A method for stress buffering of chip double-bevel bonding interconnects
By constructing a double-level inclined hollow shallow cavity structure and layered wire bonding, the contradiction between high-speed signal transmission and thermal stress buffering in traditional chip packaging processes is resolved, achieving simultaneous improvement in high-frequency transmission of high-speed signals and thermal stress buffering, thereby enhancing the long-term service stability of the chip interconnect structure.
Patent Information
- Application Number
- CN202611257609.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-19
- Publication Date
- 2026-11-10
AI Technical Summary
Traditional chip packaging bonding processes struggle to balance high-speed signal transmission performance with thermal stress buffering reliability. Orthogonal wire bonding suffers from parasitic capacitance and inductance issues, while flip-chip bump bonding cannot release thermal stress, leading to solder joint cracking and delamination.
The chip double-bevel bonding interconnection method is adopted. By constructing a double-bevel hollow shallow cavity structure and layered differentiated lead bonding molding, a short-pitch bevel lead configuration is formed. Combined with high-precision cavity modeling and interference-free feed control of bonding tools, elastic deformation and stress buffering of the lead are realized.
It achieves the integrity of high-speed signal transmission and the reliability of thermal stress relief, improves the long-term service stability and high-frequency electrical performance of chip interconnect structure, and is suitable for packaging applications of high-end high-speed chips.
Smart Images

Figure CN123076298A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of semiconductor chip packaging technology, and in particular to a method for stress buffering of chip double-bevel bonding interconnects. Background Technology
[0002] Traditional chip packaging bonding processes are mainly divided into two mainstream forms: conventional wire bonding and flip-chip bump bonding. Both processes have inherent technical shortcomings in practical applications, making it difficult to balance high-speed signal transmission performance and thermal stress buffering reliability. Conventional conventional wire bonding relies on a large arch and long span lead structure to release deformation stress, resulting in relatively excellent fatigue resistance and reliability of the solder joints. However, the overall lead length is relatively long, leading to significant parasitic capacitance and inductance. Signal loss and crosstalk issues are prominent under high-frequency operating conditions, making it unsuitable for the packaging requirements of high-speed chips. Flip-chip bump bonding uses a short-pitch rigid interconnect structure, which can effectively reduce the signal transmission path and has good high-frequency transmission characteristics. However, the bump interconnect structure lacks elastic deformation capability, and the relative displacement caused by the difference in thermal expansion coefficients between the chip and the substrate cannot be effectively released. Under thermal cycling conditions, thermal stress continuously concentrates at the interconnect root, making it prone to solder joint cracking and delamination failure during long-term service. Existing improvement solutions mostly optimize single signal performance or single stress structure, failing to resolve the performance contradiction between high-speed transmission and stress buffering at the structural level. There is still considerable room for improvement in overall packaging adaptability and service stability. Summary of the Invention
[0003] The main objective of this invention is to provide a chip double-beveled inclined bonding interconnect stress buffering method. By constructing a double-beveled hollow shallow cavity assembly structure and matching it with layered differentiated lead bonding molding and parameter control processes, this method balances short-distance low parasitic signal transmission characteristics with lead elastic deformation capability, thereby achieving the goal of simultaneously improving the high-speed transmission and thermal stress relief reliability of the chip bonding interconnect structure.
[0004] To achieve the above objectives, the present invention provides a chip double-bevel bonding interconnect stress buffering method, comprising the following steps:
[0005] Fix the chip face up to the preset mounting area of the carrier substrate;
[0006] The pre-fabricated sloping substrate with a hollowed-out slot is aligned and fastened to the upper part of the support base. The inner wall of the hollowed-out slot has a continuously connected lower gentle slope and an upper steep slope. The two slopes are respectively provided with substrate pads corresponding to different signal attributes. The inner wall of the hollowed-out slot and the support base form a hollow shallow cavity with an open top. The chip pads are lower than all substrate pads in the height direction.
[0007] Based on the cavity geometry parameters of the hollow shallow cavity, the spatial positional relationship between the chip pads and the substrate pads of each layer, and taking the shortest connection between the paired chip pads and the corresponding substrate pads as the benchmark, the motion path and bonding sequence of the bonding tool are planned.
[0008] The control bonding tool extends into the hollow shallow cavity and sequentially connects each chip pad to the substrate pad on the corresponding layer slope surface through wire bonding according to the planned motion path and bonding sequence, forming a short-pitch oblique lead configuration. The lead configuration is used to absorb the relative displacement between the chip and the oblique substrate due to the difference in thermal expansion coefficients through its own elastic deformation.
[0009] Furthermore, the step of fixing the chip face up to the preset mounting area of the carrier substrate includes:
[0010] Full-area alignment reference marks and chip limiting contours are processed in the preset mounting area of the support substrate;
[0011] Obtain global alignment reference marks, establish a unified spatial coordinate system for the base, and determine the horizontal boundary for chip placement based on the chip's positioning contour.
[0012] Pick up the chip using a pickup device, with the side of the chip with metal pads as the front, calibrate the horizontal coordinates and angles against the global alignment reference marks, and place the chip in the installation area enclosed by the chip's limiting contour.
[0013] The back side of the chip is fixed to the carrier substrate by a retaining medium.
[0014] Furthermore, before the step of aligning and fastening the pre-fabricated beveled substrate with the hollowed-out groove onto the support base, the hollowed-out groove and double-step bevel of the beveled substrate are pre-processed, including:
[0015] Select an insulating substrate and design a continuous lower slope and upper steep slope profile based on the preset height difference between the chip pads and the substrate pads.
[0016] A hollowed-out groove is processed in the middle of the substrate using a substrate molding process, and a continuous double-step inclined structure is simultaneously formed on the inner wall of the groove.
[0017] According to the signal rate classification rules, high-speed signal substrate pads are arranged on the lower gentle slope, and power and low-speed signal substrate pads are arranged on the upper steep slope. The substrate to be fastened is obtained by metallizing the pads.
[0018] Furthermore, the step of aligning and fastening the pre-fabricated beveled substrate with the hollowed-out groove onto the support base includes:
[0019] Obtain the positioning mark at the edge of the hollowed-out groove on the inclined substrate and establish a coordinate mapping relationship with the global alignment reference mark on the supporting substrate;
[0020] Based on the coordinate mapping relationship, the inclined substrate is moved along the vertical direction perpendicular to the supporting base. The difference of the reference coordinates during the movement is compared in real time, and the horizontal position and angle of the inclined substrate are adjusted to calibrate the horizontal orientation and angle.
[0021] After alignment and calibration, the inclined substrate is lowered vertically so that the cutout completely covers the mounting area where the chip is located. The inclined substrate is attached and fixed to the edge of the supporting substrate to form the open-top cutout shallow cavity.
[0022] Furthermore, based on the shortest spatial connection between the paired chip pads and the corresponding substrate pads, the steps for planning the motion path and bonding sequence of the bonding tool include:
[0023] Establish a three-dimensional geometric model of the hollow shallow cavity and a solid contour model of the bonding tool. Combine the equipment positioning error and the swing amplitude of the lead wire to set the safety margin for obstacle avoidance of the entire cavity.
[0024] Based on a unified spatial coordinate system, a one-to-one electrical pairing relationship between chip pads and substrate pads at each level is established, and the shortest spatial connection between each pair of pads is extracted as the basic routing of the lead.
[0025] The cavity obstacle avoidance safety margin is superimposed as a constraint condition onto the perimeter of the shortest connection in space to form a path avoidance boundary. Combined with the kinematic constraints of the bonding tool, a smooth, collision-free continuous motion trajectory is generated within the path avoidance boundary. Combined with the processing logic of the layered pads, the complete bonding sequence is output.
[0026] Furthermore, after generating a collision-free continuous motion trajectory, the steps of verifying and correcting the trajectory include:
[0027] Discrete spatial points are extracted along the continuous motion trajectory, and the real-time gap value between the bonding tool profile and the double-order inclined cavity wall is calculated point by point.
[0028] By comparing the gap value at each point with the cavity obstacle avoidance safety margin, interference risk points with gaps smaller than the safety margin are identified.
[0029] The trajectory coordinates of the interference risk points are locally offset, and a corrected trajectory with a smooth transition is regenerated until the gaps between all points in the entire trajectory meet the obstacle avoidance safety margin requirements.
[0030] Furthermore, the step of controlling the bonding tool to extend into the hollowed-out shallow cavity includes:
[0031] The drive bonding tool is fed vertically downward from the top opening of the hollow shallow cavity along the preset cavity entry trajectory, and the tool axis is kept in a vertical position throughout the feeding process;
[0032] During the feeding process, the three-dimensional spatial coordinates of the bonding tool are acquired in real time, and the position of the inclined surface of the cavity wall is compared synchronously to maintain a safe clearance between the tool and the cavity.
[0033] Once the bonding tool reaches the preset height directly above the target chip pad, it switches to a low-speed feed mode to descend until the lead tip contacts the chip pad surface.
[0034] Further, the step of sequentially connecting each chip pad to the substrate pad on the corresponding layer's inclined surface via wire bonding, according to the planned motion path and bonding sequence, includes:
[0035] Select the high-speed signal substrate pads arranged on the lower gentle slope and connect them with the corresponding chip pads to form lead connections;
[0036] After completing all the lead wire connections on the lower gentle slope, adjust the travel range of the bonding tool and switch the work area to the upper steep slope.
[0037] Select the power supply and low-speed signal substrate pads arranged on the upper steep slope and connect them with the corresponding chip pads to form lead wires;
[0038] After all lead wire connections are completed, cut the lead wire tails.
[0039] Further, the step of selecting the substrate pads on the lower gentle slope and the upper steep slope and constructing lead connections with the corresponding chip pads includes:
[0040] For the high-speed signal pads on the lower gentle slope, the first lead arch height parameter is set so that the lead forms the shortest path shape after it is formed.
[0041] For the power and low-speed signal pads on the upper steep slope, set the second lead arch height parameter and reserve the lead elastic deformation margin.
[0042] After the single-level lead connection is constructed, the lead morphology parameters are collected, and the deviation between the lead morphology parameters and the preset morphology control threshold is compared.
[0043] Furthermore, the step of selecting the substrate pads on the lower gentle slope and the upper steep slope and constructing lead connections with the corresponding chip pads also includes:
[0044] Configure the first set of bonding feed rate and bonding pressure process parameters based on the inclination angle of the lower gentle slope surface.
[0045] The second set of bonding feed rate and bonding pressure process parameters are configured based on the inclination angle of the upper steep slope.
[0046] When constructing lead connections, the corresponding level of process parameters is called to carry out bonding operations, matching the lead forming process under different slope conditions.
[0047] The chip double-bevel bonding interconnect stress buffering method provided by this invention has the following beneficial effects:
[0048] This invention is based on a double-beveled substrate hollow shallow cavity molding structure, constructing a spatial height difference layout between low-position chip pads and high-position layered substrate pads, forming differentiated short-pitch beveled lead configurations, effectively overcoming the performance trade-offs of traditional bonding processes. The layered double-beveled arrangement enables independent interconnection of high-speed signals and power supplies, as well as low-speed signals. The short-pitch beveled lead structure significantly reduces signal transmission paths, lowers packaging parasitic parameters, and ensures the integrity of high-frequency signal transmission. Simultaneously, the beveled leads possess controllable elastic deformation margins, absorbing relative displacement due to thermal mismatch between the chip and substrate under hot and cold conditions, mitigating stress concentration at the interconnect root. Through high-precision cavity modeling trajectory planning, interference-free cavity feed control, and differentiated bonding process parameter adjustment adapted to the bevel angle, stress concentration problems caused by assembly deviations and process defects are effectively avoided, ensuring regular lead molding morphology and uniform deformation stress across different layers. While maintaining high-frequency electrical performance, it improves the thermal cycling reliability and long-term service stability of the chip interconnect structure, adapting to long-term packaging applications of high-end high-speed chips. Attached Figure Description
[0049] Figure 1 This is a schematic flowchart of a chip double-bevel bonding interconnect stress buffering method in one embodiment of the present invention;
[0050] Figure 2 This is a schematic diagram of the double-beveled shallow cavity elevation structure of the chip double-beveled bonding interconnect stress buffering method in one embodiment of the present invention;
[0051] Figure 3 This is a schematic diagram of the side elevation structure of the double-beveled shallow cavity of the chip double-beveled bonding interconnect stress buffering method in one embodiment of the present invention.
[0052] Figure 4 This is a three-dimensional assembly schematic diagram of a chip double-bevel bonding interconnect stress buffering method in one embodiment of the present invention.
[0053] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation
[0054] 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 and not intended to limit the invention.
[0055] Reference Figure 1The diagram below illustrates a process for a chip double-bevel bonding interconnect stress buffering method proposed in this invention, comprising the following steps:
[0056] S1, fix the chip face up to the preset mounting area of the carrier substrate;
[0057] S2, the pre-made sloping substrate with a hollowed-out slot is aligned and fastened to the upper part of the support base. The inner wall of the hollowed-out slot has a continuously connected lower gentle slope and an upper steep slope. The two slopes are respectively provided with substrate pads corresponding to different signal attributes. The inner wall of the hollowed-out slot and the support base form a hollow shallow cavity with an open top. The chip pads are lower than all substrate pads in the height direction.
[0058] S3. Based on the cavity geometry parameters of the hollow shallow cavity and the spatial positional relationship between the chip pads and the substrate pads of each layer, the motion path and bonding sequence of the bonding tool are planned with the shortest connection between the paired chip pads and the corresponding substrate pads as the reference.
[0059] S4, the bonding tool is inserted into the hollow shallow cavity and, according to the planned motion path and bonding sequence, the chip pads are sequentially connected to the substrate pads on the corresponding inclined surface by wire bonding to form a short-pitch inclined lead configuration. The lead configuration is used to absorb the relative displacement when the chip and the inclined substrate are relatively displaced due to the difference in thermal expansion coefficients.
[0060] As described in step S1 above, the chip is fixed face up in the preset mounting area of the carrier substrate. This operation completes the basic positioning and mechanical holding of the chip to establish a unified spatial reference for subsequent assembly processes. The front of the chip with metal pads is open to the top, ensuring that the chip pads have the space to achieve lead interconnection with the upper inclined substrate. The carrier substrate provides a stable support carrier for the chip and also serves as a component of the subsequent hollow shallow cavity molding.
[0061] As described in step S2 above, the pre-fabricated beveled substrate with hollowed-out slots is aligned and fastened to the support base. Relying on the layered structure of the lower gentle slope and the upper steep slope of the continuously connected inner wall of the hollowed-out slot of the beveled substrate, the partitioned layout of substrate pads with different signal attributes is realized. The inner wall of the hollowed-out slot and the support base together form a hollowed-out shallow cavity with an open top. The chip pads are lower than all substrate pads in the height direction, creating a spatial height difference structure of low-position chip pads facing high-position layered substrate pads. This spatial layout provides the necessary geometric basis for the construction of the beveled lead configuration. The hollowed-out shallow cavity also forms an internal operating space for bonding tools to enter and operate.
[0062] As described in step S3 above, based on the cavity geometry parameters of the hollow shallow cavity and the spatial positional relationship between the chip pads and the substrate pads of each layer, the shortest connection between the paired chip pads and the corresponding substrate pads is used as the reference to plan the movement path and bonding sequence of the bonding tool. This process is based on the cavity space constraints and the electrical connection requirements of one-to-one pad pairing to carry out path pre-planning, taking into account the lead wire routing reference and the internal working space limitations of the cavity, avoiding spatial interference risks in advance, determining a reasonable order of operations, and ensuring that the subsequent lead wire bonding process is carried out in an orderly manner.
[0063] As described in step S4 above, the bonding tool is inserted into the hollow shallow cavity and, according to the planned motion path and bonding sequence, sequentially connects each chip pad to the substrate pad on the corresponding inclined surface through wire bonding, thereby forming a short-pitch inclined lead configuration. This lead configuration is used to absorb the relative displacement between the chip and the inclined substrate due to the difference in thermal expansion coefficients through its own elastic deformation. The short-pitch inclined lead configuration is formed based on the high and low spatial layout established in S2. Unlike the traditional planar lead architecture, the lead does not need to be set with a large-span arch. While meeting the electrical interconnection requirements, it relies on the elastic deformation capability of the lead body to accommodate the relative displacement caused by the inconsistent deformation of the two components under cold and hot conditions, and achieves a stress buffering effect by relying on the structural characteristics itself.
[0064] In one embodiment, step S1, which involves fixing the chip face up to a predetermined mounting area on the carrier substrate, includes:
[0065] Full-area alignment reference marks and chip limiting contours are processed in the preset mounting area of the support substrate;
[0066] Obtain global alignment reference marks, establish a unified spatial coordinate system for the base, and determine the horizontal boundary for chip placement based on the chip's positioning contour.
[0067] Pick up the chip using a pickup device, with the side of the chip with metal pads as the front, calibrate the horizontal coordinates and angles against the global alignment reference marks, and place the chip in the installation area enclosed by the chip's limiting contour.
[0068] The back side of the chip is fixed to the carrier substrate by a retaining medium.
[0069] Reference Figure 2 , Figure 3These figures show schematic diagrams of the single-stage inclined shallow cavity front and side front structures for semiconductor packaging proposed in this invention. In practical implementation, conventional chip bonding processes are mainly divided into two categories: upright wire bonding and flip-chip bump bonding. Neither of these conventional structures can simultaneously meet the dual requirements of high-frequency transmission and stress resistance reliability. Traditional upright wire bonding uses a large arch and long-span lead structure, which can buffer thermal mismatch stress through lead deformation, resulting in high solder joint reliability. However, the lead length is relatively large, leading to prominent parasitic parameters and making it difficult to adapt to high-speed signal transmission scenarios. Traditional flip-chip bump bonding uses a rigid short-pitch interconnect structure, offering short paths and excellent high-frequency transmission performance. However, the bump structure lacks elastic deformation capability, and the deformation displacement caused by the difference in thermal expansion coefficients between the chip and the substrate cannot be released. Under thermal cycling conditions, stress continuously concentrates at the root of the solder joint, easily inducing fatigue cracking. To address this, the present invention employs a sunken shallow cavity combined with a double-slope structure design. By constructing a spatial height difference layout of low-position chip pads and high-position layered sloped pads, a short-pitch oblique elastic lead configuration can be formed. This retains the high-frequency transmission advantages of short-pitch interconnects while also providing stress buffering capabilities for lead elastic deformation. Unlike traditional planar bonding, which has a high tolerance for chip assembly precision, the double-sloped layered bonding and short-pitch oblique lead configuration of the present invention are highly sensitive to chip assembly precision. Even a slight horizontal offset or angular deviation of the chip will directly change the preset spatial height difference and relative orientation between the chip pads and the substrate pads of the lower gentle slope and upper steep slope, thereby causing the short-pitch lead routing, arch height, and tilt angle to deviate from the design state. This directly destroys the uniformity of lead elastic deformation and stress buffering adaptability, thus requiring a high-precision assembly process in this step. Specifically: A global alignment reference mark and a chip limiting contour are processed in the preset mounting area of the substrate. Compared to the traditional simple positioning method that relies only on single-sided alignment, the global alignment reference mark set in this invention can match the global coordinate alignment requirements of the subsequent overall snap-fit of the inclined substrate, adapting to the precise matching characteristics of the large-scale, multi-region layered pads of the double-level inclined surface, and constructing a unified coordinate calibration system for the entire packaging and assembly process. The chip limiting contour matches the outer contour size of the chip, providing full circumferential physical limitation on the chip's horizontal position, adapting to the structural characteristics of this technology—short-pitch leads with no fault tolerance and sensitivity to small deviations—and avoiding slight chip offsets. After the structural processing is completed, the equipment identifies the global alignment reference mark, establishes a unified spatial coordinate system for the substrate, and accurately defines the standard mounting range and horizontal boundary of the chip in conjunction with the chip limiting contour, ensuring that each set of chip pads maintains a preset spatial relative position with the subsequent upper and lower inclined surface pads. The chip is then picked up by the pickup device, with the functional surface where the metal pads are laid out as the front of the chip, ensuring that the chip pads face upwards. This matches the height difference structure design of the chip pads being arranged at a low position and the inclined pads being arranged at a high position in this invention. The horizontal coordinates and placement angle of the chip are calibrated by referring to the global alignment reference mark, and the chip is accurately placed in the installation area enclosed by the chip limit contour.After assembly and positioning, the connection area between the chip's back side and the substrate is solidified using a holding medium to achieve a stable, integrated connection between the chip and the substrate, preventing chip position drift during subsequent shallow cavity molding and intracavity bonding processes. This step, through global reference calibration, full-circumferential physical limiting, high-precision coordinate angle alignment, and integrated stable holding, ensures a constant preset spatial position relationship between the chip pads and the pads on each level of the inclined substrate. This avoids lead configuration distortion and uneven deformation stress caused by assembly deviations, ensuring that subsequent short-pitch inclined leads can be stably formed according to the design, and ensuring that the elastic deformation of the leads can uniformly absorb thermal mismatch displacement.
[0070] In one embodiment, before the step of aligning and fastening the pre-formed beveled substrate with a hollowed-out groove onto the support substrate, the hollowed-out groove and the double-step bevel of the beveled substrate are pre-processed, including:
[0071] Select an insulating substrate and design a continuous lower slope and upper steep slope profile based on the preset height difference between the chip pads and the substrate pads.
[0072] A hollowed-out groove is processed in the middle of the substrate using a substrate molding process, and a continuous double-step inclined structure is simultaneously formed on the inner wall of the groove.
[0073] According to the signal rate classification rules, high-speed signal substrate pads are arranged on the lower gentle slope, and power and low-speed signal substrate pads are arranged on the upper steep slope. The substrate to be fastened is obtained by metallizing the pads.
[0074] In practical implementation, an insulating substrate is selected as the overall structural carrier. The substrate has insulation performance and structural rigidity that are compatible with conventional packaging. Based on the chip pad layout height, target lead slant span, and stress deformation reserved space parameters, the preset height difference between the chip pad and the substrate pad is determined. Based on this fixed height difference parameter, the contours of the continuously connected lower gentle slope and upper steep slope are designed in a directional model. The tilt angle, vertical height, and horizontal coverage of the two slopes are matched with the wiring requirements and lead deformation design requirements of the corresponding layer pads. The appropriate slope conditions for high-speed signals, power supplies, and low-speed signals are distinguished from the structural source. After determining the contour parameters of the double-slope surface, the central area of the insulating substrate is integrally formed using conventional substrate molding technology. A hollowed-out groove is integrally formed in the central area of the substrate. During the processing, a continuous lower gentle slope and an upper steep slope are integrally formed on the inner sidewall of the hollowed-out groove. The entire process adopts an integral molding method to avoid coaxiality deviation, step misalignment, and slope angle error caused by step processing, thus ensuring the positional accuracy, dimensional accuracy, and structural continuity of the groove contour and the double-slope surface structure. After completing the processing of the slot and double-sloped structure, the pads are partitioned and laid out according to the signal rate classification rules. Based on the requirements of low loss, low crosstalk, and adaptation to small elastic deformation for high-speed signals, high-speed signal substrate pads are laid out on the lower, gentler slope with a larger deformation adaptation space and smoother traces. Considering the high load-bearing capacity and large deformation tolerance range for power and low-speed signals, power and low-speed signal substrate pads are laid out on the upper, steeper slope with stronger structural rigidity and greater vertical height, achieving independent physical partitioning of signals with different attributes. Finally, through a standardized pad metallization process, metal layer deposition, patterning, and curing are completed at the preset pad positions on each level of the slope, forming layered substrate pads with stable conductivity, regular dimensions, and precise positioning. The final result is a sloped substrate ready for assembly, with satisfactory structural forming accuracy, clear pad partitioning, and direct alignment.
[0075] In one embodiment, step S2, which involves aligning and fastening the pre-formed beveled substrate with the cutout into the support substrate, includes:
[0076] Obtain the positioning mark at the edge of the hollowed-out groove on the inclined substrate and establish a coordinate mapping relationship with the global alignment reference mark on the supporting substrate;
[0077] Based on the coordinate mapping relationship, the inclined substrate is moved along the vertical direction perpendicular to the supporting base. The difference of the reference coordinates during the movement is compared in real time, and the horizontal position and angle of the inclined substrate are adjusted to calibrate the horizontal orientation and angle.
[0078] After alignment and calibration, the inclined substrate is lowered vertically so that the cutout completely covers the mounting area where the chip is located. The inclined substrate is attached and fixed to the edge of the supporting substrate to form the open-top cutout shallow cavity.
[0079] In practical implementation, after completing the aforementioned prefabrication process of the inclined substrate, the cavity structure is formed through precise alignment and assembly, providing a spatial foundation for subsequent intracavity bonding, short-distance oblique lead forming, and stress buffer structure stability. Specifically: the positioning marks pre-set on the edge of the cutout of the inclined substrate are identified and collected, while the global alignment reference marks already formed on the surface of the supporting substrate are read. Using the unified spatial coordinate system of the supporting substrate as a reference, the coordinate data of the edge positioning marks of the inclined substrate are matched, converted, and bound one by one with the global reference coordinate data of the substrate, establishing a unique coordinate mapping relationship between the two sets of structures. This allows all inclined structures, layered pads, and slot contours of the inclined substrate to reuse the unified coordinate system of the substrate, solving the problem of inconsistent alignment references and accumulated local deviations in multiple areas of the two-level layered structure. After establishing the coordinate mapping relationship, the control transfer mechanism smoothly moves the inclined substrate along a purely vertical direction perpendicular to the support substrate, with the horizontal degree of freedom limited throughout the process. During the transfer, the equipment collects the coordinate difference between the two sets of references in real time at high frequency, capturing the horizontal offset and angular skew, and provides real-time feedback to adjust the placement posture of the substrate, continuously correcting the horizontal position and placement angle of the inclined substrate until the horizontal orientation deviation and rotation angle deviation both fall within the preset high-precision threshold range, completing the two-dimensional alignment calibration and ensuring that the overall posture of the inclined substrate is completely parallel and aligned with the support substrate. After the alignment calibration is completed and the coordinate state is stable, the inclined substrate is lowered at a uniform speed while maintaining a vertical and unbiased posture, so that the hollow slot in the middle of the substrate completely covers the entire preset installation area where the chip is located below, ensuring that the chip and the inclined pads of each layer are completely within the coverage area of the slot, without obstruction or misalignment. Finally, the bottom edge of the inclined substrate is tightly attached and fixed to the upper surface edge of the support substrate. After the two structures are closed, the hollow slot sidewall and the surface of the support substrate together form a hollow shallow cavity that is open at the top and closed in the circumference. The assembly method of this invention locks the fixed height difference between the low-position pads of the chip and the high-position layered pads of the inclined substrate, regulates the bonding operation space in the cavity, provides spatial constraints for the subsequent interference-free trajectory planning of the bonding tool, and also ensures the forming accuracy of the inclined leads from the assembly level, so as to ensure that the subsequent elastic deformation of the leads can stably adapt to the thermal mismatch displacement and achieve a continuous and effective stress buffering effect.
[0080] In one embodiment, step S3, which involves planning the motion path and bonding sequence of the bonding tool based on the shortest connection between the paired chip pads and the corresponding substrate pads, includes:
[0081] Establish a three-dimensional geometric model of the hollow shallow cavity and a solid contour model of the bonding tool. Combine the equipment positioning error and the swing amplitude of the lead wire to set the safety margin for obstacle avoidance of the entire cavity.
[0082] Based on a unified spatial coordinate system, a one-to-one electrical pairing relationship between chip pads and substrate pads at each level is established, and the shortest spatial connection between each pair of pads is extracted as the basic routing of the lead.
[0083] The cavity obstacle avoidance safety margin is superimposed as a constraint condition onto the perimeter of the shortest connection in space to form a path avoidance boundary. Combined with the kinematic constraints of the bonding tool, a smooth, collision-free continuous motion trajectory is generated within the path avoidance boundary. Combined with the processing logic of the layered pads, the complete bonding sequence is output.
[0084] In practical implementation, the bonding operation of this invention is limited to the interior of a shallow, hollow cavity with confined space. The cavity contains irregularly shaped inner walls formed by a lower gentle slope and an upper steep slope. During feeding and turning movements, the bonding tool is highly susceptible to contact and interference with the cavity walls. Furthermore, this solution employs a layered wire bonding process, with different layers of pads distributed on slopes of varying heights. An disordered bonding sequence can lead to wire crossings and scraping against the cavity slopes, disrupting the formed oblique wire configuration and weakening stress buffering capacity. Therefore, pre-planning of the motion path is necessary before the actual bonding operation begins. First, a three-dimensional geometric model of the cavity is constructed based on the actual dimensions of the shallow cavity obtained from the initial assembly. Simultaneously, the dimensions of the bonding tool are imported to establish a corresponding solid contour model. Taking into account the unavoidable positioning deviations during equipment operation and the swing range of the wires caused by airflow and tension during the bonding process, a safety margin for obstacle avoidance across the entire cavity is set. This margin serves as a hard constraint lower limit to prevent contact between the tool body, wires, and cavity walls under dynamic operating conditions. Using a unified spatial coordinate system for the substrate, the chip pad coordinate data and the coordinate data of each substrate pad on the layered slope are read. According to electrical interconnection requirements, chip pads are paired one-to-one with corresponding layer substrate pads. For each pairing, the shortest spatial connection between the interconnection points is calculated, and this connection is used as the basic direction for lead forming, ensuring that the final formed lead has short-distance characteristics. The obstacle avoidance safety margin of the entire cavity is then expanded outwards to the periphery of the shortest spatial connection, forming a closed path avoidance boundary. Simultaneously, kinematic constraints such as the bonding tool's own speed and turning radius are introduced. Within the safety space defined by the path avoidance boundary, a smooth, collision-free continuous motion trajectory is generated. Combining the layout characteristics of the pads layered on the lower gentle slope and the upper steep slope, the execution order of the points is arranged according to the bottom-up layer processing logic, outputting a complete bonding sequence. This achieves orderly bonding of layered pads, preventing the upper layer lead from blocking the lower layer bonding action.
[0085] In one embodiment, the step of verifying and correcting the trajectory after generating a collision-free continuous motion trajectory includes:
[0086] Discrete spatial points are extracted along the continuous motion trajectory, and the real-time gap value between the bonding tool profile and the double-order inclined cavity wall is calculated point by point.
[0087] By comparing the gap value at each point with the cavity obstacle avoidance safety margin, interference risk points with gaps smaller than the safety margin are identified.
[0088] The trajectory coordinates of the interference risk points are locally offset, and a corrected trajectory with a smooth transition is regenerated until the gaps between all points in the entire trajectory meet the obstacle avoidance safety margin requirements.
[0089] In practical implementation, the motion trajectory directly generated by the theoretical model is calculated based solely on static geometric dimensions, without fully considering the cumulative dimensional deviations caused by model construction and assembly. Therefore, the theoretical trajectory still carries the potential risk of insufficient gaps and interference in local areas. Thus, verification and local correction operations are necessary before the trajectory is officially used. A series of discrete spatial points are uniformly extracted along the aforementioned continuous motion trajectory. Each discrete point corresponds to a set of spatial postures during the bonding tool's movement. The real-time gap values between the bonding tool's solid contour and the double-sided inclined cavity wall are calculated sequentially under each posture. The gap value calculated at each point is compared and filtered against a pre-set cavity obstacle avoidance safety margin to locate all interference risk points where the gap value is less than the safety margin threshold. For the identified interference risk points, the trajectory coordinates are locally offset and adjusted without disrupting the lead wire foundation or exceeding the overall obstacle avoidance boundary. The adjusted trajectory curve is then regenerated based on smooth transition constraints. Point gap verification and local trajectory correction operations are continuously performed iteratively until the gap values corresponding to all discrete points along the entire trajectory meet the cavity obstacle avoidance safety margin requirements, resulting in the final motion trajectory that can be directly used for equipment execution. The verified and corrected trajectory can adapt to the spatial constraints of the hollowed-out shallow cavity's irregular inner wall, ensuring that the bonding tool travels through the cavity without touching it, thus guaranteeing the stable implementation of the layered lead wire bonding process.
[0090] In one embodiment, step S4, controlling the bonding tool to extend into the hollowed-out shallow cavity, includes:
[0091] The drive bonding tool is fed vertically downward from the top opening of the hollow shallow cavity along the preset cavity entry trajectory, and the tool axis is kept in a vertical position throughout the feeding process;
[0092] During the feeding process, the three-dimensional spatial coordinates of the bonding tool are acquired in real time, and the position of the inclined surface of the cavity wall is compared synchronously to maintain a safe clearance between the tool and the cavity.
[0093] Once the bonding tool reaches the preset height directly above the target chip pad, it switches to a low-speed feed mode to descend until the lead tip contacts the chip pad surface.
[0094] In practical implementation, the bonding operation of this invention is entirely confined within a shallow, hollow cavity with a top opening, limited space, and double-sloped inner walls. Compared to traditional open bonding environments, the cavity lacks sufficient margin for error, and tool tilting, deviation, or excessive speed can easily cause the tool to scrape against the sloping cavity walls, resulting in vibration and deviation, which in turn leads to problems such as wire forming deviation, weak solder joints, and stress structure failure. Therefore, this solution employs a cavity entry control logic with constant posture, real-time coordinate monitoring, and graded fast and slow feed to achieve safe and precise cavity entry positioning of the bonding tool. Specifically, the pre-planned cavity entry trajectory is invoked, driving the bonding tool to feed vertically downwards from the top center opening of the shallow hollow cavity. Throughout the entire cavity entry feed stroke, the equipment posture control system continuously locks the angular freedom of the bonding tool, always maintaining the tool axis in an absolutely vertical posture, completely avoiding the risk of the tool tilting and touching the double-sloped cavity walls on both sides. Throughout the continuous downward feed process, the three-dimensional spatial coordinate data of the bonding tool is acquired in real time, and the cavity boundary coordinates of the lower gentle slope and upper steep slope of the hollowed-out shallow cavity are retrieved simultaneously. The relative position of the tool point with the cavity wall slope is compared in real time, and the tool's trajectory is dynamically fine-tuned to maintain a preset safe clearance between the tool and the irregular inner wall of the cavity, avoiding potential interference during dynamic feed. When the bonding tool descends to the preset safe height position directly above the target chip pad, the equipment automatically switches the feed control mode from high-speed rapid feed to low-speed precision feed mode, continuing to descend in a stable manner with low displacement and low vibration until the lead end precisely contacts the chip pad surface, completing the precise alignment before bonding and providing a precise workstation foundation for subsequent stable and deviation-free lead forming operations.
[0095] In one embodiment, the step of sequentially connecting each chip pad to the substrate pad on the corresponding layer slope via wire bonding according to the planned motion path and bonding sequence includes:
[0096] Select the high-speed signal substrate pads arranged on the lower gentle slope and connect them with the corresponding chip pads to form lead connections;
[0097] After completing all the lead wire connections on the lower gentle slope, adjust the travel range of the bonding tool and switch the work area to the upper steep slope.
[0098] Select the power supply and low-speed signal substrate pads arranged on the upper steep slope and connect them with the corresponding chip pads to form lead wires;
[0099] After all lead wire connections are completed, cut the lead wire tails.
[0100] Reference Figure 4A three-dimensional assembly schematic diagram of a chip double-bevel bonding interconnect stress buffering method in one embodiment of the present invention is shown. The diagram illustrates the substrate, wafer, and the formed beveled lead structure. The beveled leads have a 45° spatial orientation, which intuitively shows the overall spatial configuration of the short-distance beveled leads between the low-position wafer pad and the high-position beveled substrate pad.
[0101] In practical implementation, based on the structural characteristics of the double-slope layered structure and the partitioned layout of high-speed signals and low-speed power signals, this invention adopts a bottom-up layered sequential bonding logic to avoid the problems of cross-interference between different levels of leads and secondary scraping and deformation of pre-formed leads. This ensures that each inclined lead can be independently formed according to the design configuration, guaranteeing a regular and uniform overall stress buffer structure. In specific implementation, the working area of the bonding tool is first locked to the lower gentle slope. All high-speed signal substrate pads on the lower gentle slope are selected sequentially, and wire bonding is completed one by one with the corresponding chip pads, constructing a short-distance inclined lead structure from the low-position chip pad to the high-position pad on the gentle slope. After all the high-speed signal leads in the lower layer are connected and formed, the equipment recalibrates the working boundary, adaptively adjusts the maximum travel range of the bonding tool, and switches the entire working area to the upper steep slope area, avoiding the obstruction and interference of the pre-formed leads in the lower layer on the upper working trajectory. Subsequently, power supply and low-speed signal substrate pads arranged on the upper steep slope are selected group by group and wire-bonded to the corresponding chip pads to form a sloping lead structure adapted to power supply and low-speed signal operating conditions. After all layered and zoned pad wire bonding operations are completed, the equipment performs a unified tail wire cutting process to remove excess lead wire residue, completing the orderly shaping of the overall lead interconnection structure. This layered operation logic is fully adapted to the characteristics of the double-stage sloping cavity structure of this invention, and can orderly complete the construction of differentiated lead configurations, laying the structural foundation for matching the deformation buffering capacity of different levels of leads.
[0102] In one embodiment, the step of selecting the lower gently sloping substrate pads and the upper steeply sloping substrate pads to construct lead connections with the corresponding chip pads includes:
[0103] For the high-speed signal pads on the lower gentle slope, the first lead arch height parameter is set so that the lead forms the shortest path shape after it is formed.
[0104] For the power and low-speed signal pads on the upper steep slope, set the second lead arch height parameter and reserve the lead elastic deformation margin.
[0105] After the single-level lead connection is constructed, the lead morphology parameters are collected, and the deviation between the lead morphology parameters and the preset morphology control threshold is compared.
[0106] In practical implementation, due to the significant differences in inclination angle, vertical height, and spatial span between the lower gentle slope and the upper steep slope of this invention, a single lead arch height parameter cannot adapt to the differentiated spatial structure of the two-stage slope. This can lead to problems such as excessively large high-speed signal lead arch height introducing parasitic parameters and insufficient deformation margin of power supply leads, resulting in unrelieved stress. Therefore, this solution configures independent lead arch height parameters for each of the two slope stages to achieve bidirectional matching of signal performance and stress buffering performance. Specifically, for the high-speed signal pads on the lower gentle slope, a preset first lead arch height parameter is used for bonding. Based on the structural characteristics of the gentle slope—small span and gentle inclination angle—the lead arch height is constrained to an extremely low range, ensuring that the formed lead strictly conforms to the shortest spatial path shape, minimizing the signal transmission path, reducing parasitic capacitance and inductance, and ensuring the integrity of high-speed signal transmission. For the power and low-speed signal pads on the upper steep slope, the equipment calls upon a preset second lead arch height parameter. This parameter is greater than the first lead arch height parameter. Utilizing the structural advantage of ample vertical space on the steep slope, a larger margin for lead elastic deformation is proactively reserved, giving the power and low-speed signal leads stronger thermal displacement adaptability and buffering large-size thermal mismatch displacement under conditions of large temperature differences. After each single-level lead bonding operation is completed, the arch height, curvature, and orientation parameters of the formed leads are collected in real time. The measured morphological parameters are compared with the system's preset level-specific morphological control thresholds to identify out-of-tolerance deformed leads, ensuring that the forming state of each level of leads matches the design requirements of the corresponding level, and guaranteeing the uniformity and stability of stress buffering effect from a morphological perspective.
[0107] In one embodiment, the step of selecting the substrate pads of the lower gentle slope and the upper steep slope and constructing lead connections with the corresponding chip pads further includes:
[0108] Configure the first set of bonding feed rate and bonding pressure process parameters based on the inclination angle of the lower gentle slope surface.
[0109] The second set of bonding feed rate and bonding pressure process parameters are configured based on the inclination angle of the upper steep slope.
[0110] When constructing lead connections, the corresponding level of process parameters is called to carry out bonding operations, matching the lead forming process under different slope conditions.
[0111] In practice, the two-stage inclined planes exhibit differences in their tilt angle structure. The stress state of the bonding contact surface and the resistance to lead wire debonding are completely different for inclined planes with different tilt angles. If the same wire feed rate and bonding pressure are used uniformly, process defects such as weak inclined plane bonding, lead wire cracking, and wire collapse during forming will occur. This step achieves stable and uniform bonding processes for different levels of inclined planes through inclined plane tilt angle matching parameter adaptation. Specifically, the measured tilt angle parameters of the lower gentle slope and the upper steep slope are pre-stored. Based on the small tilt angle and smooth contact surface structure characteristics of the lower gentle slope, the first set of bonding wire feed rate and bonding pressure process parameters are matched and set to adapt to the smooth forming conditions, meeting the low-speed lead wire forming requirements of the gentle slope, ensuring dense bonding at the bonding points and smooth lead wire forming. Based on the structural characteristics of the steep slope with its large inclination angle and significant drop, a second set of bonding wire feed and bonding pressure process parameters is matched and set to adapt to the large drop forming conditions. This compensates for the wire routing resistance and forming tension difference during the lead wire forming process on the steep slope. During the bonding operation of each layer of pads, the corresponding exclusive process parameters are automatically called to carry out the bonding operation. This ensures that the bonding wire feed speed, total wire output, and bonding pressure are fully adapted to the current slope's inclination forming conditions, eliminating process compatibility issues caused by differences in the slope structure. This guarantees that the bonding strength of the bonding points on the upper and lower slopes meets the standards, the lead wire forming shape is regular, and local stress concentration caused by process defects is avoided, further enhancing the stress buffer reliability of the overall structure.
[0112] In summary, this invention fixes the chip face-up to a pre-defined mounting area on a carrier substrate; a pre-fabricated beveled substrate with a hollowed-out slot is aligned and fastened onto the carrier substrate. The inner wall of the hollowed-out slot has a continuously connected lower gentle slope and an upper steep slope. Substrate pads corresponding to different signal attributes are arranged on the two slopes respectively. The inner wall of the hollowed-out slot and the carrier substrate enclose a shallow hollow cavity with an open top, and the chip pads are lower than all substrate pads in the height direction. Based on the cavity geometry parameters of the shallow hollow cavity and the spatial relationship between the chip pads and the substrate pads at each level, a pairing mechanism is used. Based on the shortest connection between the chip pads and the corresponding substrate pads, the motion path and bonding sequence of the bonding tool are planned. The bonding tool is controlled to extend into the hollow shallow cavity, and according to the planned motion path and bonding sequence, each chip pad and the substrate pad on the corresponding inclined surface are connected by wire bonding to form a short-pitch inclined wire configuration. The wire configuration is used to absorb the relative displacement between the chip and the inclined substrate due to the difference in thermal expansion coefficients through its own elastic deformation, so as to achieve the purpose of simultaneously improving the high-speed transmission and thermal stress relief reliability of the chip bonding interconnect structure.
[0113] The above description is only a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.
Claims
1. A method for stress buffering of chip double-bevel bonding interconnects, characterized in that, Includes the following steps: Fix the chip face up to the preset mounting area of the carrier substrate; The pre-fabricated sloping substrate with a hollowed-out slot is aligned and fastened to the upper part of the support base. The inner wall of the hollowed-out slot has a continuously connected lower gentle slope and an upper steep slope. The two slopes are respectively provided with substrate pads corresponding to different signal attributes. The inner wall of the hollowed-out slot and the support base form a hollow shallow cavity with an open top. The chip pads are lower than all substrate pads in the height direction. Based on the cavity geometry parameters of the hollow shallow cavity, the spatial positional relationship between the chip pads and the substrate pads of each layer, and taking the shortest connection between the paired chip pads and the corresponding substrate pads as the benchmark, the motion path and bonding sequence of the bonding tool are planned. The control bonding tool extends into the hollow shallow cavity and sequentially connects each chip pad to the substrate pad on the corresponding layer slope surface through wire bonding according to the planned motion path and bonding sequence, forming a short-pitch oblique lead configuration. The lead configuration is used to absorb the relative displacement between the chip and the oblique substrate due to the difference in thermal expansion coefficients through its own elastic deformation.
2. The chip double-bevel bonding interconnect stress buffering method according to claim 1, characterized in that, The step of fixing the chip face up to the preset mounting area of the support substrate includes: Full-area alignment reference marks and chip limiting contours are processed in the preset mounting area of the support substrate; Obtain global alignment reference marks, establish a unified spatial coordinate system for the base, and determine the horizontal boundary for chip placement based on the chip's positioning contour. Pick up the chip using a pickup device, with the side of the chip with metal pads as the front, calibrate the horizontal coordinates and angles against the global alignment reference marks, and place the chip in the installation area enclosed by the chip's limiting contour. The back side of the chip is fixed to the carrier substrate by a retaining medium.
3. The chip double-bevel bonding interconnect stress buffering method according to claim 1, characterized in that, Before the step of aligning and fastening the pre-formed beveled substrate with the hollowed-out groove onto the support base, the hollowed-out groove and double-step bevel of the beveled substrate are pre-processed, including: Select an insulating substrate and design a continuous lower slope and upper steep slope profile based on the preset height difference between the chip pads and the substrate pads. A hollowed-out groove is processed in the middle of the substrate using a substrate molding process, and a continuous double-step inclined structure is simultaneously formed on the inner wall of the groove. According to the signal rate classification rules, high-speed signal substrate pads are arranged on the lower gentle slope, and power and low-speed signal substrate pads are arranged on the upper steep slope. The substrate to be fastened is obtained by metallizing the pads.
4. The chip double-bevel bonding interconnect stress buffering method according to claim 1, characterized in that, The step of aligning and fastening the pre-fabricated beveled substrate with hollowed-out slots onto the support base includes: Obtain the positioning mark at the edge of the hollowed-out groove on the inclined substrate and establish a coordinate mapping relationship with the global alignment reference mark on the supporting substrate; Based on the coordinate mapping relationship, the inclined substrate is moved along the vertical direction perpendicular to the supporting base. The difference of the reference coordinates during the movement is compared in real time, and the horizontal position and angle of the inclined substrate are adjusted to calibrate the horizontal orientation and angle. After alignment and calibration, the inclined substrate is lowered vertically so that the cutout completely covers the mounting area where the chip is located. The inclined substrate is attached and fixed to the edge of the supporting substrate to form the open-top cutout shallow cavity.
5. The chip double-bevel bonding interconnect stress buffering method according to claim 1, characterized in that, The step of planning the motion path and bonding sequence of the bonding tool based on the shortest connection between the paired chip pads and the corresponding substrate pads includes: Establish a three-dimensional geometric model of the hollow shallow cavity and a solid contour model of the bonding tool. Combine the equipment positioning error and the swing amplitude of the lead wire to set the safety margin for obstacle avoidance of the entire cavity. Based on a unified spatial coordinate system, a one-to-one electrical pairing relationship between chip pads and substrate pads at each level is established, and the shortest spatial connection between each pair of pads is extracted as the basic routing of the lead. The cavity obstacle avoidance safety margin is superimposed as a constraint condition onto the perimeter of the shortest connection in space to form a path avoidance boundary. Combined with the kinematic constraints of the bonding tool, a smooth, collision-free continuous motion trajectory is generated within the path avoidance boundary. Combined with the processing logic of the layered pads, the complete bonding sequence is output.
6. The chip double-beveled inclined bonding interconnect stress buffering method according to claim 5, characterized in that, The step of verifying and correcting the trajectory after generating a collision-free continuous motion trajectory includes: Discrete spatial points are extracted along the continuous motion trajectory, and the real-time gap value between the bonding tool profile and the double-order inclined cavity wall is calculated point by point. By comparing the gap value at each point with the cavity obstacle avoidance safety margin, interference risk points with gaps smaller than the safety margin are identified. The trajectory coordinates of the interference risk points are locally offset, and a corrected trajectory with a smooth transition is regenerated until the gaps between all points in the entire trajectory meet the obstacle avoidance safety margin requirements.
7. The chip double-bevel bonding interconnect stress buffering method according to claim 1, characterized in that, The step of inserting the controlled bonding tool into the hollowed-out shallow cavity includes: The drive bonding tool is fed vertically downward from the top opening of the hollow shallow cavity along the preset cavity entry trajectory, and the tool axis is kept in a vertical position throughout the feeding process; During the feeding process, the three-dimensional spatial coordinates of the bonding tool are acquired in real time, and the position of the inclined surface of the cavity wall is compared synchronously to maintain a safe clearance between the tool and the cavity. Once the bonding tool reaches the preset height directly above the target chip pad, it switches to a low-speed feed mode to descend until the lead tip contacts the chip pad surface.
8. The chip double-bevel bonding interconnect stress buffering method according to claim 1, characterized in that, The step of sequentially connecting each chip pad to the substrate pad on the corresponding layer slope via wire bonding according to the planned motion path and bonding sequence includes: Select the high-speed signal substrate pads arranged on the lower gentle slope and connect them with the corresponding chip pads to form lead connections; After completing all the lead wire connections on the lower gentle slope, adjust the travel range of the bonding tool and switch the work area to the upper steep slope. Select the power supply and low-speed signal substrate pads arranged on the upper steep slope and connect them with the corresponding chip pads to form lead connections; After all lead wire connections are completed, cut the lead wire tails.
9. The chip double-bevel bonding interconnect stress buffering method according to claim 8, characterized in that, The step of selecting the lower gentle slope and the upper steep slope substrate pads and constructing lead connections with the corresponding chip pads includes: For the high-speed signal pads on the lower gentle slope, the first lead arch height parameter is set so that the lead forms the shortest path shape after it is formed. For the power and low-speed signal pads on the upper steep slope, set the second lead arch height parameter and reserve the lead elastic deformation margin. After the single-level lead connection is constructed, the lead morphology parameters are collected, and the deviation between the lead morphology parameters and the preset morphology control threshold is compared.
10. The chip double-bevel bonding interconnect stress buffering method according to claim 8, characterized in that, The step of selecting the substrate pads of the lower gentle slope and the upper steep slope and constructing lead connections with the corresponding chip pads further includes: Configure the first set of bonding feed rate and bonding pressure process parameters based on the inclination angle of the lower gentle slope surface. The second set of bonding feed rate and bonding pressure process parameters are configured based on the inclination angle of the upper steep slope. When constructing lead connections, the corresponding level of process parameters is called to carry out bonding operations, matching the lead forming process under different slope conditions.
