True Current Feedback N-channel Mosfet Amp

Status
Not open for further replies.
Jump to Latest
#161
Christer said:


Do you mind taking a trip to a -10 deg. C. Sweden? I can serve you
an espresso to warm you up from the cold while you have a
look in my old textbook to see that I am indeed not lying to you.
You can have a cognac with the coffee too if you wish. 😉
Click to expand...

Maybe one day...thanks for the invitation 😉

For -10 deg C. why not some tubes???😀
 
#162
Ouroboros said:
The term 'current feedback' in its modern use is perfectly reasonable because the feedback summation is done by subtracting a feedback current (flowing through the feedback network), from an input current (from the input transconductance buffer). The result of this subtraction is an output current which is then the internal error signal feeding into the current-mirror stage.

I'm biased I guess, because I never have used the term to refer to an amp that performs an overall VCCS function!
Click to expand...

Yes I agree there is some logic to the term as you say, but it is
not necessarily more logical than the other way of using it. It
all depends on our assumptions and previous experiences. We
are all biased in one way or another, as you say. The important
thing is not to confuse this personal bias with universal truth. It
would however be very nice if there was one single authorative
terminology that everybody used.


Tube_Dude said:


Maybe one day...thanks for the invitation 😉

For -10 deg C. why not some tubes???😀
Click to expand...

Well, it is the outdoor temperature fortunately, but it is a bit
cold indoors too. Maybe I should consider a big class A amp
after all, just to use in the winter. 🙂
 
#163
Ouroboros said:
Apart from the fact that this link points to a current DRIVE amplifier. ie, a VCCS. (in spite of what the web site says!)
Click to expand...


I like this....:
This is indeed a CF amplifier according to the original classic definition: the output current is sampled and fed back to the input.

I also noted that he says: CF amplifiers will oscillate. That is not automatically true, there is no difference between a VF or CF amplifiers, or any other amplifier, for the criteria and conditions for stability. But taking a random audio amp and wrapping it into a CF loop is asking for trouble.

Jan Didden
 
#164
Jan;

Thanks for confirming my suspicion. I wish I had gotten some samples myself :bawling:

Just how many do you have, anyway 😉

mlloyd1

janneman said:
... After several years of trying to interest the big guys, Doug sold the patent and remaining stock to Wadia. They used it in a zero-feedback I/V converter...Jan Didden
Click to expand...
 
#165
Jan is correct, the true classical CFA is an invertering amplifier were currect is feedback connects to a virual ground at the input.
Here is a good link to several different types current FB. http://www.eecg.toronto.edu/~kphang/ece1371/cfa.pdf

There are also other types of CFB applications noted in this PIC.
 

Attachments

  • cfb 13di5fg1.gif
    cfb 13di5fg1.gif
    6.4 KB · Views: 412
#167
Tube Dude

The classical and original definition defining the CFB amps are for inverting designs. While that what meant I did say there were no other methods are ways to design a CFA that is why I added the picture. I just assumed you would understand that.😕
 
#168
The classical and original definition defining the CFB amps are for inverting designs. While that what meant I did say there were no other methods are ways to design a CFA that is why I added the picture. I just assumed you would understand that.
Click to expand...

Excuse me it should read: While that what meant, I did NOT state that there were no other methods are ways to design a CFA that is why I added the picture.😱
 
#169
Re: Tube Dude

jewilson said:
The classical and original definition defining the CFB amps are for inverting designs. While that what meant I did say there were no other methods are ways to design a CFA that is why I added the picture. I just assumed you would understand that.😕
Click to expand...

If you read what Janeman say about the classical current feedback in post #163, he says that a sample of the output current is feedback at the input..but .that can be done with a inverting or a non inverting amplifier...and that make the amplifier a transconductance or a current anplifier.


The link you gave is a very good paper about Current Feedback Amps! And i hope that some people here can finally learn about the subject!
 
#170
There are not such a thing as a current feedback type, in the essencial types of amplifiers ,only in the way the feedback signal is handled at the input..
Click to expand...

Exactly. :nod:

In any case transconductance amps are not of much use for driving hifi loudspeakers so we can forget them.

Now, how about focussing on what the pros and cons are of a low Z input to the differentiator as opposed to a high Z input. Related issue to using an op-amp in inverting mode or non-inverting mode.

Just an idea.:cheerful:
 
#172
Greetings,

First, I would immensely enjoy a trip to Sweden, even at -10C. I enjoy traveling and meeting other people, especially in new cultures and surroundings. I would expect Sweden to be a very inviting and friendly place to visit. Even better if someone is offering Cognac.

Second, I would like to mention that I have enjoyed Jan’s writings, especially his observations on Walt Jung’s power supply regulator.

I agree with the “classical” definition of a VCCS. If a signal representing current is used as the feedback loop, it is a current feedback system. In this case, a voltage signal is used to represent current. I can also agree that it is , indeed, a VCCS.

