[paul_houle]

Class-D amps aren't a common project for the electronic enthusiast: although the theory of the Class-D amp is beautiful and simple, there are a lot of details in the construction that are hard to master. Using an evaluation board, as the author does, is a good answer to the problem.

[dskhatri]

I agree. The author just picked up a reference design kit. It's a whole different matter to design the circuit itself. Picking the right MOSFETs, gate driver IC, feedback system are crucial to great performance. I'm not an audiophile but I know a great Class D engineer/audiophile who's designed a system with a THD+N of 0.003% (The Zetex amplifier the author chose has a THD+N of just 0.1%). It's cool to hear the engineer talk about the "color of sound".

[Addendum] THD+N is a measure of the purity of the output i.e how well the amplified audio signal matches the smaller input audio signal that's fed into the amplifier (from a CD player for example). So the lower the number the better. Typically, the amplifier circuit itself adds some distortions and noise to the actual audio signal. For example, if you fed a pure 1V 1kHz sine signal into the amplifier, your output might be a 20V 1kHz sine wave but with additional signals mixed in. These signals will include inherent noise generated by the amplifier and harmonics generated by the switching nature of the Class-D amplifer (if you look at your audio signal at the output of the MOSFET switches and before the low pass filter, it will look like a square wave or more specifically, a PWM).

http://en.wikipedia.org/wiki/Total_harmonic_distortion