Glen,
Thanks for the post, it's a pleasure to have you inform the debate.
Actually it is not important if I am taken seriously; I'm not looking for that at all, you completely misunderstand. Your attitude is completely one-sided, so I discount it.
I am very serious about trying to get the best sound possible, and all else is merely a distraction. This is applied in the many commercial designs I now have and strive to improve upon with each release. I try to collaborate with people whose opinions I respect; your response shows this is personal for you. I would suggest you email me privately if you really wish to show your feelings; you know my email, so don't use the forum.
How about another 12W Class A to keep us interested in what is really the point of this forum?
Have a good day!
Hugh
Thanks for the post, it's a pleasure to have you inform the debate.
Actually it is not important if I am taken seriously; I'm not looking for that at all, you completely misunderstand. Your attitude is completely one-sided, so I discount it.
I am very serious about trying to get the best sound possible, and all else is merely a distraction. This is applied in the many commercial designs I now have and strive to improve upon with each release. I try to collaborate with people whose opinions I respect; your response shows this is personal for you. I would suggest you email me privately if you really wish to show your feelings; you know my email, so don't use the forum.
How about another 12W Class A to keep us interested in what is really the point of this forum?
Have a good day!
Hugh
AKSA said:
Gee, thank you, Hugh.
1) My attitude isn’t completely one-sided.
2) I’m making a perfectly valid constructive criticism.
In your previous post you stated that you have never seen an “improvement in linearity without a commensurate improvement in sound quality”. That’s a very clear-cut and one-sided statement and I don’t see how it is reconcilable with your previous criticism and dismissal of those who objectively design for the best linearity. There are a lot of people here who objectively design for the best linearity, and none, as far as I'm aware, dismiss every facet of subjective evaluation or only ever “listen to test tones”.
Is that supposed to contain an underlying gripe? This is DIY audio, so I posted one of my DIY designs. Why don't you post one of yours?
Cheers,
Glen
Back 40 years ago, in 1967, I achieved .005% SMPTE IM distortion at 1W from my first home grown power amp, (equal to 104spl from my K-horn), and I thought that it was good enough. I have learned how to do even better over the years. How about the rest of you? How low in static distortion is low enough? .001%, .00001%, Why don't you folks tell me what you think is low enough, or are you just making a contest out of it?
john curl said:
G’day John.
I think that it is you who has just initiated a competition
, although I think that this is beside the point. The debate is wether or not the job can be done better with JFET’s and/or MOSFET’s instead of BJT’s. Borbely asserts that BJT’s are a limiting factor in amplifier performance, and must be designed out. I think that is a load of baloney.
I cited the ExtremA amplifier as a reference all-BJT design. Some people argue that JFET’s are better due to their distortion characteristics. I argue that that is a gross over-simplification which glosses over the fact that the performance of an amplifier is the net product of the sum of all of its parts and the manner in which they are applied. I also argue that that at 0.0001% THD, the harmonic composition of the distortion doesn’t matter a jot. You advocate FET’s over BJT’s for superior amplifier performance as Borbely does. Have you got an all-FET or a mostly-FET design that betters the ExtremA’s distortion performance?
Cheers,
Glen
Glen: I suspect that with enough circuit complexity and ingenuity, a low static distortion could be attained by nearly any technology. But is that the only objective criterion? Could you outline what you'd consider a sufficient test regime to remove the amp as an audible defect in the electronics chain?
Should actual loudspeakers (or at least good analogs) be thrown into the measurement mix?
Should actual loudspeakers (or at least good analogs) be thrown into the measurement mix?
SY said:
G'day SY
I think that practically any well thought out SS design that does better than ~0.03% THD 20Hz-20kHz at full power with a decent damping factor can be ruled out as an audible defect in the electronics chain with a comfortable safety margin. THD, admittedly, doesn’t define the whole story – an amplifiers slew rate, TIM, and IMD performance amongst others, for example, are also important, but these are all inter-related to a degree, and 20Hz-20kHz THD is a pretty good indicator to generalise on. 0.03% THD isn’t an unreasonable demand either, as it is quite easily bettered without too much trouble.
There is a lot argument over the importance of the harmonic composition of the distortion, namely that musically related even harmonics are more desirable than odd harmonics. With this I agree completely in principle, but at the ultra-low distortion levels easily achievable with modern SS components and design techniques, I think that it is mostly an academic concern.
I think that the harmonic composition of the distortion is mostly applicable to “Hi-End” design, as opposed to “Hi-Fi” design. Designers of "Hi-End” audio typically defy convention and insist, for example, for various reasons, on using vacuum tubes or designing with zero global negative feedback or low open-loop gain. Now that is all fine and dandy, but such design trade-offs, allegedly done the quest for the best sonic performance, place unavoidable limitations on the distortion performance that is practically achievable.
There are lots exorbitantly priced “Hi-End” amplifiers out there with full-power THD and IMD figures in the one or two percent range. Such distortion products are definitely audible, and, from a “Hi-Fi” perspective, abysmal. Now I’m not saying that such an amplifier will necessarily sound bad, but with 1-2% THD, as a designer, you would want to get the harmonic composition of the distortion right.
With regards to speakers, they are definitely the weakest link in the modern audio chain. Contrasted them, I think that typical squabbling over wether or not an amplifier with 0.01% THD would sound better than one with 0.001% THD is a bit inane.
Cheers,
Glen
Glen, anyone can design mid fi. Why even bother to talk about it? There are 100's of completed amp designs, both new and used, that should suit your needs. Perhaps, PA applications are where you are really going to be helpful. Even then, we would put tube amps on the tweeter array, when we could. It just sounded better.
