Design Method of Asymmetric Dual-Passband Filter Based on Frequency-Variable CT Topology
By adopting a design method based on frequency-varying CT topology, the problems of complex topology and insufficient transmission zeros in asymmetric dual-passband filters are solved, enabling more transmission zeros and simplifying the design process, thereby improving frequency selection characteristics and design efficiency.
Patent Information
- Application Number
- CN202310693279.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-06-12
- Publication Date
- 2026-11-10
- Estimated Expiration
- 2043-06-12
AI Technical Summary
Existing technologies for asymmetric dual-passband filter design suffer from complex topologies, insufficient transmission zeros, and require numerous matrix transformation operations to obtain the frequency-varying CT coupling matrix analytically, making the design process challenging.
A design method based on frequency-varying CT topology is adopted. The transmission zeros and reflection zeros are determined by the frequency transformation function. The frequency-varying coupling matrix is optimized by the optimization method, which simplifies the topology, avoids complex matrix transformations, and achieves more transmission zeros and better frequency selectivity.
Without increasing cross-coupling paths, this method achieves more transmission zeros, simplifies the topology, improves frequency selectivity, reduces tuning difficulty, and enhances the design efficiency of dual-passband filters.
Smart Images

Figure CN116722338B_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of electromagnetic field and microwave technology, and particularly relates to the design of asymmetric dual-passband filters, specifically a design method for an asymmetric dual-passband filter based on a frequency-varying CT topology. Background Technology
[0002] In the design of dual-passband filters, a sufficient number of transmission zeros are required to ensure frequency selectivity. Common CT and CQ structures are typically used to achieve transmission zeros. However, common CT and CQ structures cannot achieve N-1 transmission zeros, where N is the order of the filter. This means they do not have quasi-regularity, and excessive cross-coupling can also cause difficulties in the design of dual-passband filters and increase the difficulty of final simulation and tuning.
[0003] With the increasing demand for miniaturization and high selectivity in filters, frequency-variable coupling structures have attracted growing attention. These structures can achieve more transmission zeros, possess quasi-canonical characteristics, and facilitate topology simplification. Frequency-variable coupling topologies are mainly divided into two types: 1) in-line topology; 2) frequency-variable coupling blocks (such as frequency-variable CT and frequency-variable CQ). From an analytical perspective, the matrix transformations involved in the simplification process for in-line and frequency-variable CT topologies differ significantly: The first method starts from a wheel-shaped topology, utilizing similarity and scaling transformations to ultimately achieve an in-line topology; this topology is simple and does not contain any cross-coupling paths. The second method also starts from a wheel-shaped topology, using different similarity and scaling transformations to ultimately introduce frequency-variable coupling into the CT and CQ structures.
[0004] However, in-line topology presents a challenge in designing dual-passband filters: achieving inter-band transmission zeros is difficult. Because asymmetric dual-passband filters are designed using frequency transformation, achieving inter-passband transmission zeros with frequency-variable coupling structures requires extremely sensitive frequency changes. Typically, the coupling coefficient must change from a large constant to zero with a frequency change of around 10MHz, which is difficult to achieve in HFSS. Therefore, from a design perspective, in-line topology is currently only applicable to single-passband filter design; the industry typically uses frequency-variable coupling block topology for dual-passband filter design. Methods for obtaining the frequency-variable CT coupling matrix are mainly divided into analytical methods and optimization methods. Due to the diversity of frequency-variable CT structures, obtaining different forms of frequency-variable CT coupling matrices using analytical methods requires numerous modifications.
[0005] Currently, when designing dual-passband filters using frequency transformation methods, there is no good method in the industry to make the designed filters have good frequency selectivity characteristics. Summary of the Invention
[0006] To address the aforementioned problems and shortcomings, and to resolve the issues of complex topologies, limited transmission zeros, and the need for extensive matrix transformations when obtaining different forms of frequency-varying CT coupling matrices using analytical methods in existing asymmetric dual-passband filter designs, this invention provides a design method for an asymmetric dual-passband filter based on a frequency-varying CT topology. On one hand, this method achieves more transmission zeros without increasing cross-coupling paths, resulting in better rectangular coefficients for the dual-passband filter, simplifying the topology, and reducing the tuning difficulty. On the other hand, the optimization method avoids the need for extensive matrix transformations when obtaining different forms of frequency-varying CT coupling matrices.
