In my experience, it were the inexpensive 'technically correct' amplifiers which gave the least satisfaction to the user.
Ok take my words apart in statements and feed it to Wordfeud.
Fact is th at amplifier design moves very little in terms of structure and methods.
Same concept is reinvented over and over again..
Damir, you're right much inspired from your circuit. Paired with my OPS I would make one hell of an amplifier. Where HEC only really works with loads close to ohm(ic) mine is much more resilient, but of course also based on feedback
Fact is th at amplifier design moves very little in terms of structure and methods.
Same concept is reinvented over and over again..
Damir, you're right much inspired from your circuit. Paired with my OPS I would make one hell of an amplifier. Where HEC only really works with loads close to ohm(ic) mine is much more resilient, but of course also based on feedback
Yes, the very fact that we hear differences clearly shows that we are unable to measure amps in a way meaningful enough.
That's not an article. The claims are false, largely the one about TIM, because transient intermodulation distortion cannot exist without slew rate limiting.
This article is suspect. It lacks any references from case studies to support its claims. There is also an erroroneous omission on the first page. The primary application of feedback is to extend the bandwidth by trading gain, and in that way, feedback stabilizes linear and non-linear distortion. This "article" waits until the 9th page to clarify what should have been stated from the onset.cwtim01 said:
"Accusations are occasionally made that objectivists can't hear, and conversely that subjectivists hear things
that aren't there... Given the complaints of audiophiles over the sound of high-feedback type amplifiers, it is reasonable
to examine non-linear distortion in greater depth than is possible with a single number."
It's not examining if he's not conducting a qualified group study.
"To get lower order harmonic character we want smoother “bends” on the device's transfer
curve. This usually means Class A operation..."
This is where it really gets murkier. The comparison is between class A and B, which is a smoke and mirrors act. Nobody uses a class B amplifier for audio. They're class AB.
He then goes on to perform an experient that shows NFB increases high order harmonics. Is this a joke? If the intention was to design a poor amplifier, he succeeded. Look at the levels at 1W into 8 ohms load. This is perfect for confusing readers and making think what he wants.
is it your assertion that nfb does *not*increase higher order harmonics at the expense of lower order ones?
Are you saying that "you" can hear things that is not possible to measure?
I don't believe for a second that there is anything a human being can hear that can not be captured and explained by proper analysis.
There is not one single evidence of the contrary afaik.
Explaining "good sound" with a single number we will never be able to do. Detect (with measurement gear) any changes in voltage, current and SPL that "we" can hear is totally possible though.
This is not magic, it was physics ten eyars ago, fifty years ago and still is.
But it sounded so nice!!! It's not only me, the store I went to, use that as the reference amp in the room for their highest end speakers. No speaker in that room was less than $4000 or $5000.....in 1998.
I never even gave it as second look as it did not even look expensive, kind of small. (but it was over $3000!!!).
BTW, so which IPS/VAS gave you the warmer and more musical between the CFA and VFA? That's the very thing I love about the YBA. My Acurus sounded cold and thin next to the YBA. That's what I am looking for, one that is more warm and musical. Please don't let me hanging!!!😀
That does not reflect my thinking. At Pass Labs I use moderate amounts of
feedback. No PL product has "no feedback", nor do I advocate that as a
general principle.
Thank you.
The two amps that I found sterile were Krell and Mac. I did not know the ins and outs of amps at the time, I was just a customer. But I clearly remember thinking "what's all the fuzz about them, so expensive!!" So, that's my preference, and more important, it's my money!!!
Now I studied their schematics, both really go out of their way to lower distortion, all the bells and whistles, all the global and local feedback.............
Music recordings are all distortion to start out with. The biggest distortion is the mic. I was a musician for over 10 years, we recorded music. It sounded nothing like live. Absolutely nothing. You listen to people playing violin or sax, you'll know what I mean. Particular the violin, they sound scratchy and harsh, but in recording, they sound silky smooth. Even human voice, listen to the singing without mic and listen to the recordings, they are no where similar. I think people are fooling themselves thinking those CDs are fidelity. It's all about effect.
