The First Speech Synthesis ICs Issue Acoustic Alerts From Silent EVs
The first set of dedicated speech synthesis ICs addresses the limitations of microcontroller-based voice systems.
Electric vehicles (EVs) are much quieter than traditional gasoline and diesel-powered vehicles, and that can spell trouble for pedestrians who use audio cues to help avoid traffic. The auto industry has come up with acoustic vehicle alerting system (AVAS) standards to mitigate this issue.
The ML22120TO and ML22120GP are synthesis chips from Lapis Technology.
Lapis Technology, a Rohm Group company, recently introduced the industry’s first dedicated ICs designed for AVAS applications. The electric vehicle warning speech synthesis ICs, the ML22120xx series (datasheet linked) (ML22120TB, ML22120GP), can easily produce a wide variety of external vehicle alert sounds, including a warning for an approaching intersection, a notification for an open or closed sliding door, a sound for charging completion, and other regulatory imposed and non-regulatory sounds.
What Are Acoustic Vehicle Alerting Systems?
Any vehicle powered by electric motors, be it an electric streetcar, battery-powered bicycle, or passenger EV, is inherently quieter than those powered by internal combustion engines (ICE). While low noise is usually a desirable feature, for pedestrians, it can be deadly. Because pedestrians listen often more than they look for vehicle traffic, silent EVs can lead to a number of tragedies.
The transportation industry and regulatory institutions have looked to artificial sound generation to mitigate this danger—namely, in the form of acoustic vehicle alerting systems (AVAS). AVAS has been required by several nations since 2019 and is being widely adopted by the motor vehicle industry. The standard dictates, among other things, that vehicles emit a continuous sound when moving between 0 and 30 km per hour/18.6 mph (20 km per hour, 12 mph in the EU). The sound must vary in volume and pitch proportional to its speed so that pedestrians can sense speed based on the tone and intensity.
Sounds must emit between 56 decibels and 75 decibels and fit a frequency range hearable to most humans. AVAS allows for variations in the type of sound, and car manufacturers can and do implement custom sounds to give their cars a unique personality. As long as the basic requirements are met, manufacturers have a lot of flexibility.
Limitations of MCU-Based AVAS Sound
Many microcontrollers are very capable of generating realistic sounds. However, with AVAS, there is more involved than just making the sound. First, it must be characterized and programmed into the MCU. Second, it has to be tested for reliability under real-world conditions. For example, you don’t want a warning sound to be stopped by a high-level interrupt coming from elsewhere in the MCU system. Finally, the sound needs to be verified for compliance with AVAS and certified as such.
Lapis Technology's AVAS ICs mitigate each of those limitations. The ML22120TB and the ML22120GP integrate hardware functions that include dedicated four-channel synthesis hardware along with a built-in, five-band equalizer and fader.
Application circuit of the ML22120.
A synthesizer is essentially a multi-channel waveform generator. By varying the various frequencies, amplitudes, and waveforms, high-fidelity sound can be generated within the chip without the need for MCU oversight.
Design Benefits of the New AVAC ICs
A developer does not need to be an expert in creating sound. They only need to use the chip’s GUI to select, listen, and adjust the tones. As a dedicated synthesis chip, the chip cannot be interrupted by anything else going on in the MCU once a tone is triggered in the field. The chip is designed to make AVAS sounds, so compliance, verification, and certification are significantly easier.
Basic specifications of the ML22120xx family.
The dedicated chips, along with a GUI configuration system, reduce the design time. Since the chip has AVAS compliance built-in, the validation time is significantly reduced over microcontroller-based software solutions. For products in the field, the chips shorten reaction time and provide a greater variety of available tones.
All images used courtesy of Lapis Technology/Rohm Group.
