Design Article
Signal Chain Basics #39: Portable audio designers are excited about Class-G audio amplifier architecture
ShreHarsha Rao,
Product Marketing Manager,
Portable Audio, Texas Instruments
3/15/2010 6:00 AM EDT
Audio is an integral part of portable consumer electronic devices. An integrated headphone audio power amplifier helps amplify the low-power baseband audio signals to drive crisp and clear audio when using headphones. These amplifiers also need to be highly efficient for longer battery life. To meet this challenge, designers are using Class G, a unique audio amplifier topology.
Typical linear audio-amplifier topologies are Class A, B, C, and AB. These audio amplifiers are linear; however, they are inefficient, see Table 1 and Figure 1.

Table 1: Linear audio amplifier topologies.
(Click on image to enlarge)

Figure 1: Conduction angle for various amplifier topologies; vertical axis is zero-centered amplitude, horizontal axis is conduction angle from 0 to 360°.
(Click on image to enlarge)
Efficiency is defined as the ratio of output power (power delivered to the load) to input power (power drawn from the battery) and is expressed as a percentage. Higher efficiency implies that less power from the battery is wasted as dissipated heat and losses. To improve a portable audio device's battery life, an amplifier needs higher efficiency.
A Class-AB (linear) amplifier has a fixed supply rail and consumes a fixed amount of supply current for the desired output voltage. The supply current equals the output current in a bridge-tied-load (BTL). The current flowing from the supply through the load causes a voltage drop across all the output MOSFETs. These currents multiplied by the voltage drop across the MOSFETs create the large power dissipation in the amplifier, which is why Class-AB amplifiers are only 50 percent efficient.
What is Class G?
At a very high level, Class G topology is a variant of Class AB topology with multiple supplies. Class G topology takes advantage of the fact that typical audio/music sources have a very high crest factor (10 to 20 dB). This means that the peak audio signal is higher than the average audio signal (RMS). Most of the times the audio signals are at lower amplitudes and very rarely exhibit a higher peak.
The new Class-G topology uses adaptive step-down converters to create a supply voltage that moves with the audio signal. It creates a lower supply voltage with sufficient headroom for most average audio signals, and switches to a higher supply voltage to cater to occasional peak voltages.
Due to the adaptive nature of the power supply, power dissipation is greatly reduced for typical music/audio sources with high crest factors. This results in reduced current consumption from the battery, resulting in higher efficiency than Class AB architecture.
This supply voltage is adaptive. It increases for louder audio signals, thus preventing distorting large peak voltages, and decreases to reduce power dissipation during small audio peaks.
Class-G Operation
The Class-G amplifier operation is illustrated in Figure 2, where the supply voltage is 1.3 V for lower audio voltage peaks and adaptively increases to 1.8 V to accommodate higher peaks.

Figure 2: G topology adaptively moves amplifier supply for power savings.
(Click on image to enlarge)
A step-down DC/DC converter is used to generate these lower supply rails, Figure 3.

Figure 3: Class G headphone amplifier block diagram.
(Click on image to enlarge)
Class G amplifiers use adaptive supply rails and include a built-in step-down converter to create the headphone-amplifier positive-supply voltage (HPVDD). A charge pump inverts the HPVDD and creates the amplifier negative supply voltage (HPVSS).
This allows the headphone amplifier output to be centered at 0 V. When audio signal amplitude is low, the step-down converter generates a low HPVDD voltage (HPVDDL) (again, Figure 2). This minimizes power consumption of the Class G amplifier while playing low-noise, high-fidelity audio.
If audio amplitude increases due to louder music or a transient peak, the step-down converter generates a higher HPVDD voltage (HPVDDH). The HPVDD rise rate is faster than the audio peak's rise time. This prevents audio distortion or clipping. Audio quality and noise floor are not affected by HPVDD.
This adaptive HPVDD minimizes supply current while avoiding clipping and distortion. Because normal listening levels are below 200 mVRMS, HPVDD is most often at its lowest voltage, HPVDDL. Thus, Class G amplifiers have higher efficiency than traditional Class AB headphone amplifiers.
Next month we will cover frequency planning with high-speed ADCs.
About the Author
ShreHarsha Rao is a Systems Engineer at Texas Instruments where he researches application development for low-power wireless and RFID systems. He has a Masters Degree in Electrical Engineering, likes hiking, sports, and is in a poker league.



