In Marshall Nack's excellent review of Kharma's Matrix MP150 switching amplifier, he makes a very interesting statement. It's well worth extracting for its implications: " Bruno Putzeys, the lead engineer behind the Class D project at Dutch giant Phillips Electronics, and the man who developed the patent on which the MP150s are based, now works for Kharma and headed up the MP150 design team. He would like to dispel some of the dense fog enshrouding the subject of Class D design and open the window for some much-needed fresh air.
"First, it is not the case that everyone is employing one of the three or four basic, off-the-shelf, chip sets (Tripath, ICE, UdC, etc), and just plops it onto a circuit board. When you decide to work in Class D, the starting point is (or should be) a concept on paper. You are faced with two areas of major design decision points: 1) Control Strategy, which has some 14 parameters, each of which requires a decision, and 2) Power Conversion Strategy, with an additional three. Bruno figures these work out to 432 thinkable 'concepts', or combinations. About 100 of these have actually been attempted. Then, once beyond the schematic, there are hardware and circuit board level decisions. There is enormous variation - Class A amp designs actually present fewer permutations!"
If there's one guy who'd know, it's Putzeys, Mr. UdC. It's engineers like him who helped create the above referenced turnkey solutions for Class D OEM applications. But the genre in general often gets little luv for innovation or design chops. That's because many Class D amps on the market are repackaging jobs of prefab hardware in shiny boxes. By insisting on in-house design -- this is a Krell amp; I expect a Dan D'Agostino-designed circuit -- we fail to acknowledge that someone designed the prefab bundles we find in repackaged amps. If that someone -- or far more likely, the 'somefew' of a sizeable design team -- worked for giant think tanks at B&O, Sony, Yamaha or Phillips, the available resources of brain power and R&D funds there rather exceeded those available from a small cottage industry player. It's not that design expertise has diminished with Class D amps. It's that it's often outsourced. It benefits from engineering expertise that's well off the scale for hi-end consumer HiFi.
Popular currents sway us to automatically equate Class D with but a handful of core solutions that are endlessly cloned and repeated. As Bruno reminds us, that's selling things short and oversimplifying them. Complexities of design choices and engineering executions inherent in this amplifier class are just as real as they are in traditional circuits. For example, Kharma amps eschew an oscillator altogether. That could be an ingredient many until now believed to be an absolutely essential ingredient for high-performance Class D. Apparently not. Thanks to Marshall Nack for his Putzeys quote. It sets the record straight and points forward. If 332 possible Class D variants haven't been tried yet, who knows what the future holds? It seems this particular story is nowhere near fully told yet.