Keantoken, I dont think this would make it a CFA.
In your case, you are considering the current into the -input as the defining difference but a CFA has a very specific input structure, and the way the input current to the TIS is derived is also different to VFA.
In a VFA, the LTP current sets the maximum TIS input current, while in a CFA its the feedback resistor (and the lower feedback resistor of course also plays a role here, since its shunting away feedback current to GND) that ultimately determines the TIS input current. In a CFA, this means you can get high peak currents into the TIS, and is the main reason you can get high slew rates when compared to classic MC comp'd VFA's. I did come sims, and got 8x the standing current into the TIS. In a VFA, you can only get 2x the standing current of one half of the LTP.
If you are feeding back information from the controlled variable in current form, it follows that the - input impedance should ideally be very low - which it is in a CFA.
In your case, you are considering the current into the -input as the defining difference but a CFA has a very specific input structure, and the way the input current to the TIS is derived is also different to VFA.
In a VFA, the LTP current sets the maximum TIS input current, while in a CFA its the feedback resistor (and the lower feedback resistor of course also plays a role here, since its shunting away feedback current to GND) that ultimately determines the TIS input current. In a CFA, this means you can get high peak currents into the TIS, and is the main reason you can get high slew rates when compared to classic MC comp'd VFA's. I did come sims, and got 8x the standing current into the TIS. In a VFA, you can only get 2x the standing current of one half of the LTP.
If you are feeding back information from the controlled variable in current form, it follows that the - input impedance should ideally be very low - which it is in a CFA.
Last edited:
For others who didnt read the earlier discussions -- I also showed this affect as 'compressive' and 'expansive' -- my terms.
see #54.
THx-RNMarsh
When you overdrive a CFA , it's saturation aspects are more "tame"
"current on demand" seems to be doing this , same VFA/VAS combo,
"bombed out " clipping .. 😀😀
OS
"current on demand" seems to be doing this , same VFA/VAS combo,
"bombed out " clipping .. 😀😀
OS
Yes agree OS and Richard.
Seems with CFA because of the CoD behavior you don't get the hard stop you see in VFA in response to very fast rising or falling edges. Good old fashion large signal overdrive of course is a different matter, but I also noticed that on simple CFA, the clipping behavior is quite benign.
That said, I use input filters to limit the bandwidth not just for RF ingress, but also as part of the comp. On a simple CFA you are asking for trouble if you don't.
Seems with CFA because of the CoD behavior you don't get the hard stop you see in VFA in response to very fast rising or falling edges. Good old fashion large signal overdrive of course is a different matter, but I also noticed that on simple CFA, the clipping behavior is quite benign.
That said, I use input filters to limit the bandwidth not just for RF ingress, but also as part of the comp. On a simple CFA you are asking for trouble if you don't.
posts pulled out due to copyright issues, if you post links to books, please make sure the original publisher has granted you rights....Bonsai,
I don't see any connection between what I said and what you are saying here. I did not discuss compensation in my comment. My point about the CFA vs. VFA spectrum is the dependence on the relative impedances between the feedback network and the impedance seen looking into the emitter(s) of the input stage.
BTW, the current-on-demand feature of many CFAs is nice, but I'm not sure it is a fundamental and distinguishing feature of the CFA topology. For example, I believe that a non-complimentary series-feedback amplifier that has the feedback connected to the input emitter is a CFA, but it does not have the current-on-demand property in one of the two signal polarity directions. Maybe I'm over-defining here by suggesting the CoI property must be there for both signal polarities.
Cheers,
Bob
you're in trouble anyway if you actually need more linear ips current/Verror range than a few mA bias, ~ 100 Ohm degenerated diff pair for linear audio feedback amps fed actual audio signals
maybe Putzeys with his self oscillating Class D
or if you ignore real feedback theory and swallow Otala's flat gain perscription, using low loop gain as a consequence
maybe Putzeys with his self oscillating Class D
or if you ignore real feedback theory and swallow Otala's flat gain perscription, using low loop gain as a consequence
Last edited:
Bob, the point I was making is you can comp a CFA so it behaves somewhat like a VFA wrt to BW, ULGF etc - hence continuum between CFA and VFA. Apologies if this came across the wrong way or your post was misconstrued.
I would agree that CoD is more relavant to complementary CFA designs
I would agree that CoD is more relavant to complementary CFA designs
Hi Bob,
This is where I said pls use the convention that has been in place and established by industry. A certain topology gives a MODE of operation which has a name with defined mathematical characteristics. New computer models have been developed for it.
In the beginning of this question - what is it and how do you know - I called it by what was used in many of the books ---- CM or CMA. Current-Mode of operation. And, CMAmplifier. Maybe I will stick to CMA from now on to emphasis it is operating in full current mode... and not a maybe situation or a could be this or it could be that.
The topology put up is Not a maybe situation. We have not talked about nor designed other topologies. The topology you want to debate is one of those maybe situations. Like a tube circuit with the same topology... such as a Dynaco or other's have used since the dawn of electronics.
Time to move on and get some CMAmps built/designed.
If and when we go to the 'other' topology - we can also compare it in various modes. That isnt what this thread is about, however. It should be its own else-where in IMO. As it can lead to interesting designs on its own and is very flexible. Seems OS wants to try them in his SIM as well starting with ADevices IC design info. Direct coupling could eliminate the input/output coupling caps. Soon there after the dc servo followed as a natural course of development.
