Thank you. You answer is absolutely counterconstructive regarding your wish :
Precise references to posts of this thread would be very useful to establish an overview of the topic.
Revealing untold truths by tautologous circularity, vanishing derivatives, indefinite integration and algorithmic randomness of infinite sequences.
Functional analysis is fraudulently defined as a mathematical model to find the response of a physical system to a given excitation. It is based on the bold assumption that complex system response can be given by a linear superposition of logical constants, under ideal conditions and at a minimum of functionals, that is, absolute, uneffected states can be added together and the result will be another valid state.
Now, it is not possible to express a response mathematically, response defined as a reaction to influence. It is, to put it mildly, frivolous to call fixed-valued relationships a response. The "stimulus" does not stimulate anything, does not act as a loading and does not excite any response; there is no receptive mechanism for influence and no evaluation mechanism for response artifacts. Distortion is attributed to linear elements having linear relationships, the analytical system "responding" exactly the same way to all "impulses". The independent "dynamic entries" (coefficients) must be chosen arbitrarily, calculated separately and inserted manually in a veritable static fashion.
Mathematics can only handle resolved, settled relationships with no connection to anything external, not producing more information than what is implicitly contained in the independently applied initial data and premises. The same amounts are shifted back and forth. A state never evolves to another state. The arbitrarily chosen values are not the attribute of electrical quantities. There are not any new "properties" to derive from self-contained data, the repetitive tautologies do not give any factually new information but psychologically new information. The measured outcome (aggregate statistical mean or expectation value) is predetermined. Events are not representable. Electrical quantities are not properties of physical things or events. Functional analysis is an investigation of mathematical properties.
Functional analysis does not describe behavior. Functional analysis does not measure performance. Functional analysis does not predict anything. The prophetic faculty consists in telling most enigmatically how the terms are related to themselves definitionally by the unsexy methodology of linear probabilistic approximation.
Although linear theory cannot deal with and explain nonlinear phenomena, the empirical data based arguments (propositional variables) associated with a function and determining its value ensure good simulation matches with anticipated behavior and established beliefs. What is computed is the expectation whereby the self-fulfilling prophecy as dull statistical average comes true. The premises, assertions and assumptions applied in the form of preexisting values make the measurements "realistic". Otherwise it is necessary to go back and suitably modify the assumptions. Even though the outcome is nothing but arrant nonsense, as long as it meets the expectations why care?
Functional analysis is fraudulently defined as a mathematical model to find the response of a physical system to a given excitation. It is based on the bold assumption that complex system response can be given by a linear superposition of logical constants, under ideal conditions and at a minimum of functionals, that is, absolute, uneffected states can be added together and the result will be another valid state.
Now, it is not possible to express a response mathematically, response defined as a reaction to influence. It is, to put it mildly, frivolous to call fixed-valued relationships a response. The "stimulus" does not stimulate anything, does not act as a loading and does not excite any response; there is no receptive mechanism for influence and no evaluation mechanism for response artifacts. Distortion is attributed to linear elements having linear relationships, the analytical system "responding" exactly the same way to all "impulses". The independent "dynamic entries" (coefficients) must be chosen arbitrarily, calculated separately and inserted manually in a veritable static fashion.
Mathematics can only handle resolved, settled relationships with no connection to anything external, not producing more information than what is implicitly contained in the independently applied initial data and premises. The same amounts are shifted back and forth. A state never evolves to another state. The arbitrarily chosen values are not the attribute of electrical quantities. There are not any new "properties" to derive from self-contained data, the repetitive tautologies do not give any factually new information but psychologically new information. The measured outcome (aggregate statistical mean or expectation value) is predetermined. Events are not representable. Electrical quantities are not properties of physical things or events. Functional analysis is an investigation of mathematical properties.
Functional analysis does not describe behavior. Functional analysis does not measure performance. Functional analysis does not predict anything. The prophetic faculty consists in telling most enigmatically how the terms are related to themselves definitionally by the unsexy methodology of linear probabilistic approximation.
Although linear theory cannot deal with and explain nonlinear phenomena, the empirical data based arguments (propositional variables) associated with a function and determining its value ensure good simulation matches with anticipated behavior and established beliefs. What is computed is the expectation whereby the self-fulfilling prophecy as dull statistical average comes true. The premises, assertions and assumptions applied in the form of preexisting values make the measurements "realistic". Otherwise it is necessary to go back and suitably modify the assumptions. Even though the outcome is nothing but arrant nonsense, as long as it meets the expectations why care?
We have spent the last few hundred pages trying to to sort this out.
The divide between the accepted engineering principles of CFA operation and your single dimensional view on how they operate is too wide to bridge at this juncture.
