In this article http://www.chipcenter.com/analog/c014.htm (Are Op Amps really linear?) the great Barrie Gilbert mentions an AFA - active feedback amplifier:
"...Later, another fundamental building block, which I call the Active Feedback Amplifier, or AFA, will be discussed. In my opinion, this structure, which, unlike the OPA (vfb opamp), has high open-loop linearity and excellent closed-loop linearity due to its distortion-canceling topology, and a very high degree of versatility arising from its dual fully-differential inputs, has the potential to eclipse the OPA in all applications involving the manipulation of purely voltage-mode signals. Being a super-set, it can do anything that a conventional op amp can do, plus a whole lot more."
Well, if that doesn't sound promising...
However, I can't seem to find this later column he is referring to. He mentions that he will discuss the AFA in a later column, in two more of his articles (All of his articles are great read - recommended). But a google search on this AFA thing did not help at all. Has anybody any info on this? Or seen anything published?
Nothing's new under the sun (in amlifier electronics at least), so there must be some implementation of this, obviously under a different name... (I guess Barrie Gilbert loves to invent names!). Any info would be greatly appreciated!
Regards,
Kyriakos Georgiadis
"...Later, another fundamental building block, which I call the Active Feedback Amplifier, or AFA, will be discussed. In my opinion, this structure, which, unlike the OPA (vfb opamp), has high open-loop linearity and excellent closed-loop linearity due to its distortion-canceling topology, and a very high degree of versatility arising from its dual fully-differential inputs, has the potential to eclipse the OPA in all applications involving the manipulation of purely voltage-mode signals. Being a super-set, it can do anything that a conventional op amp can do, plus a whole lot more."
Well, if that doesn't sound promising...
However, I can't seem to find this later column he is referring to. He mentions that he will discuss the AFA in a later column, in two more of his articles (All of his articles are great read - recommended). But a google search on this AFA thing did not help at all. Has anybody any info on this? Or seen anything published?
Nothing's new under the sun (in amlifier electronics at least), so there must be some implementation of this, obviously under a different name... (I guess Barrie Gilbert loves to invent names!). Any info would be greatly appreciated!
Regards,
Kyriakos Georgiadis
Zombie Thread!!1!!! 
OMG! Zombies!!!!11!
I just found this doc ... and was gonna post ... but this thread already exists ... from way, way back!
There is an interesting trend here: operational amplifiers started out (back in analog computing days) with a totally different set of objectives to those surrounding amplifiers for ac amplification with low distortion. In the 'op amp paradigm,' the idea was to make the open-loop gain so very high that the function is determined solely by the external components. However, as we've seen, that is far from the case for a typical OPA operating at moderate gains and frequencies. In the second design style, typified by audio power amplifiers, the objective is to achieve almost perfect linearity without feedback, and then use a small amount of feedback to squeeze out the last gram of linearity. Designers of modern high-performance op amps now understand the critical importance of pursuing the latter paradigm in all high-frequency applications. This equates almost completely to achieving extremely high slew rates; value of over 5,000 V/ms are nowadays available in monolithic OPAs.
In another approach to improving HF linearity, the gm input stage is replaced by a current-conveyor, which is another type of Class-AB cell. (In fact, high-slew OPAs often used what is effectively two such cells in a differential arrangement.) A current conveyor is the starting point of a current-feedback amplifier, also called a transimpedance amplifier (TZA). This type of amplifier can, in principle, be designed to be totally free of slew-rate limitations, one of its main attractions. Another is that, unlike an OPA, in which the so-called 'virtual ground' or 'summing node' exhibits a low impedance only by the action of global feedback, becoming very high at frequencies close to w1, the TZA provides a low summing-node impedance (ohms) even with the feedback completely removed, from dc to many hundreds of megahertz.
This useful and interesting special-purpose amplifier topology will be the subject of my next column. Later, another fundamental building block, which I call the Active Feedback Amplifier, or AFA, will be discussed. In my opinion, this structure, which, unlike the OPA, has high open-loop linearity and excellent closed-loop linearity due to its distortion-canceling topology, and a very high degree of versatility arising from its dual fully-differential inputs, has the potential to eclipse the OPA in all applications involving the manipulation of purely voltagemode signals. Being a super-set, it can do anything that a conventional op amp
can do, plus a whole lot more.
