Defining feedback mode completely requires...
both the output sampling and input mixing configurations to be defined, as per Dave's post. You cannot simply say an amp has current or voltage feedback. In what manner is the signal then applied to the input stage?
In my Audio Engineering class, there were also 4 (theoretical)configurations:
1. Voltage-series feedback (as in this Elektor design)
2. Voltage-shunt feedback
3. Current-series feedback
4. Current-shunt feedback.
The first term refers to the take-off point, and the second term refers to how the the feed-back signal is applied to the input stage.
Regards
both the output sampling and input mixing configurations to be defined, as per Dave's post. You cannot simply say an amp has current or voltage feedback. In what manner is the signal then applied to the input stage?
In my Audio Engineering class, there were also 4 (theoretical)configurations:
1. Voltage-series feedback (as in this Elektor design)
2. Voltage-shunt feedback
3. Current-series feedback
4. Current-shunt feedback.
The first term refers to the take-off point, and the second term refers to how the the feed-back signal is applied to the input stage.
Regards
Re: Defining feedback mode completely requires...
Shaun,
In the book I have they define it as follows:
- If the feedback signal goes zero if you short the output, (output voltage goes zero) it is voltage feedback;
- If the feedback signal goes to zero if the load is disconnected (output current goes zero) it is current feedback.
This is consistent with all other definitions of VF and CF that I found, e.g. the figures I put in post # 22.
I agree that the definition of the four cases (combinations of shunt & series) gives a more formal classification, but as we were discussion this Elektor contraption, by the definitions I have found, it uses VF. (I know, I'm stubborn).
You can talk about VF or CF whatever way you use to combine the input and feedback signal at the input.
Cheers, Jan Didden
Shaun said:
Shaun,
In the book I have they define it as follows:
- If the feedback signal goes zero if you short the output, (output voltage goes zero) it is voltage feedback;
- If the feedback signal goes to zero if the load is disconnected (output current goes zero) it is current feedback.
This is consistent with all other definitions of VF and CF that I found, e.g. the figures I put in post # 22.
I agree that the definition of the four cases (combinations of shunt & series) gives a more formal classification, but as we were discussion this Elektor contraption, by the definitions I have found, it uses VF. (I know, I'm stubborn).
You can talk about VF or CF whatever way you use to combine the input and feedback signal at the input.
Cheers, Jan Didden
Is this a "stubborness contest"?
Sure, but feedback means: "feeding back". Now what are you feeding back? Voltage or current (or a combination). If yo were talking about "feedinto" we had a different story. In any case, the educational documentation agrees with my point of view. Some manufacturer datasheets "seem" to agree with you, but that basically is sales literature.
Jan Didden
Sure, but feedback means: "feeding back". Now what are you feeding back? Voltage or current (or a combination). If yo were talking about "feedinto" we had a different story. In any case, the educational documentation agrees with my point of view. Some manufacturer datasheets "seem" to agree with you, but that basically is sales literature.
Jan Didden
Strictly speaking...
Strictly speaking any feedback signal (except perhaps for 100% feedback) is merely a signal analogous to the output signal. Looking at it this way, I may tend to agree with your text book.
So , in your circuit description :" Here we have a signal analogous to the current in the load..."
Sorry - just had to say it!!
Strictly speaking any feedback signal (except perhaps for 100% feedback) is merely a signal analogous to the output signal. Looking at it this way, I may tend to agree with your text book.
So , in your circuit description :" Here we have a signal analogous to the current in the load..."
Sorry - just had to say it!!
Jens wrote:
" Now if the output of the system is a current and the input is a voltage, the H function must convert the current to a voltage. If the output is a voltage, the H block is not needed for converting. "
Not necessarily. It doesn't matter what the output of the system is. What matters is what you "measure" with the feedback network. A power amp is a voltage source (normally), but if you pick off a voltage representing the output current from a resistor in series with the speaker, you are using CF.
Then at the other end you can have a high impedance so that there is a predominantly a voltage into the summing node, or you have a low impedance, where there is predominantly a current into the summing node.
Note that there is always current AND voltage into the summing node. It is quite arbitrary where you draw the line between voltage and current injection. That in itself is already a clear sign that this is not a good way to make a distinction betwee CF and VF. However using the established definition (CF is when the feedback signal goes to zero if you cut the load, VF is when the feedback signal goes to zero if you short the output) is clear and unambiguous. QED.
Jan Didden
" Now if the output of the system is a current and the input is a voltage, the H function must convert the current to a voltage. If the output is a voltage, the H block is not needed for converting. "
Not necessarily. It doesn't matter what the output of the system is. What matters is what you "measure" with the feedback network. A power amp is a voltage source (normally), but if you pick off a voltage representing the output current from a resistor in series with the speaker, you are using CF.
