Single, dual, and triple microphone configurations
Audio Front-End Solution
Support for small form factor designs

Voice control is increasingly becoming a preferred way for consumers to interact with electronics, from wearable devices to smart lighting.

As developers develop products of various sizes and designs to meet consumer demands, they face challenges with scalability to meet the diverse bill of materials (BOM) and form factor requirements of end applications.

Microchip Technology Inc., through its subsidiary Microsemi, today announced the AcuEdge ZLK38AVS Development Kit for Amazon Alexa Voice Service (AVS), which adds flexibility to device development with far-field voice pick up capability and multi-microphone array configurations. This solution meets small form factor development requirements while reducing BOM costs.
An audio front-end development kit with Amazon-approved single-mic hands-free configurations and dual- and triple-mic far-field configurations, the solution features the ZL38063 audio processor with a 300MHz DSP and dedicated hardware accelerators for voice processing.

It also supports barge-in recognition to enhance the consumer experience even in extreme audio playback environments with the Stereo Acoustic Echo Cancellation (AEC) feature.

With excellent flexibility across a variety of microphone options and configurations, the development kit enables ODMs and third-party developers to cost-effectively meet design criteria across a wide range of applications, regardless of form factor constraints.

Supported microphone configurations are as follows:

One-mic configuration: Provides hands-free voice coverage in products where form factor or cost constraints prevent a multi-mic array.

Two- and three-mic linear configurations: For sound bars, smart plugs and switches, set-top boxes, and other applications that are typically located in front of a wall and benefit from 180-degree far-field coverage. design.

Three-mic triangular configuration: This configuration features Amazon-approved long-range coverage and is ideal for applications requiring user interaction, such as portable speakers, wide-angle IP cameras, and voice-activated Wi-Fi routers.

This development kit supports far-field speech recognition in environments with interfering noises, such as device playback and external noise sources. This field-programmable and upgradable solution features proven signal processing algorithms that improve local trigger detection performance and cloud speech recognition accuracy, while the multi-microphone configuration uses Direction of Arrival (DOA) estimation to localize the primary voice source.

The development kit includes a ZL38063 audio processor that can be directly connected to the Raspberry Pi 3B with plastic and mounting hardware to simulate a microphone and speaker configuration for commonly recommended end applications.

All microphone options are supported, and it also includes an LED ring that indicates the location of the detected sound source, and a high-quality speaker representative of typical smart speaker applications.

Microchip operates audio labs and tuning facilities worldwide to support customers at every stage, from design through Amazon certification.

The ZLK38AVS Development Kit for Amazon AVS is available now.

“Voice-enabled devices are incredibly popular and continue to gain traction across a wide range of end applications, from home automation controllers to set-top boxes,” said Steve Caldwell, vice president of Microchip’s Wireless Solutions Group. “The ZLK38AVS development kit provides the flexibility needed to scale the audio front end across our product lines while enabling exceptional performance in challenging audio environments.”