Please make the distinction between the output controlled quantity and the output sampled quantity.
"Current feedback" was referred by Black as a feedback amplifier that samples the output current.
A classic CFA samples the output voltage, so it comes in direct collision with Black's original classification.
This extra hilarious my first job ever as an EE was working for Walter at the Center for Space Research. IIRC Walter was kicked off of Quora for not tolerating fools, ironic.
EDIT - Watched it, Wiki says at the very beginning and he quotes it, KVL comes from conservation of energy i.e. no energy lost in radiated fields. He starts out wrong and keeps on going. This relates tangentially to the two capacitor problem.
Last edited:
It's about the transfer function of the IPS. LTP has tanh, "diamond" CFA has sinh.
H bridge IPS also has sinh transfer function.
A valid point, often overlooked.If there is anything I find really frustrating in this thread is the lack of understanding of the small vs. large signal analysis and models. But then, I suppose it’s part of the learning curve for some.
This is what Wiki tells about the subject:
Differences between Small Signal and Large Signal
A small signal model takes a circuit and based on an operating point (bias) and linearizes all the components. Nothing changes because the assumption is that the signal is so small that the operating point (gain, capacitance, etc.) doesn't change.
A large signal model, on the other hand, takes into account the fact that the large signal actually affects the operating point, as well as that elements are non-linear and circuits can be limited by power supply values. A small signal model ignores simultaneous variations in the gain and supply values.
Hans
The first stage of a CFA has a non-inverting input and an inverting input which asts as a differential stage. The inverting stage is said to be of low impedance. The engima you have now to resolve is why, in closed loop, this low impedance does not load the feedback divider and why the same values of resistors Rf and Rg (let's say 1 kOhm or less for Rg) give similar voltage gains in CFA and VFA.
We are way off topic here, but I'm not sure what you mean, KVL is perfectly valid for non conservative fields, without violating Maxwell. However, KVL (as par of circuit theory) relies on 3 basic assumptions:
1. The circuit dimensions relative to the speed of light and the time scales of interest are small
2. The net charge inside any circuit element is negligible
3. The magnetic flux outside any circuit element is negligible
As long as the field is zero outside of the boundary of the circuit element, then it no longer matters that the fields inside is non conservative, the energy crossing the boundary is always equal to I x V so it can be treated as a standard circuit element and KVL applies.
The guy has an engineering view, which is correct. There is a follow up video in which he deals with the contour and surface integrals of the non conservative fields, but it is obvious it is not his specialty.
When do amplifier input stages work in large signals ?
BTW, note the idea circa 1985 of working with minimal signals which consists of minimizing any voltage and current variations across the active devices in charge of the feedback.
What can it be then ? It has never been contested in this thread that it was one.
If you restrict CFA's to the category of amplifiers having complementary push-pull input pairs with their feature of current on demand which is never requested in normal audio usage, it is sure that you create a dead end.
I did not claim that COD is the defining property of a CFA. I simply elucidated RNM’s expansive and compressive remark from a few years ago.
Ok, let’s agree to disagree on this (although literature is there to witness, you cold ask for references). I have to admit though, I have barely seen over the years an educated person so entrenched in his own beliefs as yourself on this CFA topic.
Could you please explain how COD is related to “compressive” and “expansive” (whatever that means)? And BTW, COD is a defining (or a native, if you prefer) property of the CFA topology; I would appreciate if anybody could show a CFA design with the small signal constant bandwidth property and without the COD property.
Last edited:
Hi Ian,
You haven't responded so far on my posting where I mentioned that using a 200K Ro was less correct, because Ro changes with alternating input voltage, in this very case between 199K and 201K with Vin=1V and Ic=1mA.
Instead of +/-5uA, current through R0 only changes by 4.4uA.
In the model below I have both versions, one with a fixed 200K resistor and the other with a correctly fluctuating resistor value.
AC current as can be seen in one model is 5uA and in the other 4.4uA.
Difference in ic/vbe is not that big, but it is simply more correct to my opinion.
Hans
You haven't responded so far on my posting where I mentioned that using a 200K Ro was less correct, because Ro changes with alternating input voltage, in this very case between 199K and 201K with Vin=1V and Ic=1mA.
Instead of +/-5uA, current through R0 only changes by 4.4uA.
In the model below I have both versions, one with a fixed 200K resistor and the other with a correctly fluctuating resistor value.
AC current as can be seen in one model is 5uA and in the other 4.4uA.
Difference in ic/vbe is not that big, but it is simply more correct to my opinion.
Hans
Attachments
Generally these terms are pretty clear based on the sign of the derivatives of the basic behavior.
Sure (1dB gain compression comes to mind) but, case in point, I never heard of a -100V/uS vs. 100V/uS slew rate
P.S. just occurred to me: “CFAs are defined by what they do, rather than by what they are. VFAs are defined based on what they are, rather than based of what they do”. Hence the dichotomy
Last edited:
Sure (1dB gain compression comes to mind) but, case in point, I never heard of a -100V/uS vs. 100V/uS slew rate
Right tell me you never got that minus sign, factor of 2, or pi/2 wrong.
You must be confusing me with someone else. I’ve always been scrupulously even handed when discussing the merits or not of the various topologies. I’ve got power amps (DIY) of both types and I think excellent amps can be built with either topology. If you are looking for someone to make unfounded claims about an amplifier based on subjective snake oil, you won’t find it with me.
Syn08, I simply said I was expanding on what RNM said and not making claims either way. Nothing else.
Please don’t read into statements things that are not there. Ok?
I am well aware of the COD property of CFA’s but it’s not an exclusive property of CFA’s since VFA’s can be designed with similar behavior - a point you raised a few weeks ago on this very thread.