lower distortion? It's already below the threshold. Wider bandwidth? My goal on anything I build on my own is 200kHz minimum, preferably 250. At the moment I'm running twice that (I have measured the bandwidth--it's actually over .5Mhz, but I'm rounding down). Do I really want to go to 2 or 3Mhz? Higher still? Nah. Maybe next time. I just don't see why I would want to claw my way up to where I'm able to put something together that performs that well (albeit by standing on John, Nelson, and Charles's shoulders--thanks guys) then gunk it up by adding bells and whistles. It just ain't elegant, you know?
Grey, this is absolute baloney . . . but its entertaining, so keep on writing this stuff. It just a pity that some people believe it.
well, post the alternative argument and let us judge on the basis of a well argued case.Bonsai said:
I already have . . . so has Bob Cordell, Janneman, Greg . . . et al and ad infinitum. Read the posts prior to this to pick up the arguments thread.
We're waiting for the zero feedback gang to explian why sophisticated listeners (their phrase) also give amps with feedback just as high ratings as those with zero feedback.
The silence is deafening.
BTW, I've just been on the Krell website. Dan Agostino uses 8dB of global feedback in his Evolution series amps. Low enough I'd say that, if it made his amps sound worse, he could have removed it all together. Of course, I think he used it to make his amps sound better.
We're waiting for the zero feedback gang to explian why sophisticated listeners (their phrase) also give amps with feedback just as high ratings as those with zero feedback.
The silence is deafening.
BTW, I've just been on the Krell website. Dan Agostino uses 8dB of global feedback in his Evolution series amps. Low enough I'd say that, if it made his amps sound worse, he could have removed it all together. Of course, I think he used it to make his amps sound better.
Only 8dB of feedback, I can live with that! I would even go for 14dB of feedback, if I could get decent specs. I wish that my amp could be made with 14dB instead of 34dB (or so) and my open loop bandwidth could go from 6KHz (or so) to 60KHz. Looks like a good goal to pursue. I suspect that high open loop bandwidth, coupled with feedback is as non-problematic as local feedback, because both have equal feedback over the audio range and more. I bet that the fundamental problem is low open loop bandwidth.
Why does Krell only use 8db of feedback? Surely, 80 dB is possible or something in between.
It may not be just feedback that is the real problem. I think it is OPEN LOOP BANDWIDTH, get it above 25KHz or so, and there should be no problem with modest amounts of feedback.
Why does Krell only use 8db of feedback? Surely, 80 dB is possible or something in between.
It may not be just feedback that is the real problem. I think it is OPEN LOOP BANDWIDTH, get it above 25KHz or so, and there should be no problem with modest amounts of feedback.
Some thoughts on other variables that may or may not affect the sound of an amp as much, or more, than the level of feedback (or absence therefof)
Front end topology - single ended differential, balanced differential, singleton; bipolar, JFET. Mirror loading, resistove loading; resistor derived current source (a la Leach) or active current source; level of diff amp degeneration; diff-amp current; use of cascodes (to accomodate high rail voltage and/or ameliorate early effects - also proposed by Self for latter issue on current source)
VAS
Balanced, single ended, differential; folded cascode etc etc; VAS standing current; use of cascodes; buffered (from diff amp side) using darlington or similar; VAS amplifer transistor emitter degeneration levels; use of Cdom and abolute value
. . . . etc etc
Output stage
Long list, I won't go into other than mosfet output, bipolar output, triples, emitter follower, CFP et al
Compensation
. . . single pole, double pole, lead, lag, non-global feedback (if you are running an open loop design), global feedback, TMC techniques
Voltage feedback vs Current feedback
Use of output inductor, non use of output inductor
Power Supply - main supply - toroid, E core etc
😱
Front end topology - single ended differential, balanced differential, singleton; bipolar, JFET. Mirror loading, resistove loading; resistor derived current source (a la Leach) or active current source; level of diff amp degeneration; diff-amp current; use of cascodes (to accomodate high rail voltage and/or ameliorate early effects - also proposed by Self for latter issue on current source)
VAS
Balanced, single ended, differential; folded cascode etc etc; VAS standing current; use of cascodes; buffered (from diff amp side) using darlington or similar; VAS amplifer transistor emitter degeneration levels; use of Cdom and abolute value
. . . . etc etc
Output stage
Long list, I won't go into other than mosfet output, bipolar output, triples, emitter follower, CFP et al
Compensation
. . . single pole, double pole, lead, lag, non-global feedback (if you are running an open loop design), global feedback, TMC techniques
Voltage feedback vs Current feedback
Use of output inductor, non use of output inductor
Power Supply - main supply - toroid, E core etc
😱
Bonsai said:
Deafening silence, eh?
