Who cares whether it is current or voltage.......
So-called "current-feedback" op-amps have the distinction of maintaining their bandwidth regardless of gain. More or less........(yeah, it will vay some at the extremes).
And since we have P-A's attention, it seems appropriate to point out......again.........that I think that they sound worse than so-called "voltage feedback" amps.
Why? I have no idea. I stopped worrying why, and made my life easier by not using either. And yes, I liked the results better.
Why? Again, I have no idea, and stopped worrying why. The point is to make money in this silly business, so I took the easy way out. Running a business is hard enough without the mental angst of wondering what something sounds like, and why. I do not begrudge any designer that takes the nested feedback loop appraoch. If they can make it work, and it sells, then that is all that counts in the end.
But I don't need no steenking Bode plots or simulators. [joke]
Jocko
So-called "current-feedback" op-amps have the distinction of maintaining their bandwidth regardless of gain. More or less........(yeah, it will vay some at the extremes).
And since we have P-A's attention, it seems appropriate to point out......again.........that I think that they sound worse than so-called "voltage feedback" amps.
Why? I have no idea. I stopped worrying why, and made my life easier by not using either. And yes, I liked the results better.
Why? Again, I have no idea, and stopped worrying why. The point is to make money in this silly business, so I took the easy way out. Running a business is hard enough without the mental angst of wondering what something sounds like, and why. I do not begrudge any designer that takes the nested feedback loop appraoch. If they can make it work, and it sells, then that is all that counts in the end.
But I don't need no steenking Bode plots or simulators. [joke]
Jocko
I couldn't manage without SPICE! It doesn't get you all the way there, but it helps a lot.
Just a word of warning for anyone using a CFB input stage based on taking signals from the power pins of an op-amp (like the Alexander power-amp does).
Many manufacturers' SPICE models do not model the dynamic current taken through the power pins accurately. Many op-amp SPICE models don't model the device power at all!
Just a word of warning for anyone using a CFB input stage based on taking signals from the power pins of an op-amp (like the Alexander power-amp does).
Many manufacturers' SPICE models do not model the dynamic current taken through the power pins accurately. Many op-amp SPICE models don't model the device power at all!
janneman said:
Thanks for this Jan.....
Like you and Baxandal, i think this 'current-feedback' stuff is one of the most unfortunate misnomers in linear electronics...
Take a non-inverting 'current feedback' configuration for instance.
One should find immediately that this is infact ordinary shunt (voltage) derived, series (voltage) applied negative feedback.
The key difference being that with so-called 'current feedback', the feedback voltage is applied to directly to the emitter of the subtracting transistor.....
...while with the usual diff. input stage, the feedback voltage is buffered from the emitter of the subtracting transistor by an emitter follower...
mikeks said:
while it is how "current feedback" is used in most audio amplifiers, not all cfb amps require the feedback (voltage) to be applied to the emitter of the subtracting transistors. In fact, a lot of cfb opamps for example use differential input pairs.
as you have correctly pointed out, both cfb and vfb amps feedback a "voltage" signal. the difference is in the internals of the amps.
a cfb amp's input stage generates an error current, which is amplified through a transimpedence stage and then buffered out. a vfb amp, on the other hand, generates an error voltage, which is amplified by a voltage stage and then buffered out.
perander's links give quite extensive difference between the two. and Mr. Homo is right that the gain / brandwidth of a current feedback amp is quite independent.
as to their sound, well, I liked my jlh very much,
mikeks said:
Is is just me, or is this guy on a one man crusade to redefine industry standard concepts and definitions?
CFB amplifiers are variable bandwidth as a function of the impedance of the feedback loop- the lower the feedback loop impedance, the greater the current feedback into the input stage, and the greater the net amplifer bandwidth.
Voltage feedback amplifiers don't work anything like this....
~Jon
JonMarsh said:
Jon,
If it is a crusade, it's more than one man's. Nobody disputes that the bandwidth thing is the result of the low impedance at the feedback point. What IS disputed is to call it therefore a CFB amp. There is a fundamental difference in pov. For me, VFB means that the effective input voltage is build from the difference of the system input signal and the feedback signal, via voltage subtraction.
