Current Feedback Amplifiers, not only a semantic problem?

I am sure that some here will find a way to refute Sergio Franco's, a university professor, fine article. And also find fault with the audio industry selling many dollars worth of CFA amps to lots of people. I will happily build and enjoy one or three, since my ears will enjoy the pleasure of telling my brain, if VFA or CFA is the best they can hear. It beats listening to imaginary amps, schematics and other builds, of esoteric UN-sharable designs and dissertations. I will set my computer to automatically go to the EDN article if I ever click on this thread again.
 
Now, Now Richard ,
You know that's not true.
😉
but i see your point .

mlloyd1

(who has a book case and file cabinets FULL of great materials pointed to by many folks here)

thank you walt. I have said same thing from other sources.... excerpts from books in fact but no one here wants to read them. I have said similar to last paragraph re low internal impedances. Just argue about something they don't know much about. So, I'm with Bonsai on this. Cheers.

THx-RNMarsh
 
Actually, MK's argument that the inverting input of a CFA was not really zero impedance, and therefor it could not be 'just' current feedback because there would always be a voltage, was the only argument I thought was debatable.

However, the very clear and eloquent way this issue is addressed - and resolved! - in part 3 of Franco's article convinced me that it is a non-issue and a non-argument.

Jan
 
Dr Franco should be embarrassed, EDN should've applied real editorial standards

I found the EDN article leading and closing paragraphs disappointing in the attack on the unnamed "they" which is bad rhetoric at best

then the unneeded polemic is cranked to simply unacceptable level by the
some self-proclaimed prophet comes forward trying to discredit decades of well-accepted analysis for what seems to be a desperate search for notoriety:

I guess you have to do something to essentially reprint in 2017 a 90's white paper on "How CFA works" from any of the monolithic op amp manufacturers making them

EDN has long lost editorial competence but allowing such hermetic argumentative framing likely confuses more of the readers than there are readers having trouble understanding "CFA"

If anything Mike's tagging the article with the ref to the audioxpress exchange at least gives context
 
If anything Mike's tagging the article with the ref to the audioxpress exchange at least gives context

For the record there was a large behind the scenes amount of discussion around this, I decided not to be involved by name. The commented quote is out of line and totally unnecessary, such a venomous disagreement about nomenclature, etc. I don't get. I wish people would understand the purpose of writing a data sheet is to simply inform average users and stating things in purely academic terms does NOT help the customer get their job done.
 
Last edited:
I am sure that some here will find a way to refute Sergio Franco's, a university professor

The first reply to M. Kiwanuka's article begins with a list of supporting signatories. This can impress but it can also make some readers to become immediately cautious, fearing the argumention from authority prevails over the pure analysis. It has to be remembered there has been a great authority thirty years ago to not agree with the concept defended today by S. Franco et al.

Seeing the CFA and VFA terms as meaning the effect of feedback on the load at output of the amplifer has been has been used for probably eighty years by now and is still in every day usage. Per se, it is not controversial.

About thirty five years ago, another meaning for these CFA and VFA terms has been projected on the electronics scene, related to the way the feedback is supposed to work inside a kind of input stage of amplifier circuits.

So the first meaning is "well established" since much longer than the second one. It would be fair not to ignore it.

Let's speak now about the second meaning and the the August 23, 2017 text by S. Franco.

The CFA circuit shown is based on an input circuit often called diamond. According to the accepted meaning in the context of the article, a CFA does not require at all to be equiped with such a diamond input. To simplify the analysis, the same function can be carried by a single bipolar (and a Constant Current Source) without the circuit looses its CFA status.

Then you can't escape the physical principle which governs the current across this transistor and therefore determines the behavior of the whole circuit. This principle is the transconductance of the transistor in function of its base-emitter voltage, Vbe for short.

The absence of any consideration to this point in the comments of M. Kiwanuka's text is strange as his arguments are merely based on it... and not disproved.

In single or push-pull CFA inputs stages, the base(s) is the non-inverting input and the emitter(s) is the inverting input, they both present physical values compatible for a subtraction, i.e. voltages. In this interpretation, the fundamental differential process of feedback is respected, which is not the case with the concept of CFA.
 
Last edited:
A current output amplifier is not the same as a current feedback amplifier. How many more times must it be stated, or the canonical feedback forms trotted out and the error of this way of thinking pointed out?

There is no excuse for not understanding this fundamental point. None. Or using it to perpetuate a ridiculous argument about the semantics of the CFA term.

I've been reading Prof Franco's articles in EDN since the very early 1990's. For heavens sake, are we now even trashing the work of a respected, well published, academic in the field simply to perpetuate a falsehood?
 
I found the EDN article leading and closing paragraphs disappointing in the attack on the unnamed "they" which is bad rhetoric at best

then the unneeded polemic is cranked to simply unacceptable level by the

I guess you have to do something to essentially reprint in 2017 a 90's white paper on "How CFA works" from any of the monolithic op amp manufacturers making them

EDN has long lost editorial competence but allowing such hermetic argumentative framing likely confuses more of the readers than there are readers having trouble understanding "CFA"

If anything Mike's tagging the article with the ref to the audioxpress exchange at least gives context






Well the article with its plethora of PSPICE investigation does I think delve a bit deeper than the typical manufacturers application note. I think an inquisitive engineering student who is capable of maintaining a level head whilst reading should find it an informative technical resource, despite perhaps being a bit puzzled over the context of the defensive stance of the opening paragraph(s).


Such a snobbish dismissive in a post decrying a lapse in academic etiquette is a bit ironic, don't you think? And just to convince me that I am not going insane, could you please spell out exactly what is so egregious in the closing paragraphs? :




Closure

In a well-designed CFA application, the designation current feedback is indeed quite appropriate: it indicates the predominant nature of the signal fed back to the summing node, and it also implies the inherently fast current-mode operation of the internal circuitry (exclusive of the gain node).
Using two-port techniques to manipulate a CFA circuit into a series-shunt configuration is a popular alternative for the paper-and-pencil calculation of the loop gain T, but it fails on other vital issues, such as the correct representation of the type of feedback actually taking place. (If we were to bring into the picture also the nonzero output impedance of the output buffer, then the CFA would exhibit feedthrough from input to output through the feedback network and around the gain node; two-port analysis fails to account for this fact [3], [4], whereas return-ratio analysis provides exact results.)
 
yes conflating trashing trash talking by however a respected figure and trashing their technical contribution is a sign that rational argument won't be winning the day

and yes I misnumbered the 2 offending paragraphs - which a competent and empowered Editor should have axed
 
How skinny people is, as soon as an article doesn't correspond to their views.


And how uncommented such offenses like "salesmen language" and worse, thrown to well respected engineers and scientists, are leaved by the same group of people who claim higher moral integrity.


Btw, with or without "offenses" the article from S. Franco is hardly disputable on a technical level.
 
yes conflating trashing trash talking by however a respected figure and trashing their technical contribution is a sign that rational argument won't be winning the day

Okay, I guess I wasn't wearing my rational-thinking cap when I read the quoted (again) sentence as being dismissive of the articles technical value.

I guess you have to do something to essentially reprint in 2017 a 90's white paper on "How CFA works" from any of the monolithic op amp manufacturers making them