The problem with a question like that is it's almost endless as it depends heavily on what the individual will consider better.
If we take the loudspeaker as the end point or rather the most limiting part of the system, then this helps. We basically need no more bandwidth then the loudspeakers can reproduce and also we don't need distortion to be significantly lower then that of the loudspeakers either.
Now not all class D amplifiers are created equal and neither are loudspeakers, but on the whole class D is capable of the technical specifications that would allow it to be insignificant compared to that of the loudspeakers.
Blameless class AB is capable of better specifications then that of class D, but then blameless class AB is also better spec wise then any other class AB amp that isn't blameless and better then anything with valves too.
If we take the loudspeaker as the end point or rather the most limiting part of the system, then this helps. We basically need no more bandwidth then the loudspeakers can reproduce and also we don't need distortion to be significantly lower then that of the loudspeakers either.
Now not all class D amplifiers are created equal and neither are loudspeakers, but on the whole class D is capable of the technical specifications that would allow it to be insignificant compared to that of the loudspeakers.
Blameless class AB is capable of better specifications then that of class D, but then blameless class AB is also better spec wise then any other class AB amp that isn't blameless and better then anything with valves too.
Despite the fact that I am in love with class D, I do not see better sonic performance than class A and class AB.
-The dead time distortions (leading to two areas of non linearity, approximately at current levels where the filter ripple current equals the load current) give similar headache as the cross over distortions (leading to one area of non linearity around zero) but in the mean time many developers learned to handle dead time very good.
-The switching residuals that are injected back into the modulator through feedback form a distortion mechanism that has no counter part in class A or AB.
-The output filter naturally makes class D amps comparably slow, as long as the power devices do not allow us to operate with multi MHz in a reasonable way.
-The output filter can influence the transfer function substantially, except the design engineer has spend pretty some love to an optimized post filter feedback loop.
Fortunately class D has evolved in a manner that top level class D amps can play in the same sonic league as decent class A or AB amps.
The high switching speed (I mean the bandwidth) is needed to ensure the slew rate must be transient. another thing is the frequency response. If we have to design an output stage (audio generator), it is necessary that the bandwidth is very high compared to the frequency generated. for the audio amplifier is the same concept. The music is composed of transient, non-sinusoidal signals as well when we measure. with a lot of things that do not measure, it is obvious that an amp goes better than another, and this is justified by the esoteric mystery.
for D class, i not agree that perfect solution is huge nfb, even with slope filter in transfer function. but this is my opinion.
for D class, i not agree that perfect solution is huge nfb, even with slope filter in transfer function. but this is my opinion.
yes, know. but no one said (to those who have decided on a table called business) that is necessary 1 mhz min. of f/s
Your question can not be answered logically because you are asking for objective reasons for a subjective quality.
Clearly, based on objective specs, ncore from Hypex leaves most amps, of any class, in the dust. Based on subjective reviews, it has also left nearly all it has been compared to (a long list of Pass class a amps, top flight tube amps, and everything in between) behind.
The naysayers who put down class d on principle are like those that say bumblebees can't fly. Instead of tossing out the well worn dogma about how class d amps can not possibly sound as good as class a, go listen to ncore and then at least you will have an informed opinion to share.
Yes but the problem here is that we could argue objective qualities all day long simply comparing class AB amplifiers and never get anywhere with regards to the relevance of those objective measurements towards which or what sounds better.
OPC may have stated a dislike for comments where people simply state their subjective opinion, but surely in this case it does at least offer a very real truth that lots of people are very happy with their class D amplifiers.
If you take the current best of class D amplifiers, such as the ncore you will see that they do compare very favourably and in most cases better, objectively, then a lot of other amplifier classes or types.
OPC may have stated a dislike for comments where people simply state their subjective opinion, but surely in this case it does at least offer a very real truth that lots of people are very happy with their class D amplifiers.
If you take the current best of class D amplifiers, such as the ncore you will see that they do compare very favourably and in most cases better, objectively, then a lot of other amplifier classes or types.
cab:
These metrics have existed for a very long time, I'm not asking for anything out of the ordinary.
If I make an amplifier for you that rolls off at 8kHz and has very high 5th, 7th and 9th order harmonics, you might tell me that it sounds terrible, and I would say, "Exactly, because it has limited bandwidth and a very undesirable distortion profile". These are simple examples of technical deficiencies that result in poor sonic performance that is easily identifiable by a listener.
