Part of the problem is that many people are desirous of
being able to use the term "no feedback"
in association with their amplifier for marketing or
bragging rights.
Taking the position that "there's no such thing as no
feedback" doesn't move us forward in the real world,
so a compromise position seems to have arisen in which
no feedback is an allowed claim for amplifiers which
have degeneration only, and local feedback describes
circuits where a loop exists only around a single gain
element.
I could be wrong, but it appears that this is the consensus
among myself, Curl, Cordell, Hansen and others.
Anyone is free to chime in here....
Ohhhhww..! That foul, all-powerfull marketing dept
About 30 years ago, a home in Holland that had a garage build on one of the outside walls, was called, well, a home with a garage.
We also had homes that had a garage actually as part of the home itself. These were called 'drive-in homes'. Drive-in homes sold for more than a mere home with a garage, so through the years the realtors agreed to the 'compromise' to call any home that had a garage fixed to it in some form, a 'drive-in home' which apparently did move us forward in the real world.
So, Nelson, having eventually acquired a 'drive-in home' myself, how can I protest?
jan didden
Last edited:
"there is nothing so practical as a good theory"
of course it helps when language Aids understanding
I could be fine with quote marks: "no feedback" being a purely Marketing term - it could also help to use the trademark symbol on that too
but as you can see from the discussion here it generates considerable cognitive dissonance in non-experts who (rightly, IMHO) feel betrayed by the marketing language that they likely 1st experience and internalized when they then have to deal with real feedback theory discussions
it seems wrong to argue against using language consistent with the engineering definitions - particularly in a diy forum where we hope "correct" circuit understanding could have some value to builders - or at least using the words to mean the same things could help their communication with formally trained EE circuit designers and in reading for self education
the informal language doesn't always have to put every important formal distinction in the way - but the word "feedback" shouldn't mean the opposite in the ad copy vs what engineers designing the circuit use (at least in textbooks, class rooms, IEEE peer reviewed papers... despite whatever Charles may actually think while using Blackman's formula to calculate input/output impedances)
some others here have as Jan points out apparently have gotten hung up on detailed "semantic analysis" of simplified "feedback" definitions that require "feeding back" the OUTPUT signal to the input ("in anti-phase" seems to be another stumbling block)
I believe a more useful definition for identifying "Feedback" is as system with a "command input" and a "controlled variable"
the "command input" - the value you want to make the controlled variable "follow"(often directly 1:1, or by a constant scale factor in amplification or more complicated function in a filter)
the command and controlled variable do not have to be the same quantities - that is where the 2x2 matrix of series-shunt combinations of I, V as command "input" and controlled (and often directly "output") variables come in - admittedly it takes a while to appreciate the distinctions - I certainly don't use the series/shunt formalism often - but it does help identify more diverse feedback systems - so I know where to find it in books or on the web when I need it
the "Feedback" part comes from calculating a controlling input for your gain device from "the difference" between the command input and the variable (again often the direct "output" but not always) controlled by your active gain device
"the difference" can be calculated implicitly by the circuit topology - I think most can identify the "input" and the "fed back" signals in the popular JLH Class A amplifier - which doesn't have a differential pair "subtractor"
I have seen some books suggest that Blackman's formulas are so fundamental to feedback that you might as well redefine "feedback" to include any circuit where impedances, gains follow the relations
the engineering definition of feedback to include degeneration, followers isn't some awkward "over-reaching" "patched on" to the theory just to make the "feedback" concept look better, more "universal"
engineers, textbooks, professors describe degeneration and followers as feedback circuits for very practical reasons - the theory explains their circuit behavior - using the same rules as for more complicated feedback system
it can be seen as consistent in a very fundamental way when you apply Blackmann's relations to calculate input and output impedance as a function of "loop gain", when you use Bode's sensitivity analysis, calculate bandwidth, or the distortion reduction/harmonic order multiplication (as explained yet again, at length in the Cordell Feedback thread)
then you can go o the lab and see these behaviors in real circuits do follow the theory
its worth repeating: "there is nothing so practical as a good theory"
the scientific and engineering communities have found Black's invention (and Nyquist, Bode, Blackman... many others explanations/extensions) to be the basis for a Really Good Theory
of course it helps when language Aids understanding
Part of the problem is that many people are desirous of
being able to use the term "no feedback"
in association with their amplifier for marketing or
bragging rights.
Taking the position that "there's no such thing as no
feedback" doesn't move us forward in the real world,
so a compromise position seems to have arisen in which
no feedback is an allowed claim for amplifiers which
have degeneration only, and local feedback describes
circuits where a loop exists only around a single gain
element.
I could be wrong, but it appears that this is the consensus
among myself, Curl, Cordell, Hansen and others.
Anyone is free to chime in here....
