A back-drilling PCB measurement method and system based on short-wave infrared OCT

Abstract

The present application relates to the technical field of PCB measurement, and particularly relates to a PCB back-drilling structure measurement method based on short-wave infrared OCT and a use method thereof. The method utilizes the relative penetration of epoxy resin dielectric base material to short-wave infrared light and the high reflectivity of copper layer, and makes the measurement light beam incident to the back-drilling hole position along the axial direction to obtain the axial reflection profile. The reflection peaks corresponding to the residual copper top end copper section and the target layer horizontal copper surface are identified, and the residual copper length is converted by the optical path difference and the dielectric material group refractive index. The dielectric layer thickness can also be measured by the adjacent copper interface. The method is a non-destructive, non-contact single-sided reflection measurement, and the axial resolution can reach the micron level. The method can locally measure deep holes with a depth-width ratio of fourteen, has online measurement capability, and effectively solves the problems of large destructiveness, low efficiency or insufficient resolution in the prior art.

Description

Technical Field

[0001] This invention relates to the field of PCB measurement technology, and in particular to a method and system for measuring PCB back-drilled structures based on short-wave infrared OCT. Background Technology

[0002] In high-speed, high-frequency multilayer printed circuit boards, excess copper plating extending beyond the signal transmission layer after via plating can form branched segments. These segments generate impedance discontinuities and resonance effects under high-speed signals, severely degrading signal integrity. To eliminate this effect, the industry commonly uses back-drilling to remove excess copper plating along the depth direction. Because the drill tip is tapered, a section of copper plating remains between the bottom of the back-drilled hole and the target signal layer. Its axial length is the residual copper length, typically ranging from 0.01 to 0.25 mm. The control precision of this residual copper length directly affects the quality of high-speed signals, making its accurate measurement extremely critical.

[0003] Existing methods for measuring residual copper length mainly include metallographic sectioning, X-ray or computed tomography (CT) scanning, and time-domain reflectometry (TDAR), among other electrical measurement methods. Metallographic sectioning is a destructive testing method, time-consuming sample preparation, and unsuitable for online full inspection. X-ray / CT methods are expensive, have radiation protection requirements, and suffer from insufficient contrast at the copper-dielectric interface at certain angles, making it difficult to match measurement speed with production line cycles. TDAR is an indirect estimation method with limited spatial resolution, requires contact fixtures, and struggles to accurately locate the physical length of residual copper at individual holes. Therefore, the industry urgently needs a technical solution that is non-destructive, optically non-contact, has micron-level axial resolution, and allows for localized measurement of specific holes. Summary of the Invention

[0004] In view of the problems existing in the prior art, the present invention is proposed.

[0005] Therefore, the present invention aims to address the problems of existing back-drilled residual copper length measurement methods being destructive, inefficient, or lacking spatial resolution, by providing a non-destructive, non-contact measurement method with micron-level axial resolution.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a PCB back-drilled structure measurement method based on short-wave infrared OCT, comprising: providing a short-wave infrared broadband light source, and dividing the light beam of the light source into a measurement beam and a reference beam via an interferometric structure; focusing the measurement beam axially onto the hole to be measured on the printed circuit board processed by back drilling; acquiring the interference signal after the interference of the measurement beam and the reference beam reflected from the printed circuit board, and obtaining the axial reflection profile of the hole to be measured from the interference signal; identifying reflection peaks corresponding to the interface group to be measured in the axial reflection profile, the interface group to be measured including: the top and bottom interfaces of the residual copper, and / or the upper and lower horizontal interfaces of the dielectric layer to be measured; and using the refractive index of the dielectric material, converting the optical path difference between the reflection peaks of the interface group to be measured into a solid dimension to obtain the residual copper length and / or the dielectric layer thickness of the dielectric layer to be measured.

[0007] As a preferred embodiment of the PCB back-drilling structure measurement method based on short-wave infrared OCT described in this invention, wherein: when the length of the residual copper is obtained, the top interface of the residual copper is the copper top surface formed by cutting the residual copper through back drilling, the bottom interface is the horizontal copper surface of the target signal layer, and both the copper top surface and the horizontal copper surface are copper interfaces.

[0008] As a preferred embodiment of the PCB back-drilling structure measurement method based on short-wave infrared OCT described in this invention, the axial position of the reflection peak on the copper top surface is determined by a fixed criterion, which is the leading edge or intensity centroid of the bright band of the reflection peak, and the reference point of the copper top surface corresponding to the fixed criterion is corrected by a standard sample with a known copper length.