Then you have the mixing and enhancement of the recording before putting on the CD.
Then the "sound stage" people talking about. What sound stage? A lot of the music recorded in a tiny little room, recorded separately one track at a time. What stage? It's an illusion.
Last but not the least, I listen to modern jazz and singing. I like piano and sax, I watch tv with the stereo. I don't like string music music like violin etc. I don't care about a system that reproduce fantastic sound for those. AND the more important, it's my money.
Last edited:
NFB does increase higher order harmonics. Say you have a circuit that produces second order harmonic distortion, then by applying NFB you will eliminate* the second harmonic at the output, but now the circuit produces second harmonic distortion of 2nd order correction signal producing a 4th harmonic. Then you get a 2nd harmonic of the 4th harmonic correction signal producing an 8th harmonic and so forth and so forth.
*only if you have infinite open loop gain and therefore can apply infinite feedback. In practical circuits the open loop gain is finite so while you get a great reduction in the magnitude of distortion components, they will not be completely eliminated. Therefore a small amount of the original harmonic distortion will remain along with some small amounts of higher order components.
That said, with proper implementation NFB is better. A circuit with only local feedback producing a 2nd order harmonic at -30dB, 3rd order at -40dB and 5th order at -50dB is far more objectionable than a circuit with NFB which produces harmonics from 2nd to 30th at -120dB.
Here is an amp that I modified by bringing the output stage into the GFB loop. Tell me which one has more objectionable harmonics, the before (local feedback only) or the after (LF+GFB):
Before:
After:
*only if you have infinite open loop gain and therefore can apply infinite feedback. In practical circuits the open loop gain is finite so while you get a great reduction in the magnitude of distortion components, they will not be completely eliminated. Therefore a small amount of the original harmonic distortion will remain along with some small amounts of higher order components.
That said, with proper implementation NFB is better. A circuit with only local feedback producing a 2nd order harmonic at -30dB, 3rd order at -40dB and 5th order at -50dB is far more objectionable than a circuit with NFB which produces harmonics from 2nd to 30th at -120dB.
Here is an amp that I modified by bringing the output stage into the GFB loop. Tell me which one has more objectionable harmonics, the before (local feedback only) or the after (LF+GFB):
Before:
After:
Last edited:
I don't quite agree. The Baxandall classic diagram, which Bob uses in 24.3
and is reproduced below starts out with the assumption that the distortion
without feedback is 10%, which I think we can agree is exaggerated. If the
distortion without feedback is reduced, then we see less of an effect,
certainly illustrated by figure 24.5 which is about 0.3% without feedback.
However it appears that 24.5 is a Class AB follower, in which case lots of
degenerative feedback is already in play.
Bob does go on to say about Baxandall's graphic, "...the audio community
read too much into it, wrongly generalizing the results..."
My view: The more linear the amplifier, the better feedback works, but
also the less need for it. Basically Bart Locanthi's point back in 1966.
Attachments
Don't forget the effect on output impedance and, hence, frequency response into a loudspeaker load. Distortion isn't the only thing out there.
It all depends on what the purpose is. If it is to reproduce music, then sufficiently good measurements are necessary. If it is to please the owner/builder then anything is possible because there is no accounting for taste.owdeo said:
NO! The issue is finding out what sounds most like the original sound, for those who want sound reproduction. Much of the work on this was done 50 years ago, which is why we know that we need at least a certain frequency response and not too much distortion.
An excellent example of 'if you can't argue with someone's position then misstate it and then ridicule that'.
Are they really "hifi" mags, or just audio mags pretending to be about sound reproduction? Who says they are the "top" mags? Who rates the amps?
