A calibration structure and calibration method for D-band glass vias
By optimizing the calibration structure and measurement method of the D-band glass via, the signal reflection and loss problems in the TGV and transmission line conversion structure were solved, achieving smooth transition and parameter matching of high-frequency signals and improving signal transmission performance.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- QIANYUAN NATIONAL LABORATORY
- Filing Date
- 2026-04-07
- Publication Date
- 2026-06-09
AI Technical Summary
In TGV and transmission line conversion structures, signals are prone to reflection and loss at the conversion point, especially at high frequencies. Existing technologies have not been able to effectively solve the problems of signal transmission matching and process parameter optimization.
A calibration structure for D-band glass vias is designed. By optimizing the number, location, and spacing of vias, and combining them with a coplanar waveguide structure, a smooth signal transition and electrical parameter matching are achieved. A vector network analyzer is used to measure scattering parameters, eliminate the influence of non-intrinsic factors, and extract the intrinsic parameters of the glass vias.
This achieved efficient transmission of D-band signals, reduced reflection and loss, improved signal quality, and provided a stable testing basis for subsequent process optimization.
Smart Images
Figure CN122171913A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of microwave / millimeter-wave interconnect and packaging testing technology, and in particular to a calibration structure and calibration method for D-band through glass vias (TGV). Background Technology
[0002] In the development of modern electronic systems, TGV and transmission line converter structures play a crucial role. With the continuous improvement of chip integration and the increasing demand for high-speed, high-frequency signal transmission, traditional two-dimensional interconnect methods are gradually becoming insufficient to meet performance requirements, leading to the emergence of three-dimensional interconnect technology, of which TGV is an important component.
[0003] As a conductive channel that vertically penetrates a glass substrate, the core function of the TGV (Transient Transistor) is to achieve electrical connections between different planes. This core function breaks through the limitations of traditional planar interconnects, providing a path for signal transmission in three-dimensional space. Glass material has excellent insulating properties; its dielectric constant is only about one-third that of silicon, and its loss factor is 2-3 orders of magnitude lower than silicon. This gives the TGV a significant advantage over silicon-based adapters in signal transmission, effectively reducing substrate loss and parasitic effects, and largely ensuring the integrity of the transmitted signal.
[0004] Transmission lines play a crucial role in signal transmission within electronic systems. Common examples include coplanar waveguides (CPWs), which efficiently transmit signals within a plane. The key to TGV-to-transmission-line (TGV) conversion structures lies in smoothly and efficiently transitioning signals from the planar transmission line to the vertical TGV, and then back to the transmission line, achieving interconnection between different levels. This conversion process presents numerous challenges to complete signal transmission. Due to the differences in structure and characteristics between transmission lines and TGVs, mismatched electrical parameters can easily lead to signal reflection at the conversion point, resulting in energy loss, amplitude attenuation, and phase delay, severely impacting signal quality. Furthermore, these reflections and losses become more pronounced at higher signal frequencies.
[0005] To address the aforementioned issues, the conversion structure requires careful design. On one hand, a smooth transition between the transmission line and the TGV must be achieved through a well-designed geometry. A gradual connection structure can be employed, allowing the electromagnetic field distribution to adjust progressively during signal transmission, reducing abrupt changes and thus minimizing reflections. On the other hand, precise matching of electrical parameters is necessary. This includes adjusting the characteristic impedances of the transmission line and the TGV to ensure they are as consistent as possible. However, characteristic impedance is affected by various factors such as conductor size, shape, and surrounding medium. Optimizing these parameters is crucial to maintaining good signal transmission characteristics during conversion. Furthermore, adding matching circuit components, such as capacitors and inductors, can further compensate for differences in electrical parameters and improve the signal transmission quality.
[0006] In terms of manufacturing processes, the TGV's fabrication process is also crucial to the performance of the conversion structure. Excessive sidewall roughness increases resistance and inductance during signal transmission, leading to increased signal loss. Therefore, selecting a suitable via forming process and strictly controlling process parameters are essential to ensuring the performance of the TGV and transmission line conversion structure.
[0007] Existing technologies rarely address the impact of the number of vias on the signal line (S line) and ground line (G line) on the conversion structure. Furthermore, in most cases, the structure at low frequencies may generate other transmission modes at high frequencies, resulting in higher losses.