[0009] As a preferred embodiment of the PCB back-drilled structure measurement method based on short-wave infrared OCT of the present invention, the measured hole is a back-drilled hole, the back-drilled hole has a back-drilled hole diameter and a center hole diameter smaller than the back-drilled hole diameter, and the reflected signal of the copper top surface and the horizontal copper surface originates from an annular region surrounding the center hole with a radial width equal to half the difference between the back-drilled hole diameter and the center hole diameter; the method further includes: illuminating the bottom of the back-drilled hole with the measurement beam, and using the coherence gate of optical coherence tomography to separate the reflection peak of the copper top surface and the reflection peak of the horizontal copper surface in the axial direction, thereby obtaining the residual copper length from the axial distance between the two reflection peaks.

[0010] As a preferred embodiment of the PCB back-drilling structure measurement method based on short-wave infrared OCT described in this invention, when the reflection peak corresponding to the top copper surface and the reflection peak corresponding to the horizontal copper surface are close to each other or partially overlap, peak fitting or deconvolution processing is performed on the signal segments where the two reflection peaks are located, and / or numerical dispersion compensation is performed on the interference signal to estimate the respective center positions of the two reflection peaks and restore the optical path difference between them, so that the identifiable residual copper length is less than the nominal axial resolution of the light source.

[0011] As a preferred embodiment of the PCB back-drilled structure measurement method based on short-wave infrared OCT described in this invention, the measured hole is a back-drilled hole with an aspect ratio of at least 10, and the back-drilled hole is an unfilled hole. The measurement beam is focused by an objective lens, the working distance of the objective lens is greater than the depth of the back-drilled hole, and the numerical aperture of the objective lens is selected so that the measurement beam is not blocked by the hole wall within the depth range of the back-drilled hole. By utilizing the characteristic of decoupling the axial resolution of optical coherence tomography from the numerical aperture, micron-level axial resolution is maintained even at low numerical apertures.

[0012] As a preferred embodiment of the PCB back-drilling structure measurement method based on short-wave infrared OCT described in this invention, the center wavelength of the short-wave infrared broadband light source is selected at the penetration window to avoid the overtone absorption band of the dielectric material, and the center wavelength is approximately 1310 nanometers.

[0013] As a preferred embodiment of the PCB back-drilling structure measurement method based on short-wave infrared OCT described in this invention, the refractive index is the group refractive index of the dielectric material in the measurement band; the method further includes: correcting the group refractive index using a standard sample of known thickness.

[0014] As a preferred embodiment of the PCB back-drilling structure measurement method based on short-wave infrared OCT described in this invention, the step of obtaining the axial reflection profile includes: taking multiple measurement points in the same area to be measured and performing statistical averaging, or aligning the measurement points with specific morphological positions of the fiberglass texture to suppress measurement variations caused by local material inhomogeneity.

[0015] The second objective of this invention is to provide a PCB back-drilled structure measurement system based on short-wave infrared OCT, comprising the following steps: a short-wave infrared broadband light source; an interference architecture that divides the light beam of the light source into a measurement beam and a reference beam, and causes the measurement beam reflected from the printed circuit board under test to interfere with the reference beam; a probe configured to focus the measurement beam axially onto the back-drilled hole of the printed circuit board under test; a detector configured to acquire the interference signal; and a signal processing unit configured to execute the above method.

[0016] The beneficial effects of this invention are as follows: This invention utilizes the penetrability of short-wave infrared light to epoxy dielectric substrates and the high reflectivity of copper layers. Axial reflection profiles are obtained through optical coherence tomography (OCT). The length of the residual copper is accurately calculated from the optical path difference between the strong reflection peaks of the copper section at the top of the residual copper and the horizontal copper surface of the target layer. The dielectric layer thickness can also be measured through adjacent horizontal copper interfaces. This method is entirely non-destructive and non-contact, allowing for hole-by-hole inspection of mass-produced products. Its axial resolution, determined by the light source bandwidth, reaches the micrometer level, superior to the spatial resolution of electrical measurement methods. Furthermore, the axial resolution is decoupled from the focusing numerical aperture, enabling the measurement of back-drilled holes with aspect ratios up to 14 even at low numerical apertures. The measurement speed is fast when using a swept-frequency light source architecture, possessing the potential for online measurement. Blind structure reflective measurement can be completed with only single-sided incident light from the board surface, eliminating the need for double-sided alignment. Multiple copper interfaces can be simultaneously resolved within the material's light transmission range in a single axial scan, enabling single-layer dielectric thickness measurement. This method is more efficient and applicable than known metallographic sections, X-rays, and electrical measurement methods. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic diagram of the measurement system architecture of the present invention.

[0019] Figure 2 This is a cross-sectional view of the back drill structure and a schematic diagram of the measurement optical path of the present invention.

[0020] Figure 3 This is a schematic diagram of the axial reflection profile (A-scan) signal of the present invention. Detailed Implementation

[0021] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0022] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0023] Secondly, the term "an embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places throughout this specification does not necessarily refer to the same embodiment, nor is it an embodiment that is mutually exclusive, either alone or selectively, with other embodiments.

[0024] See Figure 1 This is a schematic diagram of the architecture of an embodiment of the measurement system of the present invention. The measurement system includes: a short-wave infrared broadband light source (210), an interference architecture (fiber optic coupler / interferometer, 220), a probe (including objective lens, 230), a detector (240), and a signal processing unit (250). The short-wave infrared broadband light source (210) is used to emit broadband short-wave infrared light; the interference architecture (220) divides the light source beam into a measurement beam and a reference beam, the reference beam pointing to a reference mirror, and the measurement beam is axially focused by the probe (230) onto a test hole of a printed circuit board (100) under test. After the measurement beam reflected from each interface in the printed circuit board returns, it is recoupled with the reference beam returning from the reference mirror in the interference architecture to generate interference. The interference signal is captured by the detector (240) and then processed by the signal processing unit (250) to obtain an axial reflection profile (A-scan) of the test hole.

[0025] In a preferred embodiment, the system employs a swept-source optical coherence tomography (SS-OCT) architecture. The output wavelength of the short-wave infrared broadband light source (210) is scanned over time, and the axial reflection profile is obtained by Fourier transforming the time-series interference signal captured by the detector (240). This architecture has the advantages of fast measurement speed, low sensitivity decay with depth, and a measurement range of several millimeters to tens of millimeters, making it particularly suitable for online measurement scenarios involving aperture scanning. In another embodiment, a spectral-domain architecture can also be used, employing a beam splitter and a linear array detector to obtain spectral interference signals. The detector (240) is preferably an indium gallium arsenide (InGaAs) detector, whose sensitive band matches the output wavelength of the short-wave infrared broadband light source.

[0026] See Figure 2This is a cross-section of the back-drilled structure of the printed circuit board under test and a schematic diagram of the measurement optical path. The printed circuit board under test (100) includes multiple copper layers and epoxy resin dielectric layers (140) located between the copper layers. The L0 copper layer (110), L1 copper layer (120), and L2 copper layer (130) are shown sequentially from top to bottom in the figure. The back-drilling process uses a larger diameter drill bit to drill away excess copper plating in the vias to a controlled depth from the board surface (L0 side). Because the drill bit tip is conical, a conical surface (161) is formed at the bottom of the back-drilled hole. This conical surface (161) does not reach the target signal layer L1 (120), so copper is plated on the section between the conical surface (161) and L1, which is the residual copper (160). The length of the residual copper along the axial direction is the stub length, which is typically between about 0.01 mm and 0.25 mm. As shown in the figure, the back-drilled hole diameter ( (e.g., 0.35 mm) is larger than the original center aperture ( (e.g., 0.2 mm), so the reflective features available for measuring residual copper—namely, the conical surface (161) and the L1 annular horizontal copper surface (land, 162) below it—exist only around the central hole, with a radial width of approximately Within a ring-shaped area (e.g., 0.075 mm). Diameter The center hole is an open hole. If the measuring beam is incident along the center, it will pass straight to the bottom of the board and the reflected signal of the residual copper cannot be obtained.

[0027] The core principle of the measurement method of this invention is based on the comparison of the optical properties of the following materials: epoxy resin dielectric substrates have relatively good transmittance to short-wave infrared light, while copper layers have extremely high reflectivity (up to 95%) for short-wave infrared light (wavelength greater than 1 micrometer). Accordingly, when a short-wave infrared beam is incident axially, horizontal copper interfaces at different depths will generate strong reflection signals, appearing as discrete reflection peaks in the axial reflection profile. By measuring the optical path difference between each reflection peak and converting it to the refractive index of the dielectric material, the corresponding solid depth or thickness can be obtained. When using a broadband short-wave infrared light source, the axial resolution can reach several micrometers, sufficient to resolve variations in residual copper length and dielectric layer thickness on the order of 10 to 250 micrometers.

[0028] The axial resolution Δz of optical coherence tomography is related to the center wavelength λ and bandwidth Δλ of the light source, and the approximate relationship is as follows:

[0029] (1)

[0030] Since optical coherence tomography measures the optical path (optical path = refractive index × solid thickness), the conversion of solid thickness must be divided by the refractive index of the dielectric material under test at the measurement wavelength:

[0031] (2)

[0032] Where OPD is the measured optical path difference. This represents the group refractive index of the dielectric material in the measurement wavelength range. When using a broadband light source, due to the influence of material dispersion, the conversion should use the group refractive index rather than the phase refractive index to improve measurement accuracy.

[0033] In the embodiment for measuring the length of residual copper, the measuring beam of the probe (230) is incident axially onto the back-drilled hole. Since the residual copper segment is mainly distributed on the vertical hole wall of the via, the specular reflection of the incident light from the vertical wall surface is not easily returned to the probe. Therefore, this method does not directly measure the reflection of the vertical copper wall, but indirectly defines the upper and lower endpoints of the residual copper segment by measuring the depth of the top and bottom interfaces of the residual copper. The bottom interface of the residual copper is a horizontal copper surface (162, i.e., land / pad) of the target signal layer L1. This horizontal copper surface is a horizontal copper feature that can be irradiated by the axial beam and exhibits a strong reflection peak in the axial reflection profile.

[0034] Regarding the top interface of the residual copper, in a preferred embodiment of the present invention, since back drilling cuts through the copper plating on the original central hole wall with a larger diameter, a cut copper top surface (163) will be formed at the top of the residual copper. Because copper has high reflectivity to short-wave infrared light, both the copper top surface (163) and the horizontal copper surface (162) of the target signal layer L1 are copper interfaces, exhibiting strong reflection peaks in the axial reflection profile. Their signal-to-noise ratio is significantly better than that of the case where a dielectric conical surface (161) is used as the upper reference interface. Therefore, the present invention preferably uses the reflection peak of the copper top surface (163) as the reflection peak of the residual copper top interface, and the axial distance between the reflection peak of the L1 horizontal copper surface (162) and the reflection peak to determine the length of the residual copper; while the dielectric reflection peak of the conical surface (161) can be used as an auxiliary or alternative. The measured value is the actual axial length of the residual copper plating, which is physically closer to the actual amount of residual copper that needs to be controlled.

[0035] It should be noted that the copper top surface (163) is an annular copper strip cut obliquely along the conical surface. Its corresponding depth is not a single peak, but is distributed within a range of tens of micrometers. Moreover, the specular reflection component of the forward incident light deviates from the optical axis and is mainly recovered by diffuse reflection from the copper cut surface (the total diffuse reflection power of copper is still much higher than that of the dielectric surface). Therefore, in a preferred embodiment, when determining the axial position of the reflection peak of the copper top surface, a fixed criterion is used for judgment. For example, the leading edge or intensity centroid of the bright band of its reflection peak is taken as the axial position reference of the copper top surface, and a standard sample with a known residual copper length (the residual copper length can be accurately calibrated by metallographic sectioning) is used to correct the residual copper top reference corresponding to the fixed criterion to ensure the consistency and accuracy of the measurement.

[0036] See Figure 3 This is a schematic diagram of the captured axial reflection profile (A-scan) signal. The horizontal axis represents the optical path length in the depth direction, and the vertical axis represents the reflection intensity. Several reflection peaks can be seen in the figure, corresponding to the plate surface (L0 surface), the copper top surface (163) at the top of the residual copper, and the horizontal copper surface (162) of the target signal layer. The positional difference between the two peaks of the copper top surface (163) and the horizontal copper surface (162) of the target signal layer is the optical path difference corresponding to the residual copper segment, which is then expressed as... The length of the remaining copper can be obtained by conversion.

