Zhao Yulong, School of Automation, Beijing University of Aeronautics and Astronautics; Fang Liqing, Xue Deqing, Department of Artillery Engineering, School of Ordnance Engineering; and Chen Donggen, through analyzing the basic principles and electrical performance characteristics of a large-scale equipment servo system, and based on the requirements of dynamic signal detection, adopted a modular design scheme and a method based on synchronous sampling and real-time data processing to realize online data acquisition, signal conditioning, and filtering analysis of the dynamic precision signal, stable aiming speed, maximum tracking speed, and error signal of the equipment servo system. 1 Introduction The servo system of a large-scale equipment is a typical electromechanical servo system. Its performance, especially its dynamic performance indicators, directly affects the efficiency of the equipment. In general testing and maintenance, this device can only measure the static parameters of the system and diagnose some typical faults. For the dynamic performance of the system, there is a lack of means and methods for dynamic detection; only qualitative visual observation is possible. This paper, based on a modular design scheme and a method based on synchronous sampling and real-time data processing, realizes online detection of the dynamic parameters of the equipment servo system by acquiring, adjusting, amplifying, filtering, and analyzing dynamic signals online. The system can detect the accuracy of dynamic parameters such as stable aiming speed, maximum turning speed, and tracking error of the equipment, diagnose system faults, and facilitate rapid equipment maintenance. 2. Dynamic Performance Analysis of the Servo System The equipment servo system includes an azimuth servo system and an elevation servo system. The azimuth system controls the rotation mechanism to rotate horizontally around the vertical ground axis, while the elevation system controls the rotation mechanism to perform pitch motion. Both are closed-loop control systems with synchronous measuring elements, motor amplifiers, and DC servo motors. To improve coordination accuracy and increase turning speed and power, the design employs a dual-reading high-precision synchro, a two-stage amplification system (comprehensive amplification and motor amplification), and a high-power actuator motor. The position follow-up system has high power, complex circuitry, high dynamic performance requirements, and is difficult to test. The circuits of the two servo systems are basically the same; the working principle is explained using the equipment azimuth servo system as an example. 2.1 Servo System Composition A typical equipment servo system consists of a synchro, signal selection circuit, AC amplification stage, phase-sensitive rectifier stage, DC amplification stage, amplification motor, actuator motor, and reducer. Its system block diagram is shown in Figure 1. [img=Figure 1 Servo System Block Diagram]/uploadpic/THESIS/2007/11/2007111613134582219R.jpg[/img] Figure 1 Servo System Block Diagram 2.2 Main Dynamic Technical Indicators The main dynamic technical indicators of the servo system include the system's maximum aiming speed and acceleration, maximum turning speed, and the quality indicators of the system's transient process. The quality indicators of the system's transient process mainly include: overshoot, adjustment time, oscillation count, etc., which are difficult to measure in practice. Usually, for this type of system, the time required to overcome a certain misalignment angle is used to verify the system's transient quality. 3 Dynamic Test System Design The task of dynamic testing of the servo system is to detect the dynamic performance indicators of the servo system, diagnose the equipment's working condition, and perform timed and random tests as needed. The system's testing process is online, requiring simple installation, easy operation, and reliable operation. This system adopts a virtual instrument form based on a data acquisition and processing system to construct a dynamic test system. For different test items, appropriate sensors and signal conditioning units are selected and combined with DSP data acquisition modules to form corresponding test systems. Its main components include: a PC computer, various sensors, a DSP data acquisition and processing module, a programmable signal conditioning unit, a signal simulation module, and system testing software. The test system block diagram is shown in Figure 2. [IMG=Figure 2 Test System Structure]/uploadpic/THESIS/2007/11/20071116131455403392S.jpg[/IMG] Figure 2 Test System Structure 3.1 Dynamic Testing Methods There are two main types of commonly used dynamic testing methods: the transient process method and the frequency response method. Considering the needs of practical work, we adopt the transient process method. This testing method measures the output transient process of the system under a step input, and then determines the dynamic quality of the system. In equipment servo