Surely he's talking about more than just current FB opamps?
OMG! Zombies!!!!11!
I just found this doc ... and was gonna post ... but this thread already exists ... from way, way back!
There is an interesting trend here: operational amplifiers started out (back in analog computing days) with a totally different set of objectives to those surrounding amplifiers for ac amplification with low distortion. In the 'op amp paradigm,' the idea was to make the open-loop gain so very high that the function is determined solely by the external components. However, as we've seen, that is far from the case for a typical OPA operating at moderate gains and frequencies. In the second design style, typified by audio power amplifiers, the objective is to achieve almost perfect linearity without feedback, and then use a small amount of feedback to squeeze out the last gram of linearity. Designers of modern high-performance op amps now understand the critical importance of pursuing the latter paradigm in all high-frequency applications. This equates almost completely to achieving extremely high slew rates; value of over 5,000 V/ms are nowadays available in monolithic OPAs.
In another approach to improving HF linearity, the gm input stage is replaced by a current-conveyor, which is another type of Class-AB cell. (In fact, high-slew OPAs often used what is effectively two such cells in a differential arrangement.) A current conveyor is the starting point of a current-feedback amplifier, also called a transimpedance amplifier (TZA). This type of amplifier can, in principle, be designed to be totally free of slew-rate limitations, one of its main attractions. Another is that, unlike an OPA, in which the so-called 'virtual ground' or 'summing node' exhibits a low impedance only by the action of global feedback, becoming very high at frequencies close to w1, the TZA provides a low summing-node impedance (ohms) even with the feedback completely removed, from dc to many hundreds of megahertz.
This useful and interesting special-purpose amplifier topology will be the subject of my next column. Later, another fundamental building block, which I call the Active Feedback Amplifier, or AFA, will be discussed. In my opinion, this structure, which, unlike the OPA, has high open-loop linearity and excellent closed-loop linearity due to its distortion-canceling topology, and a very high degree of versatility arising from its dual fully-differential inputs, has the potential to eclipse the OPA in all applications involving the manipulation of purely voltagemode signals. Being a super-set, it can do anything that a conventional op amp
can do, plus a whole lot more.
Surely he's talking about more than just current FB opamps?
Yeah he's not talking about CFB amps, rather a completely different front-end topology for an opamp. I'm in love with these AFAs at present, they do sound so much better than traditional LTP-fronted opamps 
They do measure somewhat worse than the top-of-the-range opamps though - so much for THD measurements eh?
They do measure somewhat worse than the top-of-the-range opamps though - so much for THD measurements eh? 
Hey abraxalito!I've been searching and found all your good info and blogs etc...
More people should be trying these out!
Problem is, there's many threads each with a bit of info and they may not be easy to search for.
I really think we need one thread that references them all and brings it all together in a coherent, easy-to-find thread, something like:
"Gilbert's chips, AFA (active feedback amplifier): AD830, LT6552, TDA856X"
I'm good with info: all the spec sheets, part schematics from the specs, graphs, technical summary, suggested circuits. This would really get people going!
But I don't want to steal any of your thunder, you're doing a fantastic job!
What do you think?
PS Are there anymore AFA chips we could add to the list in the thread title (hard to edit later)
Last edited:
I reckon there's more than enough thunder to go around, so why not start the thread yourself? Then I'll chip in if I find some new stuff out or if you have specific questions. I agree these chips should be more widely known, but then if too many people find out they might be harder to find on the second-hand market (which is where I get a lot of my supplies). Many of the really interesting NXP/Philips parts have been obsolete a while, but who knows if they undergo a renaissance perhaps NXP will bring them out of retirement? That would be a win all round
Anyway, glad that you're a 'convert' now
Anyway, glad that you're a 'convert' now
I'm on it! Sweeeeet
Any other part numbers (or 'key' words) I should reference in the thread title?
hmmm, maybe a section for power amp chips and then one for line level chips. Any ideas?
For me it's not so much whether a part is listed as obsolete. If there's a few dozen actually around still, then us DIYers are good-to-go!