[0037] Regarding the acquisition of signals in the narrow annular region, as mentioned earlier, the reflected signals from the copper top surface (163) and the L1 horizontal copper surface (162) are limited to the area surrounding the central hole, with a radial width of approximately [missing information]. The annular region. If the traditional method of precisely aligning the focused light spot is used, there is an operational dilemma: if the light spot is centered, the focus falls into the through-hole and the reflected signal of the annular region cannot be obtained; if the light spot is off-center and located in the annular region, the edge of the beam is easily blocked by the back-drilled hole wall, and it is difficult to achieve both. To solve this problem, in a preferred embodiment of the present invention, the bottom of the back-drilled hole is flooded with the measurement beam, that is, the measurement beam covers the entire bottom area of ​​the hole, including the annular region. Then, the coherence gate of optical coherence tomography is used to separate the echoes from different depths in the axial direction, that is, the reflection peak of the copper top surface (163) and the reflection peak of the L1 horizontal copper surface (162) appear at different depth positions in the axial reflection profile according to their optical path difference. Thus, the residual copper length can be obtained by the axial distance between the two reflection peaks, and the measurement process does not rely on lateral resolution for precise positioning. Because the effective reflective area of ​​the conical surface (161) in the annular region is small and the surface is tilted, the specular reflection component of the forward incident light deviates from the optical axis and is mainly recovered by diffuse reflection from the rough surface, so its reflection peak is relatively weak. In a preferred embodiment, this method improves the signal-to-noise ratio of the conical surface reflection peak by at least one of the following means: increasing the power of the reference arm, extending the integration time, and / or performing multiple axial scans on the same aperture and taking the average.

[0038] When the length of the residual copper to be measured is extremely small (e.g., close to 0.01 mm, about 10 micrometers), and approaches or falls below the nominal axial resolution of the system, the reflection peak corresponding to the copper top surface (163) and the reflection peak corresponding to the L1 horizontal copper surface (162) will be close to each other or even partially overlap in the axial reflection profile, making it difficult to distinguish them directly by their peak positions. In response to this situation, in a preferred embodiment of the present invention, at least one of the following methods is employed to improve the minimum identifiable residual copper length: First, a short-wave infrared broadband light source with a wider bandwidth is selected to reduce the axial resolution Δz, for example, reducing Δz to about 3 micrometers or less, so that the two reflection peaks can be clearly separated; Second, peak fitting is performed on the signal segment where the two reflection peaks are located, using the known point spread function (i.e., the shape of the reflection peak of a single interface) as a basis, and the center positions of the two peaks are estimated by least squares or maximum likelihood fitting; Third, the signal is deconvolutioned or super-resolution reconstructed using the point spread function to sharpen the overlapping reflection peaks and restore the true optical path difference between the two peaks. Furthermore, in a preferred embodiment, to further suppress dispersion broadening caused by the medium imbalance between the measuring arm and the reference arm, this method also performs numerical dispersion compensation on the obtained interference signal, making each reflection peak sharper, which is beneficial for the separation of the two reflection peaks and the accurate estimation of the residual copper length. Through one or more of the aforementioned means, the minimum residual copper length that this invention can identify can be lower than the nominal axial resolution of the light source, thereby achieving effective measurement of extremely small residual copper.

[0039] For deep hole measurement, other optical ranging methods such as confocal or chromatic confocal are available, but they have inherent limitations in deep holes with high aspect ratios. The axial resolution δz of the confocal method is inversely proportional to the square of the focusing numerical aperture (NA), approximately equal to...

[0040] (3)

[0041] Therefore, to achieve good axial resolution, a higher numerical aperture is required. However, the converging cone corresponding to a high numerical aperture will be blocked by the hole wall when entering a deep hole with a small aperture, thus failing to reach the bottom of the hole, creating a contradiction. Taking a back-drilled hole with a diameter of about 0.35 mm and a depth of about 5 mm (depth-to-width ratio of about 14) as an example, since the measurement needs to be performed through this 0.35 mm back-drilled hole diameter, the hole wall limits the available numerical aperture to an extremely low value of about 0.03. Under this numerical aperture, according to the above formula, the axial resolution of the confocal method will deteriorate to the sub-millimeter level, which is much larger than the length of the residual copper to be measured (0.01 to 0.25 mm), and therefore cannot distinguish between the conical surface (161) and the horizontal copper surface (162) of the target signal layer. In contrast, the optical coherence tomography used in this invention has an axial resolution determined by the bandwidth (coherence length) of the light source, which is independent of the focusing numerical aperture, i.e., the axial resolution and the numerical aperture are decoupled from each other. Accordingly, in a preferred embodiment, when measuring deep holes, a low numerical aperture, long working distance objective lens is used to focus the measurement beam. The working distance of this objective lens is greater than the depth of the back-drilled hole, and its numerical aperture is selected so that the measurement beam will not be blocked by the hole wall within this depth range (for example, a numerical aperture of approximately 0.02 to 0.035, corresponding to a spot diameter of approximately 20 to 24 micrometers). This allows the beam to be successfully delivered to the bottom of the hole up to 5 millimeters deep, while maintaining micrometer-level axial resolution through a broadband light source, thus clearly distinguishing the two reflection peaks. This characteristic of "decoupling axial resolution and numerical aperture" is the key technology of this invention in using optical coherence tomography instead of confocal scanning in high aspect ratio deep hole measurements.

