Jump to content

Sir Sanders Zingmore

Advance in Class D amps?

Recommended Posts

Arg   

Bill, 'delusion levels' applies to the tendency to a belief in (1) the infinitely subtle ear, (2) one's ability to change multiple factors and hear the contribution of any one of them within the chaos, and (3) one's ability to conduct sighted listening and somehow be immune to the well established principles of conscious and unconscious bias.

The all time classic would be "I trust my ears".

Pretty normal stuff on audio discussion boards.

Share this post


Link to post
Share on other sites

Finally, intermodulation distortion test:

For this test, burst signal is used

burst.png

I tested both amps at 2khz, 5khz, 10khz, and 20khz.

The result shows that at the loudspeaker both ClassD and Class AB amps acoustic intermodulation distortion is the same.

2khz:

Burst%25202khz.png

5khz:

Burst%25205khz.png

10khz:

Burst%252010khz.png

20khz:

Burst%252020khz.png

Summary

So from this little study I summarise:

- Assume that the acoustic output of the system is what matters

When measured at the loudspeakers:

- Class D distortion sweeps equals Class AB

- Class D tone test equals Class AB

- Class D intermodulation distortion test equals Class AB

- Class D is load-sensitive and care must be taken to avoid LPF behaviour

Measuring the Amplifier themselves?

- When measured at the Amplifier, Class D may show higher distortion, however this difference is insignificant compared to loudspeaker / transducer distortions

So as per OP's original question on Hypex's claims:

Hypex presents a decisive leap forward in class D audio performance. Ncore® is the first class D amplifier not just to nudge the best linear amplifiers, but to surpass them in every aspect relevant to sound quality

The studay can be taken to the other way: Would these Class D claim also manifest at the system's output? I doubt it.

Further improvement of this study can be extended by utilising

- Other types of transducer (plasma, AMT, xstat)

- Other types of Amplifiers (Class A)

The study can be downoaded as .pdf here:

http://www.scribd.co...lone-amplifiers

Share this post


Link to post
Share on other sites

When measured at the loudspeakers:

- Class D distortion sweeps equals Class AB

- Class D intermodulation distortion test equals Class AB

- Class D is load-sensitive and care must be taken to avoid LPF behaviour

Measuring the Amplifier themselves?

- When measured at the Amplifier, Class D may show higher distortion

This is a very worrying statment. That Class D only exhibits the distortions (below link) when a speaker is not attached but just a 8ohm test resistor, and then you say class D is load sensitive????? What more benign load than a 8ohm test resistor compared to the wildly fluctuating resistive load of speakers not to mention their capacitive, inductive loads and phase angles as well.

"Measuring the Amplifier themselves?"

What like these that were taken on "Audio Precisions Analyser" of the Nuforce 9 V2 far more accurate. http://www.hardwareanalysis.com/content/article/1831.2/nuforce-reference-9-02-amplifier/

Cheers George

Share this post


Link to post
Share on other sites

This is a very worrying statment. That Class D only exhibits the distortions (below link) when a speaker is not attached but just a 8ohm test resistor, and then you say class D is load sensitive????? What more benign load than a 8ohm test resistor compared to the wildly fluctuating resistive load of speakers not to mention their capacitive, inductive loads and phase angles as well.

"Measuring the Amplifier themselves?"

What like these that were taken on "Audio Precisions Analyser" of the Nuforce 9 V2 far more accurate. http://www.hardwarea...9-02-amplifier/

Cheers George

Even on cheaper Class D which I had, the amp's distortion is negligible at the acoustic output.

That's not an R load, that's what we listen to and clearly the best explanation is transducer distortion outweighs class D distortions.

re: load, I used them outside their design range (2 ohms). Class AB is happy, Class D is not.

Share this post


Link to post
Share on other sites
Arg   

I would just like to point out that it is not in the spirit to post poor test results for poor or mediocre class D amps. You could do that for any amp type.

More to the point is the test measurement for the best-of-class class D amps. e.g. the ncore 400 that the OP identified.

Share this post


Link to post
Share on other sites
eras   

I would just like to point out that it is not in the spirit to post poor test results for poor or mediocre class D amps. You could do that for any amp type.

More to the point is the test measurement for the best-of-class class D amps. e.g. the ncore 400 that the OP identified.

I agree. Whatsmore, posting measurements for amplifiers that have now been revised, and with data that was obtained over 6 years ago- is slightly misleading.

Thanks for all the good work, gainphile. Keep it up :)

Share this post


Link to post
Share on other sites

I would just like to point out that it is not in the spirit to post poor test results for poor or mediocre class D amps. You could do that for any amp type.

More to the point is the test measurement for the best-of-class class D amps. e.g. the ncore 400 that the OP identified.

it would be wise to wait until the ncore amp is independently reviewed and tested before getting too excited. Testbench results from the manufacturer are not usually repeatable.

Share this post


Link to post
Share on other sites

Is distortion the same as harmonics ? (genuine question, I would have thought distortion at 10kHz was the 10kHz output signal being different to the 10kHz input signal)

I think the confusion here is between THD and noise.

Distortion and Total Harmonic Distortion are perhaps used interchangeably. When reading amplifier specs then THD+N, THD and SNR need to be seen together.

I haven't thought up a solution but another interesting addition to gainphile's tests would be to have a passive loudspeaker. Reduced IMD is one of the benefits of an active speaker and that's what we're looking at in these results. On the other hand tests at low freq wouldn't be comparable to high freq tests because the speaker driver is different. Which is why I guess he's stuck to the frequencies that can be tested with the XT25.

Share this post


Link to post
Share on other sites
BradC   

Not sure if it has been mentioned in this thread, but one of the main distinguishing features of the Ncore amp is its very low output impedance (~milliohms). Further this output impedance does not vary with load impedance - one of the main parameters that causes different amps to sound different.

This makes ncore isgnificantly different from previous class D amps that are mentioned in this thread.

Early comparison reports of the ncore with other amps are very favourable for the ncore.

Share this post


Link to post
Share on other sites

It appears this thread has drifted again.

WRT measurements of class D, most of the above distortion measurements are not that useful.

Since class D amps have a fairly aggressive LPF at the OP, previous multi tone IMD tests are not that relevant because a) the tones are spaced too far apart and will be masked by the OP LPF b ) they don't fully represent the HF

linearity of the class D amp.

As I stated a few pages ago the correct test, specifically developed by Audio Precision are two very closely spaced tones up close to 20kHz. Since sidebands will show up on -both- sides of the tones, we will see a true

picture of the HF linearity and it will not be masked by the OP LPF. Ref Ncore HF IMD test below, even at 400W it is close to the measurement baseline.

Another useful measurement is OPZ versus freq, specifically how much the OPZ tilts up at higher frequencies. The flatter the better. Ncore OPZ versus freq attached.

There's not much arguing that this amps linearity is quite a landmark compared to other class D amps.

post-110074-0-69291700-1334919410_thumb.

post-110074-0-61584600-1334919419_thumb.

Edited by zenelectro

Share this post


Link to post
Share on other sites

I must confess that I stopped understanding (and indeed caring) about the measurements quite a number of posts back. Perhaps some kindly soul can

a) summarise what they all mean and,

B) tell me whether or not these purported advances in class D actually*sound* better?

Share this post


Link to post
Share on other sites

Yes that too, and also this looks a little sus to me, why does the THD measurments stop at 10khz where's everything above that to 20khz and beyond?

Cheers George

post-106386-0-28594100-1334966770_thumb.

Edited by georgehifi

Share this post


Link to post
Share on other sites

It appears this thread has drifted again.

WRT measurements of class D, most of the above distortion measurements are not that useful.

Since class D amps have a fairly aggressive LPF at the OP, previous multi tone IMD tests are not that relevant because a) the tones are spaced too far apart and will be masked by the OP LPF b ) they don't fully represent the HF

linearity of the class D amp.

As I stated a few pages ago the correct test, specifically developed by Audio Precision are two very closely spaced tones up close to 20kHz. Since sidebands will show up on -both- sides of the tones, we will see a true

picture of the HF linearity and it will not be masked by the OP LPF. Ref Ncore HF IMD test below, even at 400W it is close to the measurement baseline.

Another useful measurement is OPZ versus freq, specifically how much the OPZ tilts up at higher frequencies. The flatter the better. Ncore OPZ versus freq attached.

There's not much arguing that this amps linearity is quite a landmark compared to other class D amps.

Indeed, but the frequency response is pretty poor. -0.5dB @ 20kHz places it below the cheapest linear (SS) amp I can imagine. AND, as Arty has intimated, the data that Hypex DON'T show, is more telling. It would be nice to see impulse response data. The figures shown by Hypex are impressive, but I'd like to see more.

Share this post


Link to post
Share on other sites
Rob181   

WOW!! I am a new member...what an interesting read. More opinions here than a room full of lawyers.

I have recently purchased a Spectron Misician MK 3 with VCap & Bybee upgrades, based soley on the amount of positive reports I read....I have to say it is by far the best power amp I have heard. Makes the Ambience 1600's sing like no other amp I have heard before including a Mac. To be fair, the Spectron does not use switching power which appears to be an issue in this forum.

I have read many threads with designers stating that tests and achieving benchmark specifications are important but if it sounds RS than all the best results mean nothing.

I trust my ears..though this is a bigger risk now than it was 10 years ago...bugger!!

Share this post


Link to post
Share on other sites

WOW!! I am a new member...what an interesting read. More opinions here than a room full of lawyers.

I have recently purchased a Spectron Misician MK 3 with VCap & Bybee upgrades, based soley on the amount of positive reports I read....I have to say it is by far the best power amp I have heard. Makes the Ambience 1600's sing like no other amp I have heard before including a Mac. To be fair, the Spectron does not use switching power which appears to be an issue in this forum.

I have read many threads with designers stating that tests and achieving benchmark specifications are important but if it sounds RS than all the best results mean nothing.

I trust my ears..though this is a bigger risk now than it was 10 years ago...bugger!!

The Ambience 1600 is a very nice loudspeaker. It has a very benign load impedance (almost purely resistive above 400Hz) and a relatively 'tame' HF response. As such, I'd expect no problems for almost any Class D amp.

Where Class D amps suffer, is when they are connected to difficult loads and/or highly revealing loudspeakers.

Share this post


Link to post
Share on other sites

To be fair, the Spectron does not use switching power which appears to be an issue in this forum.

The issue is in this thread, rather than the forum at large. I wonder how many people have actually heard a nice d-class set up. I don't think I have the best system in the world, but I have yet to have anyone come over, listen and not enjoy the music* - and that includes more than a few sound engineers, producers and musicians.

* Isn't that what it is all about? No offence intended, but I believe I have heard the phrase 'measurebating' bandied about on this forum before...

I trust my ears..though this is a bigger risk now than it was 10 years ago...bugger!!

You and me both.

Edited by Vitruvian

Share this post


Link to post
Share on other sites

Measurements are interesting but you ultimately listen to it.

You makin' it sound like the 'measurement' people are stupid enough to ignore what they like to hear.

Once somebody has listened to it ... then they (could) measure it (again).

Share this post


Link to post
Share on other sites
Arg   
....Where Class D amps suffer, is when they are connected to difficult loads and/or highly revealing loudspeakers....

All amps suffer under difficult loads, it's in the definition.

eg an amp that is designed to perform well with 0.5 ohm speakers eg full range ribbons, is 'going to suffer' when put on a 16 ohm speaker because it can't deliver the voltage.

And vice versa.

So, the misapplication of an amp to an environment that it wasn't designed for, is not a criticism of the amp. It's just poor judgement.

As for 'highly reveaing loudspeakers' ...... give me a break.

Your first post in this thread said 1MHz switching frequency is not good enough, that it needs to be 5MHz, that "it's all about switching frequency".

Well the ncore 400 is 0.5 MHz, so I think you want to make it fail on principle.

What if it is an outstanding amp and your principle is wrong?

Share this post


Link to post
Share on other sites

All amps suffer under difficult loads, it's in the definition.

eg an amp that is designed to perform well with 0.5 ohm speakers eg full range ribbons, is 'going to suffer' when put on a 16 ohm speaker because it can't deliver the voltage.

And vice versa.

Not necessarily.

So, the misapplication of an amp to an environment that it wasn't designed for, is not a criticism of the amp. It's just poor judgement.

Never an issue in doubt.

As for 'highly reveaing loudspeakers' ...... give me a break.

OK. So, you've never listened to a highly revealing pair of speakers? Whatever you may think, some loudspeakers are brutally revealing, whilst others are much easier on amplifiers and sources.

Your first post in this thread said 1MHz switching frequency is not good enough, that it needs to be 5MHz, that "it's all about switching frequency".

Yep.

Well the ncore 400 is 0.5 MHz, so I think you want to make it fail on principle.

Nup. I have yet to hear one. It might have solved all the problems that afflict all the other Class D amps I've had on the bench or in a system. Or not.

What if it is an outstanding amp and your principle is wrong?

Then it's wrong.

Share this post


Link to post
Share on other sites
Nada   

Has anyone got in the cue for ordering the DIY 400 nCore, so Zaph can test out his 5MHz predictions?

The 400nCore seem to be consistently sold out on the site and the DIY thread is buried in 377 pages and counting.

Is this an aberration or a game changer?

NC400-3.jpg

Share this post


Link to post
Share on other sites

I wasn't prepared to wait for the new NCORE module to make its way into vendors' products, so instead have opted to try Hypex's existing product in the form of Channel Islands Audio D200 mk2 monoblocs.

I'll post something regarding these, compared to the Valvet Class A monoblocs in due course. Initial impressions are quite favourable.

I was convinced by the use of a traditional linear PSU in the CI Audio products, as opposed to SMPS as found in the B&O based units from the likes of Bel Canto and W4S.

Edited by Lil Caesar

Share this post


Link to post
Share on other sites

I think the power supply board in the Bel Canto M series makes a difference over the stock ICEpower module.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now


  • Recently Browsing   0 members

    No registered users viewing this page.



×