[serengetiplains]

The noise is already so slight that I can't help but wonder how any real improvement will be audible.

Cab, because I'm unsure what you mean by noise, and just to be clear, what I mean by noise is any deviation from perfect signal reproduction.  It's easy to mistake a single large instance of noise reduction, such as any audible upgrade---whether of an amplifier or DAC---as giving perfect reproduction.  Imo, the Ncore amps undoubtedly produce layers of noise that interfere with musical enjoyment.

One important source of noise is the power supply.  Every supply has, among other things, non-zero non-linear impedance which will create dynamical voltage fluctuations that will necessarily mix with the signal on the amplifier output.  From what I can discern, Bruno has created a more truly differential class D amplifier.  To the extent it operates in mirror-differential mode, it will amplify the signal aikido-like without mixing signal with noise---a kind of free passage through the thicket of power supply grunge.  I can see a benefit from operating two Ncores in bridge mode.  What noise does exist on the output of a single Ncore amp should reduce audibly further by operating in bridge mode.

[cab]

Imo, the Ncore amps undoubtedly produce layers of noise that interfere with musical enjoyment.

I think the data sheets (and Bruno) say otherwise....

[Freo-1]

The limitation with noise measurement data sheets are that they generally tied to a given single frequency into a resistive load, not a total musical signal reproduction into an actual speaker load (which can vary wildly). 

[serengetiplains]

I'm also interested in what data sheets do not say and what current knowledge does not know.  Fwiw I recall Bruno using pretty much the same language to describe the quality of both his UcD and Ncore lines.

But back to the data sheets.  They show distortion, yes?

[Freo-1]

I'm also interested in what data sheets do not say and what current knowledge does not know.  Fwiw I recall Bruno using pretty much the same language to describe the quality of both his UcD and Ncore lines.

But back to the data sheets.  They show distortion, yes?

Not quite sure what you are driving at. Are you saying that the residual noise will be amplified by the Class D operation?

 

[cab]

Look at the data sheets and you will see that the THD+N is already so low that any possible improvement will be extremely difficult, if possible at all, to discern. The distortion produced by this amplifier is already extremely low. I believe Bruno has already said that he didn't expect any audible sonic improvement from bridging. Try it and let us know.

[serengetiplains]

Freo, I'm saying the Ncore distorts.  Ergo, it can be improved.

Cab, low is not zero.  One reason for the Ncore's non-zero distortion rating, I'm sure, is the amp uses electrolytics in its power supply.  Electrolytics generate considerable non-linear distortion.  Some of that distortion is finding its way to the output---necessarily, according to basic electronics theory.

Bruno has said that THD plots do not reveal power supply anomalies of the types I'm naming here.  Measurements are helpful.  They just don't say everything.  Classical physics of, say, year 1800 could not measure what it didn't know to exist, like quantum phenomena.  Does that analogy make sense?  Progress always moves into what was previously unknown.  If something is unknown, by definition it cannot be measured---even assuming measuring capability for such.

[Freo-1]

Freo, I'm saying the Ncore distorts.  Ergo, it can be improved.

Cab, low is not zero.  One reason for the Ncore's non-zero distortion rating, I'm sure, is the amp uses electrolytics in its power supply.  Electrolytics generate considerable non-linear distortion.  Some of that distortion is finding its way to the output---necessarily, according to basic electronics theory.

Bruno has said that THD plots do not reveal power supply anomalies of the types I'm naming here.  Measurements are helpful.  They just don't say everything.  Classical physics of, say, year 1800 could not measure what it didn't know to exist, like quantum phenomena.  Does that analogy make sense?  Progress always moves into what was previously unknown.  If something is unknown, by definition it cannot be measured---even assuming measuring capability for such.

 
OK, so how would compare the distortion characteristics of the Ncore as opposed to something along the lines of a Pass Labs XA 30.5 or XA 60.5?

[serengetiplains]

Btw, I do plan to build a bridge Ncore amplifier.  I will drive each bridge with a single Hypex smps.  I'll report back after I receive the modules to say whether bridging does in fact work as I think it should.  I'm putting my money where my intuition is.

Freo, I'm unsure how to compare distortion spectra of two amps.  I don't doubt the Ncores are good amps (see above).  My sense is the Ncores are very good, but I'm expecting they'll suffer somewhat from NFB dullness.  I say "somewhat," here, because ideally I want a zero-distortion amp, which NFB helps attain, but my experience with NFB is it diminishes something subtly vital in music reproduction.  But the Ncore is a digital amp (refutations noted), which seems to me to open new possibilities for using NFB in ways that augment more than diminish.  And it seems to me the high switching frequency (500KHz thereabouts) should ameliorate some of the HF strangeness of class D, of which I'm very familiar (I own a Tact, which drives my ears wild after long listening sessions and which I attribute to VHF nastiness).

In audio, because the very parts we use to build amplifiers are solely responsible for distorting the signal, we're left with the choice of designing very simple amplifiers (like SETs and SS varieties like L'Amp, which I'm also building), or of using those distortion-creating components in a complexity that is finely adjusted to self-defeat that distortion.  Bruno's new design looks to me to employ complexity in rather innovative ways.  I'm impressed with what he has accomplished and look forward to hearing these critters.  Bridged, of course.

[OzarkTom]

If low distortion figures are so important to actual sound quality, Japan would have put out all of the tube amp companies out of business 40 years ago.

But new tube amps in the $80-100K were introduced at the Munich show in Germany last month. It seems that Japan failed.

[cab]

But new tube amps in the $80-100K were introduced at the Munich show in Germany last month. It seems that Japan failed.

Or it might be further proof that PT Barnum was correct..

[Freo-1]

Btw, I do plan to build a bridge Ncore amplifier.  I will drive each bridge with a single Hypex smps.  I'll report back after I receive the modules to say whether bridging does in fact work as I think it should.  I'm putting my money where my intuition is.

Freo, I'm unsure how to compare distortion spectra of two amps.  I don't doubt the Ncores are good amps (see above).  My sense is the Ncores are very good, but I'm expecting they'll suffer somewhat from NFB dullness.  I say "somewhat," here, because ideally I want a zero-distortion amp, which NFB helps attain, but my experience with NFB is it diminishes something subtly vital in music reproduction.  But the Ncore is a digital amp (refutations noted), which seems to me to open new possibilities for using NFB in ways that augment more than diminish.  And it seems to me the high switching frequency (500KHz thereabouts) should ameliorate some of the HF strangeness of class D, of which I'm very familiar (I own a Tact, which drives my ears wild after long listening sessions and which I attribute to VHF nastiness).

In audio, because the very parts we use to build amplifiers are solely responsible for distorting the signal, we're left with the choice of designing very simple amplifiers (like SETs and SS varieties like L'Amp, which I'm also building), or of using those distortion-creating components in a complexity that is finely adjusted to self-defeat that distortion.  Bruno's new design looks to me to employ complexity in rather innovative ways.  I'm impressed with what he has accomplished and look forward to hearing these critters.  Bridged, of course.

 
I’ll be looking for your feedback. To date, the best overall performance from amps I’ve listened to are Class A units from Papa Pass.  They definitely (to me) sound cleaner and more detailed than any Class AB or Class D unit I’ve heard so far. 
 
I also would like to listen to these Ncore units to see just how far Class D has progressed.  They certainly seem to have a good reputation and following.

[cab]

Freo, I'm saying the Ncore distorts.  Ergo, it can be improved.

Cab, low is not zero.

At what level is it below your ability to hear it? It is already below the noise floor of the measuring equipment a good way out. Again, I doubt anyone can hear the difference between 0.0003% and 0.0001% THD, especially when there are other devices in the chain that are producing distortion at levels 1000 times greater....

[serengetiplains]

I love simple amps, Tom.  I've built a single-stage tube amp---one hot 6C45P running the show---and it sounded glorious.  I'm building a L'Amp to replicate that experience which, for practical reasons, I let go.

Here's a question I find salient.  All other things being equal, is more distortion ever desirable?  My answer is no.  Don't get me wrong, I straddle all sorts of boundaries---I love tubes and NOS DACs and tubelike SS and class D and vinyl and computer playback and single- and many-driver speakers.  I want all the qualities I adore in all these components ... in one system. 

Freo, I'm currently building a Pass-inspired SIT amp---single stage, class A, inductor loaded, Hynes regulated supply (heat!), transformer signal-input, and parallel-SE output.  The amp is essentially push-pull with a single supply, but uniquely, each SIT will drive a separate full-range driver, the latter being crosswired to allow noise cancellation.  I asked the driver manufacturer to wind parallel voice coils to allow this noise cancellation to occur in the voice coil.  Unfortunately, they didn't comply.

Cab, precisely, how can we possibly know the lower threshold at which distortion levels are inaudible?  Bruno's Ncore amps suggest we haven't hit that threshold yet (yes?).  If we haven't, we have room for improvement.  And given the nature of things audible, or imo at least, the greater musical enjoyment resides in the last few percentage points of improvement.  Enjoyment level plotted against so-called percent-improvement is imo a logarithmic function.  The last 1% probably gives a magnitudinal leap in enjoyment.  That's been my experience so far, and is evidently the experience of quite a large number of people given the $$$ audible improvements fetch.  Why, after all, are people excited about the Ncores?

[Freo-1]

Very cool!
 
Will be looking forward to your feedback. 

[OzarkTom]

We audiophiles so worried about distortion, yet you will not hardly see any of the recording studios worry about it. Most use the worst SS mics ever made, most use Monster cables thouhout, lousy mixing techniques and high distorted mixing boards. No wonder recordings are just so-so.

There is an older 1950's AKG tube mic that is better sounding than any SS or digfital mic built today. On occasion, someone will use that mic to record today, but very rarely.

[jtwrace]

OK.  This is for the builds not technical discussion.  Please talk about this there.   

http://www.audiocircle.com/index.php?topic=106188.0

[cab]

Since our hearing is imperfect and has its limits of perception, the law of diminishing returns applies to distortion in audio reproduction. Less is better up to the point where it is no longer perceptible.

[jtwrace]

Here are some NCore amps that I built. 

[Nick77]

Nice!! How do we top that?