Of the manufacturing designers present here, some of you have released CFA designs in addition to VFA designs.
What prompted you to to utilise CFA typology, and what advantages do CFA designs have in the speaker cable/loudspeaker real world.
RNM, you have commented that you prefer VFA for bass, CFA for mids (something like that)...can you elaborate please.
Dan.
What prompted you to to utilise CFA typology, and what advantages do CFA designs have in the speaker cable/loudspeaker real world.
RNM, you have commented that you prefer VFA for bass, CFA for mids (something like that)...can you elaborate please.
Dan.
Last edited:
For most participants of the thread accepted engineering principles of CFA is a sad joke, just as it was thirty five years ago for the greatest names of the audio engineering (by the way, two of them have been published by a CFA defender !). There is a large consensus :
- that, even among the defenders of the concept, CFA is badly named. In this context, your attempt to restrict the label of CFA to push-pull inputs makes your analysis more than strange.
- that the intepretation of the intimate mechanism suggested by the current feedback expression is wrong. This mechanism firstly involves voltage and transconductance.
Concerning my single dimensional view, the total lack of arguments to support the criticism it received and many posts here had the unilateral effect to reinforce it.
Note :
- nobody had his mind changed by the thread. The art of convincing is decidely not on the CFA's side.
- the fundamental fact, which escaped to many, that the low impedance of in(-) does not load the feedback network has been largely highlighted by me.
I’m anybody. There is a premise in your question, that there are consistent subjective differences between all instances of CFA and VFA circuits. There is a further problem of defining the boundaries of what a VFA and CFA are. I suggest the premise is false and therefore there is no answer to your question.
Forr,
You have certainly opened a pot of worms when starting this CFA thread.
At the same time you were well aware of the fact that there was a great controversy between the many links you gave in posting #1, not in the least also in the academic world.
So far it was a very vivid discussion, but the chance that "Audio Mortals" could come to a united vision was very small indeed right from the start.
In one of your links I read this interesting comment:
regarding the Truth about CFA,
I sense an "either" vs "or" goal.
Either "Voltage" or"Current" as the controlling entity.
From Philosophy, I have come to see the possibility
that neither inifinity nor zero actually exist.
From Math, I have come to see that
we have "agreed" on methods of handling infinity and zero.
In Practice I have come to see that
both Voltage "and"Current are always present ( OHMs Law ),
so "both" are always parts of a controlling entity.
What makes this comment interesting is that it clearly shows that you can look at the subject from at least 3 different perspectives, all using their own language.
A Babylonian confusion as the outcome is guaranteed.
Your conclusion on the other hand that nobody changed his mind, comes out of the blue.
It has certainly made may people more aware of what's going on in a so called CFA and for those who didn't have a firm opinion, I assume they look at it now in a more differentiated way.
I think what is undisputed is that a CFA has a clever engineered topology of its own in which transimpedance is one of the key factors.
And therefore this topology has been given a name, and as long as everybody understands what being meant with this name, it serves its purpose well.
Outside a magnetic field, no current can flow without first applying a voltage difference. So I can only repeat the above comment that both are always part of a controlling entity.
Whether in a feedback situation the voltage is steering the current or vice versa is to me the chicken and the egg problem, who was first.
So don't put everybody camps. Many of us see the dualism and accept that.
And that the art of convincing is "not on the CFA's side" is therefore an unjustified conclusion and can just as well be accounted to the way you have failed to convince people of your vision.
I'm sorry to say that.
Hans
I ask for real world experiences from which can be drawn trends.....perhaps.
RNM indicates preference for different topologies for different applications, it would be interesting to hear the experiences of others.
Of course the schematics upholding the individuals subjective findings would be useful, and fine examination of these schematics is required to help to understand the basis of these findings/preferences.
My understanding of the general definition of CFA (Current Feedback Amplifier) is that the return/feedback signal (significant current) acts directly at a point in the amplifier input/driver stage causing modulation of the gain of that stage (by significant current injection into emitter circuit)
VFA operation generally depends on a input differential pair to derive the drive signal, and could be described as a secondary correction as opposed to CFA direct/immediate correction.
This difference in fundamental operation will alter intrinsic/excess noise production and properties/consequences.
I expect these noise behaviours are likely the basis for preferences.
Anyone else have real world experience/findings ?.
Dan.
Last edited:
N101N is my new Hero.
Funny, I would have thought that the Electric Universe theory would have appealed to Electrical and Electronic Engineers...
It certainly appeals to me, since it predicts with accuracy some facts that the "accepted" theories were not capable to predict... Cheers,
M.
There seems to be an impossibility to clearly define by a few words what is the Current Feedback concept in a CFA
The idea of a predominant factor has been introduced here. If there is one at the input stage in a CFA, it can only be a differential voltage at least at small signals and low frequencies.
