Looks like Class D is the only option for ultra clean, high power amplification

[CleanSound]

Stay away from class D amps, look at the professional reviews those amps cannot even pass standard square wave tests, show visible and substantial distortions. Class D amps outside of their use in subwoofers is another snake oil to squeeze money out of people which think they are "audiophiles".

Can you share the links of such square wave measurements?

 

[anarchist]

Can you share the links of such square wave measurements?

NAD C 298 power amplifier Measurements

Sidebar 4: Measurements

www.stereophile.com

NAD C 3050 LE BluOS streaming integrated amplifier Measurements

Sidebar 3: Measurements

www.stereophile.com

 

[NTK]

have to say there was no trolling on my side, just pure reference to objective measurements, like those found on stereophile.com

Just FYI: someone's opinion not matching yours does not necessarily needs to be called "trolling", those which managed to mature from teenage years already understand that.

So how well does your non-class D amp work at GHz range?

 

[Killingbeans]

Hypex, Purifi, IcePower. Those are the brands that produces these OEM Class D modules.

Don't forget Pascal. Another big Danish Class-D manufacturer.

Stay away from class D amps, look at the professional reviews those amps cannot even pass standard square wave tests, show visible and substantial distortions. Class D amps outside of their use in subwoofers is another snake oil to squeeze money out of people which think they are "audiophiles".

What's the frequency content of the distortion? If it's way above 20kHz, I don't really see a problem.

 

[DonH56]

The first harmonic to create a square wave is the third, so a 20 kHz square wave needs 60 kHz bandwidth to pass that, and so forth -- you probably need 7 or more harmonics to look more like an ideal square wave. See e.g. https://www.audiosciencereview.com/.../composition-of-a-square-wave-important.1921/ for the harmonics needed to create a square wave.

The output filter (in any amp, class D or otherwise, for stability and to block RF from feeding back into the amp, as well as suppress ultra-sonics generated by the amp), may add reduce the edges, add some phase shift, and introduce some overshoot, all of which is usually audibly benign IME/IMO but will alter the shape of a square wave. Note this does not add distortion like THD or IMD, but what some consider "linear" distortion that alters the frequency response (and thus the waveshape). Class D amplifiers usually have a large'ish filter inductor on the output to help reduce the switching frequency noise and that may lead to higher peaking (overshoot) on square waves, I don't know.

I am sure we can all find excellent and poor examples of square-wave performance in any class of amplifier. A transformer-coupled output (SS or tube) into a reactive load will probably look pretty ugly unless well bandlimited.

 

[radix]

Class D amps are usually tested with a low-pass filter, like the Ap AUX, is my understanding. The AP ones have a 40 kHz cutoff, so you won't get square waves beyond a few kHz out them them.

 

[boXem]

Hypex, Purifi, IcePower. Those are the brands that produces these OEM Class D modules.

The manufacturers that uses them that I am personally looking into are Nord Acoustic and Apollon because, they actually prioritize good looks, which is VERY important to me.

You may have some other manufacturers who sees me not recommending their products and will take exception and comment (one particular manufacturer is annoyingly vocal in doing so), don't mind them. I started a thread, which you can search for where I was attempting to start a list of all known Hypex, Purifi and IcePower class D manufacturers. But the thread got shut down because one of the manufacturers that I don't know about and did not list originally threw a pissy fit.

Home

Nord Acoustics supply Hypex Amplifier NCore NC500, NCx500, NC1200, Purifi Amplifier 1ET400A, 1ET7040SA. Hand built using discrete op amps and voltage regulators, Worldwide delivery.

nordacoustics.co.uk

Apollon Audio - Hypex NCore, Purifi Based Class-D Amplifiers

Hypex NCore & Purifi Class-D mono, stereo & multichannel Amplifiers hand-built with Highest Precision and Care. High-end sound with Efficiency and Great Detail.

apollonaudio.com

Thanks for giving me the opportunity to mention that at boXem, we prioritize good engineering and appretiate good looks.

 

[ahofer]

Which is based on measurements only, normally, not "tell them apart" subjectivity.

Plus how to interpret the measurements, knowledge of human audibility thresholds, and blind comparisons. You missed those parts entirely.

Here’s an explanation of those benign square wave results:

Post in thread 'NAD M23 Stereo Amplifier Review'
https://audiosciencereview.com/foru...23-stereo-amplifier-review.45462/post-1687148

 

[CleanSound]

Forgot to attach some pictures Apollon and Nord.

Here is an Apollon. This case looks pretty hefty, love the mesh vents.

Here is Nord, Nord offers two cases, this is the SE case upgrade, I think it's extra £100, but seems like worth it. I also spoke with Colin, the owner, at Nord, great guy, never any pressures anyone to buy nor is he ever in anyone's face to blow his own horn, takes his time to answer your questions with no expectations. I think Nord is one of the most tenured Dutch and Danish Class D manufacturers out there.

