john curl said:
John, your assertion that Charles does not use error correction because "he has tried it...[but]...it did not make the amp 'sound' any better..." is devoid of merit because we do not know precisely how the worthy attempted to implement the technique-if at all.
Moreover, O Dread Majesty, how do you consistently measure "sound quality" as the "sole criterion for 'real' success"?
mikeks said:John, your assertion that Charles does not use error correction because "he has tried it...[but]...it did not make the amp 'sound' any better..." is devoid of merit because we do not know precisely how the worthy attempted to implement the technique-if at all.
Moreover, O Dread Majesty, how do you consistently measure "sound quality" as the "sole criterion for 'real' success"?
If John says Charles told him that he tried it, I think we can
accept that he tried it "-if at all"
And are you asking if John's technique is consistent, or are you
asking what the technique is?
mikeks said:
Hi Mike,
Knowing what little I do of Charles Hanson's design philosophy of no negative feedback and simpler is better, I would be very surprized if he actually tried the error correction technique. Even if he did and failed, it just means that EC is not for everyone and that not all implementations will be successful. It does not reflect on the technique that someone tried it in an unknown way and it didn't work well. I think John would do better to share his own experiences, rather than alleged experiences of third parties.
Bob
mikeks said:
How odd, i had the impression that Bruce halcro Candy cares as much about marketing or more than design.
Some may find reading that Mr Hansen has tried the EC approach as interesting as reading papers by Hawskford, Cordell, Yamaha, or the late Hephaistos.
It would have been more exillerating if C. Hansen favored error correction, but questioning Mr Curl's words is uncalled for.
Last time i read them it stated that he was just giving an example by mentioning Mr Ayre.
It can also be read as Mr Hansen not being able to put it to good use in his products, without a need for cold hard evidence to back it up.
Forcing everyone to take on an engineering stand at all times is remarkable coming from someone with a vast history of being criticised of meaningless one-liners and total lack of respect for others.
jacco vermeulen said:
Nobody has asked Mr Hansen for evidence; that's grotesque misrepresentation.
Rather, I maintained that it is unreasonable to impugn a technique on the grounds that someone tried it and found the "sound" unsatisfactory. This is manifest rubbish, given that further details were not forthcoming.
Finally, Mr Cordell's riposte here summerizes the position perfectly.
Will you folks ever progress if you doubt what other professionals say?
I simply passed on what Chas Hansen told me. Is this really worth impugning?
Personally, I don't always agree with Chas Hansen in his design methods, but I will not criticize him for it. IF he wants to make a no global feedback amp, no matter what the cost in: time, effort, and measurement, so be it. Let's see how it works in the audio marketplace.
If Bob Cordell wants to use feedback, including the Hawkford circuit, all the best to him as well. His measured results are certainly impressive.
Personally, my design philosophy is to make the most straight forward complementary topolgy, using even cascode only when absolutely necessary. Personally, I like to listen in the class A region when music is normally playing. Therefore, I try to make my designs as class A as possible, given the heatsink available, and work hard to create a smooth transition between the class A and the class B region as possible, without circuit tricks or internal feedback loops. My goal is to keep the HIGHER ORDER harmonic distortion down before using negative feedback.
With my power amps, and even most of my preamps, I use global feedback. I have done this for 40 years. The trick is to see how little global feedback I can use and still get respectiable measurements. This helps me keep a higher than average open loop bandwidth, that I suspect is more important than many here realize or appreciate.
As far as marketing is concerned, I find it a deep insult to be accused of marketing anything. I have my opinions and I am willing to share them. What else can or should you expect?
I simply passed on what Chas Hansen told me. Is this really worth impugning?
Personally, I don't always agree with Chas Hansen in his design methods, but I will not criticize him for it. IF he wants to make a no global feedback amp, no matter what the cost in: time, effort, and measurement, so be it. Let's see how it works in the audio marketplace.
If Bob Cordell wants to use feedback, including the Hawkford circuit, all the best to him as well. His measured results are certainly impressive.
Personally, my design philosophy is to make the most straight forward complementary topolgy, using even cascode only when absolutely necessary. Personally, I like to listen in the class A region when music is normally playing. Therefore, I try to make my designs as class A as possible, given the heatsink available, and work hard to create a smooth transition between the class A and the class B region as possible, without circuit tricks or internal feedback loops. My goal is to keep the HIGHER ORDER harmonic distortion down before using negative feedback.
With my power amps, and even most of my preamps, I use global feedback. I have done this for 40 years. The trick is to see how little global feedback I can use and still get respectiable measurements. This helps me keep a higher than average open loop bandwidth, that I suspect is more important than many here realize or appreciate.
