Design and research of servo drive link of high-precision workbench

　　With the development of society, the amount of information stored is increasing, and optical disc information storage technology is also developing rapidly, which puts forward higher requirements for the precision of master disc manufacturing. 　　

　　Currently, the CD National Engineering Research Center uses a high-speed digital signal processor (DSP), digital closed-loop control principles and traditional servo motor drive methods to achieve continuous large-stroke motion of a high-precision worktable.

　　Its micro-displacement positioning accuracy is 50nm, and its macro-displacement positioning accuracy is better than 150nm, which can meet the requirements of continuous speed change and 50nm control accuracy of the master recording linear feed table.

　　On this basis, this paper studies and designs the simulation drive link of the workbench to improve the stability and rapidity of the low-speed response of the control system.

　　1 Overall structure of the system

　　The base of the linear feed table of the master recording system is fixed on the vibration-isolated marble table, and a V-shaped groove with high-precision balls is installed on the base as a motion guide rail. The worktable is decelerated by a worm gear and a small pitch precision screw in two stages, and is driven by a DC servo motor.

　　The master disc recording system adopts a constant linear speed recording method. The ideal movement of the focused light spot relative to the master disc is to move from the inside to the outside at a constant linear speed along the equally spaced Archimedean spiral centered on the center of the master disc. The motion is synthesized by the high-speed rotation of the master disc and the radial linear feed of the recording optical head.

　　The normal working speed of this precision workbench for master disc burning is about 30μm/s. The mechanical transmission system using the above-mentioned large reduction ratio inevitably has transmission errors. Therefore, to achieve precise positioning, a fully closed-loop control system must be used to directly detect the position of the workbench and perform servo control based on position errors. The overall structure of the workbench control system is shown in Figure 1.

　　2 Modeling of simulation driving links

　　2.1 DC motor model

　　The workbench drive motor adopts the DC torque speed measuring motor unit 45L-CZ001 produced by Shanghai Electric Machinery Factory. If the armature inductance and viscous damping coefficient are ignored, the transfer function of a DC motor with armature voltage Ua(s) as input and speed Ω(s) as output is: F(s)=Ω(s)/Ua(s) )＝(1/Ke)/[(Tms+1)(Tes+1)]≈(1/Ke)/Tms+1 Among them, Ke is the back electromotive force coefficient of the motor, and its unit is V%26;#183;ｓs ; Tm is the mechanical time constant of the motor; Te is the electrical time constant of the motor, its value is very small and can be ignored, so the DC motor can be simplified as a first-order system. Figure 5 The measurement of the mechanical time constant of the motor in the actual PI correction step can be approximated by adding a 7V step voltage to the motor, and then using an oscilloscope to measure the time it takes for the response to reach a stable value of 0.632, as shown in Figure 2. The mechanical time constant Tm=0.06s is obtained. In the open-loop situation, the input voltage directly drives the motor after passing through the linear power amplifier. Use the tachometer HT-331 to measure the corresponding rotation speed to obtain the amplification factor. The measured data are listed in Table 1. Table 1 Measured data table Voltage/V 0.5 1.0 1.5 2.0 2.5 Speed/rpm 0 70 302 520 750 Voltage/V 3.0 3.5 4.0 4.5 5.0 Speed/rpm 993 1195 1448 1686 1930 After straight-line fitting of the data, the amplification factor is obtained: 463.25. The electrical time constant is very small, approximately taking Te=0.0012, the transfer function of the DC motor model after the power amplifier can be obtained: F(s)=Ω(s)/Ua(s)=463.25/[(0.06s+) (0.0012s+1)]

　　2.2 Drive circuit design

　　In order to improve the rapid response, stability and load capacity of the system at low speeds, the analog drive circuit must be designed. Speed ​​negative feedback is introduced from the tachometer. The voltage difference is amplified by the PI correction link and the linear power amplifier to drive the DC servo. Motor movement. The driving link scheme is shown in Figure 3. The design of the PI correction link has an important impact on the performance of the driving link. The schematic diagram is shown in Figure 4. The transfer function is: V0/Vin=Ki(1/T0is+1)(Tjs+1/Tis) Among them, Ki=Ri/R0 is the proportional amplification coefficient of the corrector, τi=RiCi is the corrector time constant, T0i= R0C0i/4 is the filtering time constant, which generally has a smaller value and is used to filter high-frequency noise interference. In order to design the speed loop as a typical second-order link, it is necessary to ensure that the selection of the zero point of the corrector can eliminate the time constant with a large adjustment time, that is, τi = Tm. If the filter time constant T0i=0.25ms, R0=100kΩ, then the filter capacitor C0i=0.01μF. Taking the proportional magnification as Ki=3, we get Ri=KiR0=300kΩ, so we get Ci=0.2μF. 　　

　　In order to ensure that the amplifier will not be saturated due to open loop when the PI correction link reaches a steady state, a large feedback resistor R1 = 1MΩ is connected in parallel to the PI feedback line. In addition, in order to facilitate adjustment, the proportional coefficient function is added to the PI corrector, but in order to prevent the adjustment from having a great impact on the time constant, it is necessary to ensure that Ri>>R1, and take R1=10kΩ and R2=1kΩ. The actual circuit diagram used is shown in Figure 5. Next, the speed measurement feedback coefficient is determined, and the data are listed in Table 2. Table 2 Table speed feedback coefficient Table speed/RPM 0 70 302 520 750 voltage/V 0.96 5.75 11.0 16.0 speed/RPM 993 1195 1486 1930 voltage/V 21.1 26.5 36.8 41.8 After the data is directly matched, the feedback coefficient is: H(s)=0.022

　　Ignoring the filter time constant T0i of the PI correction link, the speed can be finally obtained. The ring-opening transfer function is: G(s)H(s)＝3（0.06s+/0.06s）(463.25%26;#215;0.022)/（0.06s+1）(0.0012s+1) ＝509.6/ [(s0.0012s+1)]

　　3. Driving circuit simulation

　　The selected simulation environment is Matlab6.1 and the Simulink toolbox under it.

