本研究以開發可校正誤差之磁性編碼器為核心，透過將磁性編碼器以及控制單元(Microcontroller Unit, MCU)進行結合，設計一個具訊號後處理以及補償的磁性編碼器感測電路。 根據先前實驗室的研究結果，本研究將透過定點數運算取代浮點數運算來降低MCU的運算時間並且使用二次多項式的拉格朗日插值法推估建立補償模型中所需要的餘弦波模型數值，並透過將誤差模型的數據進行分段取樣後進行快速傅立葉轉換(Fast Fourier Transform, FFT），透過這步驟可以大量減少計算所需記憶體的空間使得利用MCU進行FFT可行化。並透過步進馬達以全步進的方式進行驅動，根據補償後的數值與理論數值進行性能的對比驗證。 此外本研究同時提供使用者透過通訊方式更新補償模型的參數並將更新的參數存取於非揮發性記憶體中並於開機讀取，更新參數時不需再透過PC端編譯以及燒錄，更方便使用者進行使用。 ;This study focuses on developing a magnetic encoder that can correct eccentricity. By integrating the magnetic encoder with a Microcontroller Unit (MCU), design a magnetic encoder sensing circuit with signal post-processing and compensation capabilities. Based on previous laboratory research, this study replaces floating-point operations with fixed-point operations to reduce MCU computation time. Additionally, quadratic polynomials Lagrange interpolation method can estimate the sinusoidal model for compensation modeling. Furthermore, by segmenting and sampling the data, the required memory space for Fast Fourier Transform (FFT) can be significantly reduced, making it feasible to run FFT on MCU. By driving the stepper motors in full-step mode, comparing compensated values with theoretical values for validation. Furthermore, this study allows users to update compensation model parameters via serial communication and stores parameters in non-volatile memory for boot-time retrieval. Users can update parameters without requiring PC-side compilation and burning, to make it more user-friendly.
