Application of AGILENT 34970A in product parameter verification

For high-precision products such as standard resistors , they need to be calibrated one by one after production. The manual calibration method is cumbersome and easy to introduce errors. AGILENT 34970A (data acquisition/ switch unit) is used to connect to the PC host through the HPIB bus to form an automatic test virtual instrument to test it. After the test is completed, the calibration report is automatically generated through DDE (Dynamic Data Exchange). In this way, not only the measurement process is simplified, but also the measurement accuracy is improved.

introduction

The verification of the parameters of standard resistors after production is currently generally carried out by direct measurement or substitution with the same nominal value, and the measuring instrument is usually a bridge. Although this method is very mature, it uses a lot of supporting equipment and is cumbersome to operate. In addition, the entire verification process is carried out under human intervention, which inevitably introduces deviations (such as: caused by factors such as manual readings), and some non-technical factors (such as human emotions, etc.), making it difficult to ensure the quality of verification. With the rapid development of computer technology and measurement technology today, test automation has become an inevitable trend.

Based on the above reasons, we used Agilent's AGILENT 34970A and connected it to the computer host through the GPIB bus. Under the AGILENT VEE software platform, we developed an automatic calibration system for standard resistors. It enables the calibration to be carried out in an automatic state, including the calibration report after the calibration is completed, which is automatically generated in EXCEL using DDE (Dynamic Data Exchange). Therefore, the system combines reliability and efficiency.

Test system structure

1. Discussion on the verification method

There are many methods for calibrating standard resistors, including direct measurement method, substitution method with the same nominal value, transition transfer method, etc. The advantage of the direct measurement method is simplicity, but it requires the accuracy of the resistance measuring instrument to be two levels higher than the resistance being measured. For measuring more levels of standard resistors, the price paid is to use more accurate and expensive test instruments; the transition transfer method is to transfer the resistance value by measuring two different resistance values ​​to obtain the ratio between the two. This method is often used for measuring high-precision resistors, and has high requirements for test instruments and a relatively complicated test process; the substitution method with the same nominal value is to use a resistance measuring instrument to measure the resistance values ​​of the standard resistor (standard resistor that has been measured and meets the requirements) RS and the resistor being tested RX in turn, and obtain the calibration result after the following calculation.

RX = RS + (AX - AS)

Where AS is the indication of the measuring instrument when measuring RS.

AX——the indication of the measuring instrument when measuring RX

From the above, we can see that the same nominal value substitution method has lower accuracy requirements for resistance measuring instruments than the direct measurement method, and the measurement method is also simpler than the previous two. According to the regulations JJG166-93, when the resistance measuring instrument cannot reach two levels higher than the accuracy level of the tested standard resistor RX, and there is a resistance standard RS with the same nominal value as the tested resistor RX, the resistance value of the tested resistor can be calibrated by the same nominal value substitution method. What we are now implementing is the calibration of the nominal resistance of 0.01 level, and the existing measuring instrument is AGILENT 34970A. From the accuracy of AGILENT 34970A, it can be seen that although it cannot meet the two-level higher requirements required for direct measurement of 0.01 level standard resistance, it can be fully used as a measuring instrument for the same nominal value substitution method. Based on the above analysis, this system adopts the same nominal value substitution method and uses AGILENT 34970A as the core test instrument.

2. AGILENT 34970A and its applications

AGILENT 34970A is a multi-functional measuring instrument with a 6 1/2 A/D converter and three module slots at the rear. Agilent provides eight module options, which can form different measurement modes and complete various test functions according to needs. This system is mainly used for standard resistance measurement, so AGILENT 34901 is selected. AGILENT 34901 provides 20 analog channels (40 channels for single-ended input) . We choose two of them to connect the standard resistor RS and the circuit RX under test. Since the test adopts the same nominal value substitution method, RS and RX will be measured in turn during the test, and a switch is required for switching. Therefore, AGILENT 34904 module is selected. This is a 4×8 two-wire matrix conversion switch, which can be programmed at will to switch the object under test. Figure 1 is a schematic diagram of the standard resistance test wiring (four-wire wiring).

Figure: 1

According to the requirements of JJG166-93 regulations, the error introduced by the resistance measuring instrument used for resistance substitution comparison should not be greater than 10×10-6 (1/10 of the grade index). In order to achieve this goal, a four-wire measurement method is used when measuring low-value resistance values ​​(AGILENT 34970A provides this function); in order to eliminate the error caused by zero potential such as contact potential on resistance measurement, the "offset compensation" (OFFSET COMPENSATION) technology provided by AGILENT 34970A is applied: the first time is to measure the voltage value of the resistor under the action of a constant current source, and the second time is to measure the zero voltage value of the resistor when the constant current source is turned off. The voltage value of the resistor is measured twice to eliminate the zero potential effect on the measured resistor in the substitution measurement, so that it meets the requirements of the regulations.

3. System structure

This system uses AGILENT 34970A as the measuring instrument, connects the measured resistance through AGILENT 34901 and AGILENT 34904 modules, uses AGILENT VEE as the test system software operation platform, and connects to AGILENT 34970A through the HPIB bus interface . AGILENT 34970A supports SCPI commands, and all actions of AGILENT 34970A are implemented by sending SCPI commands to it through the AGILENT VEE platform.

Software Implementation

AGILENT VEE (AGILENT Visual Engineering Environment) is a graphical programming language developed by Agilent. It has the following features:

Figure: 2

a. Easy to program;

b. Powerful GUI functions: You can create a powerful and user-friendly user interface without much effort;

c. Powerful instrument control functions: A large number of instrument driver functions enable users to freely control instruments.

Due to the above characteristics, this system uses AGILENT VEE as the test software. Figure 2 is the test program flow chart.

When the software is running, the operator will be prompted to enter some necessary data, such as the number of resistors to be tested, the desired average number of measurements, the nominal value of the resistor to be tested, etc. If the type of object to be tested is single, these interactions can be avoided. Figure 3 is the operating interface of this test program.

Figure: 3

Figure 3 shows an overview of the human-computer interaction interface. The curve dynamically reflects the real-time status of each test, and the average digital value of each test is reflected in real time above the curve; the color of these numbers corresponds to the color of the curve. Moving the small triangle on the curve can reflect the numerical difference between any two points below the curve. It provides the operator with a simple but necessary means of data analysis. After the test, the data is directly transferred to Excel ( electronic spreadsheet software) by the DDE function provided by AGILENT VEE, and EXCEL performs data analysis and synthesis and generates a test report. This process is completed in one go by the program.

Error synthesis

When the standard resistor is tested by substitution method, the main sources of error of the tested standard resistor are the standard instrument RS and the test device. The main test instrument in this system is AGILENT 34970A. According to the provisions of the regulation JJG166-93, the standard instrument RS we selected is a second-class standard resistor with a grade index of 0.001%, which fully meets the requirement of the regulation that it should not be greater than 1/4 to 1/5 of the tested resistor. A comprehensive analysis of this test device reveals its possible error sources as follows:

a. Error ΔU1 caused by the annual change of the second-class standard resistor RS, i.e. the grade index CS;

b. The transfer error ΔU2 caused when the second-class standard resistor RS is tested;

c. Error ΔU3 introduced by the repeatability of the measuring instrument;

d. Error ΔU4 caused by insufficient resolution of the measuring instrument;

e. Error ΔU5 caused by ambient temperature change;

f. Error ΔU6 caused by lead resistance, parasitic potential, zero current, etc.

Among the above six errors, the error caused by lead resistance is mainly for small resistors. The objects affected by it are mainly 100Ω and 1000Ω resistors. According to the characteristics of AGILENT 34970A, a four-wire measurement method is used to eliminate the influence of lead resistance. Due to the influence of parasitic potential and zero current, another characteristic of 34970A, offset compensation method, is used to reduce its influence on measurement, so that the error ΔU6 caused by lead resistance, parasitic potential, zero current, etc. is not greater than 0.05CX, that is, ΔU6=5×10-6. Using the 6.5-digit display capability of AGILENT 34970A, the integration time is increased to 200PLC, and the resolution reaches 0.00000022 X range, then the resolution error ΔU4=1.1×10-7.

The repeatability of the measuring instrument AGILENT 34970A is based on the 24-hour indicator of its resistance unit accuracy. Among the three ranges of 100Ω to 10kΩ used, the resistor with the lowest accuracy is the 100Ω range, and its 24-hour indicator of accuracy is 0.0065%. According to the regulations, the deviation limit of the 0.01-level standard resistor is 0.01%. According to our actual test, the repeatability error of AGILENT 34970A meets the regulations, that is, ΔU3≤10×10-6.

The error ΔU5 caused by the change in ambient temperature is considered to be in compliance with the regulations due to the limitations of experimental conditions, that is, ΔU5=8×10-6.

The comprehensive error ΔU of the above six errors is:

ΔU=√ΔU12+ΔU22+ΔU32+ΔU42+ΔU52+ΔU62 = 17.7×10-6 < 33×10-6

This value is less than the requirement of JJG166-93 that the total uncertainty caused by the calibration device and environmental conditions should not exceed 33×10-6 (1/3 of the grade index) of the resistor being measured.

The specific error distribution and comprehensive error are shown in the following table (the values ​​are expressed as relative errors):

Conclusion

The development of this system shows that it is completely feasible to calibrate standard resistors of 0.01 level and below in the range of 102~104Ω based on the same nominal value substitution method. If a measuring instrument with higher resolution is used, such as: 7 1/2 resolution, the range of resistance values ​​to be tested can be wider. The automation of the test not only improves the measurement accuracy, but also greatly improves the efficiency. From this we can also see that virtual instruments, as a brand-new concept, have very attractive application prospects.