　　3.1 Velocity loop open-loop Bode diagram

　　The open-loop transfer function of the speed loop is: G(s)H(s)=509.6/[(s0.0012s+1)] Use Matlab 6.1 to draw the Bode diagram and obtain Figure 6. Shear frequency: 416Hz Phase angle margin: 65 degrees The system has sufficient phase angle margin, which shows that the system is stable.

　　3.2 Speed ​​loop closed-loop step response simulation

　　Use the Simulink toolbox under Matlab 6.1 to build the speed loop closed-loop system structure diagram, as shown in Figure 7. Adding a 0.2V step signal and taking the feedback coefficient as 0.022, the simulation results are shown in Figure 8. It can be seen from the response curve that the rise time of the system step response is 5ms, the overshoot is 6%, the rotation speed stability value is 10rpm/s, and the system performance is good.

　　4 Experimental data processing and analysis

　　After theoretical modeling and program simulation, the simulation drive link for the precision servo workbench is designed and debugged, and time domain analysis is performed to compare the experimental results.

　　4.1 No analog driver link

　　First, there is no analog driving link, and the output signal of the DSP digital controller (through the linear power amplifier) ​​is used to directly drive the DC torque motor movement.

　　4.1.1 DSP open loop experiment

　　Add an input voltage to the DSP digital controller in an open-loop condition, and test the relationship between the applied voltage and the speed of the workbench. The speed of the workbench is obtained by processing the collected linear position grating signal through the VC++ program. The obtained data are listed in Table 3. Table 3 Relationship between input voltage and workbench speed Voltage/V 1.0 1.1 1.2 1.3 1.4 Speed/μms-1 0 0～5 5～10 10～15 15～20 It can be seen from the data in the table that the speed stability of DSP open loop is poor. The dead zone voltage is 1.1V, and the system sensitivity needs to be improved.

　　4.1.2 DSP closed-loop experiment

　　When the DSP digital controller is in closed loop, the specified workbench moves at a low speed of 20 μm/s. In Figure 9, (a) is the velocity response curve, (b) is the displacement response curve, and (c) is a partial enlarged view of the displacement response curve.

　　It can be seen from Figure 9(a) and Figure 9(c) that the system has a delay time of nearly 40ms, of which 20ms is the dead time (the system does not respond). There are two main reasons for the delay in the system: there are backlash, return stroke and other errors in the mechanical transmission system; there is a delay in the mechanical response of the motor. From Figure 9, we can get the time domain response index of the system step input without adding analog driving link as follows: Delay time: 40ms Rise time: 60ms Peak time: 100ms Overshoot: 25% Steady-state error: 15% It can be seen that without When adding analog drive links and directly using DSP closed-loop control, the low-speed response of the precision workbench has reached a certain level of rapidity and stability. However, when used for master disc burning, the stability of the workbench needs to be further improved.

　　4.2 Adding analog driver link

　　In the workbench control system, an analog drive link with speed loop, PI correction and linear power amplifier is used to drive the motor movement, and DSP digital controller open-loop and closed-loop experiments are conducted.

　　4.2.1 DSP open-loop experiment

　　After using the analog driving link, the experiment measured that when the DSP digital controller is open-loop, the system can already produce a more continuous response at a voltage of 0.2V, as shown in Figure 10. It can be seen that the system sensitivity has been improved.

　　4.2.2 DSP closed-loop experiment

　　After adding the simulation drive link, the system was subjected to a closed-loop experiment with a DSP digital controller, and the workbench was still specified to move at a low speed of 20 μm/s. In Figure 11, (a) is the velocity response curve, (b) is the displacement response curve, and (c) is a partial enlargement of the displacement response curve. It can be seen from Figure 11(a) and Figure 11(c) that the delay time of the system is 20ms, of which 10ms is the dead time. It can be seen that the delay of the system is reduced after adding the analog driving link.

　　From Figure 11, the time domain response index of the system step input after adding the analog driving link can be obtained as follows: Delay time: 20ms Rise time: 30ms Peak time: 60ms Overshoot: 7.5% Steady-state error: 7.5% Figure 11 Comparative experimental results of the closed-loop step response of the workbench with a drive ring. It can be seen that after adding a simulated drive link, the sensitivity of the precision workbench system is greatly improved, and the low-speed stability of the system is doubled. However, the speed curve still fluctuates, which is mainly due to two reasons: First, the accuracy of the mechanical transmission system affects the steady speed accuracy of the workbench; secondly, the resolution of the workbench displacement detection grating is limited, which directly affects the increase in displacement between sampling points. Quantity measurement accuracy.

　　In addition to high positioning accuracy, the feed table of the master recorder also requires continuous movement, good speed stability at low speed (about 30 μm/s), and a steady-state error of less than 10%. The lower the speed, the worse the conditions for servo table movement, the higher the requirements on the servo control system, and the greater the difficulty in ensuring speed stability. It can be seen from the experimental results and analysis that after using the analog drive link in the DSP digital control system, the low-speed response performance of the workbench has been significantly improved. Not only has the response speed been improved, but the steady-state error has also been reduced, and the speed changes The range is within 7.5%. The improvement of servo drive performance is conducive to ensuring the accuracy of master recording.