My bad, I keep slicing them off, the cutting board is the only place I can see them.
An externally hosted image should be here but it was not working when we last tested it.
I've found in my search for better amp that it's easier, in CFA, to move ULGF up in frequency and still keep very good phase and gain margin.
Attachments
I have made it very clear to use the term 'classic' VFA and 'classic' CFA simply to explain the differences and even showed the two models I used. There's no rocket science or 'subjective' beliefs at work. If you up the OLG in a CFA, you will have to deal with the additional phase shift and probably require MIC or TMC to comp the system - I've also made that abundantly clear and shown plots to explain that.
At the end: Everybody tried to explain you are totally wrong in your hate for CFA and near all your comments about.
http://www.esperado.fr/temp/VSSA/vssa-vs-vfa.html
Those plots had been made with exactly the same amp, one in CFA, one in VFA.
The only difference between the two are the input stage, both with the same gain and same currents.
And you can see that, for a "fixed compensation order", the open loop bandwidth of the CFA is slightly better in the VFA. And this at the price to unbalance the LTP with a feedback path of reduced impedance to reject the low pass filter made with it and the parasitic capacitance of the - input transistor.
If you keep the two balanced (10K is traditional, as a good charge for pre-amplifiers) and compensate both for flat closed loop banwidth, the situation is worse. And the slew rate difference is huge in favor to the CFA, while the difference of PSRR is huge too in favor to the VFA.
Now, you have the right to consider that what happens around 3-5MHz is not important for Audio, And the phases margins are good enough in the two configuration, but it will be only YOUR personal point of view: My LT Spice has no preference.
We are at least 3 people here (Dadod, Bonsai, and me) to had made and published a comparative study between the two topologies. We all (of course ;-) came to the same conclusions whatever our approaches.
Note that we all had designed several amplifiers in both topologies. And, of course we love the children of our two stepfamilies .
So for the last time: be a little scientific for a while. Please, build a CFA, measure-it and listen to it, or stop your comments about a subject you, obviously, don't learned neither experimented.
I believe somebody had once convinced-you that CFAs are snake oil, and now, you engaged a campaign against-it like a fanatical member of a strange sect.
Just one correction, I never compared two topologies and never said that CFA is better for audio, just it sims to me simpler to get the result I was after in simulation and it looks in real life too, as I never had any problem with stability on any load. I can't say what is better as I don't have possibility for comparative testing (not enough precise instruments) or in listening sessions not enough experience (I haven't had chance to do that with someone more experinced). I like my VFAs equally, but than I designed them.
You are right: What I wanted to said is you "experimented" both topologies in your designs. A matter to understand and compare the smell of fingers in the dirty oil;-).
And, you compared with your very innovative and impressive pre-amp CFA VS no global feedback at all, and you can confirm, at least, the differences are not so "dramatic" neither, right ? ;-)
And, Oh, Lord, I forgot to mention Lazy Cat and, most of all, OStripper.
http://www.diyaudio.com/forums/solid-state/248105-slewmaster-cfa-vs-vfa-rumble.html
On my side, I said I tend to globaly prefer CFAs, whith the only arbiter of elegance I trust at the end: Repeated listening impressions during my (yet) long life.
(The unique reason why i tried to make this comparative study to try to understand the reason of my preferences.)
I want to make clear that is a VERY PERSONAL taste, VERY SUBJECTIVE, that I would not impose to nobody.
I would like to add that, all the 10 years, progress in technologies were such that I prefered some VFAs of the new generation to the best CFAs I knew in the last one.
Anyway, the difference is now so little, so sublte, that it does not worth all those controversies ;-)
As I said before, we can chose our poison: no magic at work.
Last edited:
What can I add to your results, proving once again you are not even remotely understanding feedback, since you are using "Open loop X", where X=gain, phase, bandwidth instead of the "Loop Y" where Y=gain, phase and the "Unity loop gain frequency" concept, to estimate the phase margin?
I'll stop here, not before mentioning that your level of knowledge is ultimately not helping the CFA cause at all. I could probably further debate with dadod how and why would a dual pole be anywhere CFA specific, if he analyzed all the 4 or five internal loops for stability, and if what he got to qualifies as "simple", but with you, it's definitely useless. We certainly don't speak the same (technology) language. I apologize if this is not what you would like to hear.
Why I would say that a dual pole is CFA specific, this is a Loop Gain of my VFA built few years ago.
Attachments
Last edited:
- Status
- Not open for further replies.