Improve the innovative design of smart speakers and seize the next market opportunity!

Reducing system cost is a top priority for the entire smart home industry. In smart speaker design, the most likely thing designers can do is to reduce system cost through audio amplifiers. Instead of sacrificing sound quality in exchange for system cost, the audio design team of Texas Instruments (TI) has taken a different approach and developed a series of solutions around audio amplifiers that can improve performance while reducing system cost.

For example, for audio systems below 10 W, TI EMI suppression technology replaces inductors with ferrite beads. The average cost of four inductors is $0.16, while the average cost of four ferrite beads is only $0.04. This alone can save designers an average of $0.12 in a two-channel system.

Small battery size can make smart speakers smaller, lighter and more portable. Higher efficiency and optimized thermal effects not only help reduce battery size but also maintain battery life.

Microphones and system-on-chip (SoC) are major power hogs, but they need to be constantly listening for commands, so to reduce standby power and heat dissipation, designers must look elsewhere.

Starting with the audio amplifier is an ideal choice because, apart from the SoC, it has the highest standby power consumption and the largest heat dissipation. To reduce standby power consumption, the audio amplifier can adjust the duty cycle of the output pulse width modulation using a hybrid modulation module to control the current from the inductor to the speaker, thereby reducing the idle current by 50%. In this way, the overall power consumption can be reduced, the heat dissipation can be optimized, and ultimately the battery life can be extended without increasing the battery size.

The accuracy of speech recognition affects the performance of virtual assistants and is also an important indicator of smart speaker performance. One of the challenges is that microphones pick up a variety of environmental noises, including other electrical appliances, human conversations, and even the music being played by the smart speaker. In order to filter out environmental noise and more accurately recognize the user's voice commands, smart speakers should be equipped with multiple microphones and use comprehensive technologies to achieve accurate speech recognition.

One of these technologies is echo cancellation, which recognizes the digital signal of the music being played and removes it from the microphone input. Echo cancellation digitally removes any background noise, including the music. To better filter out ambient noise, audio amplifiers with echo cancellation have an SDOUT (electronic sound output) signal. As shown in Figure 1, the SDOUT signal is fed into the SoC, which filters out the music signal received by the microphone.

Figure 1: Echo cancellation system block diagram

Smart speakers are designed to be small and beautiful. However, with smaller size, the audio output power will also be reduced. An amplifier with good speaker protection function can increase the volume output of the speaker without compromising the sound quality. Using algorithms and models, the speaker protection amplifier can analyze the input music signal and then provide a peak output power of 5 times the rated power of the speaker, ultimately achieving a two to three times output volume amplification effect without affecting the sound quality of the speaker. In this way, smart speakers with a small appearance can also achieve high-quality sound effects.

In addition to adding voice interaction functions to traditional speakers, another focus of the smart speaker market is to achieve high-fidelity sound. Although most Bluetooth® speakers on the market do not have voice recognition or voice interaction functions, in the near future, Bluetooth plus voice recognition and voice interaction will become the market standard.

Portable smart speakers with their own batteries need to achieve the same power as line-powered speakers, which will become the next opportunity in smart speaker design.

Over the past 10 years, speaker design has evolved from large, home-use to smaller, portable, and smarter personal applications. In this process, Texas Instruments has provided a large number of technologies and system solutions to achieve longer battery life, better sound quality, stable connectivity, and fast charging for smart speakers, while minimizing system costs for customers. Figure 2 shows the block diagram of the Texas Instruments smart speaker system, which can be used for quick system selection and design.

Figure 2: TI smart speaker system block diagram