5.1 Introduction
Much has been written in this text regarding the efficiency of various power amplifier topologies. While class A is known for its circuit layout simplicity, it is also known for its very low efficiency. Class B and class AB, while more complex than class A, present serious improvements in efficiency. In spite of these improvements, the family of class B amplifiers can hardly be considered as exhibiting high efficiency. Although not explicitly covered in this text, class G and H topologies are variations on class B and attempt to increase efficiency through the use of multiple sets of power supply rails or output devices, and in the process, tick the complexity up to another level.
The class D amplifier is perhaps the last word in amplifier efficiency. Theoretically with ideal devices, the efficiency of the output stage approaches 100%. Unlike the other amplifier forms, the transistors used in class D amplifiers never operate in the linear region; the output devices only operate as a switch, in either saturation or cutoff. High switching speed turns out to be a huge plus as it plays a major role in efficiency.
The increase in efficiency comes at a considerable increase in circuit complexity, however, for some applications this turns out to be a very good trade-off. As odd as it might at first seem, the two areas where class D topologies have taken root are at the opposite ends of the power output spectrum. The first area is perhaps the most obvious, mainly, very high output power amplifiers. An example might be an amplifier used as part of a large public address system and capable of delivering in excess of 1000 watts into a loudspeaker. High efficiency does two things here: First, it reduces the waste heat in the amplifier itself, and second, it reduces the current draw from the AC mains. Both of these are serious issues in a PA system used to fill a stadium or large concert hall as there may be dozens of such amplifiers comprising the system. As a bonus, improved efficiency also leads to a lighter and small enclosure because the need for heat sink area and mass will be reduced, as will the size of the AC power supply transformer. These traits will also reduce production costs and help offset the design complexity cost. The advantages have become so great that, in recent years, class D designs dominate the high end professional audio power amplifier market as well as the very high power automotive audio market (here there is another system limitation working in favor of class D, and that’s the limited current capacity of the vehicle’s alternator to deliver current).
The second area where class D has found acceptance is for low power portable devices. Examples include personal music devices, cell phones and hearing aids. Output powers for these applications might range from tens of milliwatts up to a few watts, so excess heat is generally not a big problem except in the most compact of enclosures. What is a problem, though, is the energy budget. Unlike a large PA amplifier that might pump out in excess of two horsepower, these portable devices do not have the luxury of running off of the AC mains with tens or even hundreds of amps of current capacity. Instead, these devices are restricted to battery power and batteries can only store so much energy. For a given battery capacity, higher efficiency directly translates into longer battery life. Another way of thinking about this is that, given a higher efficiency, a smaller battery can be used to achieve the same battery life, and this means that the unit can be both smaller and less expensive. Of course, nothing says that we can’t opt for a little of each.
