Projection device based on gpmi super front line technology and communication method thereof

Abstract

The application provides a projection device based on GPMI super front-line technology and a communication method thereof, comprising: identifying a concentrated distribution area of high-frequency return current on a shell bonding surface by analyzing the return current path length of each buckle pressing point in the grounding return current path and the trend of the return current path along the shell bonding surface; dynamically adjusting the return crosstalk judgment of adjacent data channels according to the crosstalk evaluation basis, combining the return current path length of the grounding return current around the shell edge buckle point, and identifying the return interference caused by the shell edge suspension bonding; extracting the interference frequency band where the return interference is located, determining the frequency range that needs to be avoided by the projection device return link, and generating a communication state adjustment instruction of the projection device return link.

Description

Technical Field

[0001] This invention relates to the field of information technology, and in particular to projection devices and communication methods based on GPMI Super One-Line Communication technology. Background Technology

[0002] In the communication of projection devices using Super One-Line Communication technology, ensuring high-quality transmission of audio and video signals is a crucial factor determining device performance. As devices become thinner and lighter, the layout of internal components is becoming increasingly compact, and the physical distance between the detachable camera module and the motherboard is constantly decreasing, directly affecting the stability of the communication link. Current conventional wisdom dictates that a smaller distance between the module and the motherboard indicates shorter traces and reduced crosstalk. However, this approach ignores the interplay between mechanical structure and electrical performance in complex assembly environments and fails to detect the hidden damage to the high-frequency signal return path caused by the casing assembly. This leads to frequent and unpredictable signal fluctuations in actual mass production. When users install the detachable camera module, even if it appears fully assembled to the naked eye, the slight mechanical stress during the pressing process of the casing's snap-fit ​​structure may prevent the module edges from forming a complete fit with the device casing, resulting in gaps of a few millimeters or even smaller in some areas. Although these gaps are extremely small, they are enough to disrupt the metal casing grounding return path that high-frequency signals rely on. Under normal circumstances, when audio and video signals are transmitted from the motherboard to the camera module, the high-frequency current returns through the adjacent metal casing, forming a low-impedance loop. However, if the edge becomes suspended, this return current is forced to detour to other conductor paths, such as nearby audio and video data cables or internal wiring. This detour causes high-frequency current to enter the data channel originally used to transmit images or sound, resulting in abnormalities such as screen flickering, color distortion, or audio interruption. Accurately capturing this change in the grounding return path caused by mechanical assembly and effectively identifying audio and video backhaul interference caused by edge suspension and bonding has become a key issue in ensuring the stable communication of projection equipment. Summary of the Invention

[0003] This invention provides a projection device and its communication method based on GPMI Super One-Line Communication technology, mainly including: Obtain the module spacing between the camera module and the motherboard of the projection device, the fitting gap between the edge of the outer shell and the ground plane of the motherboard, and the edge pressing force of each buckle pressing point, and calibrate the solid-fit state or suspended-fit state of each buckle pressing point. Based on the contact state, contact gap and edge pressing force of each snap-fit ​​point, the contact resistance is extracted and the static potential is measured. Snap-fit ​​points with low contact resistance and small deviation are used as effective return nodes to construct the grounding return path. Analyze the return path length of each snap-fit ​​point in the grounding return path and its direction along the outer shell bonding surface to identify the concentrated distribution area of ​​high-frequency return current on the outer shell bonding surface; If the concentrated distribution area contains suspended clips, it is determined that the high-frequency return current leaves the surface of the metal shell and forms an alternative path through the shielding layer of the GPMI return channel. The disturbance period is extracted by comparing the level change and the original waveform is collected. The original waveform is aligned and superimposed for multiple cycles at the decision sampling time. The eye diagram opening of the returned signal is extracted. The narrowing of the eye diagram opening level relative to the time width when there is no crosstalk is determined as the basis for crosstalk evaluation of adjacent data channels. Based on the crosstalk assessment criteria and the return path length of the ground return current around the edge of the casing, the backhaul interference caused by the suspended attachment of the casing edge is identified. Furthermore, the module spacing between the camera module and the motherboard of the projection device, the bonding gap between the edge of the outer shell and the ground plane of the motherboard, and the edge pressing force of each snap-fit ​​point are obtained. The actual bonding state or the suspended bonding state of each snap-fit ​​point is calibrated, including: using the module spacing as a positioning reference for the snap-fit ​​distribution area on the edge of the outer shell, collecting bonding gaps along the edge of the outer shell to form a bonding gap sequence and locating the area of ​​each snap-fit ​​point, defining the effective collection range in combination with the module spacing, collecting the edge pressing force at each snap-fit ​​point, and establishing gap pressure matching data; for each snap-fit ​​point in the matching data, if the bonding gap is lower than the preset bonding threshold and the edge pressing force is higher than the preset pressure threshold, it is classified as the actual bonding state; if it exceeds or is lower, it is classified as the suspended bonding state, thus obtaining the bonding state of each snap-fit ​​point on the edge of the outer shell.

[0004] Furthermore, the step of extracting contact resistance and measuring static potential based on the contact state, contact gap, and edge pressing force of each snap-fit ​​point, and constructing a grounding return path using snap-fit ​​points with low contact resistance and small deviation as effective return nodes, includes: substituting the contact gap and edge pressing force into the contact resistance correspondence table of the snap-fit ​​surface for each snap-fit ​​point to obtain the contact resistance value, and establishing a contact resistance distribution according to the snap-fit ​​number; measuring the DC voltage of the metal surface of each snap-fit ​​point relative to the reference terminal of the motherboard ground plane under the motherboard's steady-state power-on condition to obtain the static potential, subtracting the static potential from the reference terminal potential and taking the absolute value to obtain the static potential deviation, forming resistance potential pairing data; for each snap-fit ​​point in the resistance potential pairing data, if the contact resistance is lower than a preset resistance threshold and the static potential deviation is lower than a preset potential threshold, it is determined to be an effective return node, starting from the reference ground terminal of the camera module backplate, sequentially connecting all effective return nodes along the metal surface of the outer shell and extending to the edge of the motherboard ground plane to obtain a grounding return path.

[0005] Furthermore, the analysis of the return path length of each snap-fit ​​point in the grounding return path and its direction along the outer shell bonding surface, identifying the concentrated distribution area of ​​high-frequency return current on the outer shell bonding surface, includes: measuring the arc length between two adjacent effective return nodes along the conductive segment of the outer shell metal surface for each effective return node in the grounding return path to obtain the node spacing, and connecting them in series according to the snap-fit ​​distribution order on the outer shell edge to form a return path length sequence; projecting the geometric position of each path segment onto the outer shell bonding surface to obtain the direction trajectory, and mapping the length sequence and direction trajectory according to the effective return node number to form a path length direction distribution along the outer shell edge; for each path segment in the distribution, if the path length is lower than a preset length threshold and the angle between the tangent directions of adjacent path segments is lower than a preset angle threshold, it is determined to be a high-frequency return current dense convergence segment, and adjacent dense convergence segments are merged into the same outer shell segment to obtain a concentrated distribution area. The interference frequency band where the return interference is located is extracted, the frequency range that the projection device's return link needs to avoid is determined, and a communication status adjustment command for the projection device's return link is generated.

[0006] Furthermore, the detection process includes checking whether the concentrated distribution area contains suspended snap-fit ​​points. If so, it is determined that the high-frequency return current leaves the metal casing surface and forms an alternative path through the GPMI return channel shielding layer. The disturbance period is extracted by comparing the level change and the original waveform is acquired. This includes: retrieving the bonding state distribution of each snap-fit ​​point on the casing edge, verifying the bonding state markings corresponding to the effective return current nodes in the concentrated distribution area, and if there are suspended bonding snap-fit ​​points, they are determined as contact interruption points, and the coordinate position of the contact interruption point on the casing bonding surface is obtained as the starting position of the alternative path. Based on the starting position of the alternative path, the backhaul signal level sequence of the most recent GPMI backhaul data channel is collected before the reference period and after the current period. The level changes are formed by subtracting the absolute values ​​point by point according to the sampling clock. The level changes are arranged along the time axis to form the level jump curve of the backhaul signal. The level change is compared with the preset jump threshold point by point for the level jump curve. If the change is continuously higher than the threshold on consecutive timestamps, it is determined to be a disturbed period. During the disturbed period, the original waveform data of the adjacent data channel is collected synchronously according to the sampling clock beat of the GPMI protocol backhaul link.

