Design and implementation of an automatic calibration system for microwave power meters

0 Introduction
Microwave power is an important parameter for measurement in industrial research and production. Microwave power meters have the characteristics of large dynamic range of power measurement and wide frequency range, and are widely used in radar systems, communication systems, and electronic countermeasure systems. For example, when using radar for tracking tests, the first thing to do is to determine the radar's range, which requires measuring the transmission power of the radar transmitter. Therefore, the accuracy of the microwave power meter is crucial. The traditional method of calibrating a microwave power meter is to use a standard microwave power meter to perform manual test calibration and comparison results. The steps are complicated, the accuracy is low, and it takes a long time. To address this problem, this paper designs and implements an automatic calibration system for microwave power meters from 0.05 to 26.5 GHz.

1 System construction, calibration principle and method
The main task of microwave power calibration is to establish power standards, transfer power values ​​and ensure the accuracy and consistency of power values. The most commonly used power value transfer methods are the alternating comparison method, the transfer standard method and the six-port method. Among them, the transfer standard method is a widely used microwave power value transfer method. The transfer standard method can ensure the accuracy of power value transfer. Therefore, this paper uses the transfer standard method to establish a measurement standard for the 0.05-26.5 GHz frequency band. The composition of the measurement standard and its test block diagram are shown in Figure 1.

The system uses a three-port device, a standard power socket and a standard power indicator to form a transfer standard. Since the operating frequency range of the power divider (0.05-26.5 GHz) is wider than that of the directional coupler and is easy to implement, and it is convenient to realize fully automated calibration tests, a power divider with good symmetry is used as a three-port device. There is no external PIN electric attenuator in the transfer standard. Instead, the voltage detected by the power meter probe is sent to the signal generator and compared with the reference voltage inside the signal generator, thereby establishing an amplitude stabilization loop and obtaining an equivalent signal source with a small reflection coefficient. The transfer standard is calibrated by a higher-level power standard to determine the calibration factor Kc, which is then used to calibrate the power sensor under test.
The calibration system mainly consists of a signal generator, a power divider, a standard power socket (power sensor), a standard power indicator, and a computer. When the test port of the power divider is connected to the power sensor to be calibrated, the microwave power incident on the calibrated power meter is:

Where: Pu is the replacement power of the calibrated power sensor, that is, the reading of the power indicator, in mW; Pe is the replacement power of the standard power socket, that is, the reading of the standard power indicator, in mW; Γge and Γu are the reflection coefficients of the equivalent signal source and the calibrated power sensor, respectively. [page]
During actual calibration, since the reflection coefficient phase in the reflection coefficient term |1-ΓgeΓu|2 cannot be determined, according to the worst-case analysis of the error limit phase combination introduced by this term, this term can be approximately analyzed as 1; this term will be considered in the uncertainty analysis. Therefore, it can be simplified to:

2 Software Design and Implementation
The microwave power automatic calibration system in this paper is designed and developed using the Visual Basic 6.0 software platform, and each device is controlled through the IEEE 488 interface. In the design of the software, a top-down tree structure is adopted, and a modular design concept is introduced.
According to the manual calibration steps and experience of the microwave power transfer system of the transfer standard method, the automatic calibration test program of the power transfer system is compiled according to the flowchart 2 on the development platform Visual Basic 6.0. By pre-setting the initial parameters (such as: starting test frequency, ending test frequency, model and serial number of the power socket to be tested, etc.), the system runs the power meter, calibration test, data acquisition, data processing, save and print subroutines under the control of the computer to realize automatic testing of the full frequency band. The program running process is shown in the program flowchart 2.

After debugging and running, the program has realized the automation of microwave power meter calibration. The structure of its main interface is shown in Figure 3. The various functional modules of the control program have strong cohesion, and the mutual communication between modules is uniformly scheduled by the main program. The human-machine interface is very friendly, and there are corresponding prompts during the test.
Compared with the original manual calibration, the automatic calibration system has three advantages: high degree of automation, high calibration efficiency, and easy operation. [page]

3 Systematic measurement uncertainty analysis
Using the transfer standard method, the main sources of measurement uncertainty for calculating the calibration factor of the calibrated power sensor are:
(1) The relative standard uncertainty component u1 introduced by the uncertainty of the standard power seat calibration factor Kc.
After the power transfer standard is calibrated with a higher-level power standard, the allowable error term error limit given is: ±3%~±4%. Assuming a uniform distribution , the corresponding relative standard uncertainty is:

(2) The standard uncertainty component u2 introduced by the inaccurate measurement of the substitute power of the standard power socket.
The substitute power is measured by the standard power indicator, and the allowable error limit of the power indicator is: ±0.1%. Assuming uniform distribution , the corresponding relative standard uncertainty is:

(3) The standard uncertainty component u3 introduced by the inaccurate measurement of the substitute power of the calibrated power meter.
The substitute power is measured by the power indicator of the calibrated power meter, and the allowable error limit of the power indicator is: ±0.5%. Assuming uniform distribution , the corresponding relative standard uncertainty is:

(4) The measurement standard uncertainty component u4 introduced
by the mismatch. The measurement relative expanded uncertainty U(M)=2|ΓgeΓu| introduced by the mismatch is evaluated according to the Class B uncertainty. The possible values ​​of M obey the inverse sine distribution, including the factor . The reflection coefficient of the equivalent signal source measured is 0.05, and the reflection coefficient of the calibrated power meter is 0.05~0.16. Its relative standard uncertainty u4 is:

(5) Standard uncertainty component u5 introduced by various random effects.
Due to various random factors such as undesirable environmental conditions and non-repeatability of joint connections, measurement standard uncertainty components are introduced, which are usually characterized by experimental standard deviation and evaluated using the Class A uncertainty assessment method. Therefore, this standard uncertainty component is: The

above is synthesized into:

Therefore, the expanded uncertainty of the calibration factor of the calibration system transfer standard is:
U = kuc = 2uc = 3.8% ~ 5.2%,
which meets the requirements of the measurement transfer standard.

4 Conclusion
The microwave power automatic calibration system has changed the traditional manual calibration method of power meters, significantly improved the efficiency of calibration work, and has stable system performance and convenient operation. It ensures measurement uncertainty, meets the requirements of measurement transfer, and realizes the automation and a certain degree of intelligence of the microwave power meter calibration process.