Abstract: This paper presents the design of an absolute photoelectric rotary encoder based on the vernier encoding principle. First, the encoding scheme is proposed, introducing the three-track vernier encoding scheme, and the solution method for absolute position encoding is derived. The subdivision error model formula and sinusoidal signal correction results are given; after sinusoidal signal correction, the amplitude error is less than 0.33 %, the zero-bias error is less than 0.2 % , and the phase shift error is less than 0.22 %. The encoder's communication interface is described, achieving a baud rate of 10Mb/s at a communication distance of 10m and 5Mb/s at a distance of 25m. Finally, the structure of the entire encoder system is presented. Experimental results show that the encoder works normally and functions stably.
Keywords: Absolute photoelectric encoder; Vernier encoding; Subdivision model; BiSSC mode
Chinese Library Classification Number: TP274+ .2 Document Identification Code: A
1 Introduction
The photoelectric rotary encoder is an angular displacement measuring device that integrates optics, mechanics, and electronics [1]. In a typical closed-loop motion control system, the photoelectric rotary encoder acts as an observer, and its performance parameters directly affect the accuracy, gain, and stability of the control system [2]. With the development of robotics and automation technology, control systems require photoelectric rotary encoders to output absolute position, have miniaturized structures, and at the same time, have higher parameter requirements for resolution, accuracy, and time response speed. In order to improve system reliability, the encoder and controller need to be connected by a digital communication interface with verification capability.
The miniaturization of encoder structure restricts the maximum resolution of encoders [3][4][5]; on the other hand, the miniaturization of encoder structure exacerbates the impact of grating disk marking errors on accuracy [6]. Early absolute encoders used binary or Gray code encoders, with each code track corresponding to one binary bit, requiring multiple code tracks to meet the high-resolution angle measurement requirements. Therefore, the code disk size was very large, which did not meet the requirements of encoder miniaturization. Vernier encoding is an encoding method that uses the vernier principle to obtain the absolute position. Encoders using vernier encoding can significantly reduce the number of code tracks, which is beneficial to encoder miniaturization. While maintaining high resolution and high absolute accuracy, vernier encoding increases the allowable tolerance of signal error; the photoelectric detection method using phase array technology reduces the impact of marking errors on accuracy; this means that the manufacturing and assembly difficulty of grating disk and encoder shaft system is reduced [7], which is very beneficial to the industrialization of small absolute encoders.
This paper proposes an absolute photoelectric rotary encoder based on vernier encoding (Figure 1). A phase modulation method is used to encode the absolute code, and electronic subdivision technology is employed to obtain high-resolution, precise code. Due to the small number of code tracks, it is suitable for hollow shaft absolute encoders used in general-purpose servo motors. Currently, the laboratory prototype can achieve 25-bit resolution, and the maximum resolution of mass-produced products is consistently maintained at 23 bits. The encoder uses a BiSS-C digital interface, with a maximum communication rate of 10MHz and a maximum time response period of <10µs. The encoder employs a 6-bit CRC checksum, which verifies the measurement value for each communication cycle. For extremely secure applications, the encoder provides a 16-bit CRC checksum and cyclic technology to ensure extreme reliability.
2. Encoding Principles
This paper uses sinusoidal signals from three code tracks M, N, and S to determine the absolute angular position. This method requires lower precision than the method using two scales. The main code track M determines the system resolution and absolute precision, while the vernier code track N and the segment code track S generate coarse code information. For any scale period d, the phase α of position x within d is:
The period of β is equal to the least common multiple of the periods of the markings on tracks M and N. Since the number of markings on tracks M and N differs by 1, 1β is uniquely determined throughout the entire encoding distance, while 2β varies periodically throughout the entire encoder distance. Due to marking errors and a series of equivalent errors, 1β cannot be directly used for encoding. Using the periodic variation of 2β, 1α, together with 1β and 2β, constructs the vernier code for the entire encoder distance, as shown in Figure 2.
In the actual circuit, the differential signal corresponding to the M channel is sin_PMNM; the differential signal corresponding to the N code channel is sin_PN, NN; the differential signal corresponding to the S code channel is sin_PS, sin_NS; the expressions of the differential signals of the M , N , and S channels are shown in equations (4), (5), and (6), respectively.
