AT32F421 Motor Controller

Version 1.1 had the following issues: 1.

LDO 12V->5V overheating was severe; the package size (SOT89) was too small, making it unusable for screens; a larger package was needed. The hardware has been fixed.

The original solution was to reduce the Buck output voltage to 10V.
2. The eg1198 can have its EN pin floating; it has an internal pull-up, saving resistors. 3. The eg2134 datasheet had an issue; the 8V power supply had no HO output, but switching to 12V allowed it to work normally. 4. Components should ideally be placed in areas where copper pouring is prohibited to avoid short circuits. In this experiment, a short circuit occurred at the Hall interface between 5V and GND. 5. The button component diagram provided by JLCPCB was incorrect; all button functions were lost, only Bluetooth or serial control was possible.
         The original solution has been fixed: disassemble all components and short-circuit them with tweezers to use as buttons. 6. SWD programming requires a common ground; otherwise, data transmission frequently fails. 7. Forgot to add a TVS diode, which damaged nearly ten of my MOSFETs. Luckily, they were all less than two cents each.
     Solution: Reduce the input voltage to below 20V.
8. Bluetooth is a bit useless; the highest baud rate is only 9600, so it's not very useful. It was removed in the second version, but it's still cheap and okay for playing around.
9. The Hall effect sensor doesn't have a pull-up resistor; you need to manually add a pull-up resistor on the MCU for it to work properly.
 
2.1 Simulink simulation and reality are roughly consistent.
2.2 Closed-loop square wave control (six-step commutation).
 
 
3.1 AT32_M&C_V1_2 preview.