My own contribution to that topology which CAN be used (and was used) as a CMA, was to compliment the non-compl series fb amp in a way that it would be direct (DC) coupled. My motivation to do that was because of the poor quality of coupling caps used at the time.... and still are (the current model Marantz is a case in point).
THx-RNMarsh
Last edited:
Care to cite a few references/books about "Current Mode of Operation" and "CMAmplifier" (Current Mode Amplifier, I infer)?
I am not aware of these, perhaps it's something I've seen under other name(s).
Richard,
I've read through your long post and still don't know what you are talking about. Please be specific, up-front. Are you are objecting to my use of the term CFA, which everybody else on this thread is using? I am not a mind-reader, and this is very frustrating.
I am not debating a topology. I am trying to understand by putting forward something and seeing if I've got it right. If I'm wrong, that's OK.
You would get a lot of cooperation to "move on" if you would help here, rather than introducing yet more acronyms into the discussion.
Help me out here.
Cheers,
Bob
Hi Bob,
Well.... I think what i said will fit into the scheme of things, in time. I just wish you would have come into this earlier... we went past this area and into SIM and then building phase is starting. I just cant do it again for each person. You have your own style of learning. However, as I have demonstrated by my replies - I start with ideas and concepts first and then break it down into details... somewhat backwards from some approaches... who take all the details and construct from those. Its all good.
So, please continue as you feel is needed. I will wait for the designs of the CMA here and work from that point forward now. I look forward to the details being filled in with clever ideas implemented in designs/SIM. Even might get to hear a few of them, I hope.
Thx-RNMarsh
Well.... I think what i said will fit into the scheme of things, in time. I just wish you would have come into this earlier... we went past this area and into SIM and then building phase is starting. I just cant do it again for each person. You have your own style of learning. However, as I have demonstrated by my replies - I start with ideas and concepts first and then break it down into details... somewhat backwards from some approaches... who take all the details and construct from those. Its all good.
So, please continue as you feel is needed. I will wait for the designs of the CMA here and work from that point forward now. I look forward to the details being filled in with clever ideas implemented in designs/SIM. Even might get to hear a few of them, I hope.
Thx-RNMarsh
Last edited:
This is a pain of any web forum discussion. Anywhere, everywhere. Individual action of any participant is needed but it takes time 😉Richard,
It would take a lot less verbiage to answer my very simple question - what terminology am I using that you object to?
Please don't answer with a bunch of condescending, pedantic jibborish.
The signal-to-noise ratio on this thread has been very low, and this kind of obtuse response is one reason why.
A few days ago you asked me to include a chapter on CFAs in my next edition, and you certainly are not making it easier for me to want to do that with this kind of behavior.
Maybe some of the other, more helpful, folks on this thread can help me out.
not so cheery right now,
Bob
Hi Bob,
I am sorry at this point in time not to go into any more details on the subject. maybe later.
I know you will need to have a clear picture and details to write about CMA operation and that is always a time consuming endevor. I dont want to sound rude nor condensenting. I just need to move onto other things and I dont think i can add much more. Bonsai seems to have a good grip on the subject. And, there are so many books out there that would fill this forum for a long time.
The details are very imprtant. I'll check in from time to time and see what I can contribute any further. I also think Scott Wurcer is a good source as ADI make both CMA and VMA opamps and surely must know the design issues up one side and down the other... far more than i could ever do. And, you do need to cover it fairly well in your book.
Not being mean... I just have to get other time consuming interests a fair shot, too.
[Kind of feels like a basket ball game... here we are -- running up and down the court and the ball being passed back and forth and then...... the buzzer.]
THx-RNMarsh
I am sorry at this point in time not to go into any more details on the subject. maybe later.
I know you will need to have a clear picture and details to write about CMA operation and that is always a time consuming endevor. I dont want to sound rude nor condensenting. I just need to move onto other things and I dont think i can add much more. Bonsai seems to have a good grip on the subject. And, there are so many books out there that would fill this forum for a long time.
The details are very imprtant. I'll check in from time to time and see what I can contribute any further. I also think Scott Wurcer is a good source as ADI make both CMA and VMA opamps and surely must know the design issues up one side and down the other... far more than i could ever do. And, you do need to cover it fairly well in your book.
Not being mean... I just have to get other time consuming interests a fair shot, too.
[Kind of feels like a basket ball game... here we are -- running up and down the court and the ball being passed back and forth and then...... the buzzer.]
THx-RNMarsh
Last edited:
So the thread ends .... with a whimper. 
Leaving the movers and shakers to build the better "CMA" ?? 😕
So be it ...
After I read all the op-amp "CMA" material ... I will go in my
own (unique) direction... 🙂
OS
Leaving the movers and shakers to build the better "CMA" ?? 😕
So be it ...
After I read all the op-amp "CMA" material ... I will go in my
own (unique) direction... 🙂
OS
I dont think its ended 😉
We've gotten a lot of designers interested - including you - who are producing some very good stuff.
I am afriad threads like this have there own life - up and down etc.
We've gotten a lot of designers interested - including you - who are producing some very good stuff.
I am afriad threads like this have there own life - up and down etc.
OS,
I think you have to look at Richard like one of those professors you would have that has the answers but won't give them to you. He just points you in a direction and makes you do the real work, including the research. Your doing great, like Bonsai says there are others who are interested in this subject, but you the man, your so damned fast nobody is keeping up with you!
I think you have to look at Richard like one of those professors you would have that has the answers but won't give them to you. He just points you in a direction and makes you do the real work, including the research. Your doing great, like Bonsai says there are others who are interested in this subject, but you the man, your so damned fast nobody is keeping up with you!