NAD M23 Stereo Amplifier Review

[Rja4000]

This is especially interesting feature of NAD C 298, square wave sent into the amplifier comes out not exactly square, despite nearly perfect SINAD. I wonder is this a problem with most class D amps?

Or, maybe, you see this because you have to measure a class D amplifier with a low-pass filter on the Audio Precision ?
This limits bandwidth, per definition.
And since this measurement is about a 10kHz square wave, you see it.

The NAD C298 anyway has an integrated low-pass filter.

Here is what JA said about it:

"The small-signal bandwidth was restricted by the low-pass filter between the amplifier's class-D stage and its output terminals. Into 8 ohms (fig.1, blue trace), the ultrasonic rolloff reached –3dB at 66kHz. This rolloff lengthened the risetimes of a 10kHz squarewave (fig.2). There is a critically damped overshoot on the tops and bottoms of the waveform, but there is no ringing."

Limiting a square wave to a few harmonics will impact its form.

Here is what looks like a square wave when limited to H5:

Good news: we can't hear it.

 

[antcollinet]

This is especially interesting feature of NAD C 298, square wave sent into the amplifier comes out not exactly square, despite nearly perfect SINAD. I wonder is this a problem with most class D amps?

Nothing to do with Sinad and everything to do with bandwidth as pointed out by @rga4000.

 

[anarchist]

Or, maybe, you see this because you have to measure a class D amplifier with a low-pass filter on the Audio Precision ?
This limits bandwidth, per definition.
And since this measurement is about a 10kHz square wave, you see it.

The NAD C298 anyway has an integrated low-pass filter.

Here is what JA said about it:

"The small-signal bandwidth was restricted by the low-pass filter between the amplifier's class-D stage and its output terminals. Into 8 ohms (fig.1, blue trace), the ultrasonic rolloff reached –3dB at 66kHz. This rolloff lengthened the risetimes of a 10kHz squarewave (fig.2). There is a critically damped overshoot on the tops and bottoms of the waveform, but there is no ringing."

Limiting a square wave to a few harmonics will impact its form.

Here is what looks like a square wave when limited to H5:
View attachment 305883

Good news: we can't hear it.

I'm not convinced we cannot hear it. yes, low pass filter is there as in any class D by design, so what? Class AB amplifiers never show you such distorted signal, because they are true analogue, rather than digitalizing.
NAD C298 uses Purify module BTW, which is not much different to Hypex.

 

[Julf]

I'm not convinced we cannot hear it. yes, low pass filter is there as in any class D by design, so what? Class AB amplifiers never show you such distorted signal, because they are true analogue, rather than digitalizing.
NAD C298 uses Purify module BTW, which is not much different to Hypex.

It has nothing to do with analog ("pure" or not). The signal is not "distorted", and it would look exactly the same if you measured a class A/B amp with a low pass filter. Oh, and by the way, the hypex and purifi class D amps are analog too, no digitizing going on.

What makes you think you can hear it?

 

[anarchist]

It has nothing to do with analog ("pure" or not). The signal is not "distorted", and it would look exactly the same if you measured a class A/B amp with a low pass filter. Oh, and by the way, the hypex and purifi class D amps are analog too, no digitizing going on.

Isn't class D by design converts original signal to PWM, and then back? That's not digitization? I think it is.
The low pass filter supposed to convert PWM back to original signal, but at least in case of Purify it fails to restore original signal fully without distortion.
I now understand better why class D is mostly used for LF amplification, like subwoofers, this type of distortion is not as pronounced in LF band.

 

[Julf]

Isn't class D by design converts original signal to PWM, and then back?

Yes. Analog PWM.

That's not digitization? I think it is.

Digitization means converting the signal into discrete numerical values at regular sample intervals. The purifi and hypex class D amps don't do that, they produce a continuously variable (analog) PWM signal.

The low pass filter supposed to convert PWM back to original signal, but at least in case of Purify it fails to restore original signal fully without distortion.

Bandwidth limiting (that *all* amplifiers do) is not distortion.

I now understand better why class D is mostly used for LF amplification, like subwoofers, this type of distortion is not as pronounced in LF band.

I understand that is your bias that seems to stem from misinformation.

 

[peng]

I'm not convinced we cannot hear it. yes, low pass filter is there as in any class D by design, so what? Class AB amplifiers never show you such distorted signal, because they are true analogue, rather than digitalizing.
NAD C298 uses Purify module BTW, which is not much different to Hypex.

You can easily find out what is the highest frequency you can hear, so no one needs to convince you.

There are many such tests available on the internet such as the one below:

How high can you actually hear? Test yourself and find out

This hearing test will reveal the highest frequency you can pick up...

www.classicfm.com

I am convinced I can barely hear 15,000 Hz, so the output filters of class D amps has no audible effects on me, ymmv.. It is highly doubtful people older than even 20 years old can be bothered by even 25,000 Hz, not with the typical 20-30 dB ambient noise in most living rooms anyway.

 

[anarchist]

Yes. Analog PWM.

Digitization means converting the signal into discrete numerical values at regular sample intervals. The purifi and hypex class D amps don't do that, they produce a continuously variable (analog) PWM signal.

"Pulse Width Modulation (PWM) is a fancy term for describing a type of digital signal"

https://learn.sparkfun.com/tutorials/pulse-width-modulation/all

 

[anarchist]

You can easy find out what is the highest frequency you can hear, so no one needs to convince you.

There are many such tests available on the internet such as the one below:

How high can you actually hear? Test yourself and find out

This hearing test will reveal the highest frequency you can pick up...

www.classicfm.com

I am convinced I can barely hear 15,000 Hz, so the output filters of class D amps has no audible effects on me, ymmv.. It is highly doubtful people older than even 20 years old can be bothered by even 25,000 Hz, not with the typical 20-30 dB ambient noise in most living rooms anyway.

I see, so class D is only good for older people, like retirees...

 

[pablolie]

"Pulse Width Modulation (PWM) is a fancy term for describing a type of digital signal"

https://learn.sparkfun.com/tutorials/pulse-width-modulation/all

Nyquist-Shannon theorem. It results in a perfect analog signal. Math doesn't lie. If you can hear it, you weren't born on planet earth. :-D Many get confused because, to explain PWM visually, they show crude steps to easily visualize things. But when you apply Nyquist-Shannon, the signal will look absolutely and completely analog.

An equivalent is the number Pi. Is it wrong to use 3.14159 in practical terms? Do you prefer to go 3.1415926535897932384626433.... and spend the rest of your life writing it down? Do you think every groove in a vinyl is *perfectly round*?

 

[Waxx]

"Pulse Width Modulation (PWM) is a fancy term for describing a type of digital signal"

https://learn.sparkfun.com/tutorials/pulse-width-modulation/all

I think you need to read more about electronics and Pulse Wide Modulation. A SMPS also uses PWM to provide the current, and digital signals are a form of PWM (PCM form), but in a class D amp it's used to get an analog signal that is a copy of the original signal but way more powerfull. Nowhere there is a digital signal involved. Digital amplifiers do exist, but they work with pcm input in the amp module and do the conversion at the exit of the amp. A class D is full analog and works only with an analog input (AC signal).

And a pwm is done at 250kHz at minimum and with most modern amps go more like 1 to 1.5MHz. No way you can hear that switching. With older with bad filtering you could, but then we are talking about designs of 20-30 years ago or even longer.

The real problem class D had in the past, was disharmonic noise. You can hear that untill about 100dB (in worst case) noisefloor and it's very annoying (the cold hard sound). Class A or tube amps have a much higher noisefloor, but it's mainly harmonic distortion and that is why many are not bothered at all, they like it as the wamrth of the device.

Modern class D amps often have a sinad above 100dB so even very critical people like me are not hearing it anymore. And to be honest, most people won't hear it at 80dB sinad neighter. Most pro audio power amps aim at 80dB Sinad as good enough. And i've never seen a pro amp going above 100dB sinad. And on high power like in a concert hall or club or festival you won't hear it, the enviroment is to noisy anyway so it's masked.

 

[peng]

I see, so class D is only good for older people, like retirees...

No one said such a thing, where did you read that class d is only for old people? I did mention people older than 20 years old, are those people (such as 21, 22, 30) old by your standard? This is too funny..

 

[anarchist]

Nyquist-Shannon theorem. It results in a perfect analog signal. Math doesn't lie. If you can hear it, you weren't born on planet earth. :-D Many get confused because, to explain PWM visually, they show crude steps to easily visualize things. But when you apply Nyquist-Shannon, the signal will look absolutely and completely analog.

So you saying the distorted output signal in the stereophile review is just our imagination?
It's not really about what can you hear at this point, we cannot hear THD below 0.05% levels, and our speakers distort signal to at least 0.5%, yet we say here that SINAD 90 is superior to SINAD 70?
The point here is that top of the line class D amplifiers are not able to amplify square wave without distortion, and A or AB class amps can.

 

[pablolie]

So you saying the distorted output signal in the stereophile review is just our imagination?
It's not really about what can you hear at this point, we cannot hear THD below 0.05% levels, and our speakers distort signal to at least 0.5%, yet we say here that SINAD 90 is superior to SINAD 70?
The point here is that top of the line class D amplifiers are not able to amplify square wave without distortion, and A or AB class amps can.

You listen to square waves too much...

 

[Rja4000]

The point here is that top of the line class D amplifiers are not able to amplify square wave without distortion, and A or AB class amps can.

The point here, exactly, is that the amplifier you're referring to limits bandwidth as from 66kHz or so.

Is that a real life concern, of any sort ?
No, it's not.
It never was and will never be for a human being.
Whatever his or her age.

<EOT>

 

[Julf]

The point here is that top of the line class D amplifiers are not able to amplify square wave without distortion, and A or AB class amps can.

No, class A or A/B amps can't reproduce a square wave perfectly either, as perfectly reproducing a square wave requires infinite bandwidth.

Are you here to learn, or are you yet another troll spreading incorrect anti-class D audiophile folklore?

 

[ahofer]

It seems like most criticisms of new audio technology are invented by audiophiles out of ignorance to "explain" a subjective (and usually resistant) reaction.

I remember one of my professors at Berklee (back in 1984) "explaining" to me that the "X Factor" of a band happened in between the samples, and that's why digital sounded so sterile. It fits with the adjectives they choose to describe the sounds they imagine with equipment, which often have this owner-and-dog similarity - 'warm' for tubes, "muscular" for big heavy amps, and 'thin' for smaller/lighter equipment.

 

[peng]

No, class A or A/B amps can't reproduce a square wave perfectly either, as perfectly reproducing a square wave requires infinite bandwidth.

You think he knows that? I don"t think so.

 

[anarchist]

I think you need to read more about electronics and Pulse Wide Modulation. A SMPS also uses PWM to provide the current, and digital signals are a form of PWM (PCM form), but in a class D amp it's used to get an analog signal that is a copy of the original signal but way more powerfull. Nowhere there is a digital signal involved.

PWM in class D is not a copy of original analog signal, it's essentially newly generated digital PWM signal (zeros and ones) modulated with source analog signal, the modulated PWM signal is then amplified. The problem is extracting the original analogue signal without distortion is nearly impossible, hence class D remains a niche amplification technology mostly for subwoofer use.
I wonder is Hypex a derivative from hype? I think hype is really the only way to sell this technology to "audiophiles" for multi thousand $.

 

[Mnyb]

No, class A or A/B amps can't reproduce a square wave perfectly either, as perfectly reproducing a square wave requires infinite bandwidth.

Are you here to learn, or are you yet another troll spreading incorrect anti-class D audiophile folklore?

That means nothing reproduce squerewaves properly but that’s ok

Bandwidth limitations shows up the same digital or analog or class whatever does not matter , it’s true that there was fad in hifi with very wide bandwidth amps ( but still not infinite) , maybe it was part and parcel of classic specmanship or to much to late ? very old amps may had to little usable bandwidth ?

Overdoing stuff seems be part of this hobby, it could be that 44.1 KHz for CD is a close shave and demands careful design of filters the solution seems to be 48kHz or maybe 96KHz but now we have DXD at 384kHz ? Just because ? Similar thinking as 500kHz bandwidth amplifiers.

No instruments or singer produce square waves anyway due to that they to have limited bandwidth