[0007] Furthermore, the process of performing multi-cycle alignment and superposition on the decision sampling time of the original waveform, extracting the eye diagram opening of the returned signal, and determining the narrowing of the eye diagram opening's level height and time width relative to the absence of crosstalk as the basis for crosstalk evaluation of adjacent data channels includes: for the original waveform of the GPMI returned signal during the interference period, marking the start and end boundaries of each symbol period according to the decision sampling time position of the GPMI protocol, dividing the original waveform into multiple equal-length waveform segments, and superimposing them along the time axis with the decision sampling time as the reference to obtain a multi-cycle aligned waveform cluster; taking each waveform at each time offset point. The lower bound of the envelope of the segment level values ​​is connected to form the upper edge of the eye diagram, and the upper bound of the envelope is connected to form the lower edge of the eye diagram. The area enclosed by the upper edge and the lower edge of the eye diagram near the decision sampling time is extracted as the eye diagram opening. The level difference between the upper edge and the lower edge of the eye diagram is used as the level height, and the time span between the intersection positions is used as the time width. The reference eye diagram recorded in advance under the crosstalk-free state is retrieved. The level height narrowing is obtained by subtracting the level height of the eye diagram opening from the reference level height, and the time width narrowing is obtained by subtracting the time width from the reference time width. The level height narrowing and the time width narrowing are used as the basis for crosstalk evaluation.

[0008] Furthermore, the step of identifying backhaul interference caused by the suspended contact of the outer shell edge based on the crosstalk assessment criteria and the length of the return path around the edge of the outer shell is as follows: extracting the level height narrowing and time width narrowing from the crosstalk assessment criteria, and weighting and summing them according to a preset weighting coefficient to obtain the crosstalk intensity value; retrieving the return path segment containing the contact interruption point in the ground return path, and accumulating the arc length between adjacent effective return nodes on both sides of the contact interruption point along the conductive segment of the outer shell metal surface to obtain the return path bypass length; querying the correspondence table between the bypass length and the judgment threshold based on the return path bypass length to read the current judgment threshold as the dynamic threshold value; if the crosstalk intensity value is higher than the dynamic threshold value, it is determined that there is backhaul interference caused by the suspended contact of the outer shell edge in the adjacent data channel and marked as a backhaul interference channel; if it is lower, it is determined that there is no interference, and the backhaul interference identification result is obtained.

[0009] Furthermore, the step of extracting the interference frequency band where the backhaul interference is located, determining the frequency range that the projection device backhaul link needs to avoid, and generating a communication status adjustment command for the projection device backhaul link includes: for the adjacent data channels marked as backhaul interference channels in the backhaul interference identification results, retrieving the original waveform of the GPMI backhaul signal collected during the interference period of the channel, using spectrum analysis to convert the original waveform from the time domain to the frequency domain to obtain the backhaul signal spectrum, and extracting the frequency segment with an amplitude higher than the preset amplitude threshold from the backhaul signal spectrum as the interference frequency band where the backhaul interference is located.

[0010] Furthermore, generating communication status adjustment instructions for the projection device backhaul link includes: defining the frequency range that the backhaul link needs to avoid based on the start and end frequency values ​​of the interference band, encapsulating the avoidance frequency range into a carrier frequency switching instruction field, and writing it together with the backhaul interference channel number into the backhaul link communication status adjustment instruction.

[0011] This application also provides a projection device based on GPMI Super One-Line Communication technology, which performs the above-described projection device communication method based on GPMI Super One-Line Communication technology.

[0012] The technical solutions provided by the embodiments of the present invention may include the following beneficial effects: This invention discloses a projection device and its communication method based on GPMI Super One-Line Communication technology. Addressing the problem of interrupted ground return path caused by the loose contact of the camera module and motherboard at the edge of the housing, leading to high-frequency return current transfer to the shielding layer of the adjacent GPMI return channel and crosstalk interference, this invention calibrates the actual contact or loose contact state and contact resistance of each snap-fit ​​point on the housing edge to screen effective return nodes and construct a ground return path. It identifies loose snap-fit ​​points in areas with concentrated high-frequency return current distribution, captures the level transition period of the return signal from the adjacent data channel when the return current detaches from the housing contact surface, acquires the original waveform during the interference period, and performs multi-cycle superposition to extract the eye diagram opening according to the GPMI protocol to determine the sampling time. The crosstalk level is dynamically assessed by combining the eye diagram narrowing amount with the return path length, thereby accurately identifying the return interference frequency band caused by the loose contact of the housing edge, generating frequency avoidance commands to adjust the communication state, effectively suppressing the impact of mechanical assembly defects on the signal integrity of the high-speed return link, and ensuring the communication reliability of the projection device. Attached Figure Description

[0013] Figure 1 This is a flowchart of the projection device and its communication method based on GPMI Super One-Line Communication technology of the present invention.

[0014] Figure 2 This is a schematic diagram of the projection device and its communication method based on GPMI Super One-Line Communication technology of the present invention.

[0015] Figure 3 This is another schematic diagram of the projection device and its communication method based on GPMI Super One-Line Communication technology of the present invention. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0017] like Figures 1-3 The projection device and its communication method based on GPMI Super One-Line Communication technology in this embodiment may specifically include: S101. Obtain the module spacing between the camera module and the motherboard of the projection device, as well as the fitting gap between the edge of the outer shell and the ground of the motherboard, and the edge pressing force at each snap-fit ​​point on the edge of the outer shell. Calibrate the solid fitting state or the suspended fitting state with gaps at each snap-fit ​​point on the edge of the outer shell.

[0018] The vertical distance between the camera module and the motherboard inside the projection device is measured using a laser rangefinder to obtain the module spacing, which is then used as a positioning reference for the buckle distribution area on the outer shell edge. A capacitive gap sensor is used to collect data point-by-point on the bonding gap along the outer shell edge and around the motherboard ground plane edge. The bonding gap value at each buckle distribution location on the outer shell edge is recorded sequentially, forming a bonding gap sequence along the outer shell edge. Based on the position corresponding to each value in the bonding gap sequence, the location of the buckle pressing points on the outer shell edge is determined. Combined with the module spacing, the effective collection range of the buckle pressing points is defined. A thin-film pressure sensor is placed at each buckle pressing point to collect the edge pressing force applied at the junction of the outer shell edge and the motherboard ground plane after the buckle engages. A mapping relationship is established between the bonding gap and edge pressing force of each buckle pressing point according to the buckle number, forming gap pressure pairing data for the buckle pressing points. For each snap-fit ​​point in the gap pressure pairing data, if the fitting gap at that point is lower than a preset fitting threshold and the edge pressing force is higher than a preset pressure threshold, then the snap-fit ​​point is classified as a solid fit state; if the fitting gap at that point exceeds the preset fitting threshold or the edge pressing force is lower than the preset pressure threshold, then the snap-fit ​​point is classified as a suspended fit state, thus obtaining the fit state distribution of each snap-fit ​​point on the edge of the outer shell.

[0019] In one embodiment, the projection device is a super-fast one-line projection device, and the camera module is a detachable camera module. Multiple sets of clips distributed along the edge of the housing engage with the metal edge corresponding to the motherboard's ground plane, forming a dual function of mechanical fixation and electrical grounding. Several clip engagement points are distributed along the perimeter of the motherboard along the edge of the housing. Each clip engagement point performs both mechanical engagement and serves as a grounding node on the high-frequency signal return path.

[0020] For example, the laser rangefinder is positioned at the edge of the motherboard pointing towards the metal plate on the back of the camera module. It calculates the vertical distance between the back plate of the camera module and the upper surface of the motherboard by measuring the flight time of the emitted laser beam and the reflected light, thus obtaining the module spacing. This module spacing is used to define the effective sampling range of the clip distribution area on the edge of the housing, preventing the sensor's sampling range from exceeding the camera module's obstruction area, which could lead to blind spots in subsequent bonding gap acquisition.

[0021] Specifically, the probes of the capacitive gap sensor are arranged close together along the junction of the inner wall of the housing and the motherboard ground plane, forming a parallel plate capacitor structure with the opposite metal surface. When a gap exists between the edge of the housing and the motherboard ground plane, the gap thickness corresponds to the change in dielectric layer thickness, causing the probe capacitance value to decrease as the gap thickness increases. The sensor reads the capacitance value and converts it into the actual gap thickness to obtain the fitting gap value at that sampling point. Continuous sampling is performed around the edge of the motherboard ground plane along the edge of the housing, and the fitting gap values ​​are recorded sequentially according to the buckle distribution order, forming the fitting gap sequence along the edge of the housing. Each value in the fitting gap sequence corresponds to a sampling position on the edge of the housing, with the peak value corresponding to the center of the buckle pressing point and the trough value corresponding to the elastic edge area between two buckles.

[0022] In one possible implementation, the thin-film pressure sensor is a piezoresistive flexible thin-film sensor with a thickness of less than 0.3 mm, which is attached to the interface between the snap-fit ​​surface and the motherboard ground plane. The vertical pressure generated after the snap-fit ​​is engaged acts on the thin film, causing a change in resistance due to the change in the spacing between the conductive particles inside the film. The output voltage increases with increasing pressure, thereby allowing the reading of the edge clamping force.