I think a lot of the feedback versus no feedback question boils down to a question of personality. More to the point, control freak personalities want order and unquestioning obedience. Anything that smacks of freedom, independence, or stepping outside the accepted borders must be suppressed at all costs. They think in terms of authority figures and rigid hierarchy.
Case in point: Bonsai says that high feedback amps get high marks from 'sophisticated reviewers,' whoever that might be. I am supposed to bow to "authority" and recognize that--since 'sophisticated reviewers' put low and high feedback circuits on an equal footing, then by implication they are peers in performance. And if 'sophisticated reviewers' accept high feedback circuits, then I must also do so, because Bonsai believes that they are--or should be--an authority figure to me.
(N.B.: Control freaks never admit to being control freaks. They can't. It's in their nature. In fact, they react badly to being called out on it.)
There are, however, other personality types.
I use audio reviews the same way I do movie reviews. If you can find a decent reviewer (granted--not always easy), they can screen out the junk for you and present you a "short list" of things that you might want to see...or hear, as the case may be.
I mean, really, can you imagine Bonsai's post on a movie web site? He tells me that I have to give movie ABCD high marks simply because Ebert--a reviewer, hence an authority figure--liked it. Tell me...does that sound reasonable? It doesn't to me.
I prefer to make up my own mind, rather than receive my marching orders from others.
Or from meters, for that matter.
bogdan_borko said:
Never posted anything? Hmmm...you haven't looked around the site much, have you?
My current circuit will be posted when I'm satisfied with it. At the moment, I'm working on the artwork for another version of the front end, as the previous one developed so many jumpers and stuff that it started getting hard to work with.
Descriptively, it goes something like this:
Start with John's JC-3 schematic. Complementary differential with 2SK389/2SJ109, cascoded with 2SK246/2SJ103. John's whatchamacallit folded cascode into 2SK2013/2SJ313. Cascode that with another 2SK2013/2SJ313 to reduce the capacitive load on the front end differential. Plain vanilla push-pull output using Fairchild MOSFETs running class A; output around 25W/8 Ohms. The differential runs at about 5mA. The folded differential runs hard at around 30 to 50mA Iq, depending on my mood. No servo at present. How do I get away with no servo? Heavily loading the folded cascode resistively. It gets rid of excessive gain and tends (if the resistance is low enough) to nail the thing to ground pretty effectively.
This version is all FET, single-ended. The next iteration will be a bridged version of the same circuit. In either that one or the next one, I'll experiment with bipolars, so as to get rid of the variable Vgs problems that crop up with the MOSFETs in the folded cascode.
I'm of two minds as to whether I'll post the single-ended circuit or the bridged one, as there will be limited interest in a 25W amp. A 100W circuit will be more useful to more people. Depending on time constraints I may have the single-ended one ready by mid-to-late January. The bridged one won't be as much of a booger as you might think, because I've already (as of the second version of the PCB) gone to a fully symmetrical front end; just leaving the unused side flopping around in mid-air at present. Building a bridged amp will hopefully only be a case of hanging another output bank off the other side, then fine-tuning the thing a bit. The bipolars will both simplify and complicate things, so that could take more fiddling...or it could just naturally fall into place if I'm lucky.
john curl said:
I've got a benchmark of 200 to 250kHz open loop bandwidth. I am willing to spend a modest amount of feedback if it will help me get there. I've used around around 7dB in the past and have exceeded that bandwidth with zero NFB this time around. No, I don't feel that there's anything magic about the number 7--it's just what it took to get the bandwidth where I wanted it.
Why do I want the bandwidth so badly? Because--and I've said this elsewhere--for some reason it takes a solid state circuit with at least that to even begin to sound as open as a decent tube circuit with, say, 100 to 150kHz bandwidth. Why? Hell if I know! It doesn't make any sense to me. For some reason, tube circuits can sound more airy, more relaxed, more open with what would nominally be an inferior bandwidth. My operating hypothesis on this one is that is has to do with the character of high frequency distortions, not the absolute distortion figures. That's just hand-waving and is as likely to be wrong as right. Operationally, I just go for wide open loop bandwidth and it works. The more the better.
This is your cue to explain via various distortion mechanisms why/how this works. Me? It just sounds better. I'm a pragmatic sort of guy. If it works, I do it, then try even more of it the next time to see if the trend continues.
