Introduction to the intelligent debugging principle and process of cavity filter and the intelligent debugging platform

Generally speaking, manual filter debugging is actually a real-time iterative optimization process. In order to facilitate debugging, there will be tuning screws for debugging on the filter structure, or other forms of tuning elements, so that the debugging technician can change the resonant frequency of the filter resonant unit and the coupling amount between the resonant units during debugging. When debugging, the debugging technician repeatedly turns the tuning screw according to the changes in the vector network analyzer graph until the performance of the filter meets the design requirements. For many debugging technicians, the manual debugging process is more like a craft than a science. Therefore, manual debugging of complex structure micro filters is generally completed by very experienced debugging technicians.

In the process of debugging and producing large quantities, power capacity, temperature effect, mechanical properties of materials, passive third-order intermodulation and size restrictions are all important factors to consider in the actual processing of filters. The debugging of microwave filters has become a bottleneck problem in the industrialization process. At present, a large number of projects still rely on vector network diagnosis and manual debugging, which is difficult to achieve fast and accurate debugging, especially for inexperienced filter debugging personnel.

1 Filter intelligent debugging principle and process

The purpose of developing the filter intelligent debugging platform is to continuously improve the debugging efficiency of microwave filters, greatly reduce the dependence of debugging on engineering experience, and reduce human labor as much as possible. The goal of the filter intelligent debugging platform is to establish an automated debugging platform with computers as the core, let the computer play the role of repetitive work and give it a certain level of intelligent judgment to guide the work of the debugger.

At present, intelligent debugging methods based on computer control are mainly divided into two categories: frequency domain method and time domain method:

(1) Time domain debugging method: This method mainly uses the frequency-time domain conversion of the signal to obtain the time domain response of the filter, find the change rules between each adjustable component and the time domain response, and perform corresponding debugging. Among them, the more prominent one is the time domain debugging method proposed by Agilent. The disadvantage of this debugging method is that it requires an ideal time domain response of a well-debugged filter as a template. Moreover, for cross-coupled filters, there is no obvious relationship between the filter debugging parameters and the time domain response curve.

(2) Frequency domain debugging method: The basic idea of ​​this method is to apply various numerical calculation methods to the frequency domain response curve of the filter S parameters, extract the filter model parameters, find out the gap with the ideal model parameters, and perform corresponding debugging. This system adopts the frequency domain debugging method.

As shown in Figure 1, both methods are based on equivalent circuit parameters. The main steps are as follows:

① Test the response of the filter to be adjusted;

②Use equivalent circuit model to extract parameters;

③ Compare the differences between the actual response extraction parameters and the ideal response ideal parameters;

④According to the above differences, the direction and amplitude of the next step of debugging are obtained, and the actual position of the adjustable components is changed;

⑤ Repeat the above steps ① to ④ until the measured response reaches the target. 2 Filter Intelligent Debugging Platform

As shown in Figure 2, the filter intelligent debugging platform is mainly composed of a computer, debugging machinery (such as a motor), a vector network analyzer, and the filter to be debugged. Its basic workflow is: first, the vector network analyzer tests the filter parameters, then collects the parameters into the computer, analyzes them through software, and obtains the physical quantity that needs to be debugged. Then, the computer controls the DC motor to drive the special debugging equipment to debug the debugging screw of the filter until the vector network analyzer tests that the filter parameters meet the design requirements.

2.1 Vector Network Analyzer

The vector network analyzer can fully evaluate RF and microwave devices. It includes integrated synthetic sources, test fixtures and tuned receivers. The built-in S-parameter test fixture provides full-range amplitude and phase measurements in both forward and reverse directions, as shown in Figure 3.

2.2 Debugging the machine

This solution uses a DC motor to drive a special debugging device to adjust the debugging screw to the best position. The current console is controlled by five motors, namely the x-axis, y-axis, z-axis, DM (motor for locking nuts), and DT (motor for tuning screws); x, y, and z use stepper motors, and DM and DT use servo motors. 2.3 Industrial Control Computer

Run the corresponding software on the industrial control computer to read the test parameters of the network analyzer, analyze and calculate the physical quantity that needs to be debugged, and then control the DC motor debugging equipment to debug. As shown in Figure 4, the user only needs to click Start Debugging in the software interface, and the debugging platform can automatically complete the debugging process, and can also provide friendly prompts for abnormalities that occur during the debugging process.

3 Conclusion

Microwave filters are widely used in fields such as communication, radar and measurement. With the development of society, the demand for them is also increasing. The debugging of microwave filters is a complex task, which requires rich practical experience. With the increase of the number of filter sections, the number of parameters involved in debugging also increases, and the difficulty of debugging also increases greatly. The introduction of intelligent computer-aided debugging technology can not only reduce the workload of debuggers, but also improve production efficiency, and has a good application prospect.

The intelligent debugging platform for cavity filters proposed in this paper can realize computer automatic debugging of filters. It can reduce the difficulty of test debugging of microwave components, shorten the debugging cycle, and greatly reduce the requirements for operator debugging experience. It is a very good way to improve the mass production capacity of microwave filters.