For you to come to some conclusion about harmonic distortion and the harmonic series shows an amazing lack of interest in the subject, and is even in contradiction to what was mentioned in the 'Radio Designers Handbook' in 1941, and even earlier. Haven't you ever been exposed to listener's fatigue from higher order distortion products present in the electronics?
For you to come to some conclusion about harmonic distortion and the harmonic series shows an amazing lack of interest in the subject, and is even in contradiction to what was mentioned in the 'Radio Designers Handbook' in 1941, and even earlier. Haven't you ever been exposed to listener's fatigue from higher order distortion products present in the electronics?
Measure properties of the whole chain
I tend to measure parameters and spectra of the whole audio chain. I attach measuring card and battery supplied notebook at binding posts of the speaker. Insert test disc into CD player and measure spectrum at the speakerbox terminals. This brings very interesting results. IMHO this is the only way how to assess properties of the whole audio chain operated under its normal working conditions.
I tend to measure parameters and spectra of the whole audio chain. I attach measuring card and battery supplied notebook at binding posts of the speaker. Insert test disc into CD player and measure spectrum at the speakerbox terminals. This brings very interesting results. IMHO this is the only way how to assess properties of the whole audio chain operated under its normal working conditions.
john curl said:
What on earth are you talking about??
I think that you need to read my post again – carefully. You’re attributing to me the opposite of my view. As I said, I AGREE that odd harmonics/high order distortion products are less desirable than even harmonics in principle. But I don’t believe that anyone is going to suffer listener fatigue from higher order distortion products produced by a modern, state-of-the-art solid state amplifier with <0.001% THD-20. Perhaps the harmonic composition of the distortion is a more important factor for you, since your JC-1 amp is rated somewhat higher at 0.15% THD-20.
Do you honestly think that listeners of the ‘ExtremA’ amplifier are going to be suffering listener fatigue from the vanishingly small high-order distortion products ultimately produced by it’s BJT’s, or for that matter listeners of Bob Cordell’s lesser-performing MOSFET design with its BJT VAS, driver stages and 0.001% THD-20?
Also, to the best of my knowledge, 0.001% THD-20 wasn’t easily achievable in 1941, so I think this design requirement was a bit more relevant then.
Cheers,
Glen
Glen, there is a good chance that some other distortion is being generated that is not directly measurable in high feedback circuits, just like SMPTE IM distortion failed to show TIM. However, if you can hear .15% distortion at 400W output, you are a better listener than I am. I look at the 1-10W level to get best performance, with a little global feedback as possible.
john curl said:
Good grief John, you have just avoided answering my question with a total furphy.
Let’s suppose than a well designed amplifier with 0.0004% THD-20 at full power (not a 1941 single-ended tube design with 15% THD) was indeed producing such distortion. Of what magnitude would you guess it might be, particularly at normal listening levels? I’d appreciate an answer, especially so since you readily dismiss 0.15% THD as inaudible.
Cheers,
Glen
PMA said:Glen,
this is lost. You will be always taught about high order harmonics produced by NFB, though they are -120dB and less. And the preferred "non-NFB" amp would measure like this:
That means high order harmonics 10-100x higher.
To mask is the case, not to amplify
NNNOOOOOoooooooooo...........NFB BBBBBAAAAAADDDDddddddddddd........
PMA said:
Hi PMA,
Actually, it wasn't my intention to start such semantic debate all over again, rather to comment on the Ayre V-3 amp.
However, I was forced, or at least invited by Bob to clarify my view on the definition of FB. Although we both agree that, starting from a strict definition of FB, it is impossible to build a good amp without FB, I'm not inclined to go any further on that matter.
There are much more interesting things to discus. That's why (in my reply to Hugh) I put a link to a variant on Cherry's NDFL amp.
Regrettably, nobody has responded until now. Assuming it has just been overlooked, I'll put it here again:
http://www.diyaudio.com/forums/show...=96634&perpage=10&highlight=ndfl&pagenumber=3
post # 29
Cheers,
Hi Edmond,
thank you for the reminder, I will read it.
The amplifier that really impressed me is this one:
http://www.hardwareanalysis.com/content/article/1842/
Regards,
Pavel
http://www.hardwareanalysis.com/content/article/1842.5/extrema_reference_class_a_diy_amplifier/
thank you for the reminder, I will read it.
The amplifier that really impressed me is this one:
http://www.hardwareanalysis.com/content/article/1842/
Regards,
Pavel
http://www.hardwareanalysis.com/content/article/1842.5/extrema_reference_class_a_diy_amplifier/
john curl said:
Hi John,
Not directly measurable? What on earth should that be?
Please, explain.
BTW, as Glen has pointed out, all kind of distortions are inter-related to a degree. So, if thd=zero, all other disturbances, including your mysterious brand of distortion, are zero too.
Cheers,
PMA said:
Hi Pavel,
I did have a look at that amp and indeed, the thd figures are pretty good. But I'm sorry to say, I'm not impressed. IMHO it is a rather 'insipid' design, as it is quite easy to obtain such a low thd by reverting to a power hungry class-A stage (hence it's named ExtremA?) and using an inverting input stage. Clearly, the designer of this amp wasn't challenged by the hurdles that other designers are willing to overcome.
Analogue to incandescent lamps, there should be a ban on class-A amplifiers and designers of such heaters should be punished by cutting off one ear.
To all,
I'm looking for vertical fet's like the 2SK1530/2SJ201 pair, but with a much lower Cgd. Any suggestions?
Cheers,