Abstract: This paper introduces the hardware and software design of an intelligent miniature DC electromagnetic relay testing system. The system can automatically construct test circuits or control suitable linear controlled power supplies according to the relay model, and can automatically test the parameters of multiple relay specifications using only a single socket. The testing of relay operating time performance parameters adopts a method primarily based on software timers and supplemented by hardware circuits. Keywords: DC electromagnetic relay; single socket; timer; automatic testing 0 Introduction Relays are high-precision electronic components widely used in industrial manufacturing and defense technology. However, due to manufacturing processes and materials, their quality is unstable. Therefore, it is necessary to test them to determine their quality. In terms of testing accuracy, the operating time needs to be accurate to ms or μs, and the contact resistance to mΩ. In addition, some relays consist of multiple coils or contact groups, such as the JHX-3F series relays. Currently, commonly used manual testing methods are inefficient, have large errors, and few test parameters. For the testing of high-precision relays, only microcomputer-based testing systems can be relied upon. This paper discusses the hardware and software design of a test system based on the JHX-3F series relays, focusing on the automatic testing of multi-specification relays using a single socket, μs-level action time testing, and relay testing with small voltage or high current coils. 1. Relay Test System Composition The relay test system is structured as shown in Figure 1. The data acquisition card is the AC6115 based on the PCI bus. The system consists of a PC and a tester, which communicate via the AC6115 to transmit A/D, D/A, and I/O signals. The system software platform is Windows XP/2000, the software development environment is Delphi 6.0, and the database is SQL Server 2000. Main test indicators and accuracy requirements: 1) Voltage accuracy <= 0.01V; 2) Current accuracy <= 0.01A; 3) Action time accuracy <= 1ms; 4) Contact resistance accuracy <= 1mΩ. 2. Hardware Design of the Relay Test System 2.1 Types of JHX-3F Miniature DC Electromagnetic Relays The JHX-3F series includes four main categories: A, B, C, and D. The coil type can be voltage-type, current-type, or a hybrid type. Category A includes 16 single-coil varieties; Category B includes 240 double-coil varieties; Category C includes 182 double-coil varieties; and Category D includes 16 single-coil varieties. There are a total of 454 varieties (further divided into 166 voltage-type, 72 current-type, and 216 hybrid types). The initial contact state of the relay is divided into three states: fully closed (2D), fully open (2H), and partially open/partially closed (1H1D). Therefore, the JHX-3F series relays have a total of 1362 varieties. This article explains the working principle of two types of relays. Type B Double-Coil Single-Position: This type of relay has one starting coil and one holding coil. When the starting coil is energized, the relay state changes. Then, energizing the holding coil and removing the starting coil energizer maintains the state. Removing the holding coil energizer releases the state. Type C Double-Coil Double-Position: This type of relay has one starting coil and one reset coil. When the start coil is energized, the relay state changes. Removing the energizer maintains this state. Applying energizer to the reset coil returns the state to its initial state. 2.2 Hardware Implementation for Automatic Testing of 1362 Types of Relays with Multiple Specifications Using a Single Socket The characteristics of the relay under test are: pins 1-4 are the relay coil, and pins 5-8 are the relay contacts. The product package conforms to the characteristics of a dual in-line package (DIP). These characteristics of the relay under test make it possible to automatically test 1362 types of relays with multiple specifications using only a single socket. The circuit working principle of the automatic test is as follows: If the relay under test is a single coil, no energizer is applied to the coil of the auxiliary relay FZJD1, and the controlled source is applied to both ends of sockets 1 and 4. If the relay under test is a dual coil, it operates in two stages. During the system test start coil stage, no energizer is applied to the coil of FZJD1, and the controlled source is applied to both ends of sockets 1 and 2; during the test hold or reset coil stage, energizer is applied to the coil of FZJD1, and the controlled source is applied to both ends of sockets 3 and 4. When the relay coil under test is voltage-type, no excitation is applied to the coil of auxiliary relay FZJD2; the controlled voltage source operates and is applied to the circuit. When it is current-type, excitation is applied to the coil of FZJD2; the controlled current source operates and is applied to the circuit. For contact state testing, simply send the contact under test to both I/O ports. Contact resistance testing uses a four-terminal method to eliminate the influence of contact resistance, and tests are performed using three excitation sources. 