As far as marketing is concerned, I find it a deep insult to be accused of marketing anything. I have my opinions and I am willing to share them. What else can or should you expect?
john curl said:
If Mr. Hansen wants to play with no global feedback uh... apparatus, that's OK, that is, as long as he don't sell these things. To be successful in the 'audio marketplace' , he has people believe that no global feedback is better, even at the expense of higher distortion and that global feedback is totally wrong. This kind of practice is just as bad as selling 'interconnects' made from oxygen-free copper with crystals aligned in the direction of the music.
Many here just don't understand that people actually listen to and compare high quality products. They just don't read ads in the audiophile magazine. In fact, most of our customers are not very technical at all, and they could care less what in in the enclosure. You people seem to equate measurements with listening quality, but that isn't always true. You should try to listen for yourself, or let others listen with you, in order to understand this.
john curl said:
They do actually; this is probably their biggest problem: believing what they read in the Hi-fi comics.
john curl said:
That, unfortunately, is their Achilles heel. This is relentlessly taken advantage of by those who sell overpriced oxygen-free interconnects, amplifiers etc.
john curl said:
No. Rather that ears and by implication "listening" are not reliable indicators of an amplifier's performance.
john curl said:
First, as for loudspeakers, I certainly do, but concerning amplifiers, I do trust an AP System Two more than my own ears.
Second, if you'are suggesting that there are certain 'properties' you can't measure, I would like to remind your own words: "If we do hear something and we can’t measure it then we try to find ways to measure what we hear. In the end we invariably find a measurement that matches what the ear hears and it becomes very obvious to everybody" (italics by me).
Cheers,
john curl said:
Hi John,
I think your remark is a bit condescending and presumptuous, and unfair to the people on this forum. I believe that most of the people here listen very carefully to what "professionals" like yourself say, but that does not mean that they will accept it at face value and never challenge it. You just don't like to be challenged.
I would very much love to have Chas participate here and share his experiences with EC himself. That way we can have some back-and-forth dialog. What you shared was hearsay and it had very little value because we have no way of finding out the details. Chas, if you are listening, please join in!
Implying that those who are not "professionals" (i.e., make a business out of audio) are not as good at making high-end amplifiers is just as bad as one saying that one cannot make a good amplifier without having an EE degree. Similarly, implying that those on this forum don't listen to their gear as ardently as you is also wrong. How is it that only "audio professionals" and audio reviewers are the only ones blessed with good hearing and knowledge of how to listen? Quite a coincidence, isn't it?
Finally, you do not have a monopoly on linearity in the open loop design of power amplifiers. To imply that those on this forum design less linear amplifiers and then fix up the problem by applying a lot of NFB is also wrong.
I always enjoy your participation on the forum and hope you will continue. Just keep in mind that you will get pummeled with tomatoes just as much as anyone else here if you say something that some don't understand or agree with. No free passes for "professionals".
Cheers,
Bob
It is a shame that someone thinks that just a low harmonic distortion measurement is going to tell you how an amplifier sounds. Or do you also mean triple tone IM, PIM, and Hirata distortion measurement? What does anyone here, except Bob Cordell, know about other measurments, besides harmonic distortion?
The problem here, on this website, is that Bob Cordell doesn't think much about exotic measurements either, and has told the rest of you that they don't matter. He is also stuck on very low harmonic distortion measurements, especially at full output.
Now, what about op amps?
My fascination with op amp theory goes back OVER 40 years. In fact, in 1965, I had a girlfriend who worked on the op amp assembly line at Fairchild Semi at that time. She got me all the engineering manuals and even a few working samples. These things were worth their weight in gold at the time! In 1966, I was put in charge of evaluating UA709 OP Amps when I had my first professional job at Friden-Singer. I built an op amp evaluator (much like a curve tracer) as instructed by Fairchild data sheets, and evaluated more than 100 devices for linearity, offset, etc, etc. I thought that these devices were example of all future design in audio, when they became cheap enough. I envisioned preamps made from a single chip and a few external components, etc, etc.
However, when in 1967, when I started working at Ampex, I found engineers struggling to work with these 709 op amps. I asked them why they were so disappointed in them. They told me that the SLEW RATE was really too low to make these devices useful for the high speed instrumentation recorders that we were making.
I had heard of slew rate limiting, and it WAS put on the data sheet, BUT it was NOT a popular or common term in low frequency design, contrary to what Bob Cordell has stated. It was an artifact of the high feedback OP Amp design concept. Discrete designs generally did not slew rate limit, but just bandwidth limited. In those days, we did not build op amps, in general, but simple 'ring of 3' or some such design usually with a single ended power supply and cap coupling. Our open loop bandwidth was much higher than the standard op amp of even today, and slew rate was generally undetectable. We just ran out of bandwidth.
In fact, both Walt Jung and I owned a textbook authored by CHERRY (you know who he is don't you?) and Hooper, called something like: 'Low Pass Amplifier Design' written in the late '60's. This was then the 'bible' of amp design, and was at least 1 inch thick and 1000 pages, crammed with all kinds of interesting information.
