I believe this is often the case. A detached, non-emotional study* would make it obvious that seen from the inverting input, that stage works as a common base stage. That means that the current flowing out of its collector, which is the stage output signal, MUST come from its emitter, which is supplied by the feedback network.
*Must admit it took me quite some time to progress to that stage ;-)
Jan
Last edited:
yep.
So many people want to simply replace currents with volts and think it's performance will be the same. They miss the unique characteristics of operating in current-mode.
As you all must know, as this CFA behavior inherently leads to very high speed circuits, the CFA has quickly advanced into RF apps. And with supply voltages dropping to 3.3v and 1.8 v , those circuits are all now current-mode operated.
Clue #1 ---> If you use high impedance circuitry as in VFA to understand CFA, you wont get it. CFA are relatively low Z circuitry. Notice that in the original CFA I published back in the 1970's there was no Z used higher than 1K. and currents were high compared to "opamp" VFA. Low Z values and higher operating currents automatically push the SR and BW higher as device and stray C's are charged quicker.
Also, the gain/BW doesnt change much with 'fb' under Current-Mode operating conditions. The main difference often pointed out.
THx-RNMarsh
You guys are killing me. Hung up on the word Feedback. The amp should be called Current-Mode Amplifier/operation.
THx-RNMarsh
Last edited:
Part of the problem precisely relies there. The feedback network is the load of the output emitters of the input stage which work as voltage followers. Firt of all, these followers impose their emitter voltage to the feedback network.
Then, with a very very very slight delay, due to phase and propagation in the circuit after the input stage, the current comes from the output stage (which is an amplified version of the current coming from the input stage) and begins to develop voltage across the bottom resistor of this network.
This has the effect to reduce the current of the emitter followers which now see a higher impedance load. We can also say that the enhanced voltage across the bottom resistor due to the current from the output has the effect to reduce the Vbe of the emitter followers of the input stage.
The Vbes control the current in the transistors of the input stage first, and then in the whole circuit.
A transistor is a device with a differential input, electrodes B and E, the difference of potential between them controls the current across it. It has been said in the first discussions about CFA on this forum.
I already asked the question : is the connection of the feedback network to the amplifier output changes the mode of operation of the input transitors ? I concede that, for long, I considered input active devices in feedback circuits as working in different modes of operation at the same time. I slightly changes my mind with time.
In donning the mantle of authority to be the arbiter of the "real" question, you employ a great deal of words to ignore a very simple and straight forward one: what source is the input stage current being fed back from?
I agree with Richard here - Current mode amplification/operation is a good term. It's applicable not only to the feedback in so-called CFA designs but also to certain current-driven stages with the modern (and not so modern) amplification approaches. Current mode amplification attracts me a lot during the last few years, providing excellent open loop performance - a combination of high linearity, wide bandwidth, very good phase response.
I feel your pain. Good luck on getting everyone to adopt new nomenclature.
But as Jan and I agreed in recent posts, this is not really about nomenclature. Its about the fact that some folks just don't seem to believe, or at least are not willing to admit, that current sourced from an amplifier output is fed back to and flows through the input stages of what the semi industry calls a CFA.
If this were solely about nomenclature, I for one wouldn't be here.
The model of bipolar transistors is known as having a lot of parameters. For the explanation of audio circuits, the leading one is usually considered, notably by very famous audio engineers, to be the base-emitter voltage as being the main controlling parameter of the curent across it. Currently I think my approach of CFA is very clear and currently does not encounter objection.
I am using 'current mode topology' (because some people hate the idea that an amplifier has feedback - but that's a different discussion). But does it really matter what you call it? Its just an amp that works a bit differently that what folks are mostly used to (VFA). What is important is that people understand it intuitively, feel comfortable working with it and can make the right trade off's when designing it.
Mark Whitney called it "near instantaneous error correction" or words to that effect
I don't think it matters what you call it. But does refusing to call it something mean that one doesn't understand it intuitively? After all, if it's just about a name, why are all of these long, detailed technical objections being raised?
As far as I can analyze, the input stage of a CFA delivers current in the feedback network as does, in greatest quantity, the output stage.
Am I absolutely wrong if I see these currents going in the same direction and add rather than subtract to give an error current as the Renesas article, recently mentionned, states ?
https://www.renesas.com/us/en/www/doc/application-note/an1993.pdf
Note : the error current exists. At the collectors of the input stage. In both C and V feedback topologies. It passes almost intact through the emitters of the input devices. If that was be sufficient to make a CFA, a VFA would be a CFA that ignores itself.