[0023] Preferably, a thin-film pressure sensor is arranged at each snap-fit ​​point, and each sensor is assigned a unique snap-fit ​​number according to the physical arrangement of the snap-fits on the edge of the housing. Simultaneously, the peak position values ​​in the bonding gap sequence are correlated according to the same snap-fit ​​number, and the bonding gap value under the same snap-fit ​​number is combined with the edge pressing force value to form the gap pressure pairing data. The gap pressure pairing data uses the snap-fit ​​number as an index to record the bonding gap and edge pressing force values ​​for each snap-fit ​​point.

[0024] It should be noted that the preset bonding threshold is set to 0.1 mm, and the preset pressure threshold is set to 2 Newtons. The bonding threshold is set to correspond to the maximum allowable gap required for the outer casing metal surface to form a low-impedance contact with the high-frequency signal return flow, and the pressure threshold is set to correspond to the minimum pressing force required to maintain tight contact between the metal surfaces after the buckle is engaged. The two thresholds are determined in parallel to avoid misjudgment of the state due to the influence of local tolerance fluctuations on a single parameter.

[0025] Specifically, for each snap-fit ​​point in the gap pressure pairing data, the fitting gap value and edge pressing force value of that point are read and compared with the preset fitting threshold and the preset pressure threshold, respectively. If the fitting gap of that point is lower than the preset fitting threshold and the edge pressing force is higher than the preset pressure threshold, then the snap-fit ​​point is classified as a solid fit state and marked as a candidate for a valid grounding node; if the fitting gap of that point exceeds the preset fitting threshold or the edge pressing force is lower than the preset pressure threshold, then the snap-fit ​​point is classified as a suspended fit state and marked as a grounding interruption point. The data items corresponding to all snap numbers in the gap pressure pairing data are processed one by one to obtain the fitting state distribution of each snap-fit ​​point on the edge of the outer shell.

[0026] It is understood that the bonding state distribution is indexed by the buckle number, and records the actual bonding or suspended bonding mark of each buckle pressing point, reflecting the grounding continuity of the outer shell edge along the motherboard ground plane, and providing a basis for the selection of effective return nodes on the subsequent grounding return path.

[0027] S102. Based on the actual or suspended bonding state of each snap-fit ​​point, the bonding gap and the edge pressing force, extract the contact resistance of each snap-fit ​​point. At the same time, measure the static potential of each snap-fit ​​point relative to the motherboard ground. Snap-fit ​​points with contact resistance below the threshold and small deviation from the static potential of the motherboard ground are used as effective return current nodes. Construct a grounding return current path from the camera module reference ground through the effective return current node to the motherboard ground.

[0028] Obtain the actual or suspended contact state of each snap-fit ​​point on the edge of the casing. For each snap-fit ​​point, substitute the contact gap value and the edge pressing force value into a pre-established table of contact resistance correspondence between snap-fit ​​surfaces. This table satisfies the inverse relationship law that the contact resistance increases with the increase of the contact gap and decreases with the increase of the edge pressing force, thus obtaining the contact resistance value of each snap-fit ​​point. Establish a contact resistance distribution according to the snap-fit ​​number. For each snap-fit ​​point in the contact resistance distribution, under the steady state of the motherboard power-on, use a high-impedance voltmeter to measure the DC voltage of the metal surface of the snap-fit ​​point relative to the motherboard ground reference terminal to obtain the static potential of the snap-fit ​​point. Subtract the static potential from the potential of the motherboard ground reference terminal and take the absolute value to obtain the static potential deviation. Correspond the static potential deviation to the contact resistance according to the snap-fit ​​number to form the resistance potential pairing data of the snap-fit ​​points. For each snap-fit ​​point in the resistance potential pairing data, if the contact resistance of the point is lower than a preset resistance threshold and the static potential deviation is lower than a preset potential threshold, the snap-fit ​​point is determined to be a valid return current node. Starting from the reference ground terminal of the camera module backplate, all valid return current nodes are connected sequentially along the metal surface of the outer shell and extended to the edge of the motherboard ground plane, thus obtaining a grounding return current path from the camera module reference ground through the valid return current nodes to the motherboard ground plane.

[0029] In one embodiment, the grounding return path is used to carry the return current generated when the high-frequency audio and video signals of the projection device camera module are transmitted back. The return current flows out from the reference ground terminal of the camera module back plate, returns to the motherboard ground through the conductive contact surface formed by the snap-fit ​​points between the outer casing metal surface and the motherboard ground, and the conductivity of the snap-fit ​​points directly determines the equivalent impedance of the return path.

[0030] Specifically, the pre-established table of contact resistance correspondences for the snap-fit ​​surfaces is obtained by calibrating snap-fit ​​samples on a test bench. The calibration process is performed on a snap-fit ​​sample test bench on a projection device. Different edge pressing forces are applied to snap-fit ​​samples of the same specification, while inserting inserts of different thicknesses to create different fitting gaps. The contact resistance values ​​at both ends of the snap-fit ​​surfaces are measured using a four-wire milliohm meter. These values ​​are stored in the table using a two-dimensional index composed of the fitting gap value and the edge pressing force value. The table is divided into 11 levels at 0.05 mm intervals along the fitting gap axis (0 to 0.5 mm) and 11 levels at 1 Newton intervals along the edge pressing force axis (0 to 10 Newtons). Each two-dimensional index position corresponds to a contact resistance value. The inverse proportional relationship is reflected in the fact that when the edge pressing force level is fixed, the contact resistance value monotonically increases with the increase of the fitting gap level; when the fitting gap level is fixed, the contact resistance value monotonically decreases with the increase of the edge pressing force level. For each snap-fit ​​point, its fitting gap value and edge pressing force value are read. The contact resistance value of that snap-fit ​​point is obtained by looking up the corresponding relationship table using bilinear interpolation. Specifically, the four adjacent positions of the actual measured value in the corresponding relationship table are first determined. The normalized distances d1 and d2 between the actual measured value and the four positions on the fitting gap axis and edge pressing force axis, respectively, are then calculated. g is the actual fitting gap, g0 and g1 are the values ​​of two adjacent fitting gaps, f is the actual edge pressing force, f0 and f1 are the values ​​of two adjacent edge pressing forces. Then, the contact resistance values ​​of the four positions are weighted and summed according to the weights (1-d1)×(1-d2), d1×(1-d2), (1-d1)×d2, and d1×d2 respectively to obtain the interpolated contact resistance values. The contact resistance distribution is formed by summing the values ​​according to the buckle number.

[0031] It should be noted that, for the buckle pressing point in the suspended bonding state, the bonding gap exceeds the preset bonding threshold, and the contact resistance value obtained by looking up the corresponding table falls into the high resistance range; for the buckle pressing point in the solid bonding state, the bonding gap is lower than the preset bonding threshold and the edge pressing force is higher than the preset pressure threshold, and the contact resistance value obtained by looking up the table falls into the low resistance range.

[0032] In one possible implementation, the motherboard ground plane has a reference copper pillar and an edge copper ring. The reference copper pillar is located in the center of the ground plane as a reference point for potential measurement. The input impedance of the high-impedance voltmeter is higher than 10 megohms. One end of the measuring probe is fixedly connected to the reference copper pillar, and the other end sequentially contacts the exposed metal surfaces of each latching contact point. After the projection device is powered on, when the motherboard power supply voltage fluctuation is less than 50 millivolts and lasts for more than 5 seconds, and the camera module enters signal standby mode, it is determined to be in a steady-state state. At this time, the static potential is obtained by reading the DC voltage values ​​of the metal surfaces of each latching contact point. Before measurement, the measured potential of the reference copper pillar should be collected. This serves as a reference value, incorporating minor offsets caused by grounding resistance. The static potential deviation is obtained by subtracting the measured potential of the reference copper pillar from the static potential and taking the absolute value. The calculation formula is: This deviation reflects the equivalent voltage drop between the metal surface of the snap-fit ​​point and the motherboard ground plane caused by poor contact or gaps. The static potential deviation and contact resistance value under each snap-fit ​​number are grouped into a binary pair, and then arranged and summarized in the snap-fit ​​number order to form resistance-potential pairing data.

[0033] Preferably, the preset resistance value is 50 milliohms, and the preset potential threshold value is 5 millivolts. The resistance threshold is set to correspond to the maximum contact resistance allowed when the high-frequency return current passes through the snap-fit ​​surface to form a low-impedance loop, and the potential threshold is set to correspond to the maximum equivalent voltage drop allowed between the metal surface of the snap-fit ​​point and the motherboard ground plane. The two thresholds are determined in parallel using logic. Only when the contact resistance value and static potential deviation of a snap-fit ​​point are both lower than the preset resistance threshold and the preset potential threshold is the snap-fit ​​point determined to be a valid return current node. If the contact resistance value or static potential deviation of a snap-fit ​​point is higher than the preset resistance threshold or the preset potential threshold is higher than the preset potential threshold, the snap-fit ​​point is determined to be an invalid node and excluded from the return current path construction.

[0034] Specifically, the grounding return path is constructed starting from the reference ground terminal of the camera module backplate, and sequentially connecting the effective return nodes along the metal surface of the outer casing according to the buckle number. Each segment connects to the conductive section of the outer casing metal surface between two adjacent effective return nodes, ultimately connecting to the copper ring at the edge of the motherboard ground plane, forming a complete grounding return path from the camera module reference ground through the effective return nodes to the motherboard ground plane. The copper ring at the edge of the motherboard ground plane is a continuous conductive copper ring structure surrounding the edge of the motherboard ground plane, and the copper post at the reference end of the motherboard ground plane is a measurement reference point set on this copper ring, with the copper post electrically connected to the copper ring. The grounding return path is recorded in the sequence of effective return nodes. Each node record includes: buckle number, the coordinate position of the buckle contact point on the metal surface of the outer casing, the contact resistance value of that point, the straight-line distance to the previous node, and the actual path length along the metal surface of the outer casing to the previous node. The starting point of the path is the coordinate position of the reference ground terminal of the camera module backplate, and the ending point of the path is the coordinate position of the connection point between the copper ring at the edge of the motherboard ground plane and the metal surface of the outer casing.

[0035] Understandably, when multiple consecutive snap-fit ​​points in a certain section of the outer casing edge are determined to be invalid nodes, the grounding return path will have node gaps in that section, and the span of the conductive segment on the outer casing metal surface between adjacent valid return nodes will increase, reflecting that the return path in that section has become detoured and extended. The grounding return path serves as a description of the actual path of high-frequency return current on the outer casing mating surface, recording the distribution of valid return nodes and the connection relationship between the conductive segments on the outer casing metal surface.

[0036] S103. By analyzing the return path length of each snap-fit ​​point in the grounding return path and the direction of the return path along the outer shell bonding surface, the concentrated distribution area of ​​high-frequency return current on the outer shell bonding surface is identified.

[0037] For each valid return node in the grounding return path, the arc length between two adjacent valid return nodes is measured along the conductive segment of the outer metal surface to obtain the node spacing between each valid return node. The node spacings of each segment are then connected in series according to the snap-fit ​​distribution order of the outer edge, forming a return path length sequence corresponding to each path segment in the grounding return path. The geometric positions of each path segment in the grounding return path are projected onto the unfolded coordinates of the outer edge using a method of sequentially arranging them along the arc length of the outer edge, obtaining the trajectory of the grounding return path along the outer edge. The return path length sequence and the trajectory are mapped one-to-one according to the valid return node number, forming a path length distribution along the outer edge. For each path segment in the path length distribution, if the path length is lower than a preset length threshold and the angle between the tangent directions of adjacent path segments is lower than a preset angle threshold, the path segment is determined to be a densely converging segment of high-frequency return current. Multiple densely converging segments adjacent to each other along the outer edge are merged into the same outer edge segment, obtaining a concentrated distribution area of ​​high-frequency return current on the outer edge.

[0038] In one embodiment, the effective return nodes in the grounding return path are arranged in the order of the buckle numbers on the edge of the housing. The conductive section of the metal surface of the housing between two adjacent effective return nodes constitutes a path segment. The entire grounding return path covers the return channel between the reference ground terminal of the camera module backplate and the copper ring at the edge of the motherboard grounding ground in a way that the path segments are connected end to end.

[0039] Specifically, for two adjacent valid return nodes in the grounding return path, the arc length along the inner wall of the casing is measured along the geometric direction of the conductive section on the metal surface of the casing to obtain the node spacing. When the casing edge has a bend or corner shape, the node spacing is obtained by summing the actual arc length of the trace on the inner wall of the casing, rather than the straight-line distance between the two nodes. The node spacings of all path segments are sequentially connected in series according to the buckle numbering of the casing edge to form the return path length sequence. The return path length sequence is indexed by the path segment number and records the arc length value of each path segment, reflecting the extension of each path segment when the return current flows along the metal surface of the casing. When multiple consecutive buckle contact points in a certain section are determined to be invalid nodes, the node spacing of the corresponding path segment increases significantly, and the return path length sequence shows a large value in that section; when all buckle contact points in a certain section are determined to be valid return nodes, the node spacing of the corresponding path segment remains near the buckle spacing design value, and the return path length sequence maintains a small value in that section.

[0040] It should be noted that the unfolded coordinates of the outer shell bonding surface are used to map the three-dimensional edge direction of the outer shell to a two-dimensional plane, facilitating a unified description of the path segment direction. The unfolding method uses the reference ground end of the camera module back plate as the coordinate origin, and the arc direction of the inner wall of the outer shell as the unfolded coordinate axis, arranging the starting and ending points of each path segment sequentially according to the arc length of the outer shell edge.

[0041] In one possible implementation, the two endpoints of each path segment in the grounding return path are projected onto the unfolded coordinate system of the outer casing's bonding surface. Each path segment corresponds to a continuous trajectory from the starting endpoint to the ending endpoint, and the trajectories of all path segments are connected end-to-end to form the directional trajectory. The directional trajectory is recorded as a sequence of two-dimensional coordinate points, with each coordinate point corresponding to the position of a valid return node in the unfolded coordinate system of the outer casing's bonding surface. The arc length values ​​in the return path length sequence are mapped to the corresponding path segments of the directional trajectory according to the valid return node numbers, resulting in the path length directional distribution. The path length directional distribution uses the valid return node number as an index, and simultaneously records the arc length value of the path segment corresponding to that number and the two-dimensional coordinates of the starting and ending endpoints.

[0042] Preferably, the preset length threshold is 1.2 times the design spacing of adjacent buckles on the edge of the outer shell. When the buckle design spacing is 10 mm, the preset length threshold is 12 mm. The preset angle threshold is 15 degrees. The setting of the length threshold corresponds to the maximum allowable node spacing between adjacent effective return nodes under the effective engagement state of the buckles. Exceeding this threshold indicates that the path segment has obvious detours. The setting of the angle threshold corresponds to the maximum allowable directional deflection for adjacent path segments to maintain the same direction of extension. Below this threshold, it indicates that adjacent path segments have the same direction. Further, for each path segment in the path length distribution, the arc length value L of the segment and the angle θ between the tangent direction of the segment and the previous path segment are read, where θ is the angle between the tangent vector at the starting point of the current path segment and the tangent vector at the ending point of the previous path segment. The angle value is obtained by extracting the coordinates of the endpoints of the path segment and calculating the direction of the tangent vector. L and θ are compared with the preset length threshold and the preset angle threshold respectively. If the arc length L of the path segment is lower than the preset length threshold and the tangent direction angle θ is lower than the preset angle threshold, then the path segment is determined to be a densely converging segment and marked as a path segment through which the high-frequency return current flows in a concentrated manner; if the arc length L of the path segment is higher than the preset length threshold or the tangent direction angle θ is higher than the preset angle threshold, then the path segment is determined to be a non-densely converging segment and marked as a path segment through which the high-frequency return current flows in a sparse manner.

[0043] It is understandable that the densely converging section reflects the section on the outer shell bonding surface where the effective return nodes are closely arranged and have a continuous direction. This type of section forms a low-impedance and directionally stable return channel on the metal surface of the outer shell, and the high-frequency return current tends to flow along this type of section.

[0044] Specifically, along the unfolded coordinate axis of the outer shell bonding surface, the densely converging segment markers in the path length distribution are scanned, and multiple adjacent densely converging segments are merged into the same outer shell segment according to the arc length direction of the outer shell edge. The merging rule is that if the positional distance between two densely converging segments on the unfolded coordinate axis of the outer shell bonding surface is less than the actual arc length distance between these two densely converging segments... and the average node spacing of the path ratio When r is less than 1.5, the two densely converging segments are classified into the same shell segment; if there is a non-densely converging segment between the two densely converging segments, the tangent vectors v1 and v2 at the start and end points of the non-densely converging segment on the shell surface are extracted, and the tangent angle is calculated. ,when When the angle is greater than or equal to 15 degrees, it is determined to cross the outer shell bending angle, and merging stops; when If the angle is less than 15 degrees, it is determined that the corner has not been crossed, and the merging process continues. After merging, several continuous shell segments are obtained along the coordinate axis of the shell bonding surface. Each shell segment records its start and end coordinate positions and the set of effective return current node numbers it contains, which constitute the concentrated distribution area of ​​the high-frequency return current on the shell bonding surface.

[0045] S104. Detect whether the concentrated distribution area contains a suspended latching point. If it exists, determine that the high-frequency return current leaves the metal shell surface at that point due to contact interruption and instead forms an alternative path through the shielding layer of the spatially adjacent GPMI return channel. Compare the level change of the return signal on the adjacent data channel before and after the high-frequency return current leaves the shell and extract the time interval in which the level jump amplitude exceeds the preset threshold as the disturbed period. Collect the original waveform of the GPMI return signal during the disturbed period.

[0046] For the concentrated distribution area of ​​the high-frequency return current on the outer shell bonding surface, the bonding status distribution of each snap-fit ​​point on the edge of the outer shell is retrieved. The bonding status mark corresponding to each valid return node number within the concentrated distribution area is checked one by one. If there is a snap-fit ​​point marked as suspended bonding within the concentrated distribution area, it is determined to be a contact interruption point. At the contact interruption point, the high-frequency return current detaches from the metal outer shell surface and forms an alternative path through the shielding layer of the spatially adjacent GPMI return channel. The coordinate position of the contact interruption point on the outer shell bonding surface is obtained as the starting position of the alternative path. Based on the starting position of the alternative path of the contact interruption point, for the GPMI return data channel closest to this position, the return signal level sequence of this data channel before the existence of the contact interruption point and the current time period after its existence are collected. The two time period level sequences are aligned according to the sampling clock, and the absolute value is obtained by subtracting each sampling point to obtain the level change of the adjacent data channel. The level change is arranged along the time axis to form a return signal level jump curve. For the backhaul signal level transition curve, the level change at each timestamp is compared with a preset transition threshold. If the level change at a certain consecutive timestamp is continuously higher than the preset transition threshold, the time interval covered by the consecutive timestamps is determined to be the disturbed period. During the disturbed period, the original waveform data of the adjacent data channel is synchronously collected along the backhaul link sampling clock beat specified by the GPMI protocol to obtain the original waveform of the GPMI backhaul signal during the disturbed period.

[0047] In one embodiment, the GPMI return channel is a high-speed serial link carrying audio and video return signals between the projection device motherboard and the camera module. The GPMI return channel consists of differential signal pairs and an outer metal braided shielding layer. Under normal conditions, the outer metal braided shielding layer is grounded at both ends, forming a high-frequency signal return ground reference together with the outer casing metal surface. When the contact of the outer casing metal surface is interrupted at a certain snap-fit ​​point, the outer metal braided shielding layer, due to its spatial proximity to the contact interruption point, becomes a secondary inflow channel for high-frequency return current.

[0048] Specifically, for the concentrated distribution area of ​​the high-frequency return current on the bonding surface of the outer shell, the bonding status distribution table of each snap-fit ​​point on the edge of the outer shell is retrieved. Based on the set of valid return node numbers contained within the concentrated distribution area, the bonding status mark corresponding to each number is checked one by one. If a number is marked as a suspended bonding state in the bonding status distribution table, the snap-fit ​​point corresponding to that number is determined to be a contact interruption point; if all numbers in the concentrated distribution area are marked as solid bonding, then there are no contact interruption points in the concentrated distribution area, and subsequent alternative path analysis is skipped.

[0049] It should be noted that the high-frequency return current detaches from the metal casing surface at the contact interruption point because the gap between the metal surfaces at that point creates an open circuit, cutting off the nearest return path. The return current then seeks an alternative path from a nearby low-impedance conductor. Inside the projection device, the spatial distance between all conductors and the contact interruption point is calculated, and the low-impedance conductor with the smallest distance is selected as the alternative path. This low-impedance conductor is the outer metal braided shielding layer of the GPMI return channel. The return current is injected into the shielding layer via spatial electromagnetic coupling and flows towards the motherboard ground plane along the shielding layer's extension direction. Furthermore, a coordinate system is established on the casing's bonding surface, and the position of the contact interruption point on the bonding surface is recorded as a two-dimensional coordinate (x0, y0), where x0 represents the horizontal coordinate value and y0 represents the vertical coordinate value. This coordinate serves as the starting position of the alternative path. The geometric position coordinates (xi, yi) of each GPMI return channel on the internal wiring diagram are obtained, and the Euclidean distance is calculated. The GPMI backhaul channel with the smallest distance is determined to be the nearest data channel.

[0050] In one possible implementation, the adjacent data channel corresponds to a runtime period before the contact interruption point, where the housing latches are fully engaged and there is no suspended contact. The feedback signal level sequence of the reference period is read from the reference level sequence recorded during the factory verification phase of the projection equipment. The current runtime after the contact interruption point corresponds to a runtime period where the suspended contact latches are present in the current assembly state. The feedback signal level sequence of the current runtime is read point-by-point by a level acquisition probe at the end of the adjacent data channel according to the sampling clock cycle specified by the GPMI protocol. This level acquisition probe is integrated at the end contact position of the data channel interface circuit board, with a sampling resolution of 12 bits and a sampling frequency of 100MHz. After aligning the level sequences of the reference period and the current runtime one by one according to the sampling clock cycle, each sampling clock cycle corresponds to a level sampling point, forming a comparable dual-sequence dataset.

[0051] Specifically, for the two aligned level sequences, the level value of the current time period is subtracted from the level value of the corresponding time period at each sampling point, and the absolute value is taken to obtain the level change value at that sampling point. The level change values ​​of all sampling points are arranged along the time axis in the order of the sampling clock beats to form the return signal level transition curve. The return signal level transition curve, with the sampling clock beats as the horizontal axis and the level change value as the vertical axis, reflects the time-varying pattern of the amplitude deviation of the return signal before and after the contact interruption point in adjacent data channels. When a stable coupling is formed by the alternative path at the contact interruption point, the level transition curve shows an amplitude increase within the corresponding sampling interval; when the micro-movement of the housing latch causes fluctuations in the state of the contact interruption point, the level transition curve exhibits a continuous rise and fall pattern within the corresponding sampling interval.

[0052] Preferably, the preset transition threshold is set to 20% of the average amplitude of the reference level sequence. This value is determined by collecting background level fluctuation data from the GPMI return channel for 1000 consecutive sampling cycles under crosstalk-free conditions, and calculating the proportion of the upper limit of the 95% confidence interval of the fluctuation amplitude to the average. Further, the return signal level transition curve is scanned point-by-point along the sampling clock beat direction, and the level change value at each sampling point is compared with the preset transition threshold. If the level change values ​​at three or more consecutive sampling clock beats are higher than the preset transition threshold, the time interval covered by that consecutive sampling clock beat is marked as a disturbed period. Here, three consecutive beats are set as the judgment threshold to exclude false judgments of occasional transient noise. If the level change value at a certain sampling clock beat is lower than the preset transition threshold, then that sampling clock beat is not included in the disturbed period. After scanning, one or more disturbed periods are obtained on the return signal level transition curve, and the start and end sampling clock beat numbers of each disturbed period are recorded.

[0053] It is understood that the raw waveform data is the differential voltage waveform acquired by the probe at the end of the adjacent data channel without decision processing, and is recorded as a sequence of instantaneous differential voltage values ​​at each sampling clock cycle during the disturbance period. During the disturbance period, waveform acquisition is synchronously triggered along the sampling clock cycle of the backhaul link specified by the GPMI protocol. One instantaneous differential voltage value is read at each sampling clock cycle, and recorded sequentially according to the clock cycle order to obtain the raw waveform of the GPMI backhaul signal during the disturbance period.

[0054] For example, in a scenario where two concentrated distribution areas exist within a projection device, the first concentrated distribution area contains a contact interruption point, while all snap-fit ​​points in the second concentrated distribution area are in a solid-fit state. The starting position of the alternative path at the contact interruption point in the first concentrated distribution area is used to locate the first adjacent data channel using Euclidean distance calculation. Two disturbance periods are identified on the return signal level transition curve of this channel, and two original waveforms of the GPMI return signal are acquired for each. Since there is no contact interruption point in the second concentrated distribution area, the alternative path analysis is skipped. The two original waveforms correspond to the actual signal disturbance conditions of the adjacent data channels under the contact interruption point condition.

[0055] S105. Perform multi-cycle alignment and superposition on the original waveform according to the decision sampling time defined by the GPMI protocol, extract the eye diagram opening of the return signal, and use the narrowing of the eye diagram opening level and time width relative to the crosstalk-free condition as the basis for crosstalk evaluation of adjacent data channels.

[0056] For the original waveform of the GPMI return signal during the disturbed period, the start and end boundaries of each symbol period are marked on the original waveform according to the decision sampling time position defined by the GPMI protocol. The original waveform is divided into multiple equal-length waveform segments along the symbol period boundaries. The equal-length waveform segments are overlapped along the time axis with the decision sampling time as the alignment reference to obtain a multi-period aligned waveform cluster of the original waveform. For the multi-period aligned waveform cluster, the lower boundary of the envelope of the level values ​​of each equal-length waveform segment is taken at each time offset point to form the upper edge of the eye diagram, and the upper boundary of the envelope of the level values ​​of each equal-length waveform segment is taken to form the lower edge of the eye diagram. The closed area enclosed by the upper edge and the lower edge of the eye diagram near the decision sampling time is extracted as the eye diagram opening. The eye diagram opening uses the level difference between the upper edge and the lower edge of the eye diagram at the decision sampling time as the level height, and the time span between the intersection positions of the upper edge and the lower edge of the eye diagram within the symbol period as the time width. A reference eye diagram pre-recorded in a crosstalk-free state is retrieved from a neighboring data channel. The reference eye diagram records the reference level height and reference time width in a crosstalk-free state. The level height reduction is obtained by subtracting the reference level height from the level height of the eye diagram opening, and the time width reduction is obtained by subtracting the reference time width from the time width of the eye diagram opening. The level height reduction and the time width reduction constitute the crosstalk evaluation basis for neighboring data channels.

[0057] In one embodiment, the original waveform of the GPMI return signal is a sequence of instantaneous differential voltage values ​​acquired by a probe at the end of a nearby data channel during the disturbance period. The differential voltage flips from high to low or vice versa with the symbol data, changing periodically according to the return link transmission rate specified in the GPMI protocol. The decision sampling time is the time point within each symbol period specified in the GPMI protocol for judging the return signal level, typically located at the midpoint of the symbol period.

[0058] Specifically, for the original waveform of the GPMI return signal during the disturbed period, the symbol clock reference specified by the GPMI protocol is retrieved, and the start and end boundaries are marked sequentially along the time axis of the original waveform according to the symbol period length to obtain the segmentation points of each symbol period on the original waveform. The waveform segment between each two adjacent segmentation points is regarded as an equal-length waveform segment, and the original waveform is sequentially divided along the time axis to obtain a sequence of equal-length waveform segments.

[0059] It should be noted that multi-cycle alignment and overlay is a standard processing procedure in eye diagram analysis. Its core action is to display the waveforms of multiple symbol cycles as coincident along the time axis with the same reference point at the same moment.

[0060] In one possible implementation, the alignment reference selects the decision sampling time position in each equal-length waveform segment as the time zero point. The equal-length waveform segments are translated along the time axis to align their respective decision sampling time positions to the time zero point. Then, all the translated equal-length waveform segments are superimposed on the same time-level coordinate system to obtain the multi-period aligned waveform cluster of the original waveform. The multi-period aligned waveform cluster records the superposition shape of all symbol period waveforms within the disturbed period, with the time offset point as the horizontal axis and the instantaneous differential voltage value as the vertical axis. Furthermore, the construction of the upper and lower edges of the eye diagram is a key process for eye diagram opening extraction. In the multi-period aligned waveform cluster, there are multiple instantaneous differential voltage values ​​corresponding to equal-length waveform segments at each time offset point. The maximum value of these values ​​is called the upper bound of the envelope at that time offset point, and the minimum value is called the lower bound of the envelope at that time offset point.

[0061] Specifically, the multi-cycle aligned waveform cluster is traversed along the time offset points. For each time offset point, the minimum value of the level values ​​of all equal-length waveform segments at that point is extracted to obtain the lower bound of the envelope. The lower bounds of the envelopes at each time offset point are sequentially connected along the time axis to form the upper edge of the eye diagram. For each time offset point, the maximum value of the level values ​​of all equal-length waveform segments at that point is extracted to obtain the upper bound of the envelope. The upper bounds of the envelopes at each time offset point are sequentially connected along the time axis to form the lower edge of the eye diagram. The upper edge of the eye diagram corresponds to the lowest limit envelope of the differential high-level topping falling segment at that time offset point, and the lower edge of the eye diagram corresponds to the highest limit envelope of the differential low-level topping rising segment at that time offset point.

[0062] In one possible implementation, the upper edge and lower edge of the eye diagram converge at the beginning and end of the symbol period due to the level flipping process, forming an intersection position. This intersection creates a maximum spacing region near the decision sampling time in the middle of the symbol period. The closed region enclosed by the upper edge, lower edge, and the intersection position at the beginning and end of the eye diagram is extracted as the eye diagram opening.

[0063] Preferably, the difference between the voltage level values ​​at the upper and lower edges of the eye diagram at the decision sampling time is measured as the voltage level height, which reflects the voltage margin that the returned signal can be read at the decision sampling time. The time span measured along the time axis between the intersections of the upper and lower edges of the eye diagram is measured as the time width, which reflects the time margin that the returned signal can be read at the decision sampling time. The size of the eye diagram opening is characterized by both the voltage level height and the time width.

[0064] It is understood that the reference eye diagram is derived from the original waveform of the adjacent data channel during the factory verification stage of the projection equipment, under conditions where the housing clips are fully engaged and there is no suspended contact, extracted using the same multi-cycle alignment and superposition process. The reference level height and reference time width of the return signal eye diagram under crosstalk-free conditions are recorded. The reference eye diagram is derived from the original waveform of the adjacent data channel during the factory verification stage of the projection equipment, under conditions where the housing clips are fully engaged and there is no suspended contact, extracted using the same multi-cycle alignment and superposition process. The reference level height and reference time width of the return signal eye diagram under crosstalk-free conditions are recorded. Subtracting the level height of the eye diagram opening from the reference level height yields the level height narrowing amount Δh, and subtracting the time width of the eye diagram opening from the reference time width yields the time width narrowing amount Δt. Here, Δh represents the degree of crosstalk influence in the level dimension, and Δt represents the degree of crosstalk influence in the timing dimension. These two factors constitute the crosstalk evaluation basis for the adjacent data channel.

[0065] For example, for a multi-cycle aligned waveform cluster of adjacent data channels within a disturbed period, the eye diagram opening level height is obtained as H1, and the time width is obtained as W1; the reference eye diagram level height is recorded as H0, and the reference time width is recorded as W0. Subtracting the two yields the level height narrowing as H0 minus H1, and the time width narrowing as W0 minus W1. The two-dimensional set of quantities reflects the degree of eye diagram narrowing of adjacent data channels relative to the crosstalk-free state within the disturbed period. The greater the degree, the stronger the crosstalk experienced by the adjacent data channels.

[0066] S106. Based on the crosstalk assessment criteria and the return path length of the grounding return current flowing back through the snap-fit ​​point on the edge of the casing, dynamically adjust the crosstalk determination of the adjacent data channels to identify the return interference caused by the suspended attachment of the casing edge.

[0067] The level height narrowing and time width narrowing are extracted from the crosstalk assessment criteria of the adjacent data channels. These two values ​​are then weighted and summed according to a pre-set weighting coefficient to obtain the crosstalk intensity value. The return path segment containing the contact interruption point in the grounding return path is retrieved. The arc length between adjacent effective return nodes on both sides of the contact interruption point is accumulated along the conductive segment of the outer metal surface to obtain the return path detour length around the outer edge latch point. Based on the return path detour length, a pre-established detour length and judgment threshold correspondence table is consulted. This table follows an inverse relationship where a larger detour length corresponds to a lower judgment threshold. The current judgment threshold corresponding to the return path detour length is read and used as the dynamic threshold value for judging the crosstalk of the adjacent data channel. Regarding the crosstalk intensity value and the dynamic threshold value, if the crosstalk intensity value is higher than the dynamic threshold value, it is determined that the adjacent data channel has backhaul interference caused by the hanging and fastening points of the outer shell edge, and the adjacent data channel is marked as a backhaul interference channel; if the crosstalk intensity value is lower than the dynamic threshold value, it is determined that the adjacent data channel does not have backhaul interference caused by the hanging and fastening points of the outer shell edge, and the backhaul interference identification result caused by the hanging and fastening of the outer shell edge is obtained.

[0068] In one embodiment, the crosstalk assessment is based on a two-dimensional set of parameters consisting of voltage level narrowing and time width narrowing, reflecting the voltage margin loss and time margin loss of adjacent data channels at the decision sampling time, respectively. Both are derived from the degree of narrowing of the eye diagram opening relative to the reference eye diagram, and are a direct quantitative representation of crosstalk caused by the overhanging and snapping points of the housing edge of adjacent data channels.

[0069] Specifically, the voltage level narrowing value and time width narrowing value are read from the crosstalk assessment criteria, and then weighted and summed according to a pre-set weighting coefficient. The value of the weighting coefficient is pre-set based on the tolerance ratio of voltage margin and time margin of the GPMI backhaul channel decision circuit. A larger weighting coefficient is used for voltage level narrowing when the voltage margin tolerance is low, and a larger weighting coefficient is used for time width narrowing when the time margin tolerance is low. The sum of the two weighting coefficients is equal to 1. The weighted sum is recorded as the crosstalk intensity value; a larger value indicates more severe crosstalk experienced by adjacent data channels.

[0070] In one possible implementation, the weighting coefficient corresponding to the level height narrowing is 0.6, and the weighting coefficient corresponding to the time width narrowing is 0.4. The crosstalk intensity value is obtained by weighted summation.

[0071] It should be noted that the return path detour length reflects the actual distance the high-frequency return current travels around the edge of the housing due to the presence of contact interruption points. The longer the return path, the longer the coupling time of the current along the GPMI return channel shielding layer on the alternative path, and the more significant the crosstalk impact on adjacent data channels.

[0072] Specifically, the grounding return path refers to the current conduction path from the signal source through the conductive section of the outer casing's metal surface back to the grounding reference surface, which is obtained by tracing the conductivity continuity of each snap-fit ​​point on the outer casing's metal surface. The return path segment containing the contact interruption point is retrieved, and along the direction of the conductive section of the outer casing's metal surface, the nearest effective return nodes on both sides are traced sequentially from the contact interruption point. The arc lengths L1 and L2 along the outer casing's metal surface between adjacent effective return nodes on both sides of the contact interruption point are accumulated to obtain the total return path detour length L_total = L1 + L2. When three or more consecutive snap-fit ​​points around the contact interruption point are marked as suspended and fitted, the distance between the effective return nodes on both sides is larger, and the corresponding return path detour length value is higher. Furthermore, the pre-established detour length and judgment threshold correspondence table is obtained by measuring the crosstalk intensity of snap-fit ​​layouts with different suspension scales on the projection equipment sample stage. The correspondence table uses the detour length value as an index, divided into 20 levels in 5-millimeter increments, ranging from 0 to 100 mm. It records the judgment threshold value corresponding to each detour length level. The inverse correspondence rule is as follows: the larger the detour length level, the lower the corresponding judgment threshold value; the smaller the detour length level, the higher the corresponding judgment threshold value. This inverse correspondence rule corresponds to the physical trend that crosstalk depth intensifies with increasing detour length, enabling the use of a more sensitive judgment standard in scenarios with larger detour lengths. The detour length value of the return path is read, and its closest level in the correspondence table is determined using rounding rules. The judgment threshold value corresponding to this level is then obtained as the dynamic threshold value for judging crosstalk in the current adjacent data channel return.

[0073] Preferably, the dynamic threshold value is set between 15% and 40% of the reference eye diagram reference level height, with a higher value for shorter detour lengths and a lower value for longer detour lengths.

[0074] In one embodiment, the crosstalk intensity value is compared with the dynamic threshold value. If the crosstalk intensity value is higher than the dynamic threshold value, it is determined that the adjacent data channel has backhaul interference caused by the overhanging and sticking points on the shell edge. The adjacent data channel is marked as a backhaul interference channel, and the set of sticking point numbers of this channel is associated with the backhaul interference mark and stored in the backhaul interference identification result. If the crosstalk intensity value is lower than the dynamic threshold value, it is determined that the adjacent data channel does not have backhaul interference caused by the overhanging and sticking points on the shell edge. The adjacent data channel is marked as a normal channel and is not included in the backhaul interference identification result. The adjacent data channels corresponding to all the affected time periods are processed one by one to obtain the backhaul interference identification result caused by the overhanging and sticking points on the shell edge.

[0075] It is understood that the backhaul interference identification result is stored in the form of a record consisting of five fields: adjacent data channel number, corresponding contact interruption point buckle number, crosstalk intensity value, dynamic threshold value, and backhaul interference mark, reflecting the actual interference situation of the GPMI backhaul channel caused by the suspended contact buckle point on the edge of the shell.

[0076] S107. Extract the interference frequency band where the backhaul interference is located, determine the frequency range that the backhaul link of the projection device needs to avoid, and generate a communication status adjustment command for the backhaul link of the projection device.

[0077] For the adjacent data channels marked as backhaul interference channels in the backhaul interference identification results, the original waveform of the GPMI backhaul signal collected during the interference period of the backhaul interference channel is retrieved. A Fast Fourier Transform is used to convert the original waveform from the time domain to the frequency domain to obtain the backhaul signal spectrum of the backhaul interference channel. Frequency segments with amplitudes higher than a preset amplitude threshold are extracted from the backhaul signal spectrum as the interference frequency band where the backhaul interference is located. Based on the start and end frequency values ​​of the interference frequency band, the frequency range that the projection device's backhaul link needs to avoid is defined. The frequency range to be avoided is encapsulated as a carrier frequency switching instruction field and written into the communication status adjustment instruction of the projection device's backhaul link along with the backhaul interference channel number.

[0078] In one embodiment, for a neighboring data channel marked as a backhaul interference channel in the backhaul interference identification result, the original waveform of the GPMI backhaul signal acquired during the interference period of the backhaul interference channel is retrieved. A Fast Fourier Transform is used to convert the original waveform from a time-domain instantaneous voltage sequence into a frequency-domain complex sequence. The modulus of the complex sequence is then taken to obtain the backhaul signal spectrum of the backhaul interference channel. The backhaul signal spectrum, with frequency as the horizontal axis and amplitude as the vertical axis, reflects the energy distribution of the backhaul signal at each frequency component during the interference period.

[0079] Specifically, the amplitude values ​​of each frequency component along the frequency axis of the returned signal spectrum are compared with a preset amplitude threshold. If the amplitude values ​​of five or more consecutive frequency points are higher than the preset amplitude threshold, the frequency segment covered by these consecutive frequency points is marked as an interference band. The basis for setting the number of consecutive frequency points to five is as follows: in a typical 10 MHz bandwidth scanning scenario, the spectral resolution is 100 kHz, and the minimum bandwidth of a single interference signal is usually 500 kHz, corresponding to a width of five frequency points. Isolated peaks below this number are considered noise fluctuations. The preset amplitude threshold is set to three times the amplitude of the background noise in the returned signal spectrum. Exceeding this threshold indicates that there is significant crosstalk energy concentration in this frequency band. Furthermore, the start and end frequency values ​​of the interference band are retrieved, and the frequency range between the start and end frequency values ​​is defined as the frequency range that the projection device's return link needs to avoid. The communication status adjustment command consists of two parts: a carrier frequency switching command field and a return interference channel number field. The carrier frequency switching command field occupies 4 bytes, using 32-bit unsigned integer encoding to represent the target carrier frequency value in kilohertz. The return interference channel number field occupies 1 byte, with a value range of 0 to 255, identifying the specific interference channel number. The message format is carrier frequency switching command field first, followed by return interference channel number field, with a total length of 5 bytes, transmitted using big-endian encoding. In the example, if the frequency range to be avoided is 2,400,000 kilohertz to 2,483,500 kilohertz, and the interference channel number is 3, then the command message content is, in sequence, 4 bytes of carrier frequency field data 0x00249F00, and 1 byte of channel number field data 0x03.

[0080] This application also provides a projection device based on GPMI Super One-Line Communication technology, which performs the above-described projection device communication method based on GPMI Super One-Line Communication technology.

[0081] The above description is merely a preferred embodiment of this application and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in this application is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the concept of this application. For example, technical solutions formed by substituting the above features with (but not limited to) technical features with similar functions disclosed in this application.

Claims

1. A communication method for projection devices based on GPMI Super One-Line Communication technology, characterized in that, include: Obtain the module spacing between the camera module and the motherboard of the projection device, the bonding gap between the edge of the outer shell and the ground plane of the motherboard, and the edge pressing force of each snap-fit ​​point. Calibrate the solid bonding state or suspended bonding state of each snap-fit ​​point. Extract the contact resistance and measure the static potential based on the bonding state, bonding gap and edge pressing force of each snap-fit ​​point to determine the effective return current node and construct the grounding return current path. Analyze the return path length of each snap-fit ​​point in the grounding return path and its direction along the outer shell bonding surface to identify the concentrated distribution area of ​​high-frequency return current on the outer shell bonding surface; detect whether the concentrated distribution area includes suspended snap-fit ​​points. If it exists, determine that the high-frequency return current leaves the metal shell surface and forms an alternative path through the GPMI return channel shielding layer. Compare the level change to extract the disturbed period and collect the original waveform. The original waveform is aligned and superimposed for multiple cycles at the decision sampling time. The eye diagram opening of the return signal is extracted. The narrowing of the eye diagram opening level relative to the time width when there is no crosstalk is determined as the crosstalk evaluation basis for adjacent data channels. Based on the crosstalk evaluation basis and the return path length of the ground return current around the buckle point of the shell edge, the return interference caused by the shell edge being suspended and attached is identified.

