Researches On The Technology Of Servo Motor Magnetic Encoder Based On TMR Sensor

The encoder is an important part of the servo motor,which has an important influence on the positioning and control performance of motor in modern servo system.Magnetic encoder is a new type of encoder which developed with the optical encoder,Compared with the optical encoder,Magnetic encoder has a strong ability to resist vibration,high reliability,strong anti-interference ability,simple structure,small volume,and so on,which has the trend of replacing the optical encoder in many application fields,magnetic encoder has been the focus of research at home and abroad in recent years.The magnetic encoder used to detect the magnetic field signal is based on Hall Effect,AMR(anisotropic magnetoresistance resistance)Effect and GMR(giant magnetoresistance resistance)Effect of magnetic sensor,these magnetic sensors exist many shortcomings like low sensitivity,high power consumption,and so on.As the newest generation of magnetic sensor,TMR(tunnel magnetoresistance resistance)magnetic sensor has a lot of advantages like the strong signal output,low power consumption,prime temperature characteristic,high sensitivity,and so on,which makes up for the deficiency of the above several kinds of magnetic sensor to a certain extent,so this paper carry out of the research on servo motor magnetic encoder based on TMR sensor.CORDIC(coordinate rotation digital computer)algorithm is widely used in magnetic rotary encoder,because of the noise and non-orthogonal signal caused by encoder manufacturing and installation process,classic CORDIC algorithm has a big output angle error.Based on this,this paper designs a new type of magnetic encoder structure to reduce encoder angle error of installation structure,and according to the characteristics of angle error produced by the non-orthogonality of signal,the classic CORDIC algorithm is improved and verified.The main working contents are as follows:(1)According to the detection characteristics of TMR sensor,a simple structure,and small volume,high integration of the magnetic encoder structure is designed.The two-pole magnet is selected to provide magnetic field,and the magnetic field distribution of the two-pole for single pole is analyzed by Maxwell software to determine the best mounting clearance between the two-pole and the TMR sensor.(2)On the acquisition of magnetic field signal,this paper studies on magnetic encoder signal generation and detection,and focus design on signal detection and signal pretreatment circuit,chose proper FPGA device according to the consumption of system hardware resources,and build the magnetic signal acquisition platform on this basis.(3)In the aspect of algorithm design,aiming at the noise in the voltage signal that converted by A/D device,the median filter,wavelet transform and mean filter three kinds of noise reduction technology and performance have been comprehensively analyzed.The CORDIC algorithm is realized by using pipelining structure of sixteen levels,this paper analyzes the calculating error of angle,according to the result that error curve is similar to the characteristics of the cosine function,uses CORDIC algorithm to process the value of the input signal,and then shift the phase and again use the CORDIC algorithm to process the value of the input signal,deals with the difference between the results of two CORDIC algorithm processing,finally calculates the angle of the servo motor.(4)Improved CORDIC algorithm is verified by Voltage signal from Signal acquisition platform,the results show that the output angle error is between-0.25°to 0.25°,so it effectively reduced the angle detection error,and improved the angle detection accuracy.