systems, since the transient process of the output angle is not easily measured directly, the system quality is determined by measuring the transient process of the error voltage (error angle). θsc = θsr - δ Where: δ is the error angle, θsr is the step input. The transient processes of θsc and δ differ only by a constant; measuring the transient process of the error angle can completely determine the system quality. Due to high precision requirements, equipment servo systems have large amplification coefficients, resulting in relatively small linear regions. System transient quality indicators such as overshoot, settling time, and oscillation frequency are difficult to measure in practice. Typically, the time required to overcome a certain offset angle is used to verify the transient quality of such systems. In actual measurements, the dynamic quality of the system is approximated by measuring the transient process of the demodulation filter output voltage UL. 3.2 Dynamic Test System Composition The test system mainly uses a PC computer as its core, completing functions such as acquisition and control of the measured signal, data processing, and display of measurement results. It features wide applicability, high reliability, strong anti-interference capability, and ease of expansion with card-type instruments. The signal under test of the device is sent to the signal processing unit of the instrument through the sensor for filtering, amplification, signal shaping and special signal optical and electrical isolation, etc., and then sent to the computer data acquisition module by A/D, I/O, etc. The servo system dynamic test system is composed of the device signal simulation unit, power sensor and DSP data acquisition and processing module. (1) Signal simulation module The dynamic test of the servo system requires the device main control signal. For this purpose, a dedicated device main control signal simulation module was designed. The device main control analog signal is mainly composed of 400Hz intermediate frequency reference signal voltage and input signal. The analog signal control device can control the phase, magnitude and output power of the two signal voltages. The controller of the intermediate frequency reference signal voltage adopts a multi-layer three-position switch to control the positive phase and negative phase signal voltages of 3V, 6V and 12V respectively. The digital simulation module generates multiple digital signals to supply the data transmission device, and then performs the corresponding tests. (2) Power Sensor The signals to be detected in the servo system of this device are mostly electrical signals. When selecting a power sensor, it is necessary to consider the type of signal to be isolated (current, voltage, DC, AC, pulsating DC, mixed AC/DC signals, etc.), the type of sensor output signal, the type and range of the isolated signal, and the various indicators to be met, such as accuracy, linearity, response time, and operating environment. Based on the input requirements of the designed acquisition module and the signal characteristics of the device's servo system, an ultra-small precision voltage sensor produced by Beijing Xingge Measurement and Control Technology Co., Ltd. was selected for signal isolation and conditioning. For AC signal isolation, the AC voltage sensor SPT304 was selected. This sensor is welded and uses a high-isolation, impact-resistant, all-resin outer seal. This type of sensor is small in size and light in weight. It can isolate and transform the input AC voltage signal into a signal suitable for acquisition by the acquisition card, and the signal after conditioning can be directly connected to the A/D sampling circuit. Figure 3 shows the circuit diagram of the SPT304 AC voltage sensor. This type of sensor requires an auxiliary power supply (±15V or ±12V) when working. After an AC signal is applied to the input terminal, no additional auxiliary circuitry is required. Its output voltage (absolute value less than 5V) and output current (less than 10mA) are both within the allowable range of the PCL-711B data acquisition card. Therefore, its output terminal can be directly connected to the input terminal of the PCL-711B data acquisition card. Furthermore, this sensor has good linearity (0.5%), which fully guarantees the needs of signal acquisition and processing. [IMG=Figure 3 SPT304 AC Voltage Sensor Circuit Schematic]/uploadpic/THESIS/2007/11/2007111613161758439W.jpg[/IMG] Figure 3 SPT304 AC Voltage Sensor Circuit Schematic The DC voltage signal of the equipment servo system is pulsating DC, and a KV50A/P type DC voltage sensor is selected. The KV50A/P DC voltage sensor is a magnetically balanced type, primarily utilizing the Hall effect and employing magnetic compensation principles. The primary and secondary circuits are insulated. When measuring voltage, the voltage being measured is connected to the sensor via a resistor, and the measuring current is proportional to the voltage being measured. It can also measure AC and pulsating voltage