2. The projection device communication method based on GPMI Super One-Line Communication technology according to claim 1, characterized in that, The process of acquiring the module spacing between the camera module and the motherboard of the projection device, the bonding gap between the edge of the outer shell and the ground plane of the motherboard, and the edge pressing force of each snap-fit ​​point, and calibrating the actual bonding state or the suspended bonding state of each snap-fit ​​point, includes: using the module spacing as a positioning reference for the snap-fit ​​distribution area on the edge of the outer shell, collecting bonding gaps along the edge of the outer shell to form a bonding gap sequence and locating the area of ​​each snap-fit ​​point, defining the effective collection range in combination with the module spacing, collecting the edge pressing force at each snap-fit ​​point, and establishing gap pressure matching data; for each snap-fit ​​point in the matching data, if the bonding gap is lower than a preset bonding threshold and the edge pressing force is higher than a preset pressure threshold, it is classified as a actual bonding state; if it exceeds or is lower than the threshold, it is classified as a suspended bonding state, thus obtaining the bonding state of each snap-fit ​​point on the edge of the outer shell.

3. The projection device communication method based on GPMI Super One-Line Communication technology according to claim 1, characterized in that, The process of extracting contact resistance and measuring static potential based on the contact state, contact gap, and edge pressing force of each snap-fit ​​point, and determining effective return current nodes to construct a grounding return current path, includes: for each snap-fit ​​point, substituting the contact gap and edge pressing force into the contact resistance correspondence table of the snap-fit ​​surface to obtain the contact resistance value, and establishing a contact resistance distribution according to the snap-fit ​​number; measuring the DC voltage of the metal surface of each snap-fit ​​point relative to the reference terminal of the motherboard ground plane under the motherboard's steady-state power-on condition to obtain the static potential, subtracting the static potential from the reference terminal potential and taking the absolute value to obtain the static potential deviation, forming resistance potential pairing data; for each snap-fit ​​point in the resistance potential pairing data, if the contact resistance is lower than a preset resistance threshold and the static potential deviation is lower than a preset potential threshold, it is determined to be an effective return current node, starting from the reference ground terminal of the camera module backplate, sequentially connecting all effective return current nodes along the metal surface of the outer shell and extending to the edge of the motherboard ground plane to obtain a grounding return current path.

4. The projection device communication method based on GPMI Super One-Line Communication technology according to claim 1, characterized in that, The analysis of the return path length and direction along the outer shell bonding surface of each snap-fit ​​point in the grounding return path, and the identification of concentrated distribution areas of high-frequency return current on the outer shell bonding surface, includes: measuring the arc length between adjacent effective return nodes along the conductive segment of the outer shell metal surface for each effective return node in the grounding return path to obtain the node spacing, and connecting them in series according to the snap-fit ​​distribution order of the outer shell edge to form a return path length sequence; projecting the geometric position of each path segment onto the outer shell bonding surface to obtain the direction trajectory, and mapping the length sequence and direction trajectory according to the effective return node number to form a path length direction distribution along the outer shell edge; for each path segment in the distribution, if the path length is lower than a preset length threshold and the angle between the tangent directions of adjacent path segments is lower than a preset angle threshold, it is determined to be a high-frequency return current dense convergence segment, and adjacent dense convergence segments are merged into the same outer shell segment to obtain a concentrated distribution area. The interference frequency band of the return interference is extracted, the frequency range that the projection device's return link needs to avoid is determined, and a communication status adjustment command for the projection device's return link is generated.

5. The projection device communication method based on GPMI Super One-Line Communication technology according to claim 1, characterized in that, The detection process involves checking whether the concentrated distribution area contains suspended snap-fit ​​points. If so, it is determined that the high-frequency return current leaves the metal casing surface and forms an alternative path through the GPMI return channel shielding layer. The disturbance period is extracted by comparing the level changes, and the original waveform is acquired. This includes: retrieving the bonding state distribution of each snap-fit ​​point on the casing edge; verifying the bonding state markers corresponding to the effective return current nodes within the concentrated distribution area; if suspended bonding snap-fit ​​points exist, they are determined as contact interruption points, and the coordinates of these interruption points on the casing bonding surface are obtained as the starting position of the alternative path. Based on the alternative path... The path starts by collecting the return signal level sequence of the most recent GPMI return data channel before the reference period and after the current period. The signal level is then aligned with the sampling clock and the absolute value of the subtraction is taken point by point to form the level change. The level change is then arranged along the time axis to form the level transition curve of the return signal. The level change is then compared point by point with the preset transition threshold. If the change is continuously higher than the threshold on consecutive timestamps, it is determined to be a disturbed period. During the disturbed period, the original waveform data of the adjacent data channel is collected synchronously according to the sampling clock of the GPMI protocol return link.

6. The projection device communication method based on GPMI Super One-Line Communication technology according to claim 1, characterized in that, The process of performing multi-cycle alignment and superposition on the decision sampling time of the original waveform, extracting the eye diagram opening of the return signal, and determining the narrowing of the eye diagram opening relative to the time width when there is no crosstalk as the basis for crosstalk evaluation of adjacent data channels includes: for the original waveform of the GPMI return signal during the interference period, marking the start and end boundaries of each symbol period according to the decision sampling time position of the GPMI protocol, dividing the original waveform into multiple equal-length waveform segments, and superimposing them along the time axis with the decision sampling time as the reference to obtain a multi-cycle aligned waveform cluster; taking the electrical signal of each waveform segment at each time offset point. The lower bound of the envelope of the average value is connected to form the upper edge of the eye diagram, and the upper bound of the envelope is connected to form the lower edge of the eye diagram. The area enclosed by the upper edge and the lower edge of the eye diagram near the decision sampling time is extracted as the eye diagram opening. The level difference between the upper edge and the lower edge of the eye diagram is used as the level height, and the time span between the intersection positions is used as the time width. The reference eye diagram pre-recorded under the crosstalk-free state is retrieved. The level height narrowing is obtained by subtracting the level height of the eye diagram opening from the reference level height, and the time width narrowing is obtained by subtracting the time width from the reference time width. The level height narrowing and the time width narrowing together form the basis for crosstalk evaluation.

7. The projection device communication method based on GPMI Super One-Line Communication technology according to claim 1, characterized in that, The step of identifying backhaul interference caused by the suspended contact of the outer shell edge, based on the crosstalk assessment criteria and the length of the return path around the edge of the casing, includes: extracting the level height narrowing and time width narrowing from the crosstalk assessment criteria, and weighting and summing them according to a preset weighting coefficient to obtain the crosstalk intensity value; retrieving the return path segment containing the contact interruption point in the grounding return path, and accumulating the arc length between adjacent effective return nodes on both sides of the contact interruption point along the conductive segment of the metal surface of the casing to obtain the return path bypass length; querying the correspondence table between the bypass length and the judgment threshold based on the return path bypass length to read the current judgment threshold as the dynamic threshold value; if the crosstalk intensity value is higher than the dynamic threshold value, it is determined that there is backhaul interference caused by the suspended contact of the outer shell edge in the adjacent data channel and marked as a backhaul interference channel; if it is lower, it is determined that there is no backhaul interference, thus obtaining the backhaul interference identification result.

8. The projection device communication method based on GPMI Super One-Line Communication technology according to claim 1, characterized in that, The steps of extracting the interference frequency band where the backhaul interference is located, determining the frequency range that the projection device's backhaul link needs to avoid, and generating a communication status adjustment command for the projection device's backhaul link include: for the adjacent data channels marked as backhaul interference channels in the backhaul interference identification results, retrieving the original waveform of the GPMI backhaul signal collected during the interference period of the channel, using spectrum analysis to convert the original waveform from the time domain to the frequency domain to obtain the backhaul signal spectrum, and extracting frequency segments with amplitudes higher than a preset amplitude threshold from the backhaul signal spectrum as the interference frequency band where the backhaul interference is located.

9. The projection device communication method based on GPMI Super One-Line Communication technology according to claim 8, characterized in that, Generate a communication status adjustment instruction for the projection device backhaul link, including: defining the frequency range that the backhaul link needs to avoid based on the start and end frequency values ​​of the interference band, encapsulating the avoidance frequency range as a carrier frequency switching instruction field, and writing it together with the backhaul interference channel number into the backhaul link communication status adjustment instruction.

10. A projection device based on GPMI Super One-Line Technology, characterized in that, The projection device communication method based on GPMI Super One-Line Communication technology as described in claim 1 is implemented.

Images