Damir I have been on business trips and to make matters worse have some problems with my email account.
Ill get back to you soon.
This is precisely where semantics of CFA becomes problematic.
There is more negative feedback relative to the current sensed by the shunt resistance.
Consider the amp like a black box whatever its internal circuit.
If its output voltage decreases when it delivers current, it means that it is a voltage source with significant output resistance or impedance.
If one wants to work with a specific driver but inadequate damping, positive or negative impedance are ways to modify it. Other ways can be "bell" equalizations or transforms as such those made popular by Linkwitz.
See Erik Stahl works and comments at DIYaudio, searching "Stahl".
If the main resonance damping and frequency response problems are solved, there are great advantages to current driven topologies.
Search Meriläinen, Hawksford, Greiner and Smith.
Nelson Pass has also promoted current driven systems.
Maintaining an output voltage independant of the current it delivers to a load which impedances of vary can be considered as a whole servo-controlled system. This is the most common case, as usual amplifiers have a fraction of an Ohm output impedance.
If conceived as a whole system, making the amp to present a significant amp impedance carefully chosen for a specific driver can be very beneficial in many ways.
Of course it is a bit more complicated that the usual way to do things.
Last edited:
If that is true, then how is it possible to have IM distortion when it is only one instantaneous signal?
It seemed to me that when there is lack of tools and methods in studying or exam the linearity of a circuit or a system in time domain, people turn to means and methods in frequency domain where we are far better equipped. IM distortion can be viewed as a perception in frequency domain of certain time domain non-linearity problem. We have become so used to frequency domain means of studying, measuring, and designing our stuff that we often forget it is time domain that all physical events progress within, and confused illusion for reality.
Last edited:
not again with the "Fourier Denial"
the complex Fourier transform of a time and bandwidth limited signal is a Mathematical Dual of the time series data - all of the "time domain information" is present in the Fourier data - they can in principle be converted back and forth without loss - only limited in practice by numerical accuracy and windowing effects
the complex Fourier transform of a time and bandwidth limited signal is a Mathematical Dual of the time series data - all of the "time domain information" is present in the Fourier data - they can in principle be converted back and forth without loss - only limited in practice by numerical accuracy and windowing effects
What!!? You lied. I couldn't see any porn
Please share what you have in the cookies
Whatever you consider, current or voltages, "if the whole voltage gain decreases." means the voltage at the output ? If it decrease with a constant input source, the output impedance of the amp will... decrease. Whatever your feedback is current or voltage taken.I believe most of us, here, knows what is a current source.
I don't feel that way. CFAs are voltage amplifiers, despite their (controversial) name.
That's all the problem.
Some use horns with a descending curve, and use serial resistance to take benefit of the selfic impedance increase to linearize more or less the response curve. I've got better results in: linearizing the impedance curve, correcting the response curve by a impedance compensated passive filter. More complicated, more expensive, i agree.
Well, what about returning in CFA's subject ? We can use them as well in a servo controlled speaker ;-)
Try this, break the feedback loop and insert a perfect voltage controlled voltage source in each type of circuit and see if they are both only a "voltage amplifier". Please do it in sim, I don't want you to blow anything up.
You will see the feedback must also control the current into the inverting input to define the GBW (in one case of course). The feedback network sets the gm of the forward transfer, oh my what is gm voltage to current.
Last edited:
How so , can you be more detailed .....
Are your amplifiers CFA Richard ?
So whats the NFB magic number , 15, 20, 50 db ..?
These are Sales figures so easily confirmed and not open to Golden Pinnae debate.
The Best seller was also designed by my R&D Dept. but not by me. If you were in the trade at that time, it should now be obvious what these models are
What time frame ....?
Indeed. Reason why we can use the feedback impedance to change-it in CFAs
. This said, as a CFA have a low output impedance comparing to the load/feedback network, we can still consider-it as a voltage amplifier, right ?
Is this one of the keys for the audio signature difference between the two topologies ?
What i wanted to say is, before the name "Current feedback was given to this topology, there was no such controversies about. Analog device (first to use this name ?) bears a heavy responsibility in global conflicts (between audio designers;-)
I believe you're one of those in the best position to explain the LTP universal fashion ?
+ effect of the input stage parasitic emitter capacitance ?Try this, break the feedback loop and insert a perfect voltage controlled voltage source in each type of circuit and see if they are both only a "voltage amplifier". Please do it in sim, I don't want you to blow anything up.
You will see the feedback must also control the current into the inverting input to define the GBW (in one case of course). The feedback network sets the gm of the forward transfer, oh my what is gm voltage to current.
In the circuit I was refering , there is a shunt resistor between the loudspeaker and the ground.If the voltage across this shunt resistor is fedback to the inverting input of an amp, it turns the amp to a current source. This kind of output control has been called a Current Feedback Amplifier for years.
I do not know how it is called by now, because this name, which is explicit about the kind of value under control, has been been usurped by a name which describes the internal function of the control (more exactly, describes the way a few people think this internal function works).
Playing with the voltage across the shunt (frequency tailoring, inverting...) allows to define a system different from what it is when the driver is submitted to a voltage source, ie an usual amp.
Of course, it can be done with VFA or CFA (maybe more predictable with VFA) it is just application of feedback theory.
*
Returning to the topology which this thread is about, were there studies or simulations about
- unbalanced currents in the complementary input pair
- about common mode distortion with this topology.
Last edited:
what happens to the distortion of a connected system with a speaker load on the VFA power amp and gnfb creating distorted power amp input bias (say single ended bjt input - worse case) and that affects the gnfb loop of the preamp.... and it's slew rates or other factors were just minimal on each amp by themselves.... is there anyway this can have any negative affects?
Does the input Z of the power amp driving a speaker ever-under any conditions (even clipping)- have an affect on the preceding FVA's gnfb loop and thus its distortion?
Thx-RNMarsh
Does the input Z of the power amp driving a speaker ever-under any conditions (even clipping)- have an affect on the preceding FVA's gnfb loop and thus its distortion?
Thx-RNMarsh
Last edited:
I agree (but the few people, unless you consider that, if it is not you, it is few people ;-)).
Ok, if you look, as you said, both amp topologies as black boxes, both let-you free to sense in the load whatever voltage/current/ or any combination of the two the way you like. You can even sense an other coil or an accelerometer to get better servo control of the speaker's membrane movement.
Pure semantic question, CFA is the name, nowadays, and we are all supposed to understand what it is about.
Last edited: