6 Efficiency

Class-D amplifiers provide a significant improvement in efficiency compared to traditional linear amplifiers. Common reasons for seeking high efficiency in an audio amplifier include seeking longer runtimes from battery-powered speakers, managing heat in space-constrained applications such as mini-smart speakers or set-top boxes (STBs), or even passing government regulations on stand-by current in always-on devices. Higher efficiency means lower energy waste, where the waste of a Class-D amplifier is in the form of heat and electromagnetic field (EMF) emissions. The more wasteful the device is, the more the system needs to accommodate that waste. For thermal accommodations, this could mean a PCB with thicker copper layers, a heatsink and thermal paste, or sometimes even a fan with ventilation, all of which can contribute to overall system cost.

While it is easy to tout “greater than 90% efficiency,” this number refers to the power efficiency of the amplifier at a specific load and output power, which may not reflect the typical use case. The efficiency of an amplifier is best shown by a graph, for example the efficiency of TPA3221 in Figure 6-1. Since Class-D amplifiers burn a base amount of power just by operating, they’re most inefficient at their lowest output power levels and most efficient at their highest output levels.

A key concern in the audio industry for efficiency is idle power loss. This refers to how much power is consumed by the audio amplifier when there is no audio playing, but the device is ready to play at any moment. This is especially important for battery-powered applications, since the battery still drains when the device is on even while the user is not actively playing audio. Sometimes the idle power consumption is specified explicitly in a datasheet for different configurations, and sometimes it has to be calculated from other values. Power is the product of voltage and current, so the total idle power consumption can be calculated by multiplying the corresponding voltages and idle currents in the amplifier and adding them together. In cases when there is an LDO or boost converter, the calculation can become more complicated since there are additional losses to be accounted for in the entire audio system. See Figure 6-2 for an example of an idle current graph for different modulation schemes on the TPA3221.

The RDS(ON), which is the resistance between the drain and source terminals of the internal MOSFETs in the amplifier, effectively relates to the amount of power that is consumed when biasing the MOSFET on. Therefore, a lower RDS(ON) results in lower idle current, making this value frequently sought out by engineers to quickly compare amplifiers for idle efficiency.

While efficiency is critical to battery powered speakers for runtime, it’s also important to all systems for thermal and electromagnetic interference (EMI) reasons. Due to the conservation of energy, all of the “wasted” power converts to other forms of energy. In audio amplifiers, it manifests as heat and EMF. Heat is a primary concernbecause all the components inside a device have thermal operating limits. If a device gets too hot it could damage the device, but since most devices have thermal protection they usually shut off. Some devices have protection features to dial back power near high temperature thresholds such as thermal foldback to keep the device running as long as possible, and others just shut down when those thresholds are exceeded. Again, the higher the efficiency, the less demanding the amplifier will be thermally.