NXP RT685 Audio Weaver / TalkTo


Leverage the dual-core architecture of an RT685 processor to run DSP Concepts’ Alexa far-field qualified TalkTo audio front-end on the HiFi4 DSP core and playback processing on the Cortex-M33 core for high performance voice-enabled smart speakers and sound bars. This full-featured design, with its built-in support for multiple USB audio modes including recording, streaming and playback processing, is extendable to support conference calling on USB cameras and speaker phones.


6m+ Far-field voice activation range

  • DSP Concepts high-performance TalkTo Audio Front-end

Seamlessly integrate Alexa to your product

  • High performance 4 mic TalkTo for far-field voice pickup
  • AVS qualified TalkTo with production-ready algorithms
  • 4 mic pre-processing on HiFi4 core

Support for multiple USB audio modes

  • High fidelity USB Audio streaming with multiple USB audio modes for recording, streaming and playback


Made for multi-core applications

  • Audio Weaver running on Arm® Cortex®-M33 and Cadence® Tensilica® HiFi 4 DSP processor core for multi-core applications
  • Playback processing and audio I/O on M33


Audio Front-end Features

DSP Concepts’ TalkTo for excellent far-field performance in noisy environments

  • • 4 mic Beamformer with Single Channel Noise Reduction
  • • State-of-the-art noise cancellation with Adaptive Interference Canceller (AIC) 
    • ○ No reference signal needed
    • ○ Eliminates 25 dB of external noise sources such as background music or a booming movie
  • • 30+ dB Acoustic Echo Cancellation (AEC) for excellent music barge-in

Stereo Audio Playback processing running on Cortex-M33


Support for multiple audio modes


USB streaming of mic signals into offline ASR

USB recording of raw and processed mic signals from external mic board

USB streaming of music from PC through speaker playback processing

Line input audio for speaker playback processing

Playback audio through line out to drive a powered speaker

Software Integrations

Software Integrations - Audio Weaver on Cortex-M33 and HiFi 4 core
Audio Weaver running on Cortex-M33 and HiFi 4 core
Software Integrations - Audio Weaver Designer Standard
Audio Weaver Designer Standard
Drag-and-drop Graphical Tool to customize & tune audio -processing models in real-time with 500+ modules
Software Integrations - Audio Weaver Core
Audio Weaver Core
Embedded audio-processing engine for Cortex-M33 and HiFi4 DSP
Software Integrations - TalkTo
DSP Concepts’ high-performance audio front-end for state-of-the-art far-field voice control performance
Software Integrations - Playback processing algorithms
Playback processing algorithms
Software Integrations - Sensory TrulyHandsfree
Sensory TrulyHandsfree ‘Alexa’ wake word engine

Hardware Specifications


  • i.MX RT685 Cortex-M33 core processor with Cadence Tensilica HiFi 4 DSP


  • 64 MB Macronix Octal SPI Flash operating at 1.8 V
  • 8 MB Apmemory PSRAM
  • Full-size SD card slot (SDIO)

Audio I/O

  • Stereo audio codec with line in/out and electret microphone
  • Stereo NXP TFA9894 digital amplifiers, with option for external +5V power for higher performance speakers
  • Support for up to eight off-board digital microphones via 12-pin header
  • Two on-board DMICS 

USB Connectivity

  • Onboard, high-speed USB, Link2 debug probe with CMSIS-DAP protocol (supporting Cortex M33 debug only) 
  • High-speed USB port with micro A/B connector for the host or device functionality
  • UART, I2C and SPI port bridging from i.MX RT685 target to USB via the on-board debug probe

Included with the demo kit

  • NXP RT685 Evaluation kit 
  • DSP Concepts 4 mic square and Trillium microphone array board
  • Voice UI Quickstart Bundle: Everything you need to know to design voice-enabled IoT products, including design guidelines, microphone placement, an Audio Weaver Designer layout (.AWD), and tools to get started.
  • USB cables
NXP RT685 Diagram

Ready to get started?