#subdivide the error

Foreword

Photoelectric shaft angle encoder, also called the photoelectric angular-position pickup, it is a kind of accurate digital angular measurement device incorporating all, machine, electricity into an organic whole, it converts the angle information of the axle to the number code, can realize dynamic measurement and real time control after connecting with computer and display device. With the wide application of the Coder in departments such as industry, national defence, spaceflight,etc., the technical indicator to the Coder has proposed the increasingly high requirement. Examine the important technical indicator that the angular error is the Coder, it is the main weight which examines the angular error to subdivide the error, subdivide the verification of the error and demand to use the delicate small angle measuring instrument | The instrument goes on under the rigorous laboratory condition, and the detection procedure is complicated, time is long. Use the scene to detect subdividing the error of the Coder in time fast in order to store the Coder, develop subdividing the fast measuring system of the error of a kind of Coder, reach the photoelectric acquisition of signal of the P-code of the Coder the computer, carry on harmonic analysis, calculate subdividing the error of the Coder.

A data collecting system

1.1 System hardware

The key device of this system is TMS320F2812 and MAX125. TMS320F2812 a DSP at a high speed of TI Company deals with the chip. This system adopts a slice of DSP chip to control 2 slices of A/ D to change the chip. Fig. 1 shows and controls 2 sheets of system block diagrams which change the chip for DSP.

1.2 TMS320F2812

C281x series DSP is the 32-bit microcomputer fixed point digital signal processor that TI Company introduces newly, are based on TMS320C2000 digital signal processor platform and developed, its code and 24x/ 240x digital signal processor are fully compatible.

C281x processor adopts C/ C ++ and writes the software, it is high in efficiency, users not merely can employ the high level language to write the system program, can also adopt C/ C ++ efficient mathematic algorithm. C281x series digital signal processor all have a higher characteristic while finishing tasks such as mathematic algorithm and system control,etc., in this way, it needs trouble of many processors in a system to avoid users. C281x processor kernel includes a multiplication of * 32 32-bit microcomputers and adds up to the unit, can finish 64's data-handling capacity. Thus make this processor finish the more high-accuracy treatment task.

1.3 MAX125

MAX125 is that a high speed 14 A/ Ds with multi-channel sampling holder changes the chip. This chip includes 14 for one, change-over time is successive approximation type A/ D converter of 3μ s, internal voltage reference of one +2.5 V. Every sampling holder corresponds to the input signals of No. 2, can receive the signal input of No. 8 to cost altogether. The input voltage is ± 5 V. The chip has ± 17 V to input voltage protection.

What MAX125 adopted is parallel data interface, the same as data access sequence of DSP, does not need to wait for cycle while communicating with DSP, can reach very high communication speed. The highest speed that the single channel changes is 250 ksps, the highest speed of No. 4 is 76 ksps.

1.4 Control flow chart of the software

Should analyze subdividing the error while working of Coder, the speed of data acquisition should be in above 100 ksps. This system adopts 2 slices of A/ D to change the chip, gathers signals of No. 4 of the Coder in timesharing, when reading the data of a slice of chip among them, start another chip to begin to be changed. Adopt 4 passways of highest sampling rates reachable above 120 ksps of this method, 2 passways of highest sampling rates reachable above 240 ksps.

Data got to gather need, set up the leaf Fu and vary, calculate subdividing the error of the Coder. When the data bulk is too large, will bring much trouble to subsequent data processing; When there is too little data bulk, can't reflect the truth of the signal. The experiment proves: It should be among 60-100 o'clock that each P-code of Coder is periodic to sample the point size best. The sampling point sets as 85 points in this system, according to the change of the Coder acceleration, the sampling point will fluctuate around. DSP is being received after sampling the order, samples the Coder signal, at first, judge the rotational speed of the Coder, presume sampling frequency, guarantee the point size of a periodic sample is at about 85 o'clock, the intact cycle of more than 3 of continuous sampling, sample and finish, pass the data to computer processing together.

2 data processing

Encoder output is the angle information, its outcoming signal is a space function taking angle as independent variable. If the sampling frequency of the data collecting card is stationary, and, the Coder rotates at the uniform velocity, can get the angle information when the corner in theory. When or work in the Coder experiment, the sampling frequency of the data collecting card can be controlled into fixed sampling frequency, but the Coder can not be a uniform motion rigorously enforced in the meaning, can be regarded as the uniform accelerated motion of very little acceleration, so, the P-code electro-optical signal gathered does not wait for the corner. When the Coder works, it is π / sine P-code electro-optical signal usin and ucos of 2 to gather the phase difference of No. 2, first of all, calculate the acceleration of the Coder, reuse the linear interpolation method in the corner of going on etc. to deal with, get a series of new data u ' sin and u ' cos, the P-code electro-optical signal composed of u ' sin and u ' coses is the electro-optical signals of a series of,etc. corners.

The phase difference of No. 2 of the Coder is π / sine P-code electro-optical signal ua and ub of 2, the electro-optical signal of P-code usually contains direct-current level, group and involves the higher order wave. The higher order wave regards third harmonic of the quadratic sum as the main weight, so its undulate equation is