Then at the other end you can have a high impedance so that there is a predominantly a voltage into the summing node, or you have a low impedance, where there is predominantly a current into the summing node.
Note that there is always current AND voltage into the summing node. It is quite arbitrary where you draw the line between voltage and current injection. That in itself is already a clear sign that this is not a good way to make a distinction betwee CF and VF. However using the established definition (CF is when the feedback signal goes to zero if you cut the load, VF is when the feedback signal goes to zero if you short the output) is clear and unambiguous. QED.
Jan Didden
carrots and potatoes
Hi Jan,
the problem is in the input section. You need to have the same units, or else the system will not work.
Try subtracting current and voltage, I can't ! QED
I can se what you mean on the feedback signal beeing a current, but if the input is a voltage, somwhere one of the two is conveted.
\Jens
Hi Jan,
the problem is in the input section. You need to have the same units, or else the system will not work.
Try subtracting current and voltage, I can't ! QED
I can se what you mean on the feedback signal beeing a current, but if the input is a voltage, somwhere one of the two is conveted.
\Jens
Redefined electrical and electronics terms
Hello all,
I agree almost entirely with Jan, but I'd like to add that "current feedback" is not the only term that seems to have been redefined. For example:
The term "linear" is used with the following, completely different, meanings:
A. complying with the superposition theorem,
B. having a flat frequency response,
C. anything analogue,
D. not switch-mode.
I believe A to be the original meaning. Therefore, according to the original meaning, a "linear" voltage regulator has to be broken down: its output voltage is by definition proportional to the input voltage.
Active device:
A. Device capable of supplying energy (in network theory),
B. Device suitable to be used as an amplifying device when properly biased by an external power source.
Active input resistance:
A. Input resistance having a negative real part for some frequency or frequency range (network theory).
B. Input resistance realised with active circuitry, regardless of whether the real part is positive or negative.
I'm sure there are more of these terms around. I frequently resort to giving a definition before using a term, so that at least everyone knows what I mean.
Marcel van de Gevel
Hello all,
I agree almost entirely with Jan, but I'd like to add that "current feedback" is not the only term that seems to have been redefined. For example:
The term "linear" is used with the following, completely different, meanings:
A. complying with the superposition theorem,
B. having a flat frequency response,
C. anything analogue,
D. not switch-mode.
I believe A to be the original meaning. Therefore, according to the original meaning, a "linear" voltage regulator has to be broken down: its output voltage is by definition proportional to the input voltage.
Active device:
A. Device capable of supplying energy (in network theory),
B. Device suitable to be used as an amplifying device when properly biased by an external power source.
Active input resistance:
A. Input resistance having a negative real part for some frequency or frequency range (network theory).
B. Input resistance realised with active circuitry, regardless of whether the real part is positive or negative.
I'm sure there are more of these terms around. I frequently resort to giving a definition before using a term, so that at least everyone knows what I mean.
Marcel van de Gevel
undeterminology
Perhaps we should refer to VIF - voltage information feedback and CIF - current information feedback to distinguish information content from signal impedances. I think the original terms VF and CF meant this in the earlier literature. Designers looked at the function of the whole circuit rather than a subset contained in a chip. The chip people have tried to re-define things in chip - centric terms. The trouble with chip - centric is that the user can drop a strategic resistor in somewhere in the overall user design that changes the whole function.
It would be nice to just refer to things as V - amp, I - amp, Transconductance - amp, and Transresistance - amp. But, there is still the ambiguity as to whether these terms refer to the chip alone or to the overall user circuit, so these need to be carefully applied too. Just my $0.02.
Checked Feucht's book "Handbook of Analog Circuit Design" and he uses the chip - centric definition, CF = current input. If enough new books get written this way, and maybe all the new ones are, we will have to be very specific about what we mean.
Don
Perhaps we should refer to VIF - voltage information feedback and CIF - current information feedback to distinguish information content from signal impedances. I think the original terms VF and CF meant this in the earlier literature. Designers looked at the function of the whole circuit rather than a subset contained in a chip. The chip people have tried to re-define things in chip - centric terms. The trouble with chip - centric is that the user can drop a strategic resistor in somewhere in the overall user design that changes the whole function.
It would be nice to just refer to things as V - amp, I - amp, Transconductance - amp, and Transresistance - amp. But, there is still the ambiguity as to whether these terms refer to the chip alone or to the overall user circuit, so these need to be carefully applied too. Just my $0.02.
Checked Feucht's book "Handbook of Analog Circuit Design" and he uses the chip - centric definition, CF = current input. If enough new books get written this way, and maybe all the new ones are, we will have to be very specific about what we mean.