2.3 Hardware Implementation for High-Current or Low-Voltage Relay Testing: The voltage and current sources used for testing are controlled by the DAC output on the data acquisition card. The output voltage of the voltage source and the output current of the current source are linearly related to the control voltage. The system automatically selects a controlled voltage source with different output ranges for testing based on the relay model under test: one from 0 to 24VDC, and the other from 0 to 110VDC. It can test relays with rated pull-in voltages from 1.5VDC to 100VDC. The output range of the controlled current source is from 0 to 2ADC, which can test the relays under test with rated pull-in current from 0.003ADC to 2.0ADC. The circuit uses auxiliary relays to achieve the purpose of automatic circuit construction. 3 Software design of relay test system 3.1 Application of various software timers The Timer control is based on Windows system timing. It can be used in places with low requirements. The timing resolution is 55ms[1], and the accuracy is only 5 to 8ms. It is used in the system to test coil parameters, pull-in value and release value. It cannot handle precise timing. The multimedia timer comes from Windows API. It can work in a high-priority thread. Its timing event is defined as a callback process. As long as message processing code is added in the process, the timing response can be completed[2]. Delphi encapsulates it in mmsystem.pas for software timing. This system uses it to detect faulty relays, such as contacts that never move. There are two methods for action time testing. 1) Use the hardware timer on AC6115. AC6115 provides a 16-bit microsecond pulse counter with a counting range of 1 to 216 microseconds [3]. Therefore, the action time test can be completed by pulse counting, but this method will increase the complexity of hardware circuit design; 2) Use a software timer. For example, Z_Timer. It can be loaded onto the Delphi control panel in the form of an ActiveX control. Its timing resolution is 0.1ms and its timing accuracy is 0.015ms. In the effective timing event, it can monopolize all system resources, such as message response and thread time slice. Therefore, using Z_Timer is a good method for high-precision software timing based on the Windows environment. This system uses Z_Timer to monopolize system resources for a short time and interrupt the operation of other threads and messages, so as to quickly and accurately test the action time. First, set its monopolization of system resources to the second level. Because the action time accuracy is <= 1ms, the relay must act within this period of time. During this monopolization period, Z_Timer collects the contact group voltage value before and after the action and loads it into the array. Finally, the action time can be calculated by analyzing the change of the value in the array. This method simplifies the design of the hardware circuit and yields good test results. 3.2 Software Implementation of Single-Item Relay Tests 3.2.1 Software Implementation of Basic Coil Parameter Tests After the Timer starts, the coils (including the starting coil, holding coil, and return coil) are tested. The AC_6115_DA function in the DLL (Dynamic Link Library) provided by AC6115 is called to linearly control the voltage output on the coils. The voltage rise (fall) gradient is 10V/5s. A short delay is first applied at the coil contact engagement point, the rated voltage point, and the contact release point using the Sleep function. After the relay state stabilizes, the AC_6115_AD function in the DLL is called to record the engagement value, rated value, or release value, and stores it in the common data module Datamodule. 