Imagine an amp with a single input transistor with the input signal at the base and the feedback signal at the emitter. The effective input signal is then the Vbe, which is obtained through voltage subtraction: Vb - Ve. Now, of course the emitter being a low impedance point, one can say, hey, there's current going into the emitter, so let's call it CFB! That's rather arbitrary, because there is also going current into the base, after all. By that definition, a standard long-tailed pair with the signal at one base and the feedback signal at the other base could also be called CFB.
It is very unfortunate, because there IS a perfect definition for CFB, which is that you obtain the effective input signal by subtracting a feedback current from an input current. That's what has been in all te textbooks until some marketing guy started using it in a different way.
But, I realise, logic has not much to do with it, if all of you want to stick to this marketing term, maybe I should end at least my part of the crusade....
Jan Didden
peranders said:
Hi, peranders, here is what I found out on the net about the JLH1969 design (which used a single transistor input stage configured in a current feedback set-up):
============paraphrased=====================
the feedback is applied to the emitter of the input transistor, and is referred to as current feedback. Why? the feedback node is of very low impedence.
==========================================
You think the author of the above doesn't know about amp design? or basic EE terminology? or you have noticed that s/he has absolutely no idea what's all about?
Just curious of your take on this.
mikeks said:
Let's just pretend that your arguments were technically valid. Would you then propose to correct all of the following?
peranders said:
And what about these "tons of documents"?
peranders said:
And that's assuming your arguments had any merit whatsoever. It seems that you believe the world revolves around you and your definitions, and that you know more about solid state design and nomenclature than (collectively) National Semiconductor, Texas Instruments, Burr-Brown, Analog Devices, Linear Technology, et cetera, et cetera.
That qualifies as "absurd" in my book.
Hi Charles,
A couple of points: this is a discussion on technical points, related to documented definitions. The fact that a couple of manufacturers chose to modify that to suit their own needs, is irrelevant for the merits of the discussion.
We all know that we cannot rewrite history, and that the world doesn't revolve around us.
But at the very least let's try to be intellectually honest.
May I quote Edward Cherry?
"Stripped of its input level shifters and complementary first stage, and with the current mirrors and voltage followers replaced by the simplest inverting amplifier (a single common-emitter transistor), the current feedback amplifier reduces to [Fig 6].
A generation ago this circuit was called a voltage-feedback pair (italics mine). It differs from [Fig 2a] only in that the current flowing into the left-hand side of the feedback network is the input current multiplied by the gain of the first transistor, as distinct from the input current itself. The feedback connection is voltage-sensing at the output and voltage-subtracting at the input, the circuit is an ordinary voltage feedback amplifier. The term ‘current-feedback amplifier’ seems a complete misnomer".
Edward M Cherry, ‘Feedback amplifier configurations’, IEE Proceedings on Circuits, Devices and Systems, Vol 147, No 6, Dec 2000.
You wouldn't call old Ed 'absurd', now would you Charles?
Jan Didden
A couple of points: this is a discussion on technical points, related to documented definitions. The fact that a couple of manufacturers chose to modify that to suit their own needs, is irrelevant for the merits of the discussion.
We all know that we cannot rewrite history, and that the world doesn't revolve around us.
But at the very least let's try to be intellectually honest.
May I quote Edward Cherry?
"Stripped of its input level shifters and complementary first stage, and with the current mirrors and voltage followers replaced by the simplest inverting amplifier (a single common-emitter transistor), the current feedback amplifier reduces to [Fig 6].
A generation ago this circuit was called a voltage-feedback pair (italics mine). It differs from [Fig 2a] only in that the current flowing into the left-hand side of the feedback network is the input current multiplied by the gain of the first transistor, as distinct from the input current itself. The feedback connection is voltage-sensing at the output and voltage-subtracting at the input, the circuit is an ordinary voltage feedback amplifier. The term ‘current-feedback amplifier’ seems a complete misnomer".