I could be fine with quote marks: "no feedback" being a purely Marketing term - it could also help to use the trademark symbol on that too
but as you can see from the discussion here it generates considerable cognitive dissonance in non-experts who (rightly, IMHO) feel betrayed by the marketing language that they likely 1st experience and internalized when they then have to deal with real feedback theory discussions
it seems wrong to argue against using language consistent with the engineering definitions - particularly in a diy forum where we hope "correct" circuit understanding could have some value to builders - or at least using the words to mean the same things could help their communication with formally trained EE circuit designers and in reading for self education
the informal language doesn't always have to put every important formal distinction in the way - but the word "feedback" shouldn't mean the opposite in the ad copy vs what engineers designing the circuit use (at least in textbooks, class rooms, IEEE peer reviewed papers... despite whatever Charles may actually think while using Blackman's formula to calculate input/output impedances)
some others here have as Jan points out apparently have gotten hung up on detailed "semantic analysis" of simplified "feedback" definitions that require "feeding back" the OUTPUT signal to the input ("in anti-phase" seems to be another stumbling block)
I believe a more useful definition for identifying "Feedback" is as system with a "command input" and a "controlled variable"
the "command input" - the value you want to make the controlled variable "follow"(often directly 1:1, or by a constant scale factor in amplification or more complicated function in a filter)
the command and controlled variable do not have to be the same quantities - that is where the 2x2 matrix of series-shunt combinations of I, V as command "input" and controlled (and often directly "output") variables come in - admittedly it takes a while to appreciate the distinctions - I certainly don't use the series/shunt formalism often - but it does help identify more diverse feedback systems - so I know where to find it in books or on the web when I need it
the "Feedback" part comes from calculating a controlling input for your gain device from "the difference" between the command input and the variable (again often the direct "output" but not always) controlled by your active gain device
"the difference" can be calculated implicitly by the circuit topology - I think most can identify the "input" and the "fed back" signals in the popular JLH Class A amplifier - which doesn't have a differential pair "subtractor"
I have seen some books suggest that Blackman's formulas are so fundamental to feedback that you might as well redefine "feedback" to include any circuit where impedances, gains follow the relations
the engineering definition of feedback to include degeneration, followers isn't some awkward "over-reaching" "patched on" to the theory just to make the "feedback" concept look better, more "universal"
engineers, textbooks, professors describe degeneration and followers as feedback circuits for very practical reasons - the theory explains their circuit behavior - using the same rules as for more complicated feedback system
it can be seen as consistent in a very fundamental way when you apply Blackmann's relations to calculate input and output impedance as a function of "loop gain", when you use Bode's sensitivity analysis, calculate bandwidth, or the distortion reduction/harmonic order multiplication (as explained yet again, at length in the Cordell Feedback thread)
then you can go o the lab and see these behaviors in real circuits do follow the theory
its worth repeating: "there is nothing so practical as a good theory"
the scientific and engineering communities have found Black's invention (and Nyquist, Bode, Blackman... many others explanations/extensions) to be the basis for a Really Good Theory
Last edited:
When I look at the schematic I see the Sources of the
Mosfets driving the primaries of the transformers, so they
are degenerated by that impedance.
If you do not want to allow degeneration to be called "no
feedback" then I don't think this would qualify.
Well said.
You are one of the few left here (that actually contributes) that I pay strict attention to. Time and again you have proven to be a source of genuine information, devoid of the usual hocus pocus.
It's really important to SOME of us that we have someone giving solid advice that is not steeped in audiophool rhetoric.
Taking the position that "there's no such thing as no
feedback" doesn't move us forward in the real world,
so a compromise position seems to have arisen in which
no feedback is an allowed claim for amplifiers which
have degeneration only, and local feedback describes
circuits where a loop exists only around a single gain
element.
Wise words, you can’t say it any better. I totally agree.
In principle, it can’t hurt if some people here come down to earth and be clear about what this is all about. It's about equipment for music playback that, except us (some manufacturers and DIY’s), and perhaps a handful of customers around the world no one else needs. Therefore, this fundamentalism is misplaced! Of course, from a strictly scientific point of view, there is no active analogue circuit without a form of feedback. Who cares? At the end of the day it comes down to the fact how that particular unit sounds when it was built this or that way. And here we see gear with massive negative feedback (or rather, multiple negative feedbacks) in the judgment of customers looking very bad. Amplifiers with immeasurable THD figures and other hyper specs never saw a lasting success. Why is that I ask the self appointed experts here?
"No feedback" or "zero feedback" is an advertising message - who would deny that. However, it is also indicative of certain tonal characteristics. As such, I find it hypocritical to criticize it.
Lastly, there are some members in this forum - especially those who use their real names instead of nicknames - participating with the hidden agenda to promote a product, a book, or a consultancy. For others it seems to do well to the ego displaying their scientific education or expertise.
This had to be said once and that in all friendship.
it is a source follower so 100% source degeneration, a feedback none the less...
what's to be disturbed about? a rose by any other name smells the same, so says Shakespeare........
this feedback thing has become a religion like thing, as usual there are catholics and protestants......my take, whatever floats your boat......i don't care what proponents of either side say.....a feedback is a feedback is a feedback and a feedback.......
Good for you, but people tend to get influenced and misinterpretations tend to spread - do you happen to know who was the guy who said:"If you repeat a lie often enough, it becomes the truth" ?
Somehow, I can't easily accept that happening again and again...
sorry Zen that really doesn't get you free of the negative feedback "debate"
although "single ended pentode" is likely a pretty good aproximation to a "no feedback amp" - as long as you don't use a cathode R (and bias up those "extra" grids like the tube designers intended)
http://www.diyaudio.com/forums/tube...e-internal-negative-feedback.html#post2390464
unless you meant the really serious qestion was triode strapped pentode vs triode? A community dedicated to helping everyone learn the art of audio. Projects by fanatics, for fanatics! - Search Results for triode strapped pentode
although "single ended pentode" is likely a pretty good aproximation to a "no feedback amp" - as long as you don't use a cathode R (and bias up those "extra" grids like the tube designers intended)
http://www.diyaudio.com/forums/tube...e-internal-negative-feedback.html#post2390464
unless you meant the really serious qestion was triode strapped pentode vs triode? A community dedicated to helping everyone learn the art of audio. Projects by fanatics, for fanatics! - Search Results for triode strapped pentode
Last edited:
