A PCB line detection method for integrated circuit manufacturing

Abstract

This invention relates to the field of PCB circuit testing technology, specifically disclosing a PCB circuit testing method for integrated circuit manufacturing, including a flying probe testing device. The flying probe testing device is equipped with an independently moving first probe and a second probe, and includes the following steps: Step S1: Obtain the position of all testing points, and configure an estimated waiting time for each point based on historical data, classify the test tasks into single-needle or double-needle tasks, and configure the estimated completion time for each task; Step S2: Filter out long-waiting testing points; for each long-waiting point performing a double-needle task, select a group of other testing points whose total single-needle task completion time does not exceed the waiting time of that point, forming a collaborative combination; Step S3: During testing, the first probe performs a double-needle task at the long-waiting point, while the second probe uses its waiting time to sequentially complete all single-needle test tasks in the corresponding collaborative combination; finally, an efficient testing path is generated.

Description

Technical Field

[0001] This invention relates to the field of PCB circuit testing technology, and more specifically to a PCB circuit testing method for integrated circuit manufacturing. Background Technology

[0002] In the printed circuit board (PCB) manufacturing process, flying probe testing equipment is a key electrical testing device. Its working principle is to complete open and short circuit detection by contacting PCB pads with high-precision probes. Traditional testing equipment uses a preset fixed probe movement path, which has significant technical defects in actual circuit testing. For example, when there are process abnormalities such as missing components or displacement on the PCB board, the equipment still performs the testing task according to the preset path, resulting in a large amount of invalid probe displacement.

[0003] Existing technical solutions mainly shorten the testing cycle by optimizing the mechanical structure or increasing the driving speed, but they fail to fundamentally solve the core problem of unreasonable path planning. Summary of the Invention

[0004] The purpose of this invention is to provide a PCB circuit testing method for integrated circuit manufacturing, and to solve the following technical problems.

[0005] The objective of this invention can be achieved through the following technical solutions:

[0006] A PCB circuit inspection method for integrated circuit manufacturing includes a flying probe testing device equipped with an independently moving first probe and a second probe, comprising the following steps:

[0007] Step S1: Obtain all detection points on the PCB board under test, and obtain the location information of all detection points. Configure the estimated waiting time for each detection point based on historical data. The estimated waiting time is the time consumed for the signal to stabilize after the probe contacts the PCB board under test.

[0008] All test tasks at the test point are obtained, and the test tasks are divided into single-needle test tasks and double-needle test tasks. Based on historical data, the estimated completion time of each test task at the test point is configured. The estimated completion time is the time consumed to complete the test task.

[0009] Step S2: Select the detection points whose estimated waiting time exceeds the preset time threshold and record them as long waiting detection points; when the long waiting detection points perform dual-needle test tasks, take the estimated waiting time of the long waiting detection points as the time window, select a group of detection points whose total estimated completion time of single-needle test tasks is less than or equal to the time window, and form a collaborative combination of long waiting detection points.

[0010] Step S3: The first probe performs a dual-needle test on each long-waiting detection point, and within the estimated waiting time of the long-waiting detection point, the second probe sequentially completes the single-needle test on each adapted collaborative detection point within the collaborative combination; the final detection path is obtained from each long-waiting detection point and its collaborative combination.

[0011] As a further aspect of the present invention: the process of obtaining the location information of all detection points includes:

[0012] Choose any detection point on the PCB board under test as the origin, establish a coordinate system, and obtain the position coordinates of the detection point in the coordinate system, which are recorded as the position information of the detection point.

[0013] As a further aspect of the present invention: the process of obtaining the estimated waiting time for each detection point based on historical data includes:

[0014] The historical data includes the historical waiting time and historical completion time of each detection point when performing several historical test tasks; based on the historical data, the historical waiting time of each detection point is obtained, and the average of the historical estimated completion time is obtained, which is recorded as the estimated waiting time of the detection point.

[0015] As a further aspect of the present invention: the single-needle test task is a test task performed independently by a single probe, and the dual-needle test task is a test task that must be performed by two probes.

[0016] As a further aspect of the present invention: the estimated completion time of the test task at the detection point is the average of several historical completion times obtained based on historical data.

[0017] As a further aspect of the present invention: the process of setting the duration threshold includes:

[0018] Obtain the estimated waiting time for all detection points, and then calculate the average value t of all estimated waiting times. ave And obtain the standard deviation σ of all estimated waiting times; set a multiplier threshold Z, wherein the duration threshold is set to T=t ave +Z×σ.

[0019] As a further aspect of the present invention: the process of selecting a set of detection points using the estimated waiting time of long-waiting detection points as a time window includes selecting a set of detection points whose estimated completion time meets the following criteria: Where time is the time window, t i Let be the estimated completion time for the i-th detection point, and n be the total number of selected detection points.

[0020] As a further aspect of the present invention: the process of obtaining the final detection path includes:

[0021] Long-waiting detection points and their coordinated combinations are denoted as a detection unit. All detection units are arranged and combined to obtain several detection unit sequences. Based on the position information of each detection point, the total displacement of the first probe and the second probe in each detection unit sequence is obtained. The detection unit sequence with the smallest total displacement is selected as the final detection path.

[0022] The beneficial effects of this invention are:

[0023] This method systematically acquires and analyzes historical test data of the test points, accurately configures the estimated waiting time and task completion time of each point, and thus identifies long-waiting test points that affect test efficiency. It transforms the inherent waiting time window of the dual-needle test task at long-waiting test points into a resource that can be used in parallel. By intelligently matching a group of collaborative test points whose total single-needle task time does not exceed the waiting window for each long-waiting point, an efficient collaborative combination is constructed. During the actual execution of the test path, the first probe executes the dual-needle task at the long-waiting point, while the second probe simultaneously utilizes the waiting time window to efficiently complete all single-needle tasks within the collaborative combination. This parallel collaboration mode of the dual probes fundamentally reduces the inherent idle waiting time of the equipment. Furthermore, by optimizing and integrating all long-waiting points and their collaborative combinations into a test path with the minimum total displacement, this method achieves systematic and refined scheduling of equipment time and space resources without changing the hardware, significantly improving test throughput. It effectively solves the core technical bottleneck of large amounts of invalid probe displacement and time wastage in the traditional fixed-path mode, providing a new and efficient solution for PCB circuit testing. Attached Figure Description

[0024] The invention will now be further described with reference to the accompanying drawings.

[0025] Figure 1 This is a schematic diagram illustrating the steps of a PCB circuit inspection method for integrated circuit manufacturing according to the present invention. Detailed Implementation

[0026] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0027] Please see Figure 1 As shown, this invention is a PCB circuit inspection method for integrated circuit manufacturing, including a flying probe testing device. The flying probe testing device is equipped with an independently moving first probe and a second probe, and includes the following steps:

[0028] Step S1: Obtain all detection points on the PCB board under test, and obtain the location information of all detection points. Configure the estimated waiting time for each detection point based on historical data. The estimated waiting time is the time consumed for the signal to stabilize after the probe contacts the PCB board under test.

[0029] All test tasks at the test point are obtained, and the test tasks are divided into single-needle test tasks and double-needle test tasks. Based on historical data, the estimated completion time of each test task at the test point is configured. The estimated completion time is the time consumed to complete the test task.

[0030] Specifically, the Gerber design file of the PCB board under test is imported into the host computer software of the flying probe testing device, and the coordinate data of all test points in the file is parsed. Taking a typical four-layer PCB board as an example, a rectangular coordinate system is established with the positioning hole at the lower left corner of the board as the origin to read the (X, Y) coordinates of a total of 1528 test points, and all position information is stored in the test task queue.

[0031] In a preferred embodiment of the present invention, the process of obtaining the location information of all detection points includes:

[0032] Select any detection point on the PCB board to be tested as the origin, establish a coordinate system, obtain the position coordinates of the detection point in the coordinate system, and record them as the position information of the detection point.

[0033] In a preferred embodiment of the present invention, the process of obtaining the estimated waiting time for each detection point based on historical data includes:

[0034] The historical data includes the historical waiting time and historical completion time of each detection point when performing several historical test tasks; based on the historical data, the historical waiting time of each detection point is obtained, and the average of the historical estimated completion time is obtained, which is recorded as the estimated waiting time of the detection point.

[0035] Specifically, the historical testing database of 500 PCBs of the same model produced in the last three months is accessed to obtain historical data; for each testing point, the signal stabilization time data in the historical data is statistically analyzed.

[0036] For example, for the power network detection point PWR_3V3_A1 in the BGA package area, calculate its historical waiting time for the past 2000 detections, remove outliers and take the average value to finally determine its estimated waiting time.

[0037] In a preferred embodiment of the present invention, the single-needle test task is a test task performed independently by a single probe, and the dual-needle test task is a test task that must be performed by two probes.

[0038] It is worth noting that, in the specific implementation process, the time required for signal stabilization refers to the entire time interval from the moment the probe of the flying probe tester physically contacts the test pad on the PCB board until the electrical signal of the circuit at that test point decays from its initial fluctuating state to within the stable threshold range required by the test. This process is mainly affected by the inherent electrical characteristics of the circuit, such as distributed capacitance, parasitic inductance, and network complexity. For example, when testing power networks or circuits with large capacitive loads, the RC charging and discharging circuit formed after the probe contacts will prolong the voltage build-up process, which may result in a stabilization time of up to hundreds of milliseconds. However, for simple signal continuity tests, the distributed parameters in the circuit are small, and the signal can reach a stable state within tens of milliseconds. Therefore, the waiting time is a necessary technical delay to ensure the accuracy and reliability of subsequent electrical performance measurement data. If sampling and measurement are performed before the signal stabilizes, it will directly lead to misjudgment of the test results.

[0039] Specifically, based on the electrical characteristics of the test tasks, tasks that can be completed independently by a single probe, such as resistance measurement and single network continuity testing, are classified as single-needle test tasks, while tasks that require two probes to simultaneously apply excitation and acquire signals, such as inter-network insulation resistance testing and precision capacitance measurement, are classified as dual-needle test tasks.

[0040] In a preferred embodiment of the present invention, the estimated completion time of the test task at the detection point is the average of several historical completion times obtained based on historical data.

[0041] For each test task, its average execution time is also calculated based on the historical data. For example, the estimated completion time of the continuity test task at point GND_B2 is set to 120 milliseconds.

[0042] Step S2: Select the detection points whose estimated waiting time exceeds the preset time threshold and record them as long waiting detection points; when the long waiting detection points perform dual-needle test tasks, take the estimated waiting time of the long waiting detection points as the time window, select a group of detection points whose total estimated completion time of single-needle test tasks is less than or equal to the time window, and form a collaborative combination of long waiting detection points.

[0043] In a preferred embodiment of the present invention, the process of setting the duration threshold includes:

[0044] Obtain the estimated waiting time for all detection points, and then calculate the average value t of all estimated waiting times. ave And obtain the standard deviation σ of all estimated waiting times; set a multiplier threshold Z, wherein the duration threshold is set to T=t ave +Z×σ;

[0045] Specifically, a deep analysis is performed on the historical detection database constructed in step S1. Based on the statistical outlier identification principle, the time threshold calculation parameter Z=1.5 is set. The average estimated waiting time of all detection points is 185 milliseconds, and the standard deviation is 82 milliseconds. Then, the time threshold T=185+1.5×82=308 milliseconds is obtained. A total of 89 detection points with estimated waiting times exceeding the threshold are selected and marked as long waiting detection points.

[0046] In a preferred embodiment of the present invention, the process of selecting a set of detection points using the estimated waiting time of long-waiting detection points as a time window includes selecting a set of detection points whose estimated completion time meets the following criteria: Where time is the time window, t i Let n be the estimated completion time for the i-th detection point, and n be the total number of selected detection points.

[0047] Taking point PWR_1V8_C5 as an example, its estimated waiting time of 420 milliseconds for the dual-needle test task is used as the available time window. From the remaining 1439 test points, points that are geographically close and are single-needle test tasks are selected for combination optimization. Through dynamic programming algorithm, the optimal collaborative combination of 3 test points is matched: point NET_A12 (single-needle continuity test, estimated completion time 135 milliseconds), point NET_B8 (single-needle resistance test, estimated completion time 155 milliseconds), and point NET_C3 (single-needle diode test, estimated completion time 125 milliseconds). The total estimated completion time of the three is 415 milliseconds, which fully meets the 420 millisecond time window requirement.

[0048] A similar collaborative combination was established for each long-waiting detection point, and a corresponding collaborative detection relationship mapping table was generated.

[0049] Step S3: The first probe performs a dual-needle test on each long-waiting detection point, and within the estimated waiting time of the long-waiting detection point, the second probe sequentially completes the single-needle test on each adapted collaborative detection point within the collaborative combination; the final detection path is obtained from each long-waiting detection point and its collaborative combination.

[0050] In a preferred embodiment of the present invention, the process of obtaining the final detection path includes:

[0051] Long-waiting detection points and their coordinated combinations are denoted as a detection unit. All detection units are arranged and combined to obtain several detection unit sequences. Based on the position information of each detection point, the total displacement of the first probe and the second probe in each detection unit sequence is obtained. The detection unit sequence with the smallest total displacement is selected as the final detection path.