[0042] In the aforementioned deep hole measurement embodiments, since the measurement range needs to cover a hole depth of several millimeters, swept-frequency optical coherence tomography (SS-OCT) is preferably used to provide a measurement range of several millimeters to tens of millimeters, and its sensitivity decreases less with depth. Simultaneously, this measurement is preferably performed after back-drilling but before resin plugging. At this stage, the hole to be measured is an unfilled hole, and the measurement beam propagates within the hole using air (refractive index approximately 1) as the medium, eliminating scattering and absorption of the dielectric material in the hole depth direction and simplifying optical path conversion. On the other hand, in embodiments where the dielectric layer thickness is measured on an open and accessible surface (such as a copper-free window area or board edge), in addition to optical coherence tomography, confocal or dispersive confocal methods can also be used to measure the thickness of the transparent dielectric layer, achieving measurement at a lower system cost. These substitutions do not depart from the essential scope of the present invention.

[0043] In a preferred embodiment, a center wavelength of the short-wave infrared broadband light source is selected at a penetration window to avoid the overtone absorption band of the epoxy resin dielectric material. Specifically, the center wavelength is preferably about 1310 nanometers (1310 nm) to avoid the C–H / O–H overtone absorption band of epoxy resin around 1450 nm and 1940 nm, while falling within the sensitive band of the indium gallium arsenide (InGaAs) detector, thereby improving the penetration depth and signal-to-noise ratio. In practice, other suitable short-wave infrared bands can also be selected according to the absorption and scattering characteristics of the material under test.

[0044] In an embodiment for measuring dielectric layer thickness, the measurement beam is incident on a region above the dielectric layer with a light-transmitting path, such as a copper-free window area, solder mask opening, or board edge. In the axial reflection profile, reflection peaks corresponding to an upper horizontal interface and a lower horizontal interface of the dielectric layer under test are identified, at least one of which is a copper interface. The solid thickness of the dielectric layer is obtained by converting the optical path difference between the two reflection peaks using the group refractive index of the dielectric material in the measurement band. When the material's light transmission depth is sufficient, a single axial scan can simultaneously resolve multiple copper interfaces located at different depths, thereby obtaining the thickness of multiple dielectric layers in a single measurement. This is a significant efficiency advantage of this invention compared to traditional single-point thickness measurement methods.

[0045] To improve absolute measurement accuracy, in a preferred embodiment, the method includes a step of correcting the refractive index of the dielectric material using a standard sample of known thickness. Specifically, the thickness of a standard sample is first precisely calibrated using metallographic sectioning or a contact thickness gauge; then, the standard sample is measured using the measurement system of this invention to obtain the corresponding optical path difference; the group refractive index of the dielectric material in the measurement wavelength range is calculated from the ratio of the optical path difference to the known thickness. When measuring the actual product later, the measured optical path difference is converted into physical dimensions using this corrected group refractive index value. Since the epoxy resin substrate and fiberglass cloth used in the same model of PCB product are relatively fixed, the group refractive index value after one correction can be applied to the measurement of the same batch or similar products.

[0046] Considering that the epoxy resin dielectric layer of printed circuit boards is usually reinforced with fiberglass cloth, the weave pattern of the fiberglass cloth can cause local non-uniformity in the equivalent refractive index of the dielectric layer and scatter incident light, resulting in some dispersion in the measured values. In a preferred embodiment, to suppress the measurement variation caused by local material non-uniformity, this method takes multiple measurement points in the same test area and statistically averages the measurement results to reduce random errors; or aligns the measurement points with a specific morphological position of the fiberglass weave (e.g., the intersection of fiberglass bundles or the resin-rich area between fiberglass bundles) to ensure that the material state of each measurement is consistent, thereby improving the repeatability and comparability of the measured values.

