Design considerations for using Class D amplifiers

The application of switching amplifiers or Class D amplifiers in consumer audio equipment has quickly risen to a very prominent level, from MP3 players, mobile phones, game consoles, LCD-TV to home theaters. The biggest competitive advantage of Class D amplifiers is the extremely high power conversion efficiency, which can be as high as 85% to 90% in practical applications, while linear Class AB amplifiers usually only reach 25% at typical power output levels.

In handheld applications, the low power consumption of Class D amplifiers allows designers to provide high audio quality while maintaining long battery charge intervals. Battery life is a key performance indicator for all personal communication and audio equipment. For mains-powered devices (such as A/V audiovisual products and game consoles), the high power efficiency of Class D amplifiers can result in lower heat dissipation. This allows designers to use smaller heat sinks to achieve a more compact shape, lower material and assembly costs. In fact, a well-designed power supply can eliminate the need for heat sinks for applications with output power of up to several watts per channel.

Class D Amplifier Solutions

The basic topology of a Class D amplifier includes a pulse width modulator, a power bridge output circuit, and a low-pass filter. Currently, the Class D amplifiers on the market can help users save a lot of design work, such as shielding the electromagnetic interference generated by the switching operation of the amplifier and selecting the most appropriate switching frequency. Increasing the switching frequency can reduce the output filtering requirements, but will cause greater power losses due to the MOSFET gate capacitance. Therefore, the switching frequency selection requires a balance between the number of external components and power conversion efficiency. The design of the power bridge depends on the desired output power of the Class D amplifier.

For example, there are Class D headphone drivers and Class D speaker drivers available on the market today, and a key difference between these different configurations is the design of the output stage. Class D amplifiers designed for speakers can produce output power from less than 1 watt to several watts without the need for a heat sink. These ICs provide a single-chip solution for many consumer applications, from portable multimedia players to game consoles and some LCD-TVs. In most of these applications, especially handheld products, a single-chip solution is critical.

However, for very high output powers, a Class D amplifier can be implemented in conjunction with an external output stage built with audio power MOSFETs. The Class D amplifier must be provided with a suitable preamplifier, and the discrete MOSFETs chosen must be optimized for digital audio operation.

[page]

Filtering

The output of the Class D MOSFET H-bridge is a square wave representing the audio signal. The switching frequency components must be attenuated to prevent interference and ensure that the final product passes electromagnetic compatibility (EMC) certification. This requires a low-pass filter with a cutoff frequency slightly above the audio band. Therefore, the higher the switching frequency, the greater the attenuation of these components. This allows the use of smaller external filtering components.

另一方面，MOSFET上的损耗会随著开关频率的提高而增加，从而降低效率，并导致更大的功耗和带来相关的热管理问题。特别是，在MOSFET栅极电容上产生的开关损耗，会随著工作频率的提高而线性增加。因此，通常认定D类放大器输出级的设计要具有低损耗的MOSFET，并正确设置了开关频率，以满足电磁干扰(EMI)的特定要求。

Connecting a filterless Class D amplifier to a speaker (such as a cell phone speaker) is a distinct advantage for applications where size and cost are both sensitive. When the Class D amplifier output is close to the speaker, the parasitic resistance and inductance of the speaker coil can form a proper LR low-pass filter. An example of a Class D amplifier that can be used in a filterless configuration is the Wolfson WM8960. If the Class D amplifier output is far away from the speaker, a ferrite bead is needed to create a small amount of additional inductance to improve EMC performance.

Power Design

Designers using Class D amplifiers must also pay closer attention to the effects of power supply performance on audio output quality than with linear Class A/B amplifiers. Because the Class D output is a switching stage, effectively connecting the power rail directly to the audio output, audio-band fluctuations on the power supply will directly modulate the output signal. Therefore, designers must ensure high load regulation over the audio band or take steps to eliminate the effects of mains or audio-band ripple.

Many manufacturers offer floating regulators that can be added to an existing power supply to improve load regulation when necessary. Using a separate regulator for each amplifier output has the added advantage of reducing crosstalk between audio channels. However, an additional regulator or pair of regulators increases the overall implementation cost. In addition, regulator power dissipation reduces efficiency gains, which is a key reason for using Class D amplifiers.

In addition, improving the amplifier's power supply rejection ratio (PSRR) reduces the impact of load regulation on the audio output signal. Adding feedback from the PWM output to the analog audio input improves the PSRR by compensating for variations in the supply voltage. This can achieve PSRRs as high as around 80dB, which is very close to the PSRR of differential Class AB amplifiers used in portable applications. However, if the Class D amplifier input signal is purely digital audio, then this technique is not applicable. The PSRR of a fully digital Class D amplifier is 0dB, and the designer must ensure regulation for small deviations in the supply voltage.

The transient performance of the power supply should also be considered. In order to accurately reproduce the PWM waveform, the power supply must be able to react quickly to sudden current changes. Linear amplifiers have less stringent requirements in this regard because the bandwidth of the output stage is limited to the audio range. For a power supply designed for Class D amplifiers, voltage fluctuations outside the audio band that cause poor transient response will modulate the PWM signal, introducing harmonic distortion that can be heard at the audio output.

A helpful technique is to place the MOSFETs of different output stages in different switches, which can reduce the peak supply current. For example, Wolfson's 5-7.1 channel digital power amplifier controller WM8608 has a built-in 'PWM output phase' function, which can introduce a 160ns delay between the PWM signals of each output channel, which has the effect of dispersing the switching transient current within the PWM cycle. In a 6-channel multi-channel system, this technology can significantly reduce the maximum transient load current and reduce crosstalk.