Sipeed Tang Nano 4K FPGA Development Board

Specifications

• FPGA Chip: GW1NSR-LV4C
• Logic Units (LUT4): 4608
• Registers (Flip-Flops): 3456
• Embedded Processor: Hardcore ARM Cortex-M3 (32-bit RISC, up to 80MHz)
• Memory: 180Kbit Block SRAM, 64Mbit HyperRAM (PSRAM), 256Kbit User Flash, 32Mbit SPI NOR Flash
• Display Interface: Onboard HDMI Connector
• Camera Interface: 24P DVP Connector (Native support for OV2640 cameras)
• Debugging/Programming: Onboard BL702 chip (Provides USB-JTAG and USB-UART via the USB-C port)
• User I/O: 38 to 44 accessible routing pins
• Onboard Peripherals: 2x User Buttons, 1x User LED
• Power & Data Interface: USB Type-C (5V input)
• Dimensions: ~60mm x 23mm

Applications

• Embedded Vision: Real-time video capture and filtering using the OV2640 interface.
• Retro Gaming & Emulation: Outputting custom generated graphics directly to monitors via the onboard HDMI port.
• Softcore/Hardcore Processor Design: Experimenting with the embedded ARM Cortex-M3 or flashing custom RISC-V softcores.
• High-Speed Logic Analysis: Building custom, multi-channel logic analyzers that outpace standard microcontroller sampling rates.

Unlock New Capabilities

• Past Projects: If you've mastered blinking LEDs, reading basic I2C sensors, or driving simple displays on an Arduino or ESP32...
• Feature Upgrade: The Tang Nano 4K unlocks true parallel processing. You can now drive an HDMI display while simultaneously capturing high-speed camera data without missing a single clock cycle.
• Effort/Requirements: Advanced. Requires knowledge of Verilog/VHDL and familiarity with the EDA. Minimum hardware needed: A PC with Gowin IDE installed and a high-quality USB-C data cable.

Virtual Technician & Wiring Guide

• ⚠️ CRITICAL VOLTAGE WARNING: The Tang Nano 4K operates on 3.3V logic levels. Do not connect 5V sensors or standard 5V Arduino signals directly to the GPIO pins, or you will instantly destroy the Gowin FPGA. Always use a logic level converter! We guarantee this board is Probots Logic-Safe when strictly paired with 3.3V components.

Quick Pinout Reference (Top Left to Bottom Right):

• VCC: 5V (from USB) and 3.3V (Regulated output)
• GND: Common Ground
• IO Pins: Refer to the EDA pin planner. The onboard LED is mapped to Pin 10. The user buttons are mapped to Pin 14 and Pin 15.

Package Includes:

• 1 x Sipeed Tang Nano 4K FPGA Development Board

Frequently Asked Questions (FAQ)

• What IDE do I use to program the Tang Nano 4K?
  You must use the  EDA (Educational or Standard edition) available from the official Gowin website.
• Does this board need an external JTAG programmer?
  No, it features an onboard BL702 chip that handles USB-to-JTAG and USB-to-UART natively via the USB-C port.
• Can I program the board using the Arduino IDE?
  No, FPGAs are programmed using Hardware Description Languages (HDL) like Verilog or VHDL, not C++ via Arduino. However, the embedded Cortex-M3 can be programmed using C/C++ via standard ARM toolchains.
• What is the maximum clock speed of the ARM Cortex-M3?
  The hardcore Cortex-M3 processor runs up to 80MHz.
• Can I drive a standard 1080p monitor with the HDMI port?
  While it has an HDMI port, the limited logic units are best suited for lower resolutions (like 720p or custom timings) rather than full 1080p 60fps rendering.
• Does the Tang Nano 4K come with a camera?
  This specific listing is for the board only. An OV2640 camera module must be purchased separately.
• What is the logic voltage of the I/O pins?
  The I/O pins operate strictly at 3.3V.
• Can I use 5V inputs on this board?
  No, applying 5V directly to the I/O pins will damage the FPGA. Use a logic level shifter.
• How much RAM does this board have?
  It features 64Mbit of onboard HyperRAM (PSRAM) alongside 180Kbit of Block SRAM within the FPGA fabric.
• Does it run Linux?
  No, 4608 LUTs and an 80MHz Cortex-M3 are not sufficient to run full Linux distributions. It is designed for bare-metal HDL and RTOS applications.
• Is the USB-C cable included?
  No, you must provide your own USB-C data cable.
• Why does my PC not recognize the board?
  Ensure you are using a data-sync USB-C cable, not a charge-only cable. Also, verify that the Gowin USB drivers are properly installed via Zadig.
• Can I use open-source toolchains like Yosys?
  Support for the GW1NSR-LV4C in open-source toolchains like Project Apicula is currently experimental/limited compared to other Gowin chips.
• What is the difference between Tang Nano 4K and 9K?
  The 9K has more logic units (8640 vs 4608), lacks the hard Cortex-M3 core, and uses a different Gowin chip architecture.
• How do I power the board in a standalone project?
  You can power it by supplying a clean 5V to the 5V pin, or via the USB-C connector.
• Are the pin headers pre-soldered?
  No, the board comes with loose headers that require soldering, giving you the flexibility to use it on a breadboard or solder it directly into a custom PCB.
• What happens if I lose power during flashing?
  The onboard flash memory might become corrupted, but you can simply re-flash the bitstream using the Gowin Programmer.
• Can I store data permanently on the board?
  Yes, you can utilize the 32Mbit SPI NOR Flash for storing persistent user data or secondary bitstreams.
• How do I access the Cortex-M3 core?
  You must instantiate the 'Empu' (Embedded Microprocessor Unit) IP core within your Gowin EDA project to bridge the FPGA fabric and the Cortex-M3.