[0007] A design method for an asymmetric dual-passband filter based on a frequency-varying CT topology, comprising the following steps:
[0008] Step 1: Starting from the transmission zeros and reflection zeros of the two low-pass prototypes, obtain the transmission zeros and reflection zeros of the asymmetric dual-passband filter function using the frequency transformation function. The frequency transformation function is:
[0009] S = S' / a1 + b1 / (S' - jZ) 12 ),S' <Z 12
[0010] S = S' / a² + b² / (S' - jZ) 12 ),Z 12 <S'
[0011] Where the unknown parameter a i and b i The expression is:
[0012]
[0013]
[0014] In the above formula, S = jΩ, S' = jΩ', and a1, b1, a2, and b2 are unknown variables; the -∞, -1, 1, and ∞ of the first low-pass prototype correspond to the -∞ and P of the Ω' domain, respectively. L1 P U1 Z 12 The -∞, -1, 1, and ∞ values of the second low-pass prototype correspond to Z in the Ω' domain, respectively. 12 P L2 P U2 ∞. The Ω domain is the normalized frequency domain of the low-pass prototype, the Ω' domain is the normalized frequency domain of the dual-passband filter, the S domain is the normalized S domain of the low-pass prototype, and the S' domain is the normalized S domain of the dual-passband filter. P Li P is the lower boundary of the i-th passband. Ui P is the upper boundary of the i-th passband.L1 =-1, P U2 =1,Z 12 It is the transmission zero point between two passbands, P U1 <Z 12 <P L2 .
[0015] Based on the correspondence, determine the transmission zero Z between the two passbands. 12 and the upper boundary P of the first passband U1 and the lower boundary P of the second passband L2 The values of a1, b1, a2, and b2 are obtained, and then the transmission zeros and reflection zeros of the asymmetric dual-passband filter function are obtained.
[0016] Step 2: Based on the design objective that the two asymmetric passbands have different return losses, adjust Z... 12 Location, until it meets the design specifications.
[0017] Step 3: Based on Step 2, obtain the S-parameters and transverse coupling matrix M of the asymmetric dual-passband filter according to filter synthesis theory. trans .
[0018] Step 4: Select the frequency-varying CT topology and obtain the corresponding frequency-varying coupling matrix. At this point, all parameters in the frequency-varying coupling matrix except for 0 are unknown. Transform the frequency-varying coupling matrix containing unknown parameters into a non-frequency-varying coupling matrix M through matrix similarity transformation. final The objective function is constructed using the eigenvalues of the fully normalized coupling matrix; finally, all unknown parameters of the frequency-varying coupling matrix are obtained by optimization using the constrained nonlinear minimization function fmincon.
[0019] This invention relates to an asymmetric dual-passband filter based on a frequency-variable coupling structure, which has the following advantages:
[0020] 1. Compared to traditional dual-passband filter design methods based on frequency transformation, this invention (step 1) uses two low-pass prototypes instead of a single synthesized asymmetric dual-passband filter. This allows for different orders, return losses, and bandwidths in different passbands.
[0021] 2. This invention is the first to propose using the frequency-varying coupling structure in the frequency-varying CT structure to generate and control the transmission zeros outside the band, and using the CT structure in the frequency-varying CT structure to generate and control the transmission zeros between the passbands. It successfully introduces frequency-varying coupling into the design of dual-passband filters, which greatly improves the frequency selectivity of dual-passband filters.
[0022] 3. This invention employs the fmincon function to design an optimization method for an asymmetric dual-passband filter based on a frequency-varying CT topology. This avoids complex matrix transformation operations. It only requires setting the unknown parameters of the coupling matrix corresponding to the frequency-varying CT topology when determining it. Then, the frequency-varying coupling matrix is converted into a frequency-invariant coupling matrix through a simple matrix transformation operation. Finally, after setting the objective function, the unknown parameters in the frequency-varying coupling matrix can be optimized. This achieves more transmission zeros without increasing the number of cross-coupling paths, ensuring that the dual-passband filter has good rectangular coefficients. Attached Figure Description
[0023] Figure 1 This is a schematic diagram of the frequency transformation of the asymmetric frequency-converting dual-passband filter of the present invention;
[0024] Figure 2 This is a schematic diagram of the S-parameters obtained from the theoretical polynomial of the asymmetric frequency-variable dual-passband filter in the embodiment.
[0025] Figure 3 This is the frequency-varying CT topology of the embodiment;
[0026] Figure 4 This is a schematic diagram of the S-parameters of the frequency-varying coupling matrix obtained through the optimized implementation example. Detailed Implementation
[0027] The present invention will now be further explained in detail with reference to the embodiments and accompanying drawings.
[0028] First, a schematic diagram of the frequency conversion in this embodiment is shown below. Figure 1 As shown, the Ω domain is the normalized frequency domain of the low-pass prototype, and the Ω' domain is the normalized frequency domain of the dual-passband filter, where P L1 =-1 and P U2 =1. The first low-pass prototype of the Ω field is converted to the (-∞, Z) field of the Ω' field. 12 The second low-pass prototype of the Ω domain is then converted to the (Z) domain. 12 ,∞), P U1 <Z 12 <P L2 .
[0029] The design goals of the dual-passband filter are as follows: the first passband has a third-order order and a return loss greater than 25dB; the second passband has a third-order order and a return loss greater than 25dB. The normalized bandwidths of the two passbands are (-1, -0.4) and (0.4, 1), respectively, Ω' TZ =-3.78,-0.1,0.1,3.78,Z 12=0. Calculations yield a1 = 0.6, b1 = 0.67, a2 = 0.6, b2 = 0.67. Then, based on the synthesis theory of filters and the frequency transformation formula, the reflection zeros of the dual-passband filter can be obtained, thus yielding the S-parameters and the transverse coupling matrix M. trans .
[0030] The asymmetric dual-passband filter to be designed has a passband width of 100MHz, a center frequency of 2000MHz, and four transmission zeros at 1820MHz, 1995MHz, 2005MHz, and 2198MHz respectively. The first passband bandwidth is 1950MHz-1980MHz; the second passband bandwidth is 2020MHz-2050MHz. Its S-parameter diagram is shown below. Figure 2 As shown; lateral coupling matrix M trans As shown in Table 1 below.
[0031] Table 1:
[0032]
[0033] Finally, based on the obtained dual-passband filter transverse coupling matrix, the frequency-varying CT coupling matrix is synthesized, and the selected topology is as follows. Figure 3 As shown, it is important to emphasize that the transmission zeros between the two passbands must be generated and controlled by the CT structure formed between the resonators, while the out-of-band transmission zeros are generated and controlled by either the CT structure or the frequency-varying coupling structure. In this embodiment, the out-of-band normalized transmission zeros -3.78 and 3.78 are generated and controlled by the frequency-varying coupling between resonators 1 and 3 and between resonators 4 and 6. The first inter-band normalized transmission zero -0.1 is generated by the CT structure composed of resonators 1, 2, and 3, and the second inter-band normalized transmission zero 0.1 is generated by the CT structure composed of resonators 4, 5, and 6. Then, the frequency-varying coupling matrix is transformed into a non-frequency-varying coupling matrix M through matrix similarity transformation. final .
[0034] The objective function of the optimization method is Cost = (λ) trans -λ final ) T (λ trans -λ final ), where λ final The corresponding coupling matrix M after matrix transformation final The relevant eigenvalues, λ trans Corresponding horizontal coupling matrix M trans The relevant eigenvalues, where T is the transpose sign. in These are the eigenvalues of the horizontal coupling matrix; The eigenvalues are the matrix formed by removing the first row, first column, last row, and last column of the horizontal coupling matrix. The eigenvalues are the matrix formed by the elements remaining after removing the first row and first column of the horizontal coupling matrix. λ represents the eigenvalues of the matrix formed by removing the last row and last column of the horizontal coupling matrix. final The elements have the same properties as λ. trans The meaning is the same as the corresponding element.
[0035] For the objective function described above, optimization using the constrained nonlinear minimization function fmincon yields the frequency-varying coupling matrix for the specified topology. In this embodiment, the error of the objective function is less than 1*10^6. -7 After obtaining the asymmetric dual-passband frequency-varying coupling matrix as shown in Table 2, the classical coupling matrix synthesis method can be applied to obtain the following... Figure 4 The S-parameters are shown.
[0036] Table 2: Frequency-variable coupling matrix of the asymmetric frequency-variable dual-passband filter in the embodiment.
[0037]
[0038] As can be seen from the above embodiments, the optimization results of this invention, when synthesizing an asymmetric frequency-variable dual-passband filter based on two low-pass prototype functions, agree well with the theoretical polynomial response. This invention applies a frequency-variable CT topology to the design of an asymmetric dual-passband filter based on two low-pass prototype functions. After converting the frequency-variable coupling matrix corresponding to the topology into a non-frequency-variable coupling matrix, all unknown parameters of the frequency-variable coupling matrix are solved by constrained nonlinear minimization function optimization. This achieves more transmission zeros without increasing cross-coupling paths and gives the dual-passband filter better rectangular coefficients, simplifying the filter topology and reducing the complexity of debugging. This invention can synthesize dual-passband filters of different orders, return losses, and bandwidths in different passbands, simplifying the difficulty of filter design and simulation, and greatly improving out-of-band rejection and inter-passband isolation. This invention provides a good optimization effect for engineering practice and has significant engineering value.
Claims
1. A design method for an asymmetric dual-passband filter based on a frequency-varying CT topology, characterized in that, The specific steps are as follows: Step 1: Starting from the transmission zeros and reflection zeros of the two low-pass prototypes, obtain the transmission zeros and reflection zeros of the asymmetric dual-passband filter function using the frequency transformation function; the frequency transformation function is: Where the unknown parameters and The expression is: In the above formula, , a1, b1, a2, and b2 are unknown variables; the first low-pass prototype -1, 1, Corresponding in sequence domain , , , The second low-pass prototype -1, 1, Then corresponding in sequence domain , , , ; The domain is the normalized frequency domain of the low-pass prototype. The domain is the normalized frequency domain of the dual-passband filter. The field is the normalized S-field of the low-pass prototype. The domain is the normalized S-domain of the two-passband filter. It is the lower boundary of the i-th passband. It is the upper boundary of the i-th passband. , =1, It is the transmission zero point between two passbands. ; Based on the correspondence, determine the transmission zero point between the two passbands. and the upper boundary of the first passband and the lower boundary of the second passband The values of a1, b1, a2, and b2 are obtained, and then the transmission zeros and reflection zeros of the asymmetric dual-passband filter function are obtained. Step 2: Adjust according to the design target of different return losses in the two asymmetric passbands. Location, until it meets the design specifications; Step 3: Based on Step 2, obtain the S-parameters and transverse coupling matrix of the asymmetric dual-passband filter according to filter synthesis theory. ; Step 4: Select the frequency-varying CT topology and obtain the corresponding frequency-varying coupling matrix. At this point, all parameters in the frequency-varying coupling matrix except for 0 are unknown. Transform the frequency-varying coupling matrix containing unknown parameters into a non-frequency-varying coupling matrix through matrix similarity transformation. The objective function is constructed using the eigenvalues of the fully normalized coupling matrix; finally, all unknown parameters of the frequency-varying coupling matrix are obtained by optimization using the constrained nonlinear minimization function fmincon. The frequency-varying CT topology: the transmission zeros between the two passbands must be generated and controlled by the CT structure composed of resonators, while the transmission zeros outside the passband are generated and controlled by the CT structure or the frequency-varying coupling structure.
2. The design method of an asymmetric dual-passband filter based on a frequency-varying CT topology as described in claim 1, characterized in that: The objective function in step 4 is: ,in Corresponding to the coupling matrix after matrix transformation Related eigenvalues, Corresponding horizontal coupling matrix The relevant eigenvalues, where T is the transpose sign; ,in These are the eigenvalues of the horizontal coupling matrix; The eigenvalues are the matrix formed by removing the first row, first column, last row, and last column of the horizontal coupling matrix. The eigenvalues are the matrix formed by the elements remaining after removing the first row and first column of the horizontal coupling matrix. The eigenvalues of the matrix formed by the elements remaining after removing the last row and last column of the horizontal coupling matrix; The elements have the same The meaning is the same as the corresponding element.
Citation Information
Patent Citations
Design method of asymmetric dual-passband filter based on frequency-variable coupling structure
CN116205082A
