WTK6900FC-M02 Voice Recognition Module
AI‑Powered Voice Recognition with 5–8m Far‑Field, 300 Offline Commands & 95dB SNR Audio
The WTK6900FC-M02 is a high‑performance voice recognition module designed for local voice trigger applications. Powered by a 220MHz CPU with a dedicated BNPU V3 neural network accelerator, it delivers 5–8 meter far‑field voice recognition with 300 offline command words and up to 98% recognition accuracy. With 95dB SNR audio codec, built‑in 2MB Flash and 640KB SRAM, and UART control, it is the ideal solution for smart home appliances, industrial controls, and automotive voice applications.
- 220MHz AI core – high‑performance processing
- 5–8m far‑field – room‑scale voice pickup
- 300 offline commands – no cloud required
- 98% accuracy – reliable recognition
- 95dB SNR – high‑quality audio input/output
Voice Recognition Selection Guide
| Model | Recognition Range | Commands | Self‑Learning | Languages | Power | Package |
|---|---|---|---|---|---|---|
| WT2606A | 3–5m | 10 wake + 300 commands | ✘ | CN, EN | 30–40mA | QFN40 |
| WTK6900FC-M02 | 5–8m | 10 wake + 300 commands + 20 self‑learn | ✓ (1 wake + 19 commands) | CN, EN, JP, KR | 50–60mA | SSOP24 |
| WTK6900HC | 5–8m | 10 wake + 300 commands | Wake only | CN, EN | 25–30mA | SOP16/SSOP24/QFN32 |
| WTK6900HA | 3–5m | 10 wake + 300 commands | ✘ | CN, EN | 15–20mA | SOP16/SSOP24/QFN32 |
| WTK6900L | 0.5–3m | 20 commands | ✘ | CN | 20–25mA | SOP8 |
WTK6900FC-M02 offers the most comprehensive feature set with far‑field recognition, self‑learning (1 wake + 19 commands), and multi‑language support (CN, EN, JP, KR).
Key Features
Technical Specifications
| Parameter | Value / Condition |
|---|---|
| Module Size | 40 × 30 × 6.55 mm |
| CPU Frequency | Up to 220 MHz |
| Flash Memory | 2 MB |
| SRAM | 640 KB |
| eFuse | 512‑bit for encryption |
| PMU Input Voltage | 3.6V – 5.5V (5V typical) |
| IO Interface Level | 3.3V TTL |
| Speaker Output | 8Ω / 1W (max 460mA) |
| Working Current (Active) | ≤ 460 mA (with speaker) |
| Working Current (Standby) | ~50–60 mA |
| Audio Formats | MP3, 16k–24kbps |
| Microphone Input | ECM (electret condenser microphone) |
| Recommended MIC | JMO-627BA283R-10TC38 (sensitivity: -28dB, S/N: 70dB) |
| UART Baud Rate | 115200 (configurable up to 3Mbps) |
| Operating Temperature | ‑10°C ~ +70°C |
| Storage Temperature | ‑55°C ~ +150°C |
Pin Functions (10‑Pin Module)
| Pin | Name | Function |
|---|---|---|
| 1 | MIC+ | Microphone positive |
| 2 | MIC− | Microphone negative |
| 3 | SPK+ | Speaker positive (8Ω 1W) |
| 4 | SPK− | Speaker negative |
| 5 | WT8002 | Onboard amplifier (1W/8Ω) |
| 6 | WTK6900FC | Voice recognition chip (SSOP24) |
| 7 | LED (SPK) | Red when idle, off when working |
| 8 | LED (TX) | Green when TX reaches 3.3V |
| 9 | VCC | Power input (3.6–5.5V) |
| 10 | GND | Ground |
Pin 5 is the WT8002 amplifier IC. The module uses 3.3V TTL for UART communication (RX/TX are on pins 2 and 3 of the module’s internal connection – please refer to the schematic).
Command Output Examples
| Command Word | UART Output (Hex) |
|---|---|
| Wake: “小蜂管家” (Xiao Feng Butler) | 7E 06 FF 06 01 01 0D EF |
| “打开灯光” (Turn on light) | 7E 06 FF 06 01 02 0E EF |
| “关闭灯光” (Turn off light) | 7E 06 FF 06 01 03 0F EF |
| “调亮灯光” (Brighten) | 7E 06 FF 06 01 04 10 EF |
| “调暗灯光” (Dim) | 7E 06 FF 06 01 05 11 EF |
| “调到黄光” (Warm light) | 7E 06 FF 06 01 06 12 EF |
| “调到白光” (Cool light) | 7E 06 FF 06 01 07 13 EF |
Protocol: 7E + Length + Extension + Command + ID + Checksum + EF. The checksum is the low byte of the cumulative sum of Length, Extension, Command, and ID.
Application Scenarios
Why Choose WTK6900FC-M02
Frequently Asked Questions
What is the WTK6900FC-M02?
WTK6900FC-M02 is a high-performance voice recognition module featuring a 220MHz CPU, BNPU V3 neural accelerator, 5–8 meter far-field recognition, 300 offline command words, 95dB SNR audio codec, and UART control. It supports Chinese, English, Japanese, and Korean.
What is the recognition range?
WTK6900FC-M02 supports 5–8 meter far-field voice recognition, making it suitable for room-scale voice control applications in smart homes and industrial settings.
How many offline commands are supported?
The WTK6900FC-M02 supports up to 300 offline command words, enabling voice control without cloud connectivity.
What languages are supported?
WTK6900FC-M02 supports Chinese, English, Japanese, and Korean, with flexible customization options for additional languages.
What AI algorithms are used?
The WTK6900FC-M02 module uses DNN-HMM deep neural network-based speech recognition, powered by the BNPU V3 neural network accelerator, supporting DNN, TDNN, RNN, CNN, and parallel vector operations.
What control interface does it use?
It uses a standard UART interface (3.3V TTL) with a 7E … EF protocol. The module outputs command IDs via UART when voice commands are recognized.
What is the operating voltage range?
The PMU input voltage ranges from 3.6V to 5.5V (5V typical). A 460mA power supply is recommended for the speaker output.
What is the recognition accuracy?
In quiet environments, the recognition accuracy reaches up to 98%, with strong anti-interference capability for real-world applications.
What microphone does it support?
It supports electret microphones (ECM). Recommended model: JMO-627BA283R-10TC38 with -28dB sensitivity and 70dB S/N ratio.
What are the typical applications?
Smart home appliances, smart lighting, smart bathrooms, smart toys, industrial controls, automotive voice systems, and any application requiring local voice control and recognition.
Important: The module requires a 3.6–5.5V power supply with 460mA capability for speaker output. UART communication is 3.3V TTL – use level shifting for 5V MCU connections. The module must be stored in a dry environment (5% RH) and soldered within 48 hours after opening. For custom command sets or firmware modifications, contact Waytronic support. The self‑learning feature (1 wake word + 19 command words) is supported via voice command “learn wake word” or “learn command word”.