Don
Yet another term used in completely different ways
That reminds me that "current amplifier" is also a nice one.
A. Amplifier with a low input and a high output impedance, having an accurate current to current transfer.
B. Voltage follower (because it can give a larger signal current to its load than flows into its input).
C. Transconductance amplifier.
It should be noted that meanings A and B are opposites, as a voltage follower has a high input and a low output impedance and a fairly accurate voltage to voltage gain.
That reminds me that "current amplifier" is also a nice one.
A. Amplifier with a low input and a high output impedance, having an accurate current to current transfer.
B. Voltage follower (because it can give a larger signal current to its load than flows into its input).
C. Transconductance amplifier.
It should be noted that meanings A and B are opposites, as a voltage follower has a high input and a low output impedance and a fairly accurate voltage to voltage gain.
OK. You got me worried there for a minute...
Jan Didden
To Marcel: Interesting examples. Of course, sometimes it is clear from the context what is meant, like the analog amplifier that "is now sampling". Still, I like to get my ducks in a row first before firing away.
smoking-amp: I am disappointed by Feucht! BTW, anybody has a copy of Feucht's book they want to sell? Anybody has it on CD-ROM? No? Thought so...
I also want to stress again that the chip-centric way of defining CF vs. VF is arbitrary: If you have a buffer output as inverting summing input with a very small standing current, it's input impedance may actually be higher than using the base of an input transitor with heavy biasing. Where do you draw the line?
Jan Didden
Jan Didden
To Marcel: Interesting examples. Of course, sometimes it is clear from the context what is meant, like the analog amplifier that "is now sampling". Still, I like to get my ducks in a row first before firing away.
smoking-amp: I am disappointed by Feucht! BTW, anybody has a copy of Feucht's book they want to sell? Anybody has it on CD-ROM? No? Thought so...
I also want to stress again that the chip-centric way of defining CF vs. VF is arbitrary: If you have a buffer output as inverting summing input with a very small standing current, it's input impedance may actually be higher than using the base of an input transitor with heavy biasing. Where do you draw the line?
Jan Didden
The current state of art...
Hi
I thougt that in my early post i have been clear enough...lets sumarise:
-Current amplifier is a amplifier that amplifie a voltage(transcomductance amp) or a current and output a current proporcional to the input (voltage in the case of a tranconductance and current in the second case).
The output of this type of amps is a more or less perfect current source(very high impedance).
This type of amps use as feedback signal a sample of the output current.
-voltage amplifier is a amp that amplifie a voltage(voltage amp) or a current (transresistence amp) and output a voltage proporcional to the input(voltage and current respectively).
The output of this types of amps is a more or less perfect voltage source.
This type of amps use as feedback signal a sample of the output voltage
A "current feedback amp" is a amp that can be one of both types(current or voltage amps) but uses current (and not voltage) in the summing node at the input.As the inverting input as very low impedance.
Current feedback amp type tell us about the type of injection of the fedback at the input and not the the type of sample of the output(voltage or current)
They have some interesting features as increased slew rate and a bandwith that is relatively independent of the closed loop gain (but dependent of the feedback resistor).
Sumary:
Current or voltage amp - depend of what is the output of a amp a voltage or a current.
Voltage feedback amp-depend of how the feedback signal is compared with the input signal at the input. In this case in current forme
For more deep inside, about current feedback amp see:
http://www.national.com/an/OA/OA-30.pdf
Regards
Hi
I thougt that in my early post i have been clear enough...lets sumarise:
-Current amplifier is a amplifier that amplifie a voltage(transcomductance amp) or a current and output a current proporcional to the input (voltage in the case of a tranconductance and current in the second case).
The output of this type of amps is a more or less perfect current source(very high impedance).
This type of amps use as feedback signal a sample of the output current.
-voltage amplifier is a amp that amplifie a voltage(voltage amp) or a current (transresistence amp) and output a voltage proporcional to the input(voltage and current respectively).
The output of this types of amps is a more or less perfect voltage source.
This type of amps use as feedback signal a sample of the output voltage
A "current feedback amp" is a amp that can be one of both types(current or voltage amps) but uses current (and not voltage) in the summing node at the input.As the inverting input as very low impedance.
Current feedback amp type tell us about the type of injection of the fedback at the input and not the the type of sample of the output(voltage or current)
They have some interesting features as increased slew rate and a bandwith that is relatively independent of the closed loop gain (but dependent of the feedback resistor).
Sumary:
Current or voltage amp - depend of what is the output of a amp a voltage or a current.
Voltage feedback amp-depend of how the feedback signal is compared with the input signal at the input. In this case in current forme
For more deep inside, about current feedback amp see:
http://www.national.com/an/OA/OA-30.pdf
Regards
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.