A voltage can exist without current. At normal temperatures, a current can't exist without voltage.
The basic effet of current reduction in the input stage by the application of feedback is as fundemantal as often forgotten.
So the thread can be as recognized as of some use.
The topology of current conveyors, diamond inputs or even simple complementary inputs belong to the CFA category, they do not define what is CFA (Bonsai's confusion on this point).
As far as I analyse, the transimpedance effect, conversion current to voltage is not done in the input stage but at its output, after the current mirrors in a current conveyor.
Everybody understands what a CFA does, the same as a VFA, But persons who understand how a CFA works are more rare.
Do not worry, "my" vision has clarified facts for some people. Would it be invalid, it would have been be torn down since long. It is not the case.
Can you give audio/RF relevant examples of these predictions please.N101N is my new Hero.
Funny, I would have thought that the Electric Universe theory would have appealed to Electrical and Electronic Engineers...
Cheers,
M.
Dan.
N101N is my new Hero.
Very astute Max, the reason I mentioned it (Electric Universe) is that if you investigate they believe in almost complete empiricism and that using math to predict results and design experiments is a fatal bias. They could use a little better knowledge of statistics.
As voltage exist without current, the problem of who was the first between the egg and the chicken has been solved.
The first was the egg, the animal that laid the egg was not yet a chicken.
Rationalism seems but a subgroup of empiricism... though I think myself more the skeptic.
I hope your ears are better at listening than your cortex.
242 pages of this.
Ok, I suppose, if it keeps people occupied.
Hi Forr
I think it would be a good idea if everyone could discuss the same circuit. So the first thing would be to agree on a common circuit.
You insist on using a single transistor input stage and the feedback connected to the emitter, this is of course a CFA, but as you know it's quite easy to convert your prefered IPS into a regular diamond IPS.
In my view it would be better if we agreed using a diamond input stage and a diamond output stage and current mirrors.
Another thing that bothers me is that some here (not you Forr) are analysing "meaningless circuits" that has nothing to do about anything, maybe it's a language problem, but to be honest it's a long time since I have seen so much nonsense.
So if everyone could agree on discussing the same circuit and that it would be a circuit that would actually work it would be a "giant leap for mankind" it is'nt rocket science.
The next question woud be:
Why do you think that the low Z inverting input would load the feedback network?
Than what to your opinion is "the CFA category" ?
What the CFA defines is it's unique and clever topology.
Why is it important where the transimpedance is located?
Is there a transimpedance from inputs to output ? In this test you can clearly see there is, thereby separating a CFA from a VFA.
Current Feedback Amplifiers, not only a semantic problem?
Are you saying that a CFA does the same as a VFA.
That would be like saying that a Deux Chevaux is doing the same as a Greyhound Bus.
Both can transport people, have wheels, an engine and a steering wheel.
Everybody on this thread knows the difference between a CFA, a VFA and a H bridged VFA and all do very well understand what can be expected from the different topologies.
That being the case, does it really matters whether there is a current or a voltage feedback taking place, no it does not the least, so what the heck.
Hans
Hi Reodor, thanks for your interest.
I prefer a circuit using a single transistor with the emitter loaded by a CCS. DC offset and quiescent current are then easily set and there is only one path to take care of. Whatever the circuit, the analysis of the principe which is the topic of this thread is the same.
I agree but it is to the moderators to judge if it is trolling or not.
Many people think the inverting input loads the feedback network, so the feedback should be called of voltage of current according to the ration of the impedance. This is incorrect. With typical values of network resistors (R to ground = 1 kOhm or less), the inverting input acts as a voltage input.
Your question can be turned: why the low Z inverting input does not load the feedback network ? To me, it is a fundamental question, it made me to come back to this thread :
Current Feedback Amplifiers, not only a semantic problem?
The negative feedback loop has the strange property to bootstrap the load of the inverting input with the effect of decreasing its AC current amplitude. Otherwise said, there is a hidden positive feedback in the loop.
Tubes amplifiers with feedback also belong to the CFA category and exist since eighty years. Once again, I ask to not restrict CFAs to modern topologies.
I am not aware of amplifier inputs working with active devices having not transconductance as main characteristic. A real current inverting input (belonging to a real CFA) should be a transimpedance circuit needing a current as feedback signal, this has been explained by Herve sometime ago. Looking at "documentation" about CFA's, the term transimpedance related to input stages appears. Why ?
Simplified circuits are more explicit.
It matters because it is the topic of this thread : what meaning can be given to CFA ? From what I see, there are quite a lot of persons interested in an essential and surprising aspect of amplifier input stages.
Last edited: