Class D and PWM
Class D amplifiers are based on the PWM principle as they produce a PWM equivalent of the analog input signal which is fed to the loudspeaker using a suitable high quality filter network that blocks the high frequency carrier signal and recovers the amplified original audio signal. These amplifiers are characterized by very good efficiency figures (≥ 90%) and compact size/light weight for large power outputs. For a few decades, industrial and military PWM amplifiers have been in common use, often for driving servo motors. Field-gradient coils in medical MRI machines are driven by relatively high-power PWM amplifiers too.
The Direct Stream Digital (DSD) sound encoding method uses a generalized form of pulse-width modulation called pulse density modulation, at a high enough sampling rate (typically in the order of MHz) to cover the whole acoustic frequencies range with sufficient fidelity. This method is used in the SACD format, and reproduction of the encoded audio signal is essentially similar to the method used in class-D amplifiers.
Variable Width Pulses on Bottom - Resulting Sine Wave Across the Top
Class-D amplifiers work by generating a train of square pulses of fixed amplitude but varying width and separation, the low-frequency portion of whose frequency spectrum is essentially the signal to be amplified. The high-frequency portion serves no purpose other than to create a two level waveform. Because it has only two levels, it can be amplified by simple on/off switching. The output of such a switch is an identical train of square pulses, except with greater amplitude. Such amplification results in a wave-form with the same frequency spectrum, but with every frequency uniformly magnified in amplitude.
A passive low-pass filter removes the unwanted high-frequency components, i.e., smooths the pulses out and recovers the desired low-frequency signal. To maintain high efficiency, the filter is made with purely reactive components (inductors and capacitors), which store any excess energy until it is needed instead of converting some of it into heat. The switching frequency is typically chosen to be ten or more times the highest frequency of interest in the input signal. This eases the requirements placed on the output filter.
Theoretical power efficiency of class-D amplifiers is 100%. In the real-world Class D amplifier efficiencies well over 90% are common. By contrast, linear A/B-Class amplifiers typically exhibit efficiencies around 50%. An ideal class-B amplifier has a theoretical maximum efficiency of 78% but we've discussed their downside above. Class A amplifiers (purely linear, with the devices always "on") have a theoretical maximum efficiency of 50% but few if any actually achieve this kind of number. Some versions have average efficiencies below 20%.
By using this high efficiency, compact Class D amplifier technology in all GoldenEar designs, GoldenEar is able to provide class-leading power and performance in small and affordable packages.
Some of the preceding is based upon information from "Wikipedia".