3 segmentation technology
The encoder main track uses electronic subdivision technology to obtain high-resolution precision code. The quality of the sinusoidal signal determines the maximum resolution and system accuracy. To achieve better interpolation, it is important to identify and compensate for relevant signal errors. Typical error sources include zero-point offset (OS and OC) of sine and cosine signals; sensitivity to changes between sine and cosine signals (amplitude AS and AC); phase shift between sine and cosine signals deviating from 90° (φSERR, φCERR); and nonlinearity of the sensor's characteristic curve (sine shape deviation FSIN, FCOS). In general, the angle can be calculated in one period division by the arctangent of the quotient of the sine and cosine signals. According to formula (11):
Therefore, encoders need to have amplitude compensation, phase shift compensation, and zero bias compensation capabilities. In actual circuits, after compensation through complex programmable calibration circuits, the amplitude error is less than 0.33 %, the zero bias error is less than 0.2 %, and the phase shift error is less than 0.22 (relative to Vpp).
4 communication interfaces
The encoder uses a BiSS-C digital interface. BiSS-C is a high-speed, secure synchronous serial communication interface standard that can transmit sensor data back to the master port and read or write sensor parameters from or to the sensor registers in a bidirectional communication manner. The hardware of the BiSS-C interface is compatible with the SSI interface. Its basic timing is shown in Figure 3.
BiSS-C adopts the standard RS-422 electrical specification, requiring both clock and data signals. BiSS-C employs bus delay compensation technology to ensure reliable long-distance, high-speed communication. Typical communication distances for BiSS-C can reach 10 meters at 10MHz and 25 meters at 5MHz. The initial clock of the BiSS-C protocol frame is used to start the encoder's internal sampling circuit. The sampled signal is compensated and vernier-based to form an absolute value code. The absolute value code is shifted onto the data line under the trigger of the synchronization clock. The absolute value code generation time is synchronized with the initial clock transmission time, with a time deviation of less than 1.25µs (ignoring line factors). This provides a strong technical guarantee for controller algorithm optimization, and the technical advantages of the BiSS-C digital interface are even more pronounced in systems requiring high-speed, high-gain control performance.
5 System Implementation
A complete absolute encoder consists of a light source, code disk, slit, photoelectric receiver array, and decoding circuit. The decoding circuit includes six independent differential gain amplifiers, with two sinusoidal signals per code track, each 90° out of phase, for zero-bias compensation performed in hardware. The compensated signals are then quantized by an A/D converter, and amplitude matching and phase shift compensation of the sampled values are performed in software. The corrected measured values approximate ideal sinuses, and the phase angle is obtained through arctangent transformation. The absolute position of the encoder is finally obtained through vernier calculation. Figure 4 shows the block diagram of an absolute encoder.
6 Conclusions
This paper proposes an absolute encoder based on three-track vernier encoding. Three-track vernier encoding offers advantages such as simple marking, ease of assembly and adjustment, high resolution, and suitability for industrial application. The paper explains the encoding principle and provides the encoding process; it also describes the subdivision circuit, provides the error formula, and explains the digital communication interface; a functional block diagram of the encoding is presented. Practical application demonstrates that phase modulation effectively improves the encoding density of the code disk and reduces the number of code tracks, thus significantly contributing to the miniaturization of absolute encoders.
[1] Ye Shengxiang, Precision Measurement Technology of Photoelectric Displacement [M] . Chengdu : Sichuan Science and Technology Press, 2003.
[2] George Ellis Control System Design Guide : 3rd Edition [M] . Beijing : Beijing Electronic Industry Press, 2006.
[3] Liu Changshun, Wang Xianjun, Han Xudong, et al.; Eight-matrix ultra-small absolute photoelectric encoder [J] . Optics and Precision Engineering, 2010 , 18(2) , 326-332 .
[4] Yang Peng; Ai Hua; Liu Changshun; Development of an ultra-miniature quasi-absolute encoder [J]; Optoelectronic Engineering; 2008 , 35(12)141-144
[5] Chen Yun, Zhao Xingguo; Research on single-turn absolute photoelectric shaft angle encoder; Acta Photonica Sinica; 2008 , 12
[6] Compilation of Photoelectric Encoder Technology [M]; Changchun No.1 Optical Instrument Factory
【7】JoachimQuasdorfPosition/Presence/ProximityTheVernierScaleGoesDigital[J];
www.sensormag.com ;
January 1 , 2009