[0008] In addition to impedance matching, the probe tip spacing, hole spacing (too dense spacing can cause breakage), and distance between the hole and the metal must also be taken into account during the design process to fabricate a device that meets the expected results. Summary of the Invention
[0009] Based on the aforementioned analysis of TGV vias in the D-band, and combined with the sample manufacturability constraints and electromagnetic simulation verification results, the purpose of this invention is to provide a calibration structure and calibration method for D-band glass vias. By designing a glass via for D-band signal transmission and measuring the scattering parameters of the glass via body, calibration is achieved.
[0010] To achieve the above-mentioned objectives, embodiments of the present invention provide a calibration structure for a D-band glass through-hole, comprising: Independent metal pads, one for each signal line and ground line, are used for RF probe contact; Bottom metal plate; At least one set of glass vias formed on a glass substrate for D-band signal transmission, so as to realize signal transmission between each metal pad and the bottom metal plate; The calibration structure is used to calibrate the scattering parameters of the glass via: by setting differences in the transmission length or termination state of the bottom metal plate, the scattering parameters of the calibration structure are tested and the bulk scattering parameters of the glass via are indirectly obtained.
[0011] In the entire calibration structure, both the metal pads and the bottom metal plate are coplanar waveguides used for signal transmission. During calibration, the calibration structure stably introduces the coplanar waveguide (CPW) signal from the probe port into the via interconnect region formed by the glass substrate, and then transitions it back to the port of another CPW to achieve the measurement of scattering parameters (S-parameters). To eliminate the influence of non-via-body factors such as probe contacts, pads, and lead segments on the measurement results, the measurement reference plane is effectively advanced to both ends of the TGV via body through a subtraction de-embedding method, thereby extracting the intrinsic S-parameters of the via and using them for subsequent modeling of electrical parameters such as equivalent resistance (R), inductance (L), conductance (G), and capacitance (C).
[0012] Since the parameters that can be changed by a single TGV on each G-line and S-line are relatively limited, increasing the number of TGVs on each G-line and S-line allows for more parameter changes, further optimizing the design. One or two vias are added to each ground line and S-line respectively to achieve this optimization. However, increasing the number of vias while keeping the radius constant reduces transmission performance because it exacerbates signal interference. Each via generates an electromagnetic field during signal transmission, and the interaction of these fields increases crosstalk. Therefore, this invention optimizes the structural geometry to minimize the impact on transmission performance while maintaining an increased number of vias.
[0013] Regarding structural geometry parameters, the mechanical strength of the glass substrate, the accessibility of the drilling process, the reliability of the vias after metallization, and the impact of high-frequency electromagnetic coupling are comprehensively considered. This embodiment limits key dimensions: the diameter of the glass vias is set to 20-40 μm; the center-to-center distance between two adjacent glass vias on each metal pad is not less than 40 μm to avoid the risk of breakage during processing due to excessively dense vias and to suppress strong coupling and crosstalk between adjacent vias; the minimum distance from the center of any glass via to the metal pad and the edge of the bottom metal plate is not less than 20 μm to reduce the additional parasitic capacitance changes caused by the concentration of electric field at the metal edge and to improve the layout manufacturing margin.
[0014] The transition dimensions of the metal pad traces are designed to accommodate the contact requirements of 100μm pitch RF probes, while ensuring that the characteristic impedance of the transmission line is 50Ω to meet the standard impedance environment of the vector network analyzer and probe station system, thereby reducing reflection errors caused by port mismatch.
[0015] Regarding the via array configuration, the number and arrangement of vias on the signal line (S-line) and ground line (G-line) were compared, designed, and iteratively optimized. In the initial scheme, three vias were placed at corresponding positions on the S-line and each side of the G-line, arranged horizontally. However, after D-band electromagnetic simulation and parameter scanning, it was found that the S-parameters of this three-via scheme failed to meet the expected targets in the target frequency band, exhibiting increased insertion loss or worsened return signal. Furthermore, at high frequencies, non-ideal transmission modes caused by structural discontinuities and enhanced coupling were more likely to occur, leading to overall transmission performance degradation. Further parameter tuning results showed that, with the via radius remaining constant, increasing the number of vias enhanced the electromagnetic field interaction between vias, increasing signal crosstalk and parasitic coupling effects, thereby offsetting or even negatively impacting the improvement effect originally expected by increasing the number of vias. Therefore, preferably, the glass via includes a signal via group and a ground via group, wherein each signal via group corresponds to the metal pad of the corresponding signal line and includes two signal vias; each ground via group includes a metal pad corresponding to the ground line on each side and includes two ground vias, so as to form a symmetrical return path and reduce high-frequency mode degradation.