We design amplifiers to meet certain technical goals, because we know that these goals are what make an amplifier sound good to the listener.
I'm asking if there is anything, from a technical perspective, that a class D amp does better than a class AB or A that might imply that it should sound better. So far I'm still not getting an answer... and it's a pretty straightforward question.
Here's a simple example:
A class AB or class A amplifier is likely to sound better than a class B amplifier because the class B amplifier suffers from crossover distortion which results in high order harmonics and does not sound good to the listener.
Cheers,
Owen
These metrics have existed for a very long time, I'm not asking for anything out of the ordinary.
If I make an amplifier for you that rolls off at 8kHz and has very high 5th, 7th and 9th order harmonics, you might tell me that it sounds terrible, and I would say, "Exactly, because it has limited bandwidth and a very undesirable distortion profile". These are simple examples of technical deficiencies that result in poor sonic performance that is easily identifiable by a listener.
We design amplifiers to meet certain technical goals, because we know that these goals are what make an amplifier sound good to the listener.
I'm asking if there is anything, from a technical perspective, that a class D amp does better than a class AB or A that might imply that it should sound better. So far I'm still not getting an answer... and it's a pretty straightforward question.
Here's a simple example:
A class AB or class A amplifier is likely to sound better than a class B amplifier because the class B amplifier suffers from crossover distortion which results in high order harmonics and does not sound good to the listener.
Cheers,
Owen
There are plenty of amps (many tube amps come to mind) that spec poorly, yet many people swear by them.
There are amps that have great specs, yet not everyone agrees they sound fantastic.
Again, you are asking for something that can not be answered.
If you are basing your definition of "sonic performance" on specs, then the ncore, a class d amp, would be hard to beat by any amp of any class.
There are amps that have great specs, yet not everyone agrees they sound fantastic.
Again, you are asking for something that can not be answered.
If you are basing your definition of "sonic performance" on specs, then the ncore, a class d amp, would be hard to beat by any amp of any class.
That's not really the case... many tube amps do sound great even though they don't measure very well, because they have a pleasant harmonic distortion profile. Our ears are very forgiving of second and third order harmonics, and you'd be surprised just how high those two harmonics can get before people start complaining. Some people even prefer to have these harmonics present, even if they weren't part of the original recording.
Conversely, if an amplifier has even a small amount of higher order harmonics, a trained ear will quickly object. These simple technical facts can explain why a tube amp at 1% distortion can sound better than an SS amp at 0.1% distortion. There's no black magic or mystery here, we've known this for a long time.
I think you know exactly what I'm asking for, and you're dodging the question.
I'm not sure what measurements you're looking at, but what I've seen in the ncore white paper is nothing to write home about. A noise floor worse than -100dB and THD only measurements at around 0.002% are both figures that a class AB amp can exceed by a factor of 30dB or more.
Be very careful when you read white papers... they're intentionally only showing you measurements that make the product look best. You'll notice there are no FFT's showing a 1kHz tone, and absolutely no THD+N measurements because both would look bad with such a high noise floor.
If you want to see what a good class AB amp can do take a look here:
http://www.diyaudio.com/forums/soli...ce-lpuhp-16w-power-amplifier.html#post2852722
If you can find me a class D that can do that, I'll concede that it's a better type of amplifier.
Cheers,
Owen
Huh? Look again.....
The distortion of the ncore is below the noise floor of the measurement equipment out a considerable way...not sure what you are talking about with the no "THD + N measurements".....they are there.....
Again, you are asking for subjective evaluations from objective specs. As I said, clearly, if you read the dozens of reviews of the ncore, you will find that an overwhelming majority prefer it to their class a, class ab, and class d amps. Don't tell me it can't sound as good as a good class a or class ab amp. I have one, I know what it sounds like compared to other amps. Have you heard one?
16watts? The ncore distortion is below 0.0005% up to 20 watts....I think most people would find that satisfactory....
Last edited:
Yes Owen, if we're talking about technical superiority then no, currently the best of class D cannot beat the best of class A/AB.
So if that's what you want people to say, or recognise, then lets just get that out in the open right now, otherwise we will be going around in circles for a very long time.
Having said that though, most class A/AB designs do not represent anything close to the best of what class A/AB is capable of. Also some class D amplifiers show superior performance to a lot of those 'less then best' class A/AB.
Isn't that enough? You've got a situation where people are preferring class D to class A/AB where there could very well be a gain in technical performance in making the swap.
