CFA Topology Audio Amplifiers

Jump to Latest
#1,361
if I'm guessing right about the focus of the question:

the advantage is that the cascode Q gives a gnd referenced V termination for the compensation C - introducing a gnd reference changes the relation

shunt compensation to gnd should have similar PSRR improving effect, again mostly independent of loop gain
 
#1,363
:cop: Please read the title of this thread. CFA Topology Audio Amplifiers

People who persist in posting nothing but put downs and statements that VFA is better will receive infractions or bin time!

Start another thread on the subject of which is better (but keep it civil) and leave this thread unpoluted for those that are interested in exploring this topology.

Discussions of potential weaknesses or issues are obviously ok but just harping on about how VFA is superior is not.

Equally people participating in this thread that put out statements about the superiority of CFA over VFA are asking for as much trouble! Keep the discussion to the topic and don't let it get religious!!
 
#1,364
On a bog standard CFA I typically get 50 dB at LF, then an improvement in the 1-20 kaHz region to about 70 or 70 dB and then at HF it degrades again.

At LF it's loop gain related ANC I presume from your input jcx, comp reference point (but at LF I expect comp cap value is too low?)
Mid band rail decoupling kicks in
At HF cap ESR is the limiting factor.
 
#1,367
About PSRR, why not design a differential input in a CFA. It will present severally advantages, at the expense of 3dB of noise.
- Equal LTP on common mode PSRR.
- A symetrical cabling is always cleaner in regard with ground leakages ( common mode).
- It offers a bridge operation for free.
- It makes an amp ok for both professional and consumer markets.

But, anyway, can-we be wrong if we filter power ripples at the source, where it matters: Input and VAS. Not complicated neither too expensive to clean <20mA.
 
#1,374
dadod said:
What exactly is sound quality?
Click to expand...

- Clarity
Clarity is the prime acoustical goal because its perfection depends on the successful attainment of all other goals. Of paramount importance is dialogue intelligibility in movies, but one must be able to understand musical lyrics, detect quiet background details, and distinguish the timbre of each instrument. Elements that affect this goal are varied including equipment quality, room decay times levels, ambient noise levels, early reflections and room modes.

- Focus
The ability to precisely locate each reproduced sonic cue or image in a three-dimensional space is defined as acoustical focus. Recordings contain many such images superimposed side to side (2 channel music) and front to back (multichannel music & movies) in every direction for 360 degrees around the listener. A system is said to have pin-point focus if, from the perspective of the listener, each of these images is properly sized, precisely located, and not wandering. Good focus also provides that individual images be easily distinguishable from amongst others within the limits of the recordings quality.

- Envelopment
An audio system should reproduce virtual images of each recorded sound presenting the listener with its apparent source location in a three-dimensional space. Each sonic image relates a part of the recorded event and together these sounds compose a wrap- around soundstage that envelopes the listener. Proper envelopment requires that the soundstage be seamless left to right (2 channel music) and front to back (multichannel music & movies) without interruption by holes or hot spots caused by speaker level imbalance or poor placement.

- Dynamics
Dynamics is simply defined as the difference between the softest and loudest sounds reproducible by a sound system. While much emphasis is placed on the loudness side, it can be shown that the audibility of the softest sounds is an equal measure of system performance. If low level signals are overwhelmed by excessive ambient noise or reverberation in a room, they will not be audible due to masking effects and will impact focus, envelopment and clarity. At a minimum, a system must be capable of reproducing loud passages with ease and without excess while soft sounds remain easily audible.

- Response
The frequency response of a system is a measurement of the relative levels of all reproduced audio frequencies. The smoothness of response can be observed in a variety of ways; as improper tonal balance including boomy bass, excessive treble, improper musical timbre, or a general lack of realism. Factors of importance include selection of high quality components, and proper system set-up including (in a small room) proper listener position, speaker position, and correct use of equalization. At a minimum, the system must be non-fatiguing all sound levels, articulate and faithful to the original signal.

Shunt compensated CFAs excel at all these attributes, they simply can't be beaten by any other topology.
 
Last edited:
#1,379
I think the most important conclusion one can draw is that when circuits are 'simple', the CFA takes lead with the least circuit complexity. Even a bare blameless is a little more complex component count wise.

However, when you start to push for THD, it no longer matters whether you employ CFA or VFA - the OPS becomes the dictating factor. From here on either VFAs and CFAs require a lot more complexity, just look at Dadod's work trying to get into the ppm region for distortion. The same can be said for VFA; my new VAS is complex, though less complex than Dadod's CFA schematic. THD performance wise there really isn't any difference - if I shoot for Dadod's Phase-/Gainmargins, I end up with similar THD, in the 1.5 ~ 2.5ppm range.

Can it be we're actually listening to compensation rather than topology and/or device? I.e. the compensation changing the sound of music?
 
You must log in or register to reply here.
Top Bottom