However, there is NOTHING about slew rate, rate distortion, rate of change, or anything else like that in this 'bible' of amp design. Gee wiz, someone overlooked something!
However, if we listen to Cordell's opinion, every designer knew just about everything about slew rate, back in 1970 or so. Sorry, that just was not the case.
The problem here, on this website, is that Bob Cordell doesn't think much about exotic measurements either, and has told the rest of you that they don't matter. He is also stuck on very low harmonic distortion measurements, especially at full output.
Now, what about op amps?
My fascination with op amp theory goes back OVER 40 years. In fact, in 1965, I had a girlfriend who worked on the op amp assembly line at Fairchild Semi at that time. She got me all the engineering manuals and even a few working samples. These things were worth their weight in gold at the time! In 1966, I was put in charge of evaluating UA709 OP Amps when I had my first professional job at Friden-Singer. I built an op amp evaluator (much like a curve tracer) as instructed by Fairchild data sheets, and evaluated more than 100 devices for linearity, offset, etc, etc. I thought that these devices were example of all future design in audio, when they became cheap enough. I envisioned preamps made from a single chip and a few external components, etc, etc.
However, when in 1967, when I started working at Ampex, I found engineers struggling to work with these 709 op amps. I asked them why they were so disappointed in them. They told me that the SLEW RATE was really too low to make these devices useful for the high speed instrumentation recorders that we were making.
I had heard of slew rate limiting, and it WAS put on the data sheet, BUT it was NOT a popular or common term in low frequency design, contrary to what Bob Cordell has stated. It was an artifact of the high feedback OP Amp design concept. Discrete designs generally did not slew rate limit, but just bandwidth limited. In those days, we did not build op amps, in general, but simple 'ring of 3' or some such design usually with a single ended power supply and cap coupling. Our open loop bandwidth was much higher than the standard op amp of even today, and slew rate was generally undetectable. We just ran out of bandwidth.
In fact, both Walt Jung and I owned a textbook authored by CHERRY (you know who he is don't you?) and Hooper, called something like: 'Low Pass Amplifier Design' written in the late '60's. This was then the 'bible' of amp design, and was at least 1 inch thick and 1000 pages, crammed with all kinds of interesting information.
However, there is NOTHING about slew rate, rate distortion, rate of change, or anything else like that in this 'bible' of amp design. Gee wiz, someone overlooked something!
However, if we listen to Cordell's opinion, every designer knew just about everything about slew rate, back in 1970 or so. Sorry, that just was not the case.
john curl said:
Hi John,
Your characterizations of others, and where they are coming from, are innaccurate, as here. You always seem to want to go to an extreme, high-contrast way of characterizing others. If they care about objective performance, they must not listen, etc. That's too bad.
I've never said that just a low THD number tells the whole story about how an amplifier sounds. There may be some who assert that, but I'm not one of them.
As far as different and unusual tests go, I invented the MIM test. I built instrumentation to measure it, and PIM and IIM. That's more than you ever did. As far as I know, you have never measured PIM at all. I am one of the very few people that do back-driven twin tone testing.
I would love it if these tests showed something that we didn't already know from other tests, but they don't always show new things. I have never told people that other tests don't matter, but I have told the truth about what they can be expected to yield in comparison to conventional tests.
I am not "stuck" on very low THD measurements, especially at high power. Indeed I do them and others like twin tone at power levels that are very low, where noise is often a problem, and use spectral analysis to get around that problem.
I do like to pursue low distortion measurements because when an amplifier has very low THD or twin tone, there is a lot less wiggle room for crossover distortion and other more unexpected nonlinearities to be there. But ultra-low THD and IM is not the be-all and end-all. Indeed, I favor the use of open-loop soft-clipping, which sacrifices ultra low net THD in exchange for a guarantee that the feedback amplifier will not clip. How many of your amplifiers incorporate adaptive soft clipping?
If you are referring to Otala's apparent ignorance of slew rate limiting in the mid seventies, I think that is pretty inexcusable. He either did not do his homework or was being stubborn in not referring to it in his papers.
Cheers,
Bob
As an amplifier is only a device for sound reproduction, and the sound reproduction is the only aim, why would one be so dull and would not listen it?
It is a common problem of many forums - in case you are interested in measurement and excellent measured results, someone comes and says you do not care about listening. Obvious nonsense.
On the other hand, excellent measured parameters are just necessary, not satisfactory condition of good sounding product. But they are necessary! I have never seen poor measuring product sounding good for my ears.
It is a common problem of many forums - in case you are interested in measurement and excellent measured results, someone comes and says you do not care about listening. Obvious nonsense.
On the other hand, excellent measured parameters are just necessary, not satisfactory condition of good sounding product. But they are necessary! I have never seen poor measuring product sounding good for my ears.