First, you would need to find someone who actually holds that view and ask them why they hold it. I will not hold my breath, because you could be away a long time searching for such a person. Possible exceptions: students who have just finished their first module on audio electronics, so they have learnt the basics but think they know all there is to know.cwtim01 said:
Who says that 0.002% of THD is important? I'm not sure even Doug Self would say that. Why bother listening to amps with horrific levels of THD? Any competent designer can do a tube amp with sufficiently low THD - if that is what you are interested in. THD has value as a rough and ready guide to whether you have done a competent design and correctly built it. Some tube amps are FX boxes; some are not. I like them because my eyesight is not what it was so big chunky components are easier to solder.AKSA said:
I don't care at all, provided that these people don't try to tell me that their preferred distortion is somehow more accurate than my preferred smaller distortion. Sadly, most of them seem to feel the need to bolster their own position by denying that they prefer distortion.Alan0354 said:
Not really a paradox, unless the amps are incompetently designed. It just shows that some (many?) listeners don't want transparency. They are not into 'hi-fi' - they just think that they are.ivanlukic said:
Who is "we"? Can you find any serious amp designer (commercial or DIY) who actually does this? Why can't you criticise what people actually do/believe, instead of knocking down a crude caricature?MiiB said:
There was a time in the 1980s when I could attend a BBC Promenade Concert one evening and then the following evening hear the same sound on BBC Radio 3. A live concert relay back then sounded very like the real thing. That is because back then the BBC sound engineers were concentrating on sound reproduction and so achieved a high degree of success - within the constraints of the technology. If a violin sounds harsh in real life (which may mean something wrong with the violin or the performer) then a hi-fi enthusiast would want it to sound equally harsh in a recording.Alan0354 said:
I agree Nelson, but that also means that when you´d have an AB (or for that matter a class A) power amp without feedback, and you start to add feedback (while increasing loop gain of course) you most probably won´t see the effect at all, surely not with appreciable feedback.
In that sense comparing ´spectral growth´ between a single FET stage and an already reasonable class (A)B amp is comparing apples with long nose pliers. 🙂
Jan
I can only speak for myself, I can't speak for anyone else. It's my money to buy/build my amp. I just join in and give my position. But I do have to ask whether people listen because it sound good to them, or they only appreciate no distortion.
In concert, everything is mic. You hear through the distortion of the stage and the mic. I was/am a guitarist, we even have long discussion what position of the mic for what effect of the recording. We measure what is the axis of the mic, how many inches off the dust cover and experiment in recording the guitar sound. Everything is artificial. I am referring to if you listen to a band without any PA and listen to the raw true music. It sounds so different.
I was into Santana back in the 70s, I went to his concert 5 times. His guitar is known for it's silky smooth sound. But one time, I managed to stand on the 5th row, in front of him and the band, I heard his guitar sound directly from his Boogie rather than through the PA. It was fuzzy, bassy and sounded exactly the same as when I tested the Boogie in the music store!!! It's the PA, mixer that did the magic!!!
Mic just do not pick up the harshness of the violin and sax. If you listen to the raw sound without any PA that goes through the mic, it sounds totally different. You will never mistake when you hear the sax and violin raw without mic than in recording. So are the human voice and drums.
Last edited:
High performance microphones are low distortion devices if with distortion we are talking about spectral contamination ie addition of new spectral components.
A real problem though and the deal breaker is geometric distortion... the positioning of mic's and the interaction of the surce-room-mic >>> speaker-room-listener. Microphone polar pattern is extremely important as well... finding/using the right one for the job at hand. And of course on axis FR in combination with overall power response (polar pattern).
Sorry you did not succed in achieving a pleasing result. 🙂 That said CD's are not a significant limitation to hifi (as red book format, if that is what you meant).
Of course it's an illusion, no matter if it's a carefully recorded live performance in a hall or a more "synthetic" studio production. A good stereo handles both the real deal and "reverb in a box" type of recordings in a good way.