1. 关于CAD辅助设计软件和网络分析仪
For high-frequency circuit design, there are already very good CAD software available. Its powerful functions are sufficient to overcome the problems of people’s lack of design experience and tedious parameter retrieval and calculation. By combining powerful network analyzers, even experienced individuals can complete higher quality RF component designs. However, in practice, this is not the case. CAD design software relies on powerful library functions, including component parameters and basic performance indicators provided by most world RF device manufacturers. Many RF engineers mistakenly believe that as long as they use tools for design, there will be no problems. However, the actual results always go against the wishes. The reason is that they have abandoned the flexible application of basic concepts in high-frequency circuit design and accumulated experience in applying basic design principles under incorrect understanding. Therefore, they often make basic application errors in the application of software tools. The CAD software for RF circuit design is a transparent visualization software that utilizes its various high-frequency basic configuration model libraries to simulate the actual working state of the circuit. So far, we have learned that there are two types of high-frequency basic configuration models for key links, one is the component model in the form of concentrated parameters, and the other is the local functional model in traditional design. Therefore, there are the following issues:
1) Component models and CAD software have been interacting and developing with each other for a long time, and have become increasingly sophisticated. In practice, the authenticity of the model is generally credible. However, the application environment considered by the component model (especially the electrical environment of the component application) is typical numerical values. In most cases, a series of application parameters must be determined based on experience, otherwise the actual results may sometimes be even farther than the design results without CAD software assistance.
2) The traditional high-frequency basic configuration models established in CAD software are usually limited to predictable aspects under current application conditions and can only be limited to basic functional models (otherwise, products do not need to hire people, and all products rely on CAD to be born).
3) It is particularly noteworthy that the establishment of typical functional models involves applying components in typical ways and using typical and perfect process structures (including PCB board structures), and their performance has also reached a high level of “typical. But in practice, this is completely imitated, far from the model state. The reason is that although the selected components and their parameters are the same, their combined electrical environment cannot be the same. In low-frequency circuits or digital circuits, the difference of a few centimeters is not a big obstacle, but in RF circuits, fatal errors often occur.
4) In the design of CAD software, the fault-tolerant design of the software does not focus on whether there have been incorrect parameter settings that contradict the actual situation. Therefore, it provides ideal results based on the running path of the software, but in practice, it is full of problematic results. It can be known that the key error link is the failure to correctly apply the basic principles of RF circuit design when using CAD software.
5) CAD software is only a design aid tool. It utilizes real-time simulation capabilities, powerful component model libraries and their function generation capabilities, typical application model libraries, etc. to simplify people’s tedious design and calculation work. But so far, it is far from replacing artificial intelligence in specific designs. The effectiveness of CAD software in assisting in the design of RF PCB boards is an important aspect of software popularity. But in practice, many RF engineers are often misunderstood by it. The reason is still the fault tolerance of parameter settings. They often use their simulation capabilities to obtain ideal models (including each functional link), but only discover after actual debugging that it is best to use their own experience to design. Therefore, in PCB design, CAD software is still only beneficial for engineers with basic RF design experience and skills, helping them engage in tedious process design (non basic principle design).
There are two types of network analyzers, scalar and vector, which are basic instruments for RF circuit design. The usual approach is to complete the circuit and PCB board design (or use CAD software) based on basic RF circuit design concepts and principles, complete the PCB board prototype processing and assemble the prototype as needed, and then use a network analyzer to design the network analysis of each link one by one, so as to achieve the appropriate state of the circuit. But the cost of this kind of work is to produce at least 3 to 5 versions of PCBs. Without basic PCB design principles and concepts, more PCB versions are needed (or the design cannot be completed). In the process of using a network analyzer to analyze RF circuits, it is necessary to have a complete concept and principles of high-frequency circuit PCB board design, and be able to clearly understand the design defects of the PCB board through the analysis results. Only this requires the relevant engineers to have considerable experience. When analyzing the network links of the prototype, it is necessary to rely on proficient experimental experience and skills to construct a local functional network. Because in many cases, circuit defects discovered by network analyzers can have multiple factors simultaneously, it is necessary to use the construction of local functional networks to analyze and thoroughly investigate the causes. The construction of this experimental circuit must rely on clear high-frequency circuit design experience and proficient PCB board construction principles.
2 Scope of this article
This article mainly focuses on the concepts and design principles of microwave level high-frequency circuits and their PCB board design, which is a cutting-edge category of communication products. The reason for choosing the PCB design principle for microwave level high-frequency circuits is because this principle has broad guiding significance and belongs to the current high-tech popular application technology. The transition from microwave circuit PCB design concepts to high-speed wireless network (including various access networks) projects also belongs to the same category, as they are based on the same basic principle, namely the dual transmission line theory. Experienced RF engineers have a high success rate in designing digital circuits or PCB boards for relatively low-frequency circuits, as their design philosophy is centered around “distributed” parameters, which are often overlooked in low-frequency circuits (including digital circuits). For a long time, many electronic product designs completed by peers (mainly communication products) have often been filled with problems. On the one hand, it is related to the lack of electrical principle design (including redundant design, reliability design, etc.) links, but more importantly, many such problems arise when people believe that all necessary links have been considered. Faced with these problems, they often focus on checking programs, electrical principles, parameter redundancy, etc., but pay little attention to PCB board design, which is often due to PCB board design defects leading to a large number of product performance issues. The design principles of PCB boards involve many aspects, including basic principles, anti-interference, electromagnetic compatibility, safety protection, etc. For these aspects, especially in high-frequency circuits (especially microwave level high-frequency circuits), the lack of relevant concepts often leads to the failure of the entire research and development project. Many people are still stuck on the basis of “connecting electrical principles and wires to play a predetermined role”, and even believe that “PCB board design is a consideration of structure, process, and improving production efficiency”. Many RF engineers are not fully aware that this stage should be a special focus of the entire RF design work. They mistakenly focus their energy on selecting high-performance components, resulting in a significant increase in cost and minimal performance improvement. It should be pointed out that digital circuits rely on their powerful anti-interference, error detection, and error correction capabilities, allowing for the arbitrary construction of each intelligent link to ensure the normal functioning of the circuit. A typical digital application circuit with various high cost “guaranteed normal” configurations is clearly a measure without product concept. However, it often leads to a series of product issues when considering aspects that are not worth it. The reason is that from the perspective of product engineering, this functional link that is not worthy of reliability assurance should be based on the working mechanism of the digital circuit itself. However, incorrect construction in circuit design (including PCB board design) leads to the circuit being in an unstable state. The reason for this unstable state is the same concept in basic applications, similar to problems in high-frequency circuits.
In digital circuits, there are three aspects that need to be taken seriously:
1) Digital signals themselves belong to broadband signals. According to the results of the Fourier function, it contains a very rich high-frequency component, so the high-frequency component of the digital signal is fully considered in the design of digital ICs. However, apart from digital ICs, any signal transition area within and between each functional link can lead to a series of problems. Especially in circuits where digital, analog, and high-frequency circuits are mixed.
2) The various reliability designs in digital circuit applications and the reliability requirements of circuits in practical applications are related to product engineering requirements, and various high cost “guarantee” components cannot be added to traditional designs to meet the requirements of circuits.
3) The working speed of digital circuits is developing towards higher frequencies at an unprecedented pace (for example, the current CPU’s main frequency has reached 1.7GHz, far exceeding the lower limit of the microwave frequency band). Although the reliability guarantee function of related equipment is also supported, they are based on the internal and typical external signal characteristics of the equipment itself.
3 Overview of the Guiding Significance of the Dual Transmission Line Theory and PCB Circuit Design Principles in Microwave Circuit Design
The concept of PCB board based on the theory of dual transmission lines
For microwave level high-frequency circuits, each adjacent wire on the PCB board (including adjacent wires) forms a characteristic that follows the basic principle of dual transmission lines (which will be clearly explained in the following text). Although common microwave RF circuits are equipped with a ground plane on one side, which makes the microwave signal transmission line on it tend to become a complex four port network, directly following the coupled microstrip theory, its foundation is still the dual line theory. Therefore, in design practice, the guiding significance of the dual line theory is more extensive. Overall, for microwave circuits, microstrip theory has quantitative guidance significance and belongs to the specific application of bilinear theory, while bilinear theory has broader qualitative guidance significance. It is worth mentioning that all the concepts provided by the dual line theory may seem unrelated to actual design work on the surface (especially digital circuits and low-frequency circuits), but this is actually an illusion. The dual line theory can guide all conceptual issues in electronic circuit design, especially the significance of PCB circuit design concepts is more prominent. Although the dual line theory is based on microwave high-frequency circuits, it is only because the influence of distributed parameters in high-frequency circuits becomes significant, making its guiding significance particularly prominent. In digital or low-frequency circuits, distributed parameters can be ignored relative to concentrated parameter components, and the concept of bilinear theory will correspondingly become blurred. However, how to distinguish between high-frequency circuits and low-frequency circuits is often overlooked in design practice. What category does a typical digital logic or pulse circuit belong to? Obviously, low-frequency circuits and low-frequency circuits with nonlinear components can easily map out some high-frequency characteristics when sensitive conditions change. The main frequency of the CPU has reached 1.7GHz, far exceeding the lower limit of the microwave frequency band, but it is still a digital circuit. Due to these uncertainties, the design of PCB boards becomes very important. In many cases, passive components in a circuit can be equivalent to specific specifications of transmission lines or microstrip lines, and can be described through the theory of dual transmission lines and their related parameters. In short, the dual transmission line theory was born on the basis of comprehensive circuit characteristics. Therefore, strictly speaking, if every step in design practice is based on the concept reflected in the dual transmission line theory, the corresponding PCB board circuit will face very few problems (regardless of what the circuit looks like under working conditions).