Re: Re: emitter follower triple
Sorry to disagree Charles
It is true there is more than THD + N in the story, audio perception has been exquisitely tuned through the evolutionary process to the point even a below average listener can easily tell appart different known individuals speaking in the midst of noise, while even the best high tech gear cannot do decently even under favorable conditions. Yet this does not mean certain carefully rationalized measuring methods cannot hope to provide performance parameters which can be reliably extrapolated.
The comparison with elementary particles is not correct either. There was once an explosion of "elementary" particles in the 50's and 60's, which led to the Standard Model whereby the existence of three an only three generations of 2 quarks and 2 leptons each (ordinary matter is made from particles of the first generation) and this has been amply confirmed in high energy particle accelerator experiments. Sorry, out of topic.
Rodolfo
Charles Hansen said:
Sorry to disagree Charles
It is true there is more than THD + N in the story, audio perception has been exquisitely tuned through the evolutionary process to the point even a below average listener can easily tell appart different known individuals speaking in the midst of noise, while even the best high tech gear cannot do decently even under favorable conditions. Yet this does not mean certain carefully rationalized measuring methods cannot hope to provide performance parameters which can be reliably extrapolated.
The comparison with elementary particles is not correct either. There was once an explosion of "elementary" particles in the 50's and 60's, which led to the Standard Model whereby the existence of three an only three generations of 2 quarks and 2 leptons each (ordinary matter is made from particles of the first generation) and this has been amply confirmed in high energy particle accelerator experiments. Sorry, out of topic.
Rodolfo
maxpou said:
That's what I used to think also. Our first amplifier, the V-3, only had three transistors in the signal path, an input transistor, a folded cascode, and an output device.
But I found that adding two more transistors to the output stage (emitter follower triple) sounded better than a single MOSFET. Then I found that adding additional transistors to the input stage to stabilize the operating point of the input transistor also improved the sound. So my designs are becoming more complex, but keep sounding better. But I still avoid the use of feedback.
Hi Pavel,
I don't think Rodolfo was disputing the importance of listening tests, but rather my assertion that we may never have measurements that are as good as listening tests.
I understood him to say that one day we will, but I am personally skeptical. And besides, in the time it took to develop new tests that accurately reflected an amplifier's actual sonic performance, I could design dozens of new products the old fashioned way. So for me, there is no motivation to try and develop such tests.
Regards,
Charlie
I don't think Rodolfo was disputing the importance of listening tests, but rather my assertion that we may never have measurements that are as good as listening tests.
I understood him to say that one day we will, but I am personally skeptical. And besides, in the time it took to develop new tests that accurately reflected an amplifier's actual sonic performance, I could design dozens of new products the old fashioned way. So for me, there is no motivation to try and develop such tests.
Regards,
Charlie
Re: Re: emitter follower triple
Hi Charles,
Good point. Indeed, almost everybody measures that (stupid) THD+N, instead of measuring the pure distortion i.e. without noise. Looking at a distortion residual, which is not cluttered or totally swamped by noise, might provide much more information and could explain why a particular amp sounds so good (or worse).
Cheers, Edmond.
Charles Hansen said:
Hi Charles,
Good point. Indeed, almost everybody measures that (stupid) THD+N, instead of measuring the pure distortion i.e. without noise. Looking at a distortion residual, which is not cluttered or totally swamped by noise, might provide much more information and could explain why a particular amp sounds so good (or worse).
Cheers, Edmond.
PMA said:
I do not Pavel, and Charles understood what I meant.
Furthermore, as far as we do not currently know for sure how well measurements reflect the full extent of relevant disturbances in reproduction - and mix in "nice" disturbances to make matters worse - we should approach listening both with and open mind and a healthy dose of suspicion. Neither deny listening if it does not match measurement, nor take it as the only standard of reference - we are too easily fooled -.
Rodolfo
Nelson Pass said:
...The argument that designers deliberately add
distortion for a particular sound is a straw man. ....
I do not care whether this is a common practice or not. Something that has been often repeated by credible people for example, is even harmonics add a pleasing coloration.
Recalling the harmonic structure of a given note is what gives timbre, this should not be a great surprise after all, as long as IM does not fly through the roof.
