Hybrid inductive proximity sensor based on semiconductor technology and method for manufacturing the same

Abstract

The embodiment of the application provides a kind of mixed inductive proximity sensor based on semiconductor process and its manufacturing method, mixed inductive proximity sensor includes: planar induction coil and electronic chip, planar induction coil is the copper coil structure formed on substrate by photoetching and plating process, planar induction coil is integrated with multiple concentric ring sub-coil, and multiple concentric ring sub-coil is used to measure the magnetic field gradient change caused by target object in section;Electronic chip is integrated with oscillator circuit, and oscillator circuit includes inductance branch with series positive and negative temperature coefficient resistance combination, capacitor branch with parallel adjustable resistance, and comparator unit with dynamic adjustment delay;The packaging structure of mixed inductive proximity sensor includes the substrate of electronic chip flip-chip interconnection through carrier plate, surface covers protective layer;The packaging pre-treatment of mixed inductive proximity sensor includes pad microprocessing, gas protection reflow soldering and composite glue buffer layer.

Description

Technical Field

[0001] This application belongs to the technical field of microelectromechanical systems (MEMS) sensors, specifically relating to a hybrid sensing proximity sensor based on semiconductor technology and its manufacturing method. Background Technology

[0002] As a core component for non-contact displacement detection in the industrial field, inductive proximity sensors have always evolved around three main directions: miniaturization, temperature stability, and integration.

[0003] Traditional sensors employ discrete negative temperature coefficient (NTC) thermistors for temperature compensation, requiring external components to adjust impedance matching. This not only occupies extra space but also limits mass production yield due to the uncontrollability of mechanical calibration. At the packaging level, the separate layout of the thick-film coil and electronic components introduces significant parasitic effects—when an aluminum target object approaches, the drift of the time constant τ=L / R (formed by the coil inductance (L) and resistance (R)) due to temperature is on the same order of magnitude as the impedance change caused by the target displacement, resulting in measurement errors reaching the micrometer level. Furthermore, existing complementary metal-oxide-semiconductor (CMOS) processes struggle to achieve thin-film resistors with both positive and negative temperature coefficients at the micrometer scale, making on-chip integration of temperature drift compensation circuitry impossible. For example, in 750nm CMOS processes, the temperature coefficient of resistance in metal layers is typically fixed at 0.18% / ℃, while that of polysilicon resistors is -0.13% / ℃. This difference in material properties renders traditional single-branch compensation ineffective.

[0004] These technological bottlenecks severely restrict the large-scale application of high-precision, low-power proximity sensors in the field of intelligent manufacturing. Summary of the Invention

[0005] This application proposes a hybrid sensing proximity sensor based on semiconductor technology and a method for manufacturing the same, in order to overcome the deficiencies of the prior art.

[0006] According to a first aspect of the embodiments of this application, a hybrid sensing proximity sensor based on semiconductor technology is provided, comprising: a planar induction coil and an electronic chip.

[0007] The planar induction coil is a copper coil structure formed on a substrate by photolithography and electroplating processes. The planar induction coil integrates multiple concentric ring sub-coils, which are used to measure the magnetic field gradient changes caused by the target object in segments.

[0008] The electronic chip integrates an oscillator circuit, which includes an inductor branch with a series combination of positive and negative temperature coefficient resistors, a capacitor branch with a parallel adjustable resistor, and a comparator unit for dynamically adjusting delay.

[0009] The hybrid proximity sensor's packaging structure includes a substrate that is flip-chip interconnected with the electronic chip via a carrier board, and the surface of the hybrid proximity sensor's packaging structure is covered with a protective layer; the pre-packaging treatment of the hybrid proximity sensor includes pad micromachining, gas-shielded reflow soldering, and a composite adhesive buffer layer.

[0010] In some embodiments, the electronic chip is manufactured using a complementary metal-oxide-semiconductor (CMOS) process.

[0011] According to a second aspect of this application, a method for manufacturing a hybrid sensing proximity sensor based on semiconductor technology is provided, for manufacturing the hybrid sensing proximity sensor based on semiconductor technology as described above, comprising:

[0012] A copper layer is electroplated on the substrate and a planar induction coil is formed by photolithography, and multiple concentric ring sub-coils are integrated.

[0013] A positive temperature coefficient resistor is fabricated in the metal layer of an electronic chip, and a negative temperature coefficient resistor is fabricated in the polycrystalline silicon layer of an electronic chip.

[0014] The resistance parameters of positive temperature coefficient resistors and / or negative temperature coefficient resistors are adjusted to the target value using laser technology.

[0015] The substrate is flip-chip interconnected with the electronic chip via a carrier plate;

[0016] A protective layer is applied to the surface of the interconnected structure.

[0017] In some embodiments, the process includes: [The following is a continuation of the previous section, but the context is unclear: "preceding the flip-chip interconnection of the substrate with the electronic chip via the carrier board:"]

[0018] A femtosecond laser is used to process microgroove structures in the pad area of ​​the carrier board to enhance solder wettability through capillary action;

[0019] Localized heating reflow soldering of the pad area is performed in a nitrogen atmosphere, and plasma cleaning is initiated.