If you want to argue instead that a blameless class A amplifier (or amplifier of your choice) is technically superior to anything else, then you're free to make that statement, but to what end? It doesn't mean that everyone is going to swap to your version of what perfection is and it most certainly doesn't make the subjective preferences of those preferring class D invalid.
So if that's what you want people to say, or recognise, then lets just get that out in the open right now, otherwise we will be going around in circles for a very long time.
Having said that though, most class A/AB designs do not represent anything close to the best of what class A/AB is capable of. Also some class D amplifiers show superior performance to a lot of those 'less then best' class A/AB.
Isn't that enough? You've got a situation where people are preferring class D to class A/AB where there could very well be a gain in technical performance in making the swap.
If you want to argue instead that a blameless class A amplifier (or amplifier of your choice) is technically superior to anything else, then you're free to make that statement, but to what end? It doesn't mean that everyone is going to swap to your version of what perfection is and it most certainly doesn't make the subjective preferences of those preferring class D invalid.
Those are THD, not THD+N... there's an important difference.
I've been reading the white paper, but there are indeed more measurements in the "bulletin". That one seems to contain a noise floor measurement, but it's in dBR A with a 2.83V reference making it appear 9dB better than it would referenced to 0dBV.
I'm not going to deny that they measure exceptionally well, and I would imagine that they also sound very good, but it is incorrect to say that they measure better than any class A or AB amplifier.
I'm glad you like the way it sounds, and I certainly don't mean to imply that it somehow doesn't sounds good. I'm just trying to get you to be objective about why you think it sounds better than other amplifiers you have heard.
Regards,
Owen
I've been reading the white paper, but there are indeed more measurements in the "bulletin". That one seems to contain a noise floor measurement, but it's in dBR A with a 2.83V reference making it appear 9dB better than it would referenced to 0dBV.
I'm not going to deny that they measure exceptionally well, and I would imagine that they also sound very good, but it is incorrect to say that they measure better than any class A or AB amplifier.
I'm glad you like the way it sounds, and I certainly don't mean to imply that it somehow doesn't sounds good. I'm just trying to get you to be objective about why you think it sounds better than other amplifiers you have heard.
Regards,
Owen
http://www.hypex.nl/docs/NC400_datasheet.pdf
Is what you should be looking at for the Ncore. No, it isn't quite as good as 'the wire' but at levels as good or as low as those one could easily argue, who cares?
Is what you should be looking at for the Ncore. No, it isn't quite as good as 'the wire' but at levels as good or as low as those one could easily argue, who cares?
Indeed. At the levels attained by the ncore, it is certain that it is not the weak link in the reproductive chain. Can one hear the difference between .0005% and .0003%?
Yes, the 16 watt amp seems to have less distortion, up to 16 watts....The ncore is just getting started at 16 watts. I wouldn't call anything limited to 16 watts "technically superior"....While I doubt anyone could hear the difference in distortion, I know everyone will hear the difference between 16 and 200 watts....
Ahh! Much better... thanks for the link to the proper datasheet.
This is exactly the crux of the argument. The title of the thread is "Class AB vs Class D Amplifier" and what I'm trying to say is that although class D amplifiers have come an immensely long way, they are still technically not superior to the best linear amps. Perhaps in another few years they will be, but not at the moment.
This comparison has to be, by definition, between the best of each class. it would be unfair to compare the best of one class to something mediocre in the other class. Of course the best class D amps are better than a poorly designed class AB, and certainly better than a chip amp if that's what's being considered as an alternative.
The opposite is also true... have you ever heard one of those low power Tripath "Class T" disasters? Even a really bad class AB will sound better than that. It would be unfair to characterize all class D amplifiers as sounding like the worst of the breed, and there are some very very bad ones.
Cheers,
Owen
This is exactly the crux of the argument. The title of the thread is "Class AB vs Class D Amplifier" and what I'm trying to say is that although class D amplifiers have come an immensely long way, they are still technically not superior to the best linear amps. Perhaps in another few years they will be, but not at the moment.
This comparison has to be, by definition, between the best of each class. it would be unfair to compare the best of one class to something mediocre in the other class. Of course the best class D amps are better than a poorly designed class AB, and certainly better than a chip amp if that's what's being considered as an alternative.
The opposite is also true... have you ever heard one of those low power Tripath "Class T" disasters? Even a really bad class AB will sound better than that. It would be unfair to characterize all class D amplifiers as sounding like the worst of the breed, and there are some very very bad ones.
Cheers,
Owen
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.