[0047] To improve the overall performance of the measurement system, in a preferred embodiment, this method performs numerical dispersion compensation on the acquired interference signal. Because the dielectric properties between the measuring arm and the reference arm may differ (for example, the beam from the measuring arm needs to penetrate a certain thickness of dielectric material, while the reference arm propagates in air), different wavelength components of the broadband light source experience different phase delays during propagation, leading to broadening of the interference signal and widening of the reflection peaks. Correcting this dispersion mismatch using a numerical dispersion compensation algorithm can make the reflection peaks in the axial reflection profile sharper, facilitating accurate identification of reflection peaks and separation of adjacent reflection peaks, especially significantly aiding in the super-resolution measurement of the aforementioned minimal residual copper length.

[0048] See Figure 1 and Figure 2 The invention will be further described below with reference to a complete measurement operation process. First, a short-wave infrared broadband light source (210) is provided, preferably a swept-frequency light source with a center wavelength of about 1310 nanometers. Its beam is split into a measurement beam and a reference beam by the fiber coupler / interferometer (220). The measurement beam is focused by a low numerical aperture, long working distance objective lens in the probe (230) and incident axially on the test hole of the back-drilled printed circuit board (100). The test hole is preferably an empty hole after back drilling and before resin plugging, and the medium inside the hole is air. The measurement beam reflected from each horizontal interface in the board—including the board surface L0, the copper top surface (163) of the residual copper top, the horizontal copper surface (162) of the target signal layer L1, and possibly deeper copper surfaces—returns to the fiber coupler (220), and is combined with the reference beam returned from the reference mirror to produce interference. The interference signal is converted into an electrical signal by the indium gallium arsenide detector (240) and sent to the signal processing unit (250). The signal processing unit (250) performs a Fourier transform on the timing interference signal to obtain the following result: Figure 3The axial reflection profile (A-scan) is shown. In this axial reflection profile, the reflection peaks of the copper top surface (163) and the horizontal copper surface (162) of L1 are identified according to their intensity and depth. The optical path difference between these two reflection peaks is calculated. The group refractive index value, pre-corrected with a standard sample, is used according to formula... Converting the optical path difference into a physical length yields the residual copper length of the hole to be measured. For measuring dielectric layer thickness, the same principle applies: identify the reflection peaks at adjacent horizontal copper interfaces and convert the result.

[0049] During the above measurement process, if the length of the residual copper is extremely small, causing the two reflection peaks to approach each other or even overlap, the signal processing unit (250) further executes algorithms such as peak fitting, deconvolution, or super-resolution reconstruction, and supplements them with numerical dispersion compensation to distinguish the two reflection peaks and estimate the true optical path difference between them, ensuring that the measurement lower limit can be lower than the nominal axial resolution. At the same time, for cases where a conical surface (161) is used as an auxiliary or alternative reference surface, the signal-to-noise ratio of the weaker dielectric reflection peak can be improved by increasing the optical power of the reference arm, extending the detector integration time, or performing multiple A-scan averaging on the same aperture position.

[0050] In summary, the printed circuit board back-drilled structure measurement method and system based on short-wave infrared optical coherence tomography (OCT) provided by this invention utilizes the optical contrast between epoxy dielectric material and copper in the short-wave infrared band, combined with the decoupling characteristics of axial resolution and numerical aperture of OCT, to achieve non-destructive, non-contact, micron-level axial resolution single-sided reflection measurement of the residual copper length and dielectric layer thickness of back-drilled holes. This invention can perform hole-by-hole online inspection of mass-produced products, effectively overcoming the destructive and time-consuming nature of known metallographic sectioning methods, the cost and insufficient contrast of X-ray / CT methods, and the shortcomings of indirect estimation and low spatial resolution of electrical measurement methods, thus possessing significant industrial application value.

[0051] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for measuring PCB back-drilling structures based on short-wave infrared OCT, characterized in that: include, A short-wave infrared broadband light source is provided, and the light beam of the light source is divided into a measurement beam and a reference beam via an interferometric structure; The measuring beam is focused axially and incident on the hole to be measured on the printed circuit board that has been back-drilled. The interference signal of the measurement beam reflected from the printed circuit board and the reference beam is obtained, and the axial reflection profile of the hole to be measured is obtained from the interference signal. In the axial reflection profile, the reflection peaks corresponding to the interface group to be tested are identified. The interface group to be tested includes: the top and bottom interfaces of the residual copper, and / or the upper and lower horizontal interfaces of the dielectric layer to be tested. as well as, By utilizing the refractive index of the dielectric material, the optical path difference between the reflection peaks of the interface group under test is converted into a physical dimension to obtain the residual copper length and / or the dielectric layer thickness of the dielectric layer under test.