[0020] A polyimide composite adhesive is printed between the carrier plate and the substrate and cured to form a buffer layer.

[0021] In some embodiments, after printing polyimide composite adhesive between the carrier and the substrate and curing it to form a buffer layer, the following steps are included:

[0022] The buffer layer is subjected to stepped heating annealing, wherein the stepped heating annealing includes a low-temperature stable matrix structure stage and a high-temperature grain boundary diffusion stage, and nitrogen protection is applied throughout the stepped heating annealing process.

[0023] In some embodiments, the method further includes:

[0024] The impedance signals of each sub-coil are acquired synchronously through a three-channel circuit;

[0025] The impedance phase angle of each impedance signal is extracted independently using the discrete Fourier transform algorithm;

[0026] Calculate the magnetic field gradient ratio based on the multiple impedance phase angles;

[0027] The magnetic field gradient ratio is compared with the pre-stored feature library and the target object material classification result is output for debugging the material recognition function of the hybrid induction proximity sensor.

[0028] In some embodiments, the method further includes:

[0029] The target position is fixed in a constant temperature environment. The oscillation frequency at the fixed target position is measured. The voltage divider resistor ratio of the capacitor branch is dynamically adjusted to make the inductance and the voltage change rate of the capacitor branch consistent.

[0030] Adjust the temperature coefficient of the comparator bias current to counteract residual linear drift and to adjust the temperature compensation function of the hybrid inductive proximity sensor.

[0031] In some embodiments, the method further includes:

[0032] A dual-output comparator is used to replace the reference voltage source;

[0033] The power consumption and waveform stability of the hybrid proximity sensor are monitored in both working and standby modes to debug the power supply function of the hybrid proximity sensor.

[0034] In some embodiments, the method further includes:

[0035] The minimum discernible displacement of the hybrid inductive proximity sensor was tested using a micro-displacement driving device.

[0036] The frequency offset of the hybrid inductive proximity sensor at a fixed location is measured during temperature cycling.

[0037] Long-term operation tests were conducted on the hybrid proximity sensor under mechanical shock and humidity conditions.

[0038] In some embodiments, the method further includes:

[0039] The vertical interconnection between the carrier plate and the substrate is achieved by using copper pillar bumps and anisotropic conductive adhesive.

[0040] The epoxy resin coverage area is defined by a mask, while the coil induction surface is preserved;

[0041] Solder reflow is achieved under localized laser heating.

[0042] The beneficial effects of the hybrid proximity sensor based on semiconductor technology and its manufacturing method according to the embodiments of this application include at least the following:

[0043] This application's embodiments utilize multi-segment concentric ring-shaped copper coils formed through photolithography and electroplating processes. These coils can segmentally capture changes in the magnetic field gradient caused by target objects, significantly improving the positioning accuracy and distance resolution for metal objects and avoiding the detection blind spots caused by edge effects in traditional single coils. By using a combination of positive and negative temperature coefficient resistors (PTC+NTC) in series with an adjustable resistor-capacitor branch in the oscillator circuit, ambient temperature drift is dynamically offset, ensuring stable output frequency over a wide temperature range and solving the problem of false triggering caused by temperature fluctuations in industrial settings. The substrate and electronic chip employ a flip-chip interconnect structure, combined with gas-protected reflow soldering and composite adhesive buffer layer encapsulation processes, making them resistant to oil, dust, and high-pressure washing, with a lifespan of millions of operations, far exceeding that of mechanical contact sensors. By combining inductive and capacitive principles, they can simultaneously identify metal workpieces, plastic containers, or liquid levels, making them suitable for complex scenarios. High-precision mass production of the coils is achieved through photolithography and electroplating processes, avoiding yield losses associated with traditional wire-winding processes. Micromachining technology for solder pads improves packaging efficiency and significantly reduces unit costs. Attached Figure Description

[0044] Figure 1 This is a schematic diagram of the structure of a hybrid proximity sensor based on semiconductor technology according to an embodiment of this application;

[0045] Figure 2 This is a schematic flowchart illustrating a method for manufacturing a hybrid proximity sensor based on semiconductor technology according to an embodiment of this application.

[0046] Figure 3 This is an integrated temperature compensation circuit diagram of the resistor combination according to an embodiment of this application;

[0047] Figure 4 This is a schematic diagram showing the displacement of the crossover point of the inductor and capacitor branches during the charging cycle as a function of temperature, without compensation.

[0048] Figure 5 This is a schematic diagram showing the displacement of the intersection point of the inductor branch and capacitor branch after compensation based on the embodiments of this application, as the temperature changes during the charging cycle.