signals. The output voltage and current values ​​are adjusted by connecting different input and output resistors in series. Its wiring diagram is shown in Figure 4. [IMG=Figure 4 Wiring Diagram of DC Voltage Sensor]/uploadpic/THESIS/2007/11/2007111613164519827V.jpg[/IMG] Figure 4 Wiring Diagram of DC Voltage Sensor The input voltage range of this type of sensor is 0～2500V. The appropriate output voltage and output current can be obtained by adjusting resistors R1 and R2. The input and output currents are directly proportional (when the input current is 10mA, the output current is 50mA, that is, the input and output currents have a 1/5 linear relationship). The linearity coefficient of the voltage changes with the change of resistors R1 and R2. The linearity coefficient can be determined by the following formula: K=U1/U2=R1/R2 Where: U1 and R1 are the input voltage and input resistance, respectively; U2 and R2 are the output voltage and resistance, respectively. The voltage output U2=I×R2 across R2 is connected to the input terminal of the acquisition card as the sampled signal. (3) The core of the DSP-based high-speed data acquisition and processing module test system is data acquisition. Considering the large data volume and high sampling frequency in dynamic testing, we designed a high-speed A/D conversion data acquisition module using a DSP as the core device. This module serves as a slave device of the PC host, employing a modular plug-in structure and inserted into the host's ISA expansion slot. The module consists of a high-speed A/D converter, a DSP, a high-speed data buffer memory, and a communication interface. The DSP high-speed signal processing unit uses the AD-SP2181, a powerful chip with flexible interfaces, capable of integrating high-speed ADCs and DSPs into a unified system structure. In the dynamic testing system, the ADSP2181 utilizes its timing function and external interfaces to implement timing and control of the high-speed A/D converter, performing real-time signal processing to solve the level triggering problem. Simultaneously, high-speed SDRAM, along with an address generation circuit and a bidirectional access interface, forms a high-speed cache to improve the DSP's signal processing and caching capabilities. The application block diagram of the ADSP2181 in the dynamic testing system is shown in Figure 5. [IMG=Figure 5 ADSP2181 Application Block Diagram]/uploadpic/THESIS/2007/11/2007111613272498027Y.jpg[/IMG] Figure 5 ADSP2181 Application Block Diagram To further reduce the size of the test system, the ADSP2181's tri-state IDMA bus is directly connected to the industrial computer's bus without bus isolation. Its reset signal RESET is provided by an RS flip-flop implemented using GAL20V8 programming. The ADSP2181 outputs an analog level as the trigger level through DA2# conversion. When the host computer provides a signal, the comparator CMP04 outputs a high-level signal. The trigger requests of the three sensor channels are used as external interrupt request signals for the ADSP2181, connected to the IRQ0, IRQ1, and IRQ2 terminals respectively, and all three terminals are set to edge-sensitive interrupt mode. This achieves the goal of simultaneously managing the triggering of three acquisition channels with a single DSP. 4. System Testing Software The testing process demands high levels of skill in waveform transformation and display, feature information extraction and analysis, and human-machine interface control and adjustment. Traditional programming methods would be extremely labor-intensive, while virtual instrument technology can conveniently solve these problems. This system utilizes a combination of NI's LabVIEW and the VC++ visual language to implement the various functions required for testing. LabVIEW software offers a wealth of instrument control buttons and output dials, allowing for interface design, signal acquisition and control signal output, and signal processing. When using a non-NI proprietary data acquisition card on the LabVIEW platform, a custom instrument driver must be developed to handle data acquisition. We use VC++ to create a dynamic link library (DLL) and implement data acquisition from the data acquisition card through the call function (CLF) provided by LabVIEW. Signal processing functions are performed using various signal processing modules provided by LabVIEW's integrated software library, such as FFT transforms and matrix operations. All eigenvalue analyses during testing are programmed using LabVIEW; programming simply involves selecting and connecting the functional modules. 5. Conclusion This paper constructs an online testing system based on the DSP data acquisition and processing module, realizing performance testing of servo systems. It features accurate test results, high automation, stable and reliable operation, ease of operation, and convenient transfer and maintenance, thus improving the reliability of the system. (Proceedings of the 2nd Servo and Motion Control Forum; Proceedings of the 3rd Servo and Motion Control Forum)