*
EMF (Electro motive force) ?
You mean it (speakers) work as well as a generator or as a motor.
EMF (Electro motive force) ?
You mean it (speakers) work as well as a generator or as a motor.
It has been the habit of most in audio to make each following stage better in every possible aspect with the idea being that only the first stage will significantly contribute to the overall performance of the whole.
This is new, and I wonder why it is not applied in other areas of analog electronics.
Why not the other way around? Why carry the first stage gunk all along the audio chain?
Some of us have ideas, so I am not speaking for every one however
the need for excess slew rate is an unknown.
It is thought to be a marker of some unknown parameter.
Many of us hear a difference.
Thanks, Mooly.
If i remember well, there is some subs using ventilators (with motors) for ultra low frequencies ? ;-)
Well, you know, to reproduce a square wave front edge, we have to accelerate the membrane, then break-it back ... waterfalls...
If i remember well, there is some subs using ventilators (with motors) for ultra low frequencies ? ;-)
Well, you know, to reproduce a square wave front edge, we have to accelerate the membrane, then break-it back ... waterfalls...
Counter-electromotive force (counter EMF, or CEMF)
After filtering out the noise:
- Speakers injecting CEMF, and speaker cables injecting EMI, in the amplifier feedback loop were proved as red herrings a long time ago (including on this forum), at least for amplifiers with a decent damping factor.
- Looking at the VAS input you won't see a damn thing, except for the error signal in the feedback loop.
- Both of the above have absolutely nothing to do with the slew rate.
- Nobody suggested 741 for audio.
- All these (reactive loads, CEMF, speaker cable, slew rate) are not at all difficult to model. It is questionable if it's required to do so, for a good SQ.
Nobody is debating the decades of sweat and effort, although the school books would probably save a lot. Unfortunately, and as with another famous designer on this forum, these decades of sweat and efforts may have little to do with the technical truth.
Waly I think you need to take a timeout from this thread. You can do it voluntarily or it can be enforced (ie sin bin). Arguing over a point continuously is only serving to disrupt the thread. Cristophe, I suggest you DO ignore any pot stirring from Waly, your responses just serve to keep the disruption going. 
Richard I don't think that this is necessarily the case. Low distortion at high audio frequencies means generally a high feedback factor at those frequencies. That does not necessitate high bandwidth. You can have high feedback factor at 20kHz and relatively low bandwidth - it does mean that the feedback factor at lower frequencies is even higher. Many opamps with exceptionally low distortion at 20kHz (or even at 100kHz or more) have quite low bandwidth.
Jan
. See ya' later 
.In general I agree as has been done many times. never-the-less... the key word is high feedback. I have said before, i have nothing bad to say about HNFB per se. It just isnt the only way to get low distortion nor at high freqs.... the CMA in my designs are a LOT easier to stabilize even though they end up high bandwidth. Its just another route to explore its strengths and see how it turns out and some new learning along the way. It is NOT a contest of VFA vs CMA.
Thx... Richard
Last edited:
🙂Its near Halloween over here ---
You can't scare me ---- I have three daughters.
Dangerously
under-medicated
The trouble with trouble is
IT STARTS OUT AS FUN
You look like I need a drink
Hard work must have killed Someone.
🙂
Don't take life too seriously - No one gets out Alive
-RNM
You can't scare me ---- I have three daughters.
Dangerously
under-medicated
The trouble with trouble is
IT STARTS OUT AS FUN
You look like I need a drink
Hard work must have killed Someone.
🙂
Don't take life too seriously - No one gets out Alive
-RNM
Last edited:
How can-it be the case, jan, when the error correction is deeply canceled by the slightest phase error ? (Real question).
Even with the fastest power amp we try, we hardly achieve 5KHz open loop bandwidth.
Last edited:
Richard, I think you need a combo of
Low open loop non- linearity
Application of suitable amounts of feedback at HF
Both VFA and CFA deliver these requirements. However, I think it is easier in CFA, primarily because of fewer gain stages ( less open loop distortion and less phase shift). Further, as we've seen in this thread, the diamond is inherently very linear without degeneration.
Low open loop non- linearity
Application of suitable amounts of feedback at HF
Both VFA and CFA deliver these requirements. However, I think it is easier in CFA, primarily because of fewer gain stages ( less open loop distortion and less phase shift). Further, as we've seen in this thread, the diamond is inherently very linear without degeneration.
Esperado said:
I perfectly agree. Bob Cordell demonstrates the need of a decent maximal slew-rate, but he certainly does not establish a correlation, above it, between the quality of an amp and its maximal slew-rate, contrarily to what has been suggested many times in this thread.
The shapes of the VAS input current in high NFB circuits look really awful and disturbing only to unwarned persons. What matters is that the NFB does correctly its job and that the input stage is always able to linearly deliver enough current to the VAS input.
Back-EMF is not seen per se by the amplifier output, it just appears as an impedance increase of the load. And then, as the current delivered is then decreased, the amp distortion also decreases...
Whatever the origin of unwanted signals at the output, internal or external (back-EMF, microphonic effects of drivers), the negative feedback reduces them in due time according to well known analyses.
It must be noted that, apart from the resonance region of drivers which needs electrical damping, current driving is beneficial in many aspects while voltage driving is not.
Difficult loads for amps are only those which present very low impedance. Reactive loads do not disturb at all correctly designed amps.
Not sure that fast and/or CFA are less immune to EMI/RFI than others.
Curiously my Blameless has no input low-pass filter, and not even a low value resistance in series with the base of the non-inverting input transistor and it never showed sensitivity to high frequency interferences.
See Cyril Bateman's papers on cables for cures RFI .
Distortion, bandwidth and slew rate are all knobs the designer can adjust in the design process. There is no correlation beyond a certain point that driving any of these parameters in excess will deliver a 'better' amplifier. This forum is full of earnest efforts to get to 1 ppm distortion and less, more often than not at the expense of huge complexity, or compromises elsewhere in the design like stability for example.
When we stop this numbers chasing game, and instead focus on more balanced designs, we will make progress again. Not before then.
I have remarked somewhere else that over the last few years here we've replaced the nonsense of subjectivism with the tyranny of numbers. A big dose of pragmatism is sorely needed.
When we stop this numbers chasing game, and instead focus on more balanced designs, we will make progress again. Not before then.
I have remarked somewhere else that over the last few years here we've replaced the nonsense of subjectivism with the tyranny of numbers. A big dose of pragmatism is sorely needed.
As myhrrhleine has helpfully reminded us, a slew rate number is a marker for something we're not too certain about in terms of subjective SQ - probably HF linearity. Ditto THD - its another marker for linearity.
Its daft (but amazingly common) to confuse the markers with the reality - the map is not the territory.
PS -- I was refering only to a CFB Amp circuit.... when obtaining low distortion via linearizing the CFA, i end up with wide band circuit by default.
--- many opamps etc --- they are very complex circuits inside and often use a combination of techniques... too much for such a broad statement to be made.
-RNM
I saw the circuit on internet about 13 years ago, and I modified it. I designed it with CADIX software on HP-UX OS at company that I worked. (But I designed it for my self 😀). Now I am not working at that company and forgot where are the schematic. Don't ask me to draw it from my amp!
It has already proved at simulation on this thread. Please someone build it on real amp and compare to VSSA or Bonsai's amp (Nx or Sx?). I think THD is not the only one that influence the linearity.
I think an amp must design for linearity and we want sound is reproduced accurately. Is amp with THD below 1 ppm sound same? What for to go below 1 ppm? What for to use CFA?