[0049] Figure 6 This is a schematic diagram showing the curve of the output frequency of the hybrid proximity sensor in this application changing with the distance to the target object;

[0050] Figure 7This is a schematic diagram of the output frequency of the hybrid proximity sensor according to an embodiment of this application as a function of temperature. Detailed Implementation

[0051] To enable those skilled in the art to better understand the technology of this application, the following detailed description of this application is provided in conjunction with the accompanying drawings and specific embodiments.

[0052] The embodiments of this application will be described in further detail below with reference to the accompanying drawings and examples. The detailed description of the following embodiments and the accompanying drawings are used to illustrate the principles of this application by way of example, but should not be used to limit the scope of this application, that is, this application is not limited to the described embodiments.

[0053] This application provides a hybrid proximity sensor based on semiconductor technology, which is manufactured using a method for manufacturing a hybrid proximity sensor based on semiconductor technology.

[0054] In some implementations, the hybrid sensing proximity sensor based on semiconductor technology includes a planar induction coil and an electronic chip.

[0055] In some embodiments, the planar induction coil is a copper coil structure formed on a substrate by photolithography and electroplating processes. The planar induction coil integrates multiple concentric ring sub-coils, which are used to measure the magnetic field gradient changes caused by the target object in segments.

[0056] In one exemplary embodiment, the planar induction coil is made of copper and is a square with a side length of 3 mm. Its linewidth is 20 μm, the gap is 5 μm, and the thickness is 25 μm. The planar induction coil is formed on a silicon substrate by photolithography and electroplating processes. The planar induction coil integrates a multi-segment heterogeneous coil structure. By segmentally measuring the local magnetic field gradient changes caused by the target object, and combining this with digital signal processing algorithms, real-time identification of the target material is achieved.

[0057] In some embodiments, the electronic chip integrates an oscillator circuit, which includes an inductor branch with a series combination of positive and negative temperature coefficient resistors, a capacitor branch with a parallel adjustable resistor, and a comparator unit with dynamically adjustable delay.

[0058] For example, this electronic chip is a 750nm complementary metal-oxide-semiconductor (CMOS) chip that integrates a differential relaxor oscillator circuit. This differential relaxor oscillator circuit includes an inductor branch, a capacitor branch, and a comparator unit. The inductor branch consists of a coil resistor. Positive temperature coefficient resistor (Resistance is 10Ω, temperature coefficient of resistance TCR = 0.18% / ℃) and negative temperature coefficient resistors (Resistance value 8Ω, resistance temperature coefficient TCR=-0.13% / ℃) connected in series; the capacitor branch is an adjustable resistor ( (NTC) and two-component voltage regulators ( PTC and The system consists of NTCs connected in parallel, and the time constant is calculated based on the following formula:

[0059]

[0060] Where represents the time constant, and represents the capacitance value in the capacitor branch; the comparator unit adjusts the bias current... The temperature characteristics dynamically compensate for the residual temperature drift of the oscillation frequency.

[0061] In some embodiments, the packaging structure of the hybrid proximity sensor includes a substrate that is flip-chip interconnected with the electronic chip via a carrier board, and the surface of the packaging structure of the hybrid proximity sensor is covered with a protective layer; the pre-packaging treatment of the hybrid proximity sensor includes pad micromachining, gas-shielded reflow soldering, and a composite adhesive buffer layer. The electronic chip is manufactured using complementary metal-oxide-semiconductor (CMOS) technology.

[0062] For example, the substrate of the packaging structure is interconnected with the complementary metal-oxide-semiconductor electronic chip via a 100μm thick printed circuit board (PCB) as a carrier, the top sensing surface is unobstructed, and the whole is coated with a 200μm epoxy resin protective layer.

[0063] Pre-packaging processing steps may include, for example, solder joint pretreatment, flip-chip bonding, and stress buffer layer. Solder joint pretreatment includes: before ball bonding on the printed circuit board carrier, using a femtosecond laser to create micro-nano-level grooves (2μm depth, 5μm spacing) in the pad area, enhancing solder wettability through capillary action, thereby increasing solder strength. Flip-chip bonding includes: using localized heated reflow soldering (peak temperature 245℃, nitrogen protection), and simultaneously activating plasma cleaning (Ar / O2 mixed gas, 0.5Pa pressure) for 10 minutes during the reflow stage to remove organic contaminants and activate surface hydroxyl groups, improving the interfacial peel strength after epoxy resin encapsulation. The stress buffer layer includes: screen printing polyimide silver nanoparticle composite adhesive between the printed circuit board carrier and the substrate, which forms a 0.5μm thick buffer layer after curing. After the stress buffer layer is cured, a stepped annealing process (150℃ / 30min to 250℃ / 15min) is added. Through the grain boundary diffusion effect of silver nanoparticles, the thermal conductivity of the buffer layer is improved, and thermal mismatch leading to warping is prevented.

[0064] In some embodiments, the aforementioned stepped annealing process includes: placing the cured stress buffer layer in a nitrogen-atmosphere oven, heating it from room temperature to 150°C at a rate of 10