Yes, I recognize that your amp has more power, meaning more output devices and more difficulty driving them. In my case, I'm running two pairs output devices and things are just peachy. If I were to go to, say, four or five pairs, then all hell would be out for lunch. The cumulative Gate capacitance would pull down my bandwidth (and increase distortion) and I might have to bite the bullet and stick a little feedback in there. That's why I'm going to use a bridged configuration for the higher power amp, as I can put the new output devices on the currently unused side of the front end and not have to bog down the one set of drivers with all that Gate capacitance and nonlinear behavior. Ask me in a month or three how it goes--right now I can only speculate.
I'm also fortunate in that I don't have to hit THX specs or worry about market acceptance among knuckleheads who think an amp has to have feedback to sound good, just because that's the way his grandpappy did it.
Grey
MikeBettinger said:
That's the advantage of having such a detailed understanding of the theory of how things work. There is never a loss for words or references to backup what you know.
On the other hand, observations of reality really stink. No credible references availible for things that don't fit the well documented world view and only one's experience, insight and gut instincts to help piece them together into any usable form.
Even then, the conclusions drawn are subjective,
Hardly enough to sustain anyone in any serious technical discussion!
Sometimes I don't understand why some people (myself included) can't accept what has already been figured out by the smart people in the world. It would save so much time and energy that is now being wasted by trying to educate the the less fortunate.
Oh well. History repeats itself, again.
I do have to apolgize for my bad attitude. I was out Christmas shopping and got a bit of a headache trying to pick out an surround sound system for my son in law and daughter. I'll be better after a bit... .
Mike
Interesting response.
I returned this afternoon from being given a Nissan Patrol and being payed to traverse some of the most remote, vast and scenic expanses of desert landscape on the planet (coal/gold exploration).
Now I'm sitting here, in front of a crappy VDU, reading through multiple paragraphs obstinate rubbish and convoluted pseudo-scientific babble about the alleged engineering evils of NFB in audio design.
I mean, really. In one post we’ve even got the pious “I was a sinner, just like you, until I saw the light” tirade concluding with a feminised characterisation of the Devil. I won’t elaborate in the parallels here, but the audio design world is clearly flat in some quarters.
BTW, no measure of reasoned argumentation would kink the armour of the wilfully ignorant and deluded and I can think of better things to do with my spare time than expending any effort within the limits of my abilities in an attempt to do such.
Good bye!
Attachments
Grey wrote
Case in point: Bonsai says that high feedback amps get high marks from 'sophisticated reviewers,' whoever that might be. I am supposed to bow to "authority" and recognize that--since 'sophisticated reviewers' put low and high feedback circuits on an equal footing, then by implication they are peers in performance. And if 'sophisticated reviewers' accept high feedback circuits, then I must also do so, because Bonsai believes that they are--or should be--an authority figure to me.
Grey, you need to go into politics. You'd do damn well.
Please show me where I wrote 'high feedback amps get high marks from 'sophisticated reviewers' etc.
I cannot believe the sloppy agummentation and down right twisting of the facts here.
You and the zero feedback camp who claim feedback 'sucks the life out of music' simply need to explain why amplifiers with feedback (any level - you choose) also get excellent reviews from 'sophisticated liseners or reviewers'. Thats all. If you cannot explain this phenomena, your whole argument falls flat - its absolutely worth zero.
Please exaplain to me, and the feedback camp how this can be?
We're waiting.
Case in point: Bonsai says that high feedback amps get high marks from 'sophisticated reviewers,' whoever that might be. I am supposed to bow to "authority" and recognize that--since 'sophisticated reviewers' put low and high feedback circuits on an equal footing, then by implication they are peers in performance. And if 'sophisticated reviewers' accept high feedback circuits, then I must also do so, because Bonsai believes that they are--or should be--an authority figure to me.
Grey, you need to go into politics. You'd do damn well.
Please show me where I wrote 'high feedback amps get high marks from 'sophisticated reviewers' etc.
I cannot believe the sloppy agummentation and down right twisting of the facts here.
You and the zero feedback camp who claim feedback 'sucks the life out of music' simply need to explain why amplifiers with feedback (any level - you choose) also get excellent reviews from 'sophisticated liseners or reviewers'. Thats all. If you cannot explain this phenomena, your whole argument falls flat - its absolutely worth zero.
Please exaplain to me, and the feedback camp how this can be?
We're waiting.
Glen,
you sound drained having defended our feedback turf over th e last few months. I'll pick up from you for a while. You need a break.
C'mon zero feedback crew (ZFC), lets hear your explanation as to why amps with f/back also get great reviews.
you sound drained having defended our feedback turf over th e last few months. I'll pick up from you for a while. You need a break.
C'mon zero feedback crew (ZFC), lets hear your explanation as to why amps with f/back also get great reviews.
