Abstract: The calibration system composed of intelligent instruments and GPIB usually consists of three parts: computer, instrument module and application software. The software controls the operator to perform the calibration sequence of the instrument under test, avoiding human error. It is suitable for measurement occasions with high accuracy requirements. 1 Introduction Pointer-type electrical measuring instruments are widely used in industrial testing processes due to their low price, high reliability and easy installation and replacement. Periodic calibration of electrical measuring instruments is an important means to ensure their normal operation. However, the traditional manual calibration method is cumbersome and labor-intensive, and has the disadvantage of reducing the calibration accuracy due to manual reading. With the continuous progress of science and technology, the automatic calibration system based on the General Purpose Interface Bus (GPIB) represents the future development direction of measuring instruments [1] [2]. The GPIB interface realizes information sharing through communication between instruments, thereby completing the automatic comprehensive analysis and evaluation of the instrument under test. A verification system utilizing intelligent instruments and GPIB typically consists of three parts: a computer, instrument modules, and application software. The software controls the operator's verification sequence of the instruments under test, avoiding human error and making it suitable for measurement applications requiring high precision. 2. System Hardware Design The automated verification system for pointer-type electrical measuring instruments, in accordance with national verification regulations, performs verification functions for direct-acting analog indicating DC and AC ammeters, voltmeters, wattmeters, and ohmmeters, as well as multimeters measuring current, voltage, and resistance. The system is divided into seven functional modules: visual inspection, basic error verification, rise and fall variation verification, position influence, power factor influence, voltage test, insulation resistance, and damping. The hardware uses a Fluke 5520A benchtop multimeter as the signal generator for verifying the submitted instruments. The communication section uses GPIB technology, enabling simultaneous connection of up to 15 instruments, which can communicate directly with each other. The system hardware block diagram is shown in Figure 1. To ensure a user-friendly and operable system interface, the controller uses a PC and a Windows 98 or later operating system; the GPIB board used is the AX4810 and equipped with an IEEE-488.2 standard cable. 3. System Software Design The software was developed using VB6.0 and is divided into three layers from top to bottom: the upper-layer metrology process control layer, the middle-layer instrument control layer, and the lower-layer driver program. The lower-layer driver program interface is provided by the hardware manufacturer and can directly call GPIB interface functions. The middle-layer instrument control layer controls the 5520A multi-functional multi-product calibrator, including instrument initialization, instrument self-test, instrument connection, and voltage generation. The upper-layer metrology process control layer is an automated implementation strictly following the verification methods stipulated in the national metrology regulations. The software flowchart for the verification part is shown in Figure 2. 3.1 Control Process In this verification system, the PC with a GPIB interface card acts as the controller, and the Fluke5520A multifunction calibrator acts as the speaker and listener. The process is as follows: The computer sends the corresponding function programmable code to the Fluke5520A multifunction calibrator, which is then transmitted to the controlled instrument via the GPIB interface card. At this time, the controlled instrument is in a command-receiving state, and its state should be the same as the state required by the sent function code. Subsequently, the computer sends a programmable code to set the standard source to generate a signal, which, via the GPIB interface card, puts the controlled instrument into a "speaking" state. The computer then receives the measurement data from the controlled instrument, processes the data according to the user's settings, and simultaneously displays the data, thus completing the verification of one function. 3.2 GPIB Programming Technology Data communication is conducted through the AX4810P GPIB interface card. During communication, this requires calling the function functions declared in the AX4810P.DLL dynamic link library. Before communication, the communication interface must be initialized to open the instrument, set parameters, and call the ibdev function. (1) The command sending subroutine sends the command to Fluke5520A through the GPIB function ibwrt. The parameter dmm is the return value during communication initialization, and MeasureCmd is the signal value that Fluke5520A needs to generate as set by the software. Public Sub SendToDevice (MeasureCmd As String) Dim cmdstr As String 'Send the signal value to be output to Fluke5520A cmdstr = "OUT " & MeasureCmd + "; ＊WAI" ret = ibwrt(dmm, cmdstr, Len(cmdstr)) 'Make Fluke5520A execute the command cmdstr = "OPER; ＊WAI" ret = ibwrt(dmm, cmdstr, Len(cmdstr)) End Sub (2) The measurement value returned by the Fluke5520A to the PC is stored in the reading variable. The parameter count that needs to be passed when calling this procedure is the number of bytes of value that the PC needs to read from GPIB. Public Sub ReadFromDevice(count As Integer) 'Sends the output data command to the instrument cmdstr = "OUT?" ret = ibwrt(dmm, cmdstr, Len(cmdstr)) 'Delay 10s delay 1000 'Reads the data ret = ibrd(dmm, reading, count) 'Delay 20s delay 2000 End Sub 3.3 Data Rounding Subroutine During metrological verification, there are strict requirements for data processing of verification results. The maximum error and actual or corrected values ​​of pointer instruments must be calculated and rounded beforehand. China formulated the "Numerical Rounding Rules" in the national standard GB8170-87 in 1987. The basic method of numerical rounding follows the even-number rounding rule. When the value of the number to be rounded is equal to the interval between the two allowable rounding values, it is rounded to an integer multiple of 1, 2, or 5. In routine rounding, it is convenient to handle it simply by following the even-number rounding rule. However, rounding to integer multiples of 1, 2, and 5 becomes more complicated. The flowchart of the data rounding subroutine written in VB in the system is shown in Figure 3. 4. Verification Results and Analysis During the software trial run, a DC ammeter with a nominal class of 0.2 submitted by the customer was specifically verified. The verification results for 15 points in its basic range are given in Table 1. Table 1: Verification Result Analysis. During the software trial run, a DC ammeter with a nominal class of 0.2 submitted by the customer was specifically verified. The verification results for 15 points in its basic range are given in Table 1. The test data shows that the accuracy of the system's readback values ​​strictly adheres to the requirements of a 0.2-class instrument, and the rounded values ​​are accurate. The system achieved good results in actual operation, with fast communication response speed, stable system operation, and good implementation of all functional modules. 5. Conclusion The verification system built using the GPIB universal interface bus changes the traditional manual operation and single-unit use approach. Through the GPIB interface, on the one hand, users can control and configure corresponding instruments using software on a PC; on the other hand, because GPIB integrates high-level language interface programs and a large number of practical library functions, users can use any high-level language to write automatic verification and control programs to perform real-time verification and monitoring of the measured objects. In this way, users can flexibly build automatic verification systems according to their own needs, meeting the different requirements of scientific research and teaching.