[0016] The glass via designed above can support D-band signal transmission, meaning it can transmit signals in the 110GHz to 170GHz operating frequency range. The calibration structure for the glass via can be used to calibrate its scattering parameters; two specific calibration methods are available: Preferably, one method involves testing and calibrating the scattering parameters of the structure by setting differences along the transmission length of the ground metal plate, and indirectly obtaining the bulk scattering parameters of the glass aperture, including: The original transmission length of the bottom metal plate is set to L, denoted as structure L. The transmission length of the bottom metal plate is extended to 2L, denoted as structure 2L. The scattering parameters of the L-structure and the 2L-structure were tested using a vector network analyzer. The bulk scattering parameters of the glass aperture are calculated by converting the matrices of the two scattering parameters into the transmission parameter matrices.
[0017] Preferably, another method involves testing and calibrating the scattering parameters of the structure and indirectly obtaining the bulk scattering parameters of the glass aperture by setting differences in the termination state at the bottom metal plate end, including: The calibration structure can be placed in a through-circuit, open-circuit, and short-circuit state by changing the bottom metal plate. The scattering parameters of the calibration structure in the through, open, and short-circuit states were tested using a vector network analyzer. The bulk scattering parameters of the glass aperture are extracted based on the three scattering parameters.
[0018] To achieve the above-mentioned objectives, this invention also provides a calibration method for a D-band glass via, wherein the calibration method employs the aforementioned calibration structure and includes the following steps: The scattering parameters of the calibration structure are tested by setting differences in the transmission length or termination state of the bottom metal plate, and the bulk scattering parameters of the glass through-hole are indirectly obtained.
[0019] Compared with the prior art, the beneficial effects of the present invention include at least the following: This invention relates to a calibration structure for glass vias. First, the glass vias are designed, and the number of vias is convergentally optimized to achieve signal transmission in the D-band. The number of vias on the metal pads corresponding to the S-line and G-line is appropriately reduced, and the via positions, spacing, and relative relationships with the CPW conductor boundaries are jointly optimized. This results in a smoother impedance transition and a clearer return path within the conversion region, thereby achieving better S-parameter performance in the target frequency band. The optimized S-parameters achieve the expected transmission characteristics and provide a stable testing basis for subsequent via body parameter extraction.
[0020] Based on this, the S-parameters of the glass vias are measured using a calibration structure to calibrate the S-parameters of the solid glass vias. The calibrated S-parameters are used to extract electrical parameters such as RLGC of the vias, enabling quantitative evaluation of the via transmission performance and subsequent structural / process optimization. Attached Figure Description
[0021] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0022] Figure 1 This is a schematic diagram of the calibration structure for the D-band glass through-hole provided in the embodiment; Figure 2 This is a comparison diagram of the S-parameter effects of two-way holes and three-way holes provided in the embodiment; Figure 3 This is a schematic diagram of the calibration structure provided in the embodiment during calibration measurement; where (a) is an L structure and (b) is a 2L structure; Figure 4 The S-shaped glass through-hole provided in the embodiment 21 parameter; Figure 5 This is a schematic diagram of the termination state of the calibration structure provided in the embodiment, where (a) is the open state and (b) is the short state; Among them, 1-the metal pad corresponding to the signal line, 2-the metal pad corresponding to the ground line, 3-the bottom metal plate, 4-the signal line through hole, and 5-the ground line through hole. Detailed Implementation
[0023] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and do not limit the scope of protection of this invention.
[0024] The technical concept of this invention includes: In high-frequency interconnect scenarios, TGVs, as a key structure for three-dimensional interconnects on glass substrates, are significantly affected by factors such as parasitic resistance, parasitic inductance, parasitic capacitance, and dielectric loss, which significantly influence the insertion loss, return loss, and phase characteristics of the device. Existing TGV testing inevitably introduces additional parameters, resulting in the measured scattering parameters (S-parameters) including various non-target contributions such as probe contact, traces, and pad / via transitions. This affects the accurate extraction of via body parameters and reduces the reliability and comparability of the evaluation of the true performance of TGVs. To address this, this embodiment provides a calibration structure for D-band glass vias. Through optimized design, the glass vias can achieve D-band signal transmission, and the entire calibration structure can be used to measure the body S-parameters of the glass vias, achieving S-parameter calibration.