The current coming from the amp output has the effect to decrease the current coming from the input stage, with a flavour of bootstrap analysed and not refuted post #698.
The current coming from the input stage is under the control of a loop the core of which is the transconductance of the active devices of the input stage.
The primary difference between a CFA and a VFA relies in the fact the current of the active devices controlling the loop passes through the feedback network in a CFA, it does not in a VFA. It has consequences on the performances of the circuits which are duly analyzed in numerous articles but that's off topic.
The inescapable principle in feedback is transconductance, current controlled by voltage.
Last edited:
One could argue that it delivers, accepts, or does both.
Can we agree that for a CFA configured for unity voltage gain (no Rg) that the input stage collector (non-bias) current is substantially the same current that flows (exclusive of any load) from (or into, if you prefer) the output stage?
And if we add an Rg, that the collector current is the same current as some fraction of that which flows from/into the output stage?
And that if we restrict ourselves to non-zero and non-infinite impedances, no current or voltage exists without giving rise to the other, and so if a phenomenon can be explained by referring to one, it can also be explained by referring to the other?
In light of this, what forces one to reject as technically incorrect the claim that current is fed from the output stage back to and through the input stage (current feedback)?
And finally, given that feedback from output to input must occur, and that one must employ transconductance to invoke voltage feedback, and in recognition of Occam's razor, isn't dealing with voltage a needlessly more complex alternative?
I feel your pain. Good luck on getting everyone to adopt new nomenclature.
But as Jan and I agreed in recent posts, this is not really about nomenclature. Its about the fact that some folks just don't seem to believe, or at least are not willing to admit, that current sourced from an amplifier output is fed back to and flows through the input stages of what the semi industry calls a CFA.
If this were solely about nomenclature, I for one wouldn't be here.
My point had nothing at all to do with renaming....... knowing the principle characteristics of current-mode amps. The use of the word FB has got people off the track.
In past episodes here about CMA or CFA - I have given literature sources which clearly explain the subject. I have seen APP engineers and Notes distort the issues to dumb down the subject and make it easy to use/apply.
But the facts are there are many books correctly discussing the subject and taking the concept up thru RF..... so why is this a question and debate here at all? It's done.
THx-RNMarsh
Last edited:
BTW --- The topology and operation was not called CFA... it was called as I said -- current mode amplifier. Somewhere along the way, I suspect an App Engineer changed it in his company's literature to FB to simplify the use and acceptance.
https://ieeexplore.ieee.org/document/100661
I will continue to call it CMA.
THx-RNMarsh
https://ieeexplore.ieee.org/document/100661
I will continue to call it CMA.
THx-RNMarsh
Call it as you wish.
The IEEE reference you cite dates from 1989. U.S. Patent US4502020A dates from 1983. its abstract begins, "A wide-band direct-coupled transistor amplifier exhibits greatly improved settling time characteristics as the result of circuitry permitting the use of current feedback rather than voltage feedback..."
The phrase "current feedback" appears not only further on in the description of the circuit, but throughout the patent claims.
Do you believe that it is technically inaccurate to claim that the CFA employs current feedback?
The IEEE reference you cite dates from 1989. U.S. Patent US4502020A dates from 1983. its abstract begins, "A wide-band direct-coupled transistor amplifier exhibits greatly improved settling time characteristics as the result of circuitry permitting the use of current feedback rather than voltage feedback..."
The phrase "current feedback" appears not only further on in the description of the circuit, but throughout the patent claims.
Do you believe that it is technically inaccurate to claim that the CFA employs current feedback?
Yes. so?
The title is under Current Mode Amplifiers. The proper term for the amplifier talked about here. Any detail like fb or diamond input, current conveyer etc etc is not the point. The amplifier is called a CMA and is not called a CFA. Why not call this type amplifier what IEEE calls it?
Pls note the characteristic performance difference of the CMA to a VFA mentioned in the first two sentences. This is even before any fb applied.
THx-RNMarsh
Last edited:
hey, I was just reading about 2GHz mobile phone CMCD amps. Current-Mode Class D amplifers. No current feedback.
The main thing here should be how a current mode amplifier works, first. Then add fb to it. Seems it was backwards here.... trying to figure how the fb works, first.
Ok. Hope we are back on track with the old 'new' topology for audio amps. Bonsai has got it. Other's too.
It has many benefits for audio.
-RNM
The main thing here should be how a current mode amplifier works, first. Then add fb to it. Seems it was backwards here.... trying to figure how the fb works, first.
Ok. Hope we are back on track with the old 'new' topology for audio amps. Bonsai has got it. Other's too.
It has many benefits for audio.
-RNM
Last edited: