2. amps-preamps
  3. 165827-20253

My NAD 3020 D proves your Class D tropes are wrong

I have a desktop integrated, the NAD 3020D which I use with custom near field monitors. It is being fed by Roon via a Squeezebox Touch and coaxial digital.

It is 5 years old and it sounds great. None of the standard myths of bad Class D sound exist here. It may lack the tube like liquid midrange of my Luxman, or the warmth of my prior Parasound but no one in this forum could hear it and go "aha, Class D!!" by itself, except maybe by the absolute lack of noise even when 3’ away from the speakers.

I’m not going to argue that this is the greatest amp ever, or that it is even a standout desktop integrated. All I am saying is that the stories about how bad Class D is compared to linear amps have been outdated for ages.

Great to see new development with GaN based Class D amps, great to see Technics using DSP feed-forward designs to overcome minor limitations in impedance matching and Atmasphere’s work on reducing measurable distortion as well but OMG stop with the "Class D was awful until just now" threads as it ignores about 30 years of steady research and innovation.
erik_squires
deadhead1000's avatar
deadhead1000
714 posts
 
I've had many Class D amps over the years, Hafler, NAD, Mcintosh, B&K. They all sounded good. I actually miss the B&K. I also had some Pioneer, Marantz, Denon and Sony amps/receivers which all sounded horrible. I even had a Hitachi Class G which actually sounded quite good.
But I finally did try a Class A (590AXII Luxman) and I can never go back to Class D. I'm not bashing D, just saying that A has been a big improvement over every D I've ever had.
dletch2
455 posts
 
Class D feedback works so differently than linear amps that I have trouble believing the amount of feedback behaves the same way, but what do I know?




How so?
erik_squires OP
16,060 posts
 
How so?


As far as I know, feedback is fundamental to Class D.  It simply can't operate without it.  Take a look at online documentation for how a triangle wave and the incoming signal are compared, and feedback is used to construct the ouptut.  The only real choices are where you take your feedback from.

Without feedback there is no output.
atmasphere's avatar
atmasphere
11,807 posts
 
Class D feedback works so differently than linear amps that I have trouble believing the amount of feedback behaves the same way, but what do I know?
As far as I know, feedback is fundamental to Class D. It simply can't operate without it.
@erik_squires You are working with a misconception! You can (and some do) operate class D amps without any feedback at all. Up until a year or so ago, that was how all our prototypes worked. Feedback isn't used to construct the output of a class D amp, and you can have output with no feedback. IIRC the Merill class D is a zero feedback design; Technics claims that their amp is zero feedback too.
erik_squires OP
16,060 posts
 
Technics is the curious case of feed-forward design. :) Kind of sort of.
More to discover