[0025] like Figure 1 As shown, the calibration structure of the D-band glass via includes metal pads 1 and 2 on the front side of the glass substrate, which independently correspond to the signal line and the ground line respectively; a bottom metal plate 3 on the back side of the glass substrate; and at least one set of glass vias formed on the glass substrate for D-band signal transmission, so as to realize signal transmission between each metal pad and the bottom metal plate.
[0026] In this embodiment, the glass vias are divided into signal via group 4 and ground via group 5. Signal via group 4 includes two signal vias, and ground via group 5 includes two ground vias on each side of the ground line of the coplanar waveguide to form a symmetrical return path and reduce high-frequency mode degradation.
[0027] In the embodiments, such as Figure 2 As shown, 2vias and 3vias represent the calibration structures with two and three through holes, respectively, while S 11 and S 21 Then S represents the calibration structure respectively 11 Parameters and S 21 Parameters. For the calibration structure, S 11 This is expressed as return loss, and the smaller the better, S 21 This is expressed as insertion loss, and the closer to 0dB, the better. (The rest of the text appears to be a continuation of the previous sentence and can be left as is.) Figure 2It is easy to see that the 2-vias structure performs better than the 3-vias structure.
[0028] In this embodiment, the diameter of the glass via is 20 μm, the center-to-center distance between two adjacent glass vias is 40 μm, and the minimum distance from the center of any glass via to the metal edge is 20 μm. The metal pad 6 is adapted to a 100 μm RF probe pitch and has a characteristic impedance of 50 Ω. This design of the glass via enables signal transmission in the operating frequency band from 110 GHz to 170 GHz.
[0029] The glass through-hole designed above, combined with the overall calibration structure, can achieve the measurement of the S-parameters of the glass through-hole. Specific measurement methods include: The wafer or chip is fixed on the probe stage (chuck) and flattened and stabilized using vacuum adsorption. Under a microscope, the CPW port pads of the structure under test are moved to the center of the field of view. A GSG RF probe with a spacing matched to the metal pads is aligned with the signal pads and the ground pads on both sides. The azimuth and pitch angles of the probe are adjusted to ensure good coplanarity between the probe tip and the metal pads. Then, a controlled downward pressure is applied to achieve stable contact. The vector network analyzer is connected to the GSG probe through a D-band spread spectrum module and a corresponding probe. Before testing, calibration is performed by moving the reference plane to the probe tip. The calibration structure is then further measured according to the subsequent testing method to obtain the S-parameters of the calibration structure under test, which can then be used for subsequent performance evaluation and parameter extraction.
[0030] Calibration Method 1 involves testing the scattering parameters of the calibration structure and indirectly obtaining the bulk scattering parameters of the glass aperture by setting a difference in the transmission length of the bottom metal plate. For example, the original transmission length of the bottom metal plate is set to L, denoted as the L structure, as shown in 3(a); a transmission line is set in the bottom metal plate at the lower part of the calibration structure to extend the transmission length to 2L, denoted as the 2L structure, as shown in... Figure 3 As shown in (b). In this method, the S-parameters are measured at transmission lengths L and 2L as described above. Then, the matrices of the two S-parameters are converted into transmission parameter matrices. The transmission parameter matrix (i.e., the ABCD matrix) is a mathematical tool used to describe the input-output relationship of a linear two-port network. After matrix conversion, the following relationship exists:
[0031]
[0032] in, For a transmission matrix of length L, The transmission matrix is of length 2L, and all of them are ABCD matrices. The bulk transmission parameter matrix of the glass through-hole is given. Based on the above two equations, the transmission parameters of the glass through-hole are calculated. :
[0033] Transforming the extracted via transmission parameter matrix into an S-parameter matrix yields the S-parameters of the via. 21 Parameters, such as Figure 4 As shown, this can be used to evaluate the role of vias in energy transmission, which is beneficial for subsequent optimization and improvement.