Edward M Cherry, ‘Feedback amplifier configurations’, IEE Proceedings on Circuits, Devices and Systems, Vol 147, No 6, Dec 2000.
You wouldn't call old Ed 'absurd', now would you Charles?
Jan Didden
It's hard for me to follow Cherry's arguments with only the text and no diagrams. But let's assume for the moment that his argument has technical merit.
And now (assuming that the pinouts are the same), do you think you could take any given circuit and substitute a "current feedback" op-amp for a "voltage feedback" op-amp, or vice-versa?
Of course not.
And that is exactly why the term "current feedback" was coined. It provides a convenient shorthand way to separate two classes of devices. It is an important distinction, and the term itself is quite descriptive (except possibly to the overly pedantic).
And, as I predicted, this "debate" has devolved into a dreary argument about semantics.
And now (assuming that the pinouts are the same), do you think you could take any given circuit and substitute a "current feedback" op-amp for a "voltage feedback" op-amp, or vice-versa?
Of course not.
And that is exactly why the term "current feedback" was coined. It provides a convenient shorthand way to separate two classes of devices. It is an important distinction, and the term itself is quite descriptive (except possibly to the overly pedantic).
And, as I predicted, this "debate" has devolved into a dreary argument about semantics.
OK Charles,
This will be my last post on this, honestly.
- Of course you cannot just swap CFB and VFB in a given circuit. Just as you cannot just swap a JFET input opamp for a BJT input and expect everything to remain the same. Or a hi-speed opamp with a precision-opamp. This is a non-argument.
- I accept that the term 'CFB opamp' makes it possible to distinguish between types of implementations of VFB opamps (sorry, couldn't resist it)
- I accept that the term CFB has become a de facto standard for that type of circuit. So be it. But I still feel cheated.
Jan Didden
PS Like the new sig better?
This will be my last post on this, honestly.
- Of course you cannot just swap CFB and VFB in a given circuit. Just as you cannot just swap a JFET input opamp for a BJT input and expect everything to remain the same. Or a hi-speed opamp with a precision-opamp. This is a non-argument.
- I accept that the term 'CFB opamp' makes it possible to distinguish between types of implementations of VFB opamps (sorry, couldn't resist it)
- I accept that the term CFB has become a de facto standard for that type of circuit. So be it. But I still feel cheated.
Jan Didden
PS Like the new sig better?
I think I have to agree with Jan Didden about the use of the phrases current and voltage feedback. For quite some time this used to be the name for feedback derived from the output voltage (shunt) or from the output current (series). But maybe because these terms are used differently now, we might better use the words shunt and series feedback. Or, if we want to include the way the feedback is influencing the input signal (as many people are doing now when they talk about current feedback or voltage feedback), we can talk about series-series feedback, series-shunt feedback, shunt-series feedback or shunt-shunt feedback. Have a look at the picture. It is coming from http://www.s2.chalmers.se/graduate/courses/analogVLSI/doc/lecture2.pdf
Steven
Steven
Attachments
Charles Hansen said:
Charles i have to say i find you extremely disappointing ....in all respects......
...You seem totaly incapable of discussing a point on it's own merits....
janneman said:
AMEN!!!!!!
janneman said:
YES!!!
P.S: Jan....i would love to have a copy of Cheery's paper....do you have a pdf??Cheers.
Steven said:
AHHH YES....another serious engineer, and audiophile has joined us in the hallowed application of common sense and reason....AMEN..
mikeks said:
I guess I had to find myself agreeing with TAFKA:Mikek eventually.
Op amps with a low impedance negative input have simply been
misnamed, probably for marketing reasons. Nevertheless the
name isn't going to go away, so we simply have to take care
not to confuse it with circuits which actually measure the output
current and take their feedback from that.
Just like "no feedback", a reasonable convention gets established,
and as long as we understand each other when we talk about
these things, there's no problem.
pass/: favorite oxymoron is "balanced single-ended"
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.