G.Kleinschmidt said:
Glen,
I'm surprised, you should know better than to pay attention to my posts. Most are placed purely for my intellectual satifaction and amusement. I don't expect anyone to actually get or, more importantly, care about, what I'm saying.
The world, according to this site, is a pretty straightforward, black and white place. And, although I'm far from an advocate of zero feedback, I know any discussion of it is pointless because it doesn't align with the published theory and the many examples sold by those who have followed it...
How can one win an argument with the bible. Anyone who thinks this is possible is dilusional. Forget what your ears tell you, obviously there's a problem there.
Mike.
Problem is, Mike, people seem to be reacting more to their preconceptions and biases than their ears. That's where the problem is. How can this be debated? There's no subjective data, just bald assertion and Superior Dances.
john curl said:
Hi John,
You raise a couple of interesting points here. I've posted some thoughts on it over on the permanat negative feedback thread.
Cheers,
Bob
Bonsai said:
Bonsai,
Where?
Here.
And here: "C'mon zero feedback crew (ZFC), lets hear your explanation as to why amps with f/back also get great reviews."
And...oh, never mind...you have no intention of taking responsibility for what you post, anyway.
I've already posted about what I regard as the proper use of reviews. I could say all the same things again, but I think what I said in my previous post pretty much covers the necessary points. It's really, really simple: I don't feel the need to have a reviewer tell me what to think. Apparently you think I do. Or else you need someone to tell you what to think and feel that everyone else is the same way. Ain't so.
Just because a reviewer thinks a higher feedback circuit sounds good does not obligate me to agree.
But maybe that concept is too complicated for you.
One of the curious things about control freaks is that they are incapable of realizing that others don't need authority figures to tell them what to do. They are convinced that because they want a rigid hierarchy based on inflexible rules, that the same applies to others. I could give numerous examples based on religion and politics, but the moderators would only edit the stuff out.
You seem unable to separate reality from your assumptions. Until you realize that others are capable of making up their own minds without being told by authority figures, I'm going to continue thinking you fit the description in my previous post to a "T."
Or perhaps it's just poor reading comprehension.
analog_guy,
Your proposal sounds like it might cover my imaging observations, but it still leaves begging the question as to why decent tube amps can accomplish much the same sonic goals with narrower bandwidth. Particularly in light of phase shift going through an output transformer. I've argued this back and forth with myself, proposing different mechanisms, but haven't reached a satisfactory conclusion.
How low do you intend to set the bandwidth?
Grey
SY said:
It's pretty much human nature, the eternal search for the Holy Grail. Someone steps up and says "I've Found It" and five others pipe in "no you didn't, we did" the experts jump in to analyze the situation. This discussion has turned into people throwing rocks at one and another. A winner/loser standoff.
Your question "how can this be debated?" Is a core question for any site like this.
There is an underlying purpose and reality to all of the passion that motivates most of the people with this affliction; there is alot to be shared/learned through discussion. If the debate does not begin with mutual respect and the premise that the other side is actually describing and discussing the topic from an intellectually honest position, there is no point continuing.
I know, I know 🙁 ... And yes Virginia, there is a Santa Claus! I've danced with him!
Mike.
Hi
This is certainly an interesting and informative thread, to some extent. What happened to the option of using the best of both worlds? Such as making a no feedback amp from multiple stages, each stage using a negative feedback scheme, a separate amplifier by itself, designed to drive the next stage input and so on. Lately I created an amp that is such in that both of the two gain stages are bridge topology, w/feedback, but there is no GNF, rather a cascade of amplifier stages. There is so much more noticeable 'sound stage' present with this amplifier. I actually heard the conductor turn a page in the score from a recording of some of Vivaldi's music. I had to backtrack to confirm what I heard, and yep...it's in there. This I never noticed before....
😎
Of course this direction would not be appropriate for those who are fortunate enough to have this super Hi-Tech 'magic' silicon that has the linear gain of a tube.😀
This is certainly an interesting and informative thread, to some extent. What happened to the option of using the best of both worlds? Such as making a no feedback amp from multiple stages, each stage using a negative feedback scheme, a separate amplifier by itself, designed to drive the next stage input and so on. Lately I created an amp that is such in that both of the two gain stages are bridge topology, w/feedback, but there is no GNF, rather a cascade of amplifier stages. There is so much more noticeable 'sound stage' present with this amplifier. I actually heard the conductor turn a page in the score from a recording of some of Vivaldi's music. I had to backtrack to confirm what I heard, and yep...it's in there. This I never noticed before....
😎 Of course this direction would not be appropriate for those who are fortunate enough to have this super Hi-Tech 'magic' silicon that has the linear gain of a tube.😀
- Status
- Not open for further replies.