[0034] Calibration method two involves setting differences in the termination states of the coplanar waveguides to test the scattering parameters of the calibration structure and indirectly obtain the bulk scattering parameters of the glass via. For example... Figure 5 As shown in (a), the calibration structure is set to an open state by removing the signal lines from the bottom coplanar waveguide; as Figure 5 As shown in (b), the calibration structure is set to a short-circuit state by connecting the ground line and the signal line of the bottom coplanar waveguide. In this way, the S-parameters of the L-structure and the open-circuit and short-circuit states are measured as described above, and then converted into a Y-parameter matrix, yielding: Matrix1 𝑌 = Thru 𝑌 - Open 𝑌 Matrix2 𝑌 = Short 𝑌 - Open 𝑌 in, Thru 𝑌 , Open 𝑌 ,as well as Short 𝑌 These represent the Y-parameter matrices for the L-structure, open-circuit structure, and short-circuit structure, respectively. Matrix1 𝑌 and Matrix2 𝑌 Let each represent a Y-parameter matrix as an intermediate variable; then the two Y-parameter matrices are... Matrix1 𝑌 and Matrix2 𝑌 Convert to a Z-parameter matrix to obtain the Z-parameter matrix of the bottom coplanar waveguide transmission. :
[0035] The Z parameter matrix It can be transformed into an ABCD matrix. Based on this, the bulk transmission parameters of the glass through-hole can be obtained. :
[0036] in, The transmission parameter matrix of the L-structure is represented by the extracted through-hole transmission parameter matrix. Transforming it into an S-parameter matrix yields the S-parameters of the through-hole. 21 parameter.
[0037] The specific embodiments described above illustrate the technical solution and beneficial effects of the present invention in detail. It should be understood that the above description is only the most preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, additions, and equivalent substitutions made within the scope of the principles of the present invention should be included within the protection scope of the present invention.
Claims
1. A calibration structure for a D-band glass through-hole, characterized in that, include: Independent metal pads, one for each signal line and ground line, are used for RF probe contact; Bottom metal plate; At least one set of glass vias formed on a glass substrate for D-band signal transmission, so as to realize signal transmission between each metal pad and the bottom metal plate; The calibration structure is used to calibrate the scattering parameters of the glass via: by setting differences in the transmission length or termination state of the bottom metal plate, the scattering parameters of the calibration structure are tested and the bulk scattering parameters of the glass via are indirectly obtained.
2. The calibration structure for the D-band glass through-hole according to claim 1, characterized in that, The glass vias include signal via groups and ground via groups. Each signal via group corresponds to a metal pad of the corresponding signal line and includes two signal vias. Each ground via group includes a metal pad of the corresponding ground line on each side and includes two ground vias to form a symmetrical return path and reduce high-frequency mode degradation.
3. The calibration structure for the D-band glass through-hole according to claim 1, characterized in that, The diameter of the glass through-hole is 20-40 μm.
4. The calibration structure for the D-band glass through-hole according to claim 1, characterized in that, The center-to-center distance between two adjacent glass vias on each metal pad shall not be less than 40 μm.
5. The calibration structure for the D-band glass through-hole according to claim 1, characterized in that, The minimum distance from the center of any glass via to the metal pad and the edge of the bottom metal plate shall not be less than 20 μm.
6. The calibration structure for the D-band glass through-hole according to claim 1, characterized in that, The metal pads are adapted to 75-150μm RF probe spacing and have a characteristic impedance of 50Ω.
7. The calibration structure for the D-band glass through-hole according to claim 1, characterized in that, The scattering parameters of the calibration structure are tested and indirectly obtained by setting differences along the transmission length of the ground metal plate, including: The original transmission length of the bottom metal plate is set to L, denoted as structure L. The transmission length of the bottom metal plate is extended to 2L, denoted as structure 2L. The scattering parameters of the L-structure and the 2L-structure were tested using a vector network analyzer. The bulk scattering parameters of the glass aperture are calculated by converting the matrices of the two scattering parameters into the transmission parameter matrices.
8. The calibration structure for the D-band glass through-hole according to claim 1, characterized in that, The scattering parameters of the calibration structure are tested and calibrated by setting differences in the termination state at the bottom metal plate end, and the bulk scattering parameters of the glass via are indirectly obtained, including: The calibration structure can be placed in a through-circuit, open-circuit, and short-circuit state by changing the bottom metal plate. The scattering parameters of the calibration structure in the through, open, and short-circuit states were tested using a vector network analyzer. The bulk scattering parameters of the glass aperture are extracted based on the three scattering parameters.
9. The calibration structure for the D-band glass through-hole according to claim 1, characterized in that, Electrical parameters, including equivalent resistance, inductance, conductance, and capacitance, are extracted based on the bulk scattering parameters of the glass through-hole.
10. A calibration method for a D-band glass through-hole, characterized in that, The calibration method employs the calibration structure described in any one of claims 1-6, and the calibration method includes the following steps: The scattering parameters of the calibration structure are tested by setting differences in the transmission length or termination state of the bottom metal plate, and the bulk scattering parameters of the glass through-hole are indirectly obtained.