My, my.... Now the debate has moved over from CFA vs VFA to the sound of the 'system' or the entire signal chain!!
Thanks Bigun.
It is quite clear that there are those who clearly belong to only one camp and cannot or will not tolerate even discussions with/of the other camp. This is not leading to anything productive. Each camp should practice some restraint and allow for a healthy, learning discussion.
Just my 2 cents.....not advice.
Thanks Bigun.
It is quite clear that there are those who clearly belong to only one camp and cannot or will not tolerate even discussions with/of the other camp. This is not leading to anything productive. Each camp should practice some restraint and allow for a healthy, learning discussion.
Just my 2 cents.....not advice.
Member
Joined 2009
Paid Member
Hi Sam, not sure if you meant 'thanks' or 'thanks for nothing' ? - but I was agreeing with you.
I would suggest again that there may be more interesting differences between CFA and VFA when looking at single ended inputs. A single LTP vs the Singleton.
I would suggest again that there may be more interesting differences between CFA and VFA when looking at single ended inputs. A single LTP vs the Singleton.
Last edited:
One thing you can take for granted is that all the people who try to go further in this CFAs exploration are familiar to VFAs, had several in their life, if not one working in their system.
As no one from the anti CFA gang never answered when they were asked witch CFA they had or have, you can make your conclusion.
You will NEVER see any of the VFA lovers going in any VFA thread to claim endlessly VFA is junk.
NEVER.
I disagree (a personal point of view).
As in audio, optical tests of lenses can only help-you to avoid very poor parts (witch can have their charm). Much modern lenses with aspheric lenses are very disappointing. flat, clinical, too much contrast for the captor (details lost in the in darks, burned whites), etc.
As in audio, you need to use your eyes with each lens *on your photographic body* to figure out what will be your preferred tool. My sharpest and favorite manual lens (57mm F:1.2, a pure Jewell) don't measure so good than some modern autofocus lenses i have and never use.
As it audio, it would be so simple to lie on the objectivist side, measurements etc... Unfortunately, measurements tell only one part of the story...
May-be because parts behind lenses (body), as well as parts after an amplifier (enclosures) are far to be perfect and need some kind of 'matching', 'compensation' etc.
After all, Hifi is made to listen to music, and our ears+ brains are our measuring instruments, as well as eyes+brain for photography.
The main difference is in photo, you can compare with the two in front of you in the same time, while in audio, you can only compare one after the other, on time related music, means your memory is involved twice.
In both stereophonic musical reproduction and photography, we play a *make believe* game and create a 2D fake universe from a 3D reality.
On the objectivist side, running after 1ppm HD to send a signal in enclosures with > 10000ppm HD, do it makes any sens ? Everybody knows it do.
Last edited:
Member
Joined 2009
Paid Member
hmmm, a bit OT but I disagree - it seems that with a very good lens the contrast is then limited by the film/digital sensor. If you rely on the defects of the lens such as low contrast, to smooth over the inadequate dynamic range of the light sensor you are doing what in audio world is using soft-clipping distortion to mask limited dynamic range. And you may prefer the result all the same.
Let me know when we can get back to designs of CFA, pls. I think we can go to the next level, now.
-RNM
-RNM
Richard maybe its time to show some Jfet input CFA designs, simlar to the topology you used in your headphone amp. Pioneer made great success in the 80s and until recently before now changing to the use of class D amps.
Oh, yes, please, i have a lot to learn with-it.
While i believe Richard will be be soon very occupied to assemble his VSSA if he find some magnifying glasses ;-)
Attachments
Please excuse me if I dunno what you are talking about. I'm after the VFA which you and wahab believe have performance that can't be achieved with a CFA of simpler or equal complexity .. in loadsa stuff including PSR.
Many people have said CFAs can't equal VFAs in stuff like THD and noise. We now know this a myth and CFAs can do usually do this .. with less complexity too.
I want to lay the PSR myth to rest but would like a 'real life' (OK SPICE world) target from you or wahab to try and meet.
Du..uuh! Strange that with this professed attitude, you have such a poor opinion of people who have done exactly that.
Alas, this Millenium, this beach bum is unlikely to do what he would have done in his previous life .. which is to
- build
- measure
- AND conduct Double Blind Listening Tests
However, as its likely you can't achieve zillion dB PSR and other parameters in 'real life' if you can't demonstrate them in SPICE world, please let me play too by showing me your VFAs with dis supa performance.
