traderbam said:
There ain't no competition, at least not from our side.
How do you know that? Do you know how much capacitance Bob's amp can tolerate vs our amp? I guess you don't. Anyhow, your speculation is completely wrong.
Of course, because nobody is interested in the 'sound quality' of an unstable amp.
Is that so? Please, look here:
http://www.diyaudio.com/forums/showthread.php?postid=1332027#post1332027
That's right, it is not a fair comparison, that is, it's a totally unfair comparison. The MOSFETs in Bob's amp have gm of 5 S, while in our amp only 3*1S = 3 S. His RL is 8 Ohm and our RL only 4 Ohm. Bob has set the Ft of the global NFB loop at 2MHz, we at only 0.7MHz.
Summing all these inequalities together: 5/3 * 8/4 * 2/0.7 = 9.52. In other words, Bob's amp has an advantage beforehand of almost a factor of 10. Indeed, not a fair comparison.
Again, we are not in competition. Besides, why should we made such 'comprehensive comparison'? It's a waist of time, as any competent designer knows since long that NDFL will reduce the distortion by an order of magnitude.
traderbam said:Edmond wrote:
Ah. You didn't say that your design was born at CERN. That makes all the difference.
BTW, why should I be impressed by 500V/us? Why does this matter?
Your prey, Halcro, uses a single LTP input. Why do you think they chose to do this?
I suppose your next question will be: THD20=1ppp? Why does this matter?
Regarding Halcro's single LTP input, just ask Bruce Candy, as I haven't the faintest idea. Perhaps they never though of CMCL.
Edmond, I certainly know at least 2 of your hot buttons. The first is the word "complexity" and the second is the word "stability". 
Well, the complexity is surely in no doubt. And the stability is indicated by the relative size of output L and output R and the pulse response you've published. Of course, I have asked you several times in the other thread to simply say what the relative stability margin is but you have so far evaded the question.
Ok. So you consider stability margin into reactive loads and output Z are a waste of time...provided the THD figure, as you are measuring it, is better? Interesting. In that case I have no more questions to ask you.
traderbam said:[snip]
Ok. So you consider stability margin into reactive loads and output Z are a waste of time...provided the THD figure, as you are measuring it, is better? Interesting. In that case I have no more questions to ask you.
Every competent designer knows since long that NDFL, provided that it is properly implemented, has no effect on the stability margin.
>In that case I have no more questions...
I hope so!
Hi Guys. You all have some reasonable points, but let me first say that this is not a competition. I do see the PGP as a fine accomplishment, and as an inspiration. I also see the achievement of 1 ppm THD-20 as a challenge. This is different from a competition.
For example, if it is asserted that one cannot achieve 1 ppm with a unipolar input stage and without resort to NDFL, that is an interesting challenge.
Although we may disagree on the details, I think we all agree that a comparison to my 25-year-old amplifier need not be fair and that does not really matter. A lot has changed since then, and different aspects of fairness can go either way. Just as the PGP may be an inspiration for me to try harder, any role that my amplifier played in inspiring the PGP is something that is very pleasing to me. The mere fact that my old amplifier is mentioned in comparisons to the PGP makes me feel great.
Moreover, if I chose to attempt a 1 ppm amplifier, I would not choose to use laterals, so if that is a necessary component of fairness in comparison, it is impossible right off the bat.
I do agree that if I did achieve 1 ppm, it would be certainly more meaningful to others if it was at at least as much power into 4 ohms, and with at least as much stability, and with about the same total bias current as the PGP.
Cheers,
Bob
For example, if it is asserted that one cannot achieve 1 ppm with a unipolar input stage and without resort to NDFL, that is an interesting challenge.
Although we may disagree on the details, I think we all agree that a comparison to my 25-year-old amplifier need not be fair and that does not really matter. A lot has changed since then, and different aspects of fairness can go either way. Just as the PGP may be an inspiration for me to try harder, any role that my amplifier played in inspiring the PGP is something that is very pleasing to me. The mere fact that my old amplifier is mentioned in comparisons to the PGP makes me feel great.
Moreover, if I chose to attempt a 1 ppm amplifier, I would not choose to use laterals, so if that is a necessary component of fairness in comparison, it is impossible right off the bat.
I do agree that if I did achieve 1 ppm, it would be certainly more meaningful to others if it was at at least as much power into 4 ohms, and with at least as much stability, and with about the same total bias current as the PGP.