Unfortunately, there are many way to communicate the same idea. There are also many control-loop feedback theories that are shared between the various topologies. This can easily lead to language and vocabulary confusion, especially when members of this forum live on different continents.

Indeed, Jan has a point: using a random audio amp in a CF loop may be asking for trouble. However, if the gain and phase margins are carefully studied, stability can be achieved. It seems that the author has struggled with this, trading frequency response for stability. Since his system uses active crossovers, the amplifier needs to function only in the desired frequency range, which he has accomplished by choosing the impedance of the stability compensation.

This type of amplifier may not be appropriate for full-range, but it may be appealing for a subwoofer by enclosing the woofer’s voice coil within the feedback loop. As mentioned, some equalization would be required to cancel the woofer’s impedance curve.
 
#173
Now, how about focussing on what the pros and cons are of a low Z input to the differentiator as opposed to a high Z input. Related issue to using an op-amp in inverting mode or non-inverting mode
Click to expand...

This can be a very interesting example. Imagine we built an inverting op amp. The +input is grounded. We want to have 10x gain. The signal comes in to -input via 100ohm resistor, and feedback also somes to this -input via 1kohm resistor (this will gives gain -10x).

Can we consider this as a CFB? The feedback is indeed comes to a low Z (remember, +input is grounded).

Anyway, this comes to my personal question. Some says (and very often), that "Inverting power amp (with+input grounded) have better sound than ordinary non-inverting amp"

What makes this happen, while the rest of the circuit is all the same?
 
#174
Re: Re: Tube Dude

jewilson said:
The classical and original definition defining the CFB amps are for inverting designs. While that what meant I did say there were no other methods are ways to design a CFA that is why I added the picture. I just assumed you would understand that.😕
Click to expand...


Tube_Dude said:


If you read what Janeman say about the classical current feedback in post #163, he says that a sample of the output current is feedback at the input..but .that can be done with a inverting or a non inverting amplifier...and that make the amplifier a transconductance or a current anplifier.


The link you gave is a very good paper about Current Feedback Amps! And i hope that some people here can finally learn about the subject!
Click to expand...

Amazing! I fully agree with Tube_Dude this time😉

Jan Didden
 
#176
lumanauw said:


This can be a very interesting example. Imagine we built an inverting op amp. The +input is grounded. We want to have 10x gain. The signal comes in to -input via 100ohm resistor, and feedback also somes to this -input via 1kohm resistor (this will gives gain -10x).

Can we consider this as a CFB? The feedback is indeed comes to a low Z (remember, +input is grounded).

Anyway, this comes to my personal question. Some says (and very often), that "Inverting power amp (with+input grounded) have better sound than ordinary non-inverting amp"

What makes this happen, while the rest of the circuit is all the same?
Click to expand...

If I understand correctly (which is not the same as agreeing) the crowd of the "modern" CFA, this would be a CFA, because the feedback goes to a low imp node. Am I right?

Your second comment: you make the jump from : "some say ...better sound" to : ...IS better sound". These are two different things, and there is no logical relation between the two. The second one is just an opinion. Get my drift?😉

Jan Didden
 
#177
janneman said:


If I understand correctly (which is not the same as agreeing) the crowd of the "modern" CFA, this would be a CFA, because the feedback goes to a low imp node. Am I right?

Your second comment: you make the jump from : "some say ...better sound" to : ...IS better sound". These are two different things, and there is no logical relation between the two. The second one is just an opinion. Get my drift?😉

Jan Didden
Click to expand...



No. Just because the feedback node is low-Z is not enough. In the case above, even though the node is low-Z, it is still a VOLTAGE difference that is being used inside the feedback loop as the error signal. A CFB amp (modern usage!) derives the error (current) signal as the difference between an input current and the feedback current.

Regarding the sound, I really can't say, but it is worth noting that because the way the error signal is derived, CFB amps have very poor control of the output DC offset and usually require a separate servo-controlled dc feedback loop.
 
#179
Ouroboros said:




No. Just because the feedback node is low-Z is not enough. In the case above, even though the node is low-Z, it is still a VOLTAGE difference that is being used inside the feedback loop as the error signal. A CFB amp (modern usage!) derives the error (current) signal as the difference between an input current and the feedback current.

Regarding the sound, I really can't say, but it is worth noting that because the way the error signal is derived, CFB amps have very poor control of the output DC offset and usually require a separate servo-controlled dc feedback loop.
Click to expand...

Well, it's all fine with me, since I think a CF amp is something entirely different anyway, but I noted that on this post many are of the opinion that there never is an input current, because the signal input is high impedance. They say that CF is input voltage and feedback current.

Your second comment, again, I don't understand why these jumps from one opinion to the other. Yes, DC offset is generally worse, and yes, that may mean you need a servo. But why on earth would that mean worse sound???

Jan Didden
 
Status
Not open for further replies.
Top Bottom