If on the other hand what drives one's effort is to achieve the most faithfull reproduction, then a healthy wariness about what our senses can do should be something to take into account when listening testing some new concotion.
Certainly most of you do give measurement a significant place in the toolkit.
Rodolfo
ingrast said:
I would not say it in this way. I would say that low order even harmonic (2nd) /and also low order odd - 3rd/ is considerably less audible than high order even or odd harmonics of the same and also lower level. So it does not make much sense to heavily suppress 2nd and 3rd and have increased 7th and 11th.
Imagine C chord: C1 = 261.6Hz, E1 = 329.6Hz, G1 = 391.9Hz. Now 2nd order difference tone E1 - C1 = 68Hz. 68Hz is just about half tone between c (65.41Hz) and d (73.42Hz). This half tone does everything but fits to original chord . It is only a question of level if it is audible .
But, undoubtly, it will make the original chord "richer" (I hope I will be forgiven )
But, undoubtly, it will make the original chord "richer" (I hope I will be forgiven
)
Pavel, you have used the wrong scale Within the well-tempered scale (based on 2^(1/12) steps) you will have this wrong tones already from the instrument itself when each chord note is not a true sine. I play electric Guitar and Bass (both with quite a bit of distortion) and there you need to bend the chord notes properly to get the right (sub-)harmonics from the distorion. And I design guitar amps and have found even harmonics do way better than odd ones, regarding "transparency" of chords with "close spacing".
BTW, these sub-harmonics, especially with bass guitar, are one of the reasons why we need a lower than 40Hz HP corner, which is often doubted by people who say "there is nothing below 40Hz in rock music".
Within the well-tempered scale (based on 2^(1/12) steps) you will have this wrong tones already from the instrument itself when each chord note is not a true sine. I play electric Guitar and Bass (both with quite a bit of distortion) and there you need to bend the chord notes properly to get the right (sub-)harmonics from the distorion. And I design guitar amps and have found even harmonics do way better than odd ones, regarding "transparency" of chords with "close spacing".BTW, these sub-harmonics, especially with bass guitar, are one of the reasons why we need a lower than 40Hz HP corner, which is often doubted by people who say "there is nothing below 40Hz in rock music".
Or you could play a six-string bass with a low-B.
There are actually very few bass cabinets that produce a low-E (42 Hz) with anything like flat response. Far too many cabinets quote "usable" low end, by which they mean -10dB. I don't know about you, but -10dB doesn't count as usable in my book. Four-string basses get by...barely. Five or six-string basses are S.O.L. when it comes to getting good low end.
And that doesn't even begin to present the case for drums and keyboard instruments. Orchestral music is another matter, entirely.
Arguing about distortion and which types are more "desirable" than others is a waste of time. Broadly speaking, it comes down to those who regard any distortion (regardless of type) as anathema, and those who will make allowances for some arbitrary amount of distortion if the compromise will allow them to achieve some other goal.
If you persist in this argument, it will run for pages and solve nothing.
Grey
There are actually very few bass cabinets that produce a low-E (42 Hz) with anything like flat response. Far too many cabinets quote "usable" low end, by which they mean -10dB. I don't know about you, but -10dB doesn't count as usable in my book. Four-string basses get by...barely. Five or six-string basses are S.O.L. when it comes to getting good low end.
And that doesn't even begin to present the case for drums and keyboard instruments. Orchestral music is another matter, entirely.
Arguing about distortion and which types are more "desirable" than others is a waste of time. Broadly speaking, it comes down to those who regard any distortion (regardless of type) as anathema, and those who will make allowances for some arbitrary amount of distortion if the compromise will allow them to achieve some other goal.
If you persist in this argument, it will run for pages and solve nothing.
Grey
Grey, I play low-B (4-string fretless, B-E-A-D), and my live cabinets roll-off at 65Hz, which is useful and choosen deliberately. You and the man on the mixing desk do never ever want to have 30Hz at full power on-stage as this gives a completely muddy sound. I have another cabinet that goes down to 35Hz which I use for small, non-PA-supported gigs.
- Klaus (sorry @all for being OT)
- Klaus (sorry @all for being OT)