 

[CleanSound]

Thanks guys, For years I stayed away from high power amps, especially class A ( I had to turn my air conditioner up to much! and my wallet wasn't fat enough) and leaned heavily toward higher efficiently speakers and smaller amps. Now looking at all these class D amp high output ( hopefully they aren't as ridicules as some solid state ratings in the past) I may start looking for another set of speakers that need more wattage to run well.
It looks like this Netherland tech is more of what I like " clean sound is important to me."
Since we have moved to a new house ( downsizing ) my new listening room is a smaller room I use for office and now listening space, so actually I don't need larger amps
Anyway I'm glad I found this place to get me moving in the right direction for Hi Fi, Thanks

Alternatively, you can buy the Hypex DIY kit directly from Hypex and assemble it yourself.

I did that for my Class D with Hypex NC500 modules. I regret it big time, because by the time it was all said and done, I really didn't saved that much and the case that comes with it is pretty cheap and ugly, I had to spend more money on a better looking case. Also, there are no remote trigger kits for the old NC500 modules (the NCx500 has the trigger kit), so I has to hack a solution myself.

And the NCx500 kits, has "diy" labeling plastered on it, a bit too tacky for my taste, but I suppose some would take pride? It's pretty straight forward to assemble one, 8th grade project at worst case, so not a whole lot of pride to take in for the simple hand assembly.

DIYClassD is the DIY branch of Hypex Electronics BV for audiophile lovers.

Aimed at the DIY market we use Hypex Class D power amplifiers, power supplies, DSP solutions and plate amplifiers for HiFi and High End applications.

www.diyclassd.com

EDIT: I forgot to mention Buckeye, if you're American and you want to support American businesses, Buckeye is probably the best value you can find. Although the look is not for me.

Buckeye Amps

Buckeye Amps utilizes the most advanced audio amplifier technologies available. Authorized seller for Hypex and Purifi Class D amplifiers. Based in the US. Free shipping and worldwide service!

www.buckeyeamp.com

 

[mocenigo]

At best, you get two zeros with them. The class D and Benchmark are getting 3 zeros or more.

Again, not audible, but for those who want ultra clean.

The zeros with those big names are added to the price.

 

[Endibol]

That would be false. 24 bits ostensibly provides more dynamic range, but that’s not exactly what defines power requirements . As ever, it ever comes down to how much power you need to reach desired output levels in your space, with your specific speakers.

Edit: As is, 16 bit provides 96dB nominally (less with dithering). That’s quite adequate for most consumer content. Realistically, there’s only so loud you can listen, and so soft you can hear without getting drowned out by background noise in your room.

View attachment 307364

With 16 bit and dither MORE (not less as you write) dynamic range can be achieved ( approx. 112 dB).

 

[BR52]

With 16 bit and dither MORE (not less as you write) dynamic range can be achieved ( approx. 112 dB).

?? Can you explain that ? For me 16 bits pure can reach 96 dB SN with additional broadband noise ( dithering) it can only be lower.

 

[Steve81]

With 16 bit and dither MORE (not less as you write) dynamic range can be achieved ( approx. 112 dB).

Dither does not actually increase dynamic range; it reduces it by intentionally adding random noise to the mix to mitigate the effects of quantization error. This is desirable of course, but 16 bit will not suddenly have 112dB of actual dynamic range.

 

[boXem]

A calculation explaining this would be welcome.

 

[Steve81]

A calculation explaining this would be welcome.

6dB rise in noise floor = loss of 1 bit of resolution.
And more on dithering.

 

[Endibol]

?? Can you explain that ? For me 16 bits pure can reach 96 dB SN with additional broadband noise ( dithering) it can only be lower.

Have a look at this article and watch the video (famous!) as well:

24/192 Music Downloads are Very Silly Indeed

people.xiph.org

 

[boXem]

6dB rise in noise floor = loss of 1 bit of resolution.
And more on dithering.

Comment was not for you, sorry if not clear.

 

[boXem]

Have a look at this article and watch the video (famous!) as well:

24/192 Music Downloads are Very Silly Indeed

people.xiph.org

It is not dither but noise shaping that allows to gain dynamic range.

 

[NTK]

Here is a thread on dither started by JJ.

Dithering is a Mathematical Process - NOT a psychoacoustic process.

I have calculated a sine wave of magnitude .4, rounded it, added uniform dither, rounded that, and added tpd, and rounded THAT, i.e. I've set a step size of '1' in the quantizer. The routine that does this will follow in a comment. The photo I'm attaching now has 4 lines. each line shows a...

www.audiosciencereview.com

Paper by Lipshitz, Wannamaker and Vanderkooy for the mathematically inclined.

https://www.researchgate.net/publication/236340550_Quantization_and_Dither_A_Theoretical_Survey