3.2.2 Software Implementation of Operating Time Tests Operating time is a key indicator for evaluating relay quality, requiring high precision. Figure 3-1 is an "ideal diagram" for testing the operating time of a relay with an initial contact state of 2H. The actual test results, as shown in Figure 3-2, show a "bounce time," which must be considered during the test. This system uses Z_Timer, which monopolizes system resources for a short period, to complete the action time test. The timing response code for its pull-in time test is as follows: Procedure TMainForm.Draw_Time_Test(); // "Pull-in time" test Begin if (Call_DLL() < 31); // Call AC6115 dynamic link library Showmessage('DLL loading failed!'); Else Begin AC_6115_AD(Datamodule.Driverio, Datamodule.IDofcard[1], 80,3,4,0, 0, 0, 0, 0, 0, 0, 0, Datamodule.Draw_Time_Data,2048); // Collect all data before and after the contact action DA_To_Value(0); // Voltage returns to 0, preparing for release time test Sleep(100); // Smooth processing IO_Out(1,0,0); // Set I/O signal, preparing for "release time" test Sleep(2000); // Prepare time for releasing resources, after which the system will "buffer" to normal state End; End; [align=center] Figure 3-1 Ideal action time test[/align] [align=center] Figure 3-2 Actual action time test[/align] The data acquisition function in the code collects channels 3 and 4, collecting 2048/2=1024 points per channel. The channel switching time is 80 * 0.25 = 20 μs. Therefore, the time interval between two adjacent points on the same channel is 20 * 2 = 40 μs. Thus, the effective voltage monitoring time for each group of contacts is 1024 * 40 = 40960 μs, and the test accuracy for the action time is 40 μs. "Bounce time" calculation: 1024 points are sampled for each group of contacts. Starting from the 1024th point, the sampled voltage values ​​of the current point and the point before it are checked sequentially. If the absolute difference between the two is found to be 5, and this occurs for the first time, the current point position T1 is immediately recorded. Simultaneously, the "bounce time" counter is incremented. This process is repeated until the last occurrence of this condition, at which point the position T2 is recorded. Therefore, T1 is the action time point, and T1-T2 is the "bounce time". 3.2.3 Software Implementation of Contact Resistance Testing: For relays with an initial contact state of 2D, the relevant channel voltage can be directly sampled and calculated. For relays with an initial contact state of 2H, first, the applied voltage to the coil jumps to the rated value. After the contacts stabilize, a similar 2D testing method is used. For 1H1D relays, the 1D contact group is first tested using a method similar to the 2D test, and then the 1H contact group is tested using a method similar to the 2H test. 3.3 Software Implementation of Automatic Testing Combining Multiple Test Items: The main software flow for automatic testing combining multiple test items is shown in Figure 4. [align=center] Figure 4 Flowchart of Automatic Testing Software Implementation[/align] Automatic testing will complete all tests of coil parameters, operating time, and contact resistance. "Basic Parameter Settings" includes setting the relay type, coil type, number of tests for each test item, and initial state of the contact group. To improve the accuracy of test data, the automatic test is designed to be repeated multiple times, and the average value of the multiple test data is finally calculated. For example, the coil parameters, action time and contact resistance are tested M, N and K times respectively (M, N and K are all greater than or equal to 0). The actual number of times is determined by the tester according to the on-site situation. The results of each test are temporarily stored in the Datamodule. Three linked lists are established in the Datamodule to store the coil parameters, action time parameters and contact resistance parameters of each test. Each node of the linked list represents one test data of a certain test item. 4 Conclusion In the on-site test, the system runs well, the human-machine interface is user-friendly, and it can complete the test of all 1362 types of relays and provide a complete report of test data. The test accuracy is satisfactory. The voltage test accuracy has been improved from <=0.01V to <=0.001V, and the pull-in time test accuracy has been improved from <=1ms to 0.04ms. The innovation of the author of this paper is: (1) Hardware design. The single socket was used to test 1362 types of relays in 4 categories; the parameters of current, voltage or hybrid relays were tested by automatically controlling the controlled power supply. (2) Software design. A method with Z_Timer as the main component and hardware circuit as the auxiliary component was proposed to test the action time, which greatly simplified the design of the hardware circuit; in the automatic test, test items can be combined to perform multiple repeated tests to obtain the average value, thereby improving the accuracy of the test data. References [1]. Li Sibao, Yao Xiaoxian. Use of kernel timer under Win32s [J]. Microcomputer Information, 2003, Vol.19 (Issue 2): 71-72 [2]. He Weili, Kang Wei. Implementing multimedia timer in Delphi [J]. Computer Learning, 2000, (2): 26-27. [3]. AC6115 User Manual [M]. Beijing Shuangnuo Company, 2004, 58-59