Bigun I started my Adventures in SPICEland early this year cos I wanted to replicate the sound of a 1969 JLH Class A amp using a Class B circuit. My contention is that if you can replicate its THD profile, you can go a long way to replicating the sound of an amp.Bigun said:
Guru Carver, whom I regard as one of the few really innovative minds in Amp design demonstrated this with some of the most Golden of Pinnae (who might actually be true golden pinnae) http://www.stereophile.com/content/carver-challenge
IIRC, a 1969 JLH appeared in at least one of my DBLTs. It wasn't accurate but it was liked on small scale music. I hesitate to call it a 'CFA' in this august company. 🙂
I'll start to put some things together I've been thinking of and do a development of the next topology level which we already know but it follows from where we have been and then-some. Gimmie a day or two cause I have to find everything/search and gather and copy/paste etc.
In the meantime, don't get into any trouble. 🙂
-Thx RNMarsh
In the meantime, don't get into any trouble. 🙂
-Thx RNMarsh
Member
Joined 2009
Paid Member
I have a 69 JLH waiting to be finished, I read only good things about it. But I would be quite happy to call it a CFA in any company 😀
With a symmetrical amp the differences between LTP and singleton are much less obvious because a) the OLG is higher so you are traversing a small part of the I-V curve and b) the main difference between LTP and singleton is in the 2nd harmonic which is nulled out in a symmetric design.
I consider Hugh Dean a guru in his area of focus - he knows how to tweak up the right distortion profile and he prefers the single ended approach, singleton CFA for his recent NAKSA.
Hugh also did a pcb for Fetzilla, CFA with JFET input that has been very much liked and widely built.
Last edited:
Just some friendly support from the sideline to keep this thread going to ultimately produce a CFA design then build it, listen to it, compare it.
For reasons bigun mentioned plus thermal and bias stability I will be implementing a single ended version with CCS. I have simmed a number of these based on symmetrical designs posted and they work very well.
Thanks to those who have contributed to this thread. I have learnt a lot and look forward to the outcome!
For reasons bigun mentioned plus thermal and bias stability I will be implementing a single ended version with CCS. I have simmed a number of these based on symmetrical designs posted and they work very well.
Thanks to those who have contributed to this thread. I have learnt a lot and look forward to the outcome!
Enclosures don't use GNFB. They produce acoustic dBs, not volts.
Well, your question have interest, thinking about EMF/CEMF in the feedback loop...
We need some measurements in real conditions. Here is, may-be, the key to this controversy.
Behind this question, I realized i had no idea of the max level transients my horns are able to produce at 16 KHz:
> 160dB for sure. Enough to kill-me.
Last edited:
Its qualifying adjective is blameless.
I cannot imagine the VFAs I built without serious matching of the transistors of the same polarity in the differential stages
I cannot imagine push-pull CFAs without serious matching of the transistors of the complementary pairs.This is more complex and time consuming than for VFAs, so I do not consider CFAs are more simple than VFAs despite their ascetic aspect.
Which reality has shown you this is not a myth ?
Hi Bigun !
I think that old & simple NAD 3020 Singleton CFA somehow fit in this right distortion sonic profile to .
Best Regards !
Member
Joined 2009
Paid Member
I ran some simulations to look at the CFA vs VFA version of my TGM7 (SKA) and found that the distortion is really sensitive to any mismatch in the feedback shunt resistor for CFA - just a 0.5% imbalance pushes it up a lot. The VFA version was far less sensitive and even a 10% imbalance between the LTP collector loads only sweetens (H2) the distortion profile. Something to ponder.
Hi ! - yes there are some good examples of singleton CFA out there. But it becomes a symmetric vs asymmetric discussion as I think this is of more significance than the CFA vs VFA. I think the CFA is giving more benefit in the single ended approach than a symmetric one.
Last edited:
Several of you need to go back and actually read what the definition is of CFA. Both mathematically and otherwise. If it meets that definition then fine... Those I read here, recently, do not qualify. Feedback to an emitter/source/cathode does not qualify all by itself.
Thx-RNMarsh
Thx-RNMarsh
VSSA is already improved, It is first using tl06x opamp input and discrete IGBT and already looks good sounded one, LC improve it from that times. Good amps are improved with listening, not just seeing some pictures.
About matching N-P or symetrical, I think it doesn't matter, I am using totally asymetrical design with very good result.
Improving with more limitation is ended in more limited result, but it all depend on the need.
The common is already improved from that jurassic time by many comercial companies, including NAD, cyrus, etc. It is better start to learn from their schematics than start from zero.