712 Kalman Filter for IMU :: Quicker, easier and cheaper to make your own chip!

How it works

This project implements a simplified 1D Kalman Filter (Complementary Filter logic) to fuse Accelerometer and Gyroscope data from an MPU-6500 sensor. It calculates Roll and Pitch angles using a CORDIC algorithm for the accelerometer and integrates gyroscope data with a steady-state Kalman gain.

The system consists of:

SPI Master: Configures and reads data from the MPU-6500 sensor (100Hz sample rate).
CORDIC Core: Calculates atan2 for Roll and Pitch estimation from accelerometer data.
Kalman Filter: Fuses the accelerometer angle with gyroscope rate.
UART Transmitter: Outputs the calculated angles (Roll, Pitch) as a binary stream at 9600 baud.

Data Format (8 bytes per packet):

Header High: 0xDE
Header Low: 0xAD
Roll High
Roll Low
Pitch High
Pitch Low

How to test

To test the design on real hardware, you can use the FPGA ASIC simulator breakout with an MPU-6500 sensor module and an FT232 serial probe.

The default configuration assumes a 10MHz system clock.

Wiring Instructions:

FPGA Breakout ui[0] -> MPU6500 ADO
FPGA Breakout uo[0] -> MPU6500 SDA
FPGA Breakout uo[1] -> MPU6500 SCL
FPGA Breakout uo[2] -> MPU6500 NCS
FPGA Breakout uo[3] -> FT232 RX
FPGA Breakout GND -> MPU6500 GND & FT232 GND
FPGA Breakout VCC -> MPU6500 VCC & FT232 VCC (make sure voltage levels are compatible, usually 3.3V)

Running on the FPGA Breakout (macOS):

Download the fpga_bitstream artifact from the latest passing GitHub Action run.
Extract the archive to find the .bin file (e.g., tt_um_intv0id_kalman.bin).
Connect the TinyTapeout demoboard to your Mac.

Clone the tt-support-tools repository:

git clone https://github.com/TinyTapeout/tt-support-tools.git tt
pip install -r tt/requirements.txt

Use the tt_fpga.py script to upload the bitstream. Specifying the serial port for Mac (typically /dev/cu.usbmodem*):

./tt/tt_fpga.py configure --port /dev/cu.usbmodem<your_port_number> --upload --name path/to/extracted/tt_um_intv0id_kalman.bin --set-default --clockrate 10000000

Running on the Demoboard ETR (using mpremote):

Connect the TinyTapeout Demoboard ETR to your computer.
Install mpremote:
```
pip install mpremote
```

Open the REPL using mpremote and enable the project:

mpremote repl

# Enable the project
tt.shuttle.tt_um_intv0id_kalman.enable()

Live Plotting: To plot the data live using Python from the FT232, you can use the provided plot_serial.py script.

Install the required dependencies: pip install pyserial matplotlib
Run the script: python3 plot_serial.py --port /dev/ttyUSB0

For simulation, run make in the test/ directory.

External hardware

MPU-6500 (or MPU-6050/9250) IMU sensor.
UART Receiver (USB-Serial adapter).

#	Output	Bidirectional
0	SPI_MOSI	SPI_MISO
1	SPI_SCLK
2	SPI_CS_N
3	UART_TX
4
5
6
7

712 Kalman Filter for IMU