[0052] Specifically, the 89 long-waiting detection points selected in step S2 and their corresponding cooperative combinations are constructed as detection units. Taking the detection unit containing point PWR_1V8_C5 and its three cooperative detection points (NET_A12, NET_B8, NET_C3) as an example, a genetic algorithm is used to optimize the global path of all detection units. By evaluating the total movement distance of the first and second probes in different detection unit sequences, the optimal unit access sequence, i.e. the final detection path, is finally determined.

[0053] During the actual testing execution phase, when the probe arrives at the testing unit according to the planned path, the first and second probes first work together to locate point PWR_1V8_C5, jointly perform the dual-needle test task, and start a 420-millisecond wait timer. While waiting for the signal to stabilize, the second probe immediately leaves the point and proceeds to points NET_A12, NET_B8, and NET_C3 in a pre-optimized order to perform single-needle test tasks. Actual monitoring data shows that the total time taken for the second probe to complete the tests at these three testing points is 408 milliseconds, and it returns to the standby state 12 milliseconds before the end of the wait at point PWR_1V8_C5.

[0054] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the present invention should still fall within the scope of the present invention.

Claims

1. A method for inspecting PCB circuits in integrated circuit manufacturing, characterized in that, The device includes a flying probe testing apparatus, which is equipped with a first probe and a second probe that move independently, and includes the following steps: Step S1: Obtain all detection points on the PCB board under test, and obtain the location information of all detection points. Configure the estimated waiting time for each detection point based on historical data. The estimated waiting time is the time consumed for the signal to stabilize after the probe contacts the PCB board under test. All test tasks at the test point are obtained, and the test tasks are divided into single-needle test tasks and double-needle test tasks. Based on historical data, the estimated completion time of each test task at the test point is configured. The estimated completion time is the time consumed to complete the test task. Step S2: Select the detection points whose estimated waiting time exceeds the preset time threshold and record them as long waiting detection points; when the long waiting detection points perform dual-needle test tasks, take the estimated waiting time of the long waiting detection points as the time window, select a group of detection points whose total estimated completion time of single-needle test tasks is less than or equal to the time window, and form a collaborative combination of long waiting detection points. Step S3: The first probe performs a dual-needle test on each long-waiting detection point, and within the estimated waiting time of the long-waiting detection point, the second probe sequentially completes the single-needle test on each compatible collaborative detection point within the collaborative combination; the final detection path is obtained from each long-waiting detection point and its collaborative combination. In step S1, the process of obtaining the estimated waiting time for each detection point based on historical data includes: The historical data includes the historical waiting time and historical completion time of each detection point when performing several historical test tasks; based on the historical data, the historical waiting time of each detection point is obtained, and the average of the historical estimated completion time is obtained, which is recorded as the estimated waiting time of the detection point. In step S2, the process of setting the duration threshold includes: Obtain the estimated waiting time for all detection points, and then calculate the average value t of all estimated waiting times. ave And obtain the standard deviation σ of all estimated waiting times; set a multiplier threshold Z, wherein the duration threshold is set to T=t ave +Z×σ; In step S2, the process of selecting a set of detection points using the estimated waiting time of long-waiting detection points as a time window includes selecting a set of detection points whose estimated completion time meets the following criteria: Where time is the time window, t i Let n be the estimated completion time for the i-th detection point, and n be the total number of selected detection points. In step S3, the process of obtaining the final detection path includes: Long-waiting detection points and their coordinated combinations are denoted as a detection unit. All detection units are arranged and combined to obtain several detection unit sequences. Based on the position information of each detection point, the total displacement of the first probe and the second probe in each detection unit sequence is obtained. The detection unit sequence with the smallest total displacement is selected as the final detection path.

2. The PCB circuit inspection method for integrated circuit manufacturing according to claim 1, characterized in that, In step S1, the process of obtaining the location information of all detection points includes: Choose any detection point on the PCB board under test as the origin, establish a coordinate system, and obtain the position coordinates of the detection point in the coordinate system, which are recorded as the position information of the detection point.

3. The PCB circuit inspection method for integrated circuit manufacturing according to claim 1, characterized in that, In step S1, the single-needle test task is a test task that is performed independently by a single probe, and the dual-needle test task is a test task that must be performed by two probes.

4. The PCB circuit inspection method for integrated circuit manufacturing according to claim 1, characterized in that, In step S1, the estimated completion time of the test task at the detection point is the average of several historical completion times obtained based on historical data.

Images