True, it would mean less chips for us but spreadin' good sound and a new (hopefully more successful) way of looking at circuits (high freq issues in our cellular/wifi world) is just toooooo important ...
Any other part numbers (or 'key' words) I should reference in the thread title?
hmmm, maybe a section for power amp chips and then one for line level chips. Any ideas?
For me it's not so much whether a part is listed as obsolete. If there's a few dozen actually around still, then us DIYers are good-to-go!
True, it would mean less chips for us but spreadin' good sound and a new (hopefully more successful) way of looking at circuits (high freq issues in our cellular/wifi world) is just toooooo important ...
Last edited:
I'd not reference them in the title, rather in the first post - that's one that can be edited by the thread starter, so you can add more if we find them. The title can just be 'Audio applications of Barrie Gilbert's AFA' or something like that.
Hard to make a hard-and-fast distinction between line level and power amp - there are a few headphone amps for example, and also at least one power amp which has a line driver mode selectable by a pin. I'd lump them all in together and let the readers sort out which they want.
Hard to make a hard-and-fast distinction between line level and power amp - there are a few headphone amps for example, and also at least one power amp which has a line driver mode selectable by a pin. I'd lump them all in together and let the readers sort out which they want.
Found this
IEEE Xplore Abstract - The active feedback amplifier. A versatile analog building block
Brunner, E.; Barrie Gilbert,, "The active feedback amplifier. A versatile analog building block," Northcon/94 Conference Record , vol., no., pp.131,136, 11-13 Oct 1994
doi: 10.1109/NORTHC.1994.643325
Abstract: The active-feedback amplifier (AFA) topology approaches the ideal voltage-controlled voltage source (VCVS) with distinct input and output common-mode levels. This topology is very useful whenever it is required to sense a differential voltage at the input whose common-mode level is different to that at the output. With the AFA no ambiguity about the input and output grounds arises; compare this to op-amp based applications where one always has to decide where the 'ground' should be referenced to - input or output? Furthermore, when the op-amp is viewed as an integrator and not as an infinite-gain, infinite bandwidth block, then at frequencies not too remote from the unity-gain frequency the op-amp will be nothing like the infinite-gain block it is supposed to be. The AFA in contrast provides accurate performance right up to its unity-gain frequency. This paper explains these benefits
keywords: {feedback amplifiers;active feedback amplifier;analog building block;common-mode level;differential voltage;unity-gain frequency;Broadband amplifiers;Differential amplifiers;Drives;Feedback amplifiers;Frequency;Instruments;Operational amplifiers;Signal processing;Topology;Voltage},
URL: IEEE Xplore Abstract - The active feedback amplifier. A versatile analog building block
IEEE Xplore Abstract - The active feedback amplifier. A versatile analog building block
Brunner, E.; Barrie Gilbert,, "The active feedback amplifier. A versatile analog building block," Northcon/94 Conference Record , vol., no., pp.131,136, 11-13 Oct 1994
doi: 10.1109/NORTHC.1994.643325
Abstract: The active-feedback amplifier (AFA) topology approaches the ideal voltage-controlled voltage source (VCVS) with distinct input and output common-mode levels. This topology is very useful whenever it is required to sense a differential voltage at the input whose common-mode level is different to that at the output. With the AFA no ambiguity about the input and output grounds arises; compare this to op-amp based applications where one always has to decide where the 'ground' should be referenced to - input or output? Furthermore, when the op-amp is viewed as an integrator and not as an infinite-gain, infinite bandwidth block, then at frequencies not too remote from the unity-gain frequency the op-amp will be nothing like the infinite-gain block it is supposed to be. The AFA in contrast provides accurate performance right up to its unity-gain frequency. This paper explains these benefits
keywords: {feedback amplifiers;active feedback amplifier;analog building block;common-mode level;differential voltage;unity-gain frequency;Broadband amplifiers;Differential amplifiers;Drives;Feedback amplifiers;Frequency;Instruments;Operational amplifiers;Signal processing;Topology;Voltage},
URL: IEEE Xplore Abstract - The active feedback amplifier. A versatile analog building block
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.