2. The PCB back-drilling structure measurement method based on short-wave infrared OCT as described in claim 1, characterized in that: When the length of the residual copper is determined, the top interface of the residual copper is the copper top surface formed by cutting the residual copper through back drilling, and the bottom interface is the horizontal copper surface of the target signal layer. Both the copper top surface and the horizontal copper surface are copper interfaces.

3. The PCB back-drilling structure measurement method based on short-wave infrared OCT as described in claim 2, characterized in that: When determining the axial position of the reflection peak on the copper top surface, a fixed criterion is used. The fixed criterion is the leading edge or intensity centroid of the bright band of the reflection peak. The reference point of the copper top corresponding to the fixed criterion is corrected using a standard sample with a known copper length.

4. The PCB back-drilling structure measurement method based on short-wave infrared OCT as described in claim 2, characterized in that: The hole to be tested is a back-drilled hole, which has a back-drilled hole diameter and a center hole diameter smaller than the back-drilled hole diameter. The reflected signals from the copper top surface and the horizontal copper surface originate from an annular region surrounding the center hole with a radial width equal to half the difference between the back-drilled hole diameter and the center hole diameter. The method further includes: illuminating the bottom of the back-drilled hole with the measuring beam and using the coherence gate of optical coherence tomography to axially separate the reflection peak of the copper top surface and the reflection peak of the horizontal copper surface, thereby obtaining the residual copper length from the axial distance between the two reflection peaks.

5. The PCB back-drilling structure measurement method based on short-wave infrared OCT as described in claim 2, characterized in that: When the reflection peak corresponding to the top copper surface and the reflection peak corresponding to the horizontal copper surface are close to each other or partially overlap, peak fitting or deconvolution processing is performed on the signal segments where these two reflection peaks are located, and / or numerical dispersion compensation is performed on the interference signal to estimate the respective center positions of the two reflection peaks and restore the optical path difference between them, so that the identifiable residual copper length is less than the nominal axial resolution of the light source.

6. The PCB back-drilling structure measurement method based on short-wave infrared OCT as described in claim 2, characterized in that: The hole to be measured is a back-drilled hole with an aspect ratio of at least 10, and the back-drilled hole is an unfilled hole. The measuring beam is focused by an objective lens, the working distance of which is greater than the depth of the back-drilled hole. The numerical aperture of the objective lens is selected so that the measuring beam is not blocked by the hole wall within the depth range of the back-drilled hole. By utilizing the characteristic of decoupling the axial resolution of optical coherence tomography from the numerical aperture, micron-level axial resolution is maintained even at low numerical apertures.

7. The PCB back-drilling structure measurement method based on short-wave infrared OCT as described in claim 1, characterized in that: The center wavelength of the short-wave infrared broadband light source is selected at the penetration window to avoid the overtone absorption band of the dielectric material, and the center wavelength is approximately 1310 nanometers.

8. The PCB back-drilling structure measurement method based on short-wave infrared OCT as described in claim 1, characterized in that: The refractive index is the group refractive index of the dielectric material in the measurement band; the method further includes: correcting the group refractive index using a standard sample of known thickness.

9. The PCB back-drilling structure measurement method based on short-wave infrared OCT as described in claim 1, characterized in that: The steps for obtaining the axial reflection profile include: taking multiple measurement points in the same area to be measured and performing statistical averaging, or aligning the measurement points with specific morphological positions of the fiberglass texture to suppress measurement variations caused by local material inhomogeneity.

10. A PCB back-drilling structure measurement system based on short-wave infrared OCT, characterized in that: include: Shortwave infrared broadband light source; An interference architecture is used to split the light beam of the light source into a measurement beam and a reference beam, and to cause interference between the measurement beam reflected from the printed circuit board under test and the reference beam. The probe is configured to focus the measurement beam axially onto the back-drilled hole in the printed circuit board under test. Detector, configured to acquire the interference signal; and A signal processing unit configured to perform the method as described in any one of claims 1 to 9.

Images