Cheers,
Bob
traderbam said:
Please, stop whining about complexity, which is so typical for guys who are completely deaf and blind for the nice and powerful features of our amp.
Regarding "the relative stability margin" I have asked you to be more specific: phase margin, gain margin or tolerance against capacitive loads. Up till now you refused to give an answer. Apparently you didn't bother to put your questions more clearly.
BTW, you cannot estimate the stability based on the size of output L and R, as you don't know exactly what our motives were, which led to the current values.
This has become an intergalactic competition of the most urinary kind, at least from Bam's POV.
Why? Who's proving what here?
This is an audio amp, and as far as I can tell, it was born out of intellectual challenge, and it covers much of the ground covered by revered amps like the Halcro.
How sterile is an argument comparing two pretty girls if neither is present?
Fer crissakes, Bam, stop this tub thumping and build the damn thing, then you will know all the answers to your inquiries about the sound, amongst other things, and moreover, you'd be able to measure it yourself!
There comes a time when the actual work of construction must begin. You have the Space Shuttle design, now get on with building it.
Hugh
Why? Who's proving what here?
This is an audio amp, and as far as I can tell, it was born out of intellectual challenge, and it covers much of the ground covered by revered amps like the Halcro.
How sterile is an argument comparing two pretty girls if neither is present?
Fer crissakes, Bam, stop this tub thumping and build the damn thing, then you will know all the answers to your inquiries about the sound, amongst other things, and moreover, you'd be able to measure it yourself!
There comes a time when the actual work of construction must begin. You have the Space Shuttle design, now get on with building it.
Hugh
Bob Cordell said:
Here we go again. But some reasonable points are pointless.Why weren't you challenged by Bruce Candy? If he did, you must have already improved your amp a couple of years ago. Apparently, he failed to impress you, right?
(italics by me)
Nobody has said that.You missed the subtlety of my carefully chosen words. I said:
"I'm afraid you have no other choice than adding a NDFL stage or combining TMC with EC"
And in another post:
"So it seems almost impossible to hit the 1ppm target without more drastic means."
Apparently I have to repeat my words:"That's right, it is not a fair comparison, that is, it's a totally unfair comparison. The MOSFETs in Bob's amp have gm of 5 S, while in our amp only 3*1S = 3 S. His RL is 8 Ohm and our RL only 4 Ohm. Bob has set the Ft of the global NFB loop at 2MHz, we at only 0.7MHz.
Summing all these inequalities together: 5/3 * 8/4 * 2/0.7 = 9.52. In other words, Bob's amp has an advantage beforehand of almost a factor of 10. Indeed, not a fair comparison. "
So, if competition is at stake, you have nothing to complain about, Bob.
Actually, almost nothing has changed since then. OK, perhaps that audio MOSFETs are getting (slightly) better, but on the other, it's hard to get good BJTs nowadays, especially duals. The same applies for dual j-fets.
Even our knowledge and arsenal of tricks to improve the performance hasn't fundamentally changed since then. Remember that NDFL and HEC were developed at the same time! (at the end of the seventies).
Hi Bob,
And in general, let's stop this fruitless and pointless discussion on false comparisons. Our amp was never meant to be compared to other amps, rather to inspire other people, that is, that it will be carefully studied in all its relevant details and perhaps that somebody will use some parts in his own design.
And when it comes to comparisons, please, only in an analytical way. For example, one should ask: what are the benefits of NDFL sec? and what are the disadvantages? regardless of what kind of output stage is used.
Cheers, Edmond.
Edmond Stuart said:
Why weren't you challenged by Bruce Candy? If he did, you must have already improved your amp a couple of years ago. Apparently, he failed to impress you, right?
(italics by me)
You missed the subtlety of my carefully chosen words. I said:
"I'm afraid you have no other choice than adding a NDFL stage or combining TMC with EC"
And in another post:
"So it seems almost impossible to hit the 1ppm target without more drastic means."
"That's right, it is not a fair comparison, that is, it's a totally unfair comparison. The MOSFETs in Bob's amp have gm of 5 S, while in our amp only 3*1S = 3 S. His RL is 8 Ohm and our RL only 4 Ohm. Bob has set the Ft of the global NFB loop at 2MHz, we at only 0.7MHz.
Summing all these inequalities together: 5/3 * 8/4 * 2/0.7 = 9.52. In other words, Bob's amp has an advantage beforehand of almost a factor of 10. Indeed, not a fair comparison. "
So, if competition is at stake, you have nothing to complain about, Bob.
Actually, almost nothing has changed since then. OK, perhaps that audio MOSFETs are getting (slightly) better, but on the other, it's hard to get good BJTs nowadays, especially duals. The same applies for dual j-fets.
Even our knowledge and arsenal of tricks to improve the performance hasn't fundamentally changed since then. Remember that NDFL and HEC were developed at the same time! (at the end of the seventies).
Hi Bob,
And in general, let's stop this fruitless and pointless discussion on false comparisons. Our amp was never meant to be compared to other amps, rather to inspire other people, that is, that it will be carefully studied in all its relevant details and perhaps that somebody will use some parts in his own design.
And when it comes to comparisons, please, only in an analytical way. For example, one should ask: what are the benefits of NDFL sec? and what are the disadvantages? regardless of what kind of output stage is used.
Cheers, Edmond.
Hi Edmond,
Lots to respond to here, but let me emphasize that I was not trying to start a battle in that post of mine. I was actually trying to establish a non-confrontational way of looking at all of this. That is the best way for all of us to learn.
I was indeed challenged by Bruce Candy's design. I think it was about a year and a half ago that I learned of the Halcros and Bruce's use of EC. And yes, I began looking at many new twists to achieving much lower distortion with extensions to my old topology. I went off in numerous different directions, and although I built some circuits, I never converged on a complete design. The main thing that got me off the tracks was the initiative that Peter and Darren and I launched in regard to the audio workshops at RMAF and then at HE2007. These took up an enormous amount of time, but were well worth it.
I'm sorry I mis-interpreted your words in regard to the unipolar input stage. I'm still not sure whether TMC would be needed to get my design to 1 ppm.
While what we all consider as aspects of fairness in a comparison differently, I can assure your that I don't agree with the way you have defined fairness in your above repeated statement. I had previously chosen not to respond to that. I'll get back to you on the specifics of that if you are interested.
Once again, you have a fine, inspiring design. Part of the fun in audio is that there are always more than one way to skin a cat, and when someone suggests that, please understand that no offense is meant toward your fine design.
Cheers,
Bob
Ovidiu, in response to your ealier post.
This particular thread is entitled "Bob Cordell Interview: Error Correction" and the discussion in this lengthy thread has revolved around HEC and Bob's implementation of it in his amplifer. So, in this thread, I have been trying to encourage you and Edmond to make comparisons with Bob's circuit because I think this is pertinent to this thread and may be of wider interest. By highlighting the differences, both advantages and disadvantages, more learning may be achieved. Therefore, to compare meaningfully I have proposed more parameters than just THD20 and suggested you and Bob specify them.
I have already talked (argued for some reason) with Edmond about stability. I have asked how you two determined acceptable stability in your design and you haven't answered me. That's fine, you don't have to.
Edmond has been very bullish in this thread about his use of NDFL and TMC as the only means to get 1ppm THD20, saying that Bob wouldn't be able to achieve this figure without using those methods. That is a bold challenge. Based on Halcro's US patent #5892398 I don't see Halcro using either method.
By "nested feedback loops" I mean simply a hierarchy of feedback loops. Cherry NDFL is a particular method, among others, to achieve a steeper rise in feedback gain than, say, a simple integrator below a certain frequency. If we examine what this overall loop gain roll-off actually is in your amp then this will invite alternative methods to achieve a similar effect.
There is a trade-off between stability and distortion reduction in any feedback system. Feed-forward doesn't have this trade-off (but has its own challenges). The THD20 figure, for example, in the way it is normally measured, is not inclusive of sonic performance issues that may arise due to stability issues. That is my opinion. Indeed, persuing only THD20 as a goal can lead to designs that are sonically compromised by the very circuitry required to reduce the THD20. So my general question for readers of this thread is how stable is stable enough and how should it be measured?
The stability is very much affected by what load you attach to the amp because the load will be reactive. I'm sure you know all about this. The magnitude and phase of a speaker can vary hugely and can become very capacitive at certain frequencies and in certain dynamic conditions. Speaker cables behave like transmission lines and introduce significant reactive effects at certain frequencies (Nelson Pass wrote a paper on this BTW). OS distortion depends on output current and output voltage and both depend on load impedance.
Brian
This particular thread is entitled "Bob Cordell Interview: Error Correction" and the discussion in this lengthy thread has revolved around HEC and Bob's implementation of it in his amplifer. So, in this thread, I have been trying to encourage you and Edmond to make comparisons with Bob's circuit because I think this is pertinent to this thread and may be of wider interest. By highlighting the differences, both advantages and disadvantages, more learning may be achieved. Therefore, to compare meaningfully I have proposed more parameters than just THD20 and suggested you and Bob specify them.
I have already talked (argued for some reason) with Edmond about stability. I have asked how you two determined acceptable stability in your design and you haven't answered me. That's fine, you don't have to.
Edmond has been very bullish in this thread about his use of NDFL and TMC as the only means to get 1ppm THD20, saying that Bob wouldn't be able to achieve this figure without using those methods. That is a bold challenge. Based on Halcro's US patent #5892398 I don't see Halcro using either method.
By "nested feedback loops" I mean simply a hierarchy of feedback loops. Cherry NDFL is a particular method, among others, to achieve a steeper rise in feedback gain than, say, a simple integrator below a certain frequency. If we examine what this overall loop gain roll-off actually is in your amp then this will invite alternative methods to achieve a similar effect.
There is a trade-off between stability and distortion reduction in any feedback system. Feed-forward doesn't have this trade-off (but has its own challenges). The THD20 figure, for example, in the way it is normally measured, is not inclusive of sonic performance issues that may arise due to stability issues. That is my opinion. Indeed, persuing only THD20 as a goal can lead to designs that are sonically compromised by the very circuitry required to reduce the THD20. So my general question for readers of this thread is how stable is stable enough and how should it be measured?
The stability is very much affected by what load you attach to the amp because the load will be reactive. I'm sure you know all about this. The magnitude and phase of a speaker can vary hugely and can become very capacitive at certain frequencies and in certain dynamic conditions. Speaker cables behave like transmission lines and introduce significant reactive effects at certain frequencies (Nelson Pass wrote a paper on this BTW). OS distortion depends on output current and output voltage and both depend on load impedance.
Brian
Edmond wrote:
The answer is clear in terms of marketing and magazine reviews but I don't think it is very clear in terms of making amplifiers that play music well. Consider plotting a dozen leading designs on a scatter graph with axes sonic performance and THD20. Or slew rate. Or damping factor.
Good question.
The answer is clear in terms of marketing and magazine reviews but I don't think it is very clear in terms of making amplifiers that play music well. Consider plotting a dozen leading designs on a scatter graph with axes sonic performance and THD20. Or slew rate. Or damping factor.
Bob wrote:
NDFL approximately mimmicks a butterworth filter response. So you could use a filter in the forward path or a variety of transfer function modifiers.
You could think outside the HEC box and look at using a different form of NFB around the OS. Your HEC implementation involves a single, "leaky integrator". It doesn't have to.
Lots of interesting possibilities.
Brian
Quite achievable though. Wouldn't you agree that to assert that a unipolar input won't cut the mustard is to assert that a unipolar subtractor cannot be designed and built with <1ppm THD20? Halcro appear to have achieved this.
NDFL approximately mimmicks a butterworth filter response. So you could use a filter in the forward path or a variety of transfer function modifiers.
You could think outside the HEC box and look at using a different form of NFB around the OS. Your HEC implementation involves a single, "leaky integrator". It doesn't have to.
Lots of interesting possibilities.
Brian
Next joke please.
dimitri said:
Dimitri,
This is a very interesting statement. I would certainly appreciate if you could share your definition of "transmission line". In my books, a transmission line is a device designed to guide electrical energy from one point to another. Transmission lines were defined and analyzed in the 19th century by Maxwell, Kelvin and Heaviside ("Telegrapher's Equations") as a result of the challanges with the first transatlantic cables. The whole point of introducing the transmission line concept is that when the input voltage changes in a time interval comparable to the time it takes for the signal to travel down the wire, the length becomes important and the wire must be treated/analyzed using different tools and equations than simply the Ohm law.
Set aside the audibility of a (say) 100KHz wave, such a wave has a wavelength of 3Km, or about 1.9 miles. This is about 500-1000 times the length of an average speaker cable.
I have to admit my ignorance in evaluating transmission lines of the size of an average speaker cable. Could you share some of your experience on how to measure the speaker cable alleged transmission line parameters, how to extract useful design information out of these measurements and what are your current results?
As you can probably see from my other posts, I am trying to keep my mind open, and try to learn as much as possible from other's experience. I personally don't need much of an impulse to "go measure" if I would know what/how and I think I have all the tools and equipments required to measure almost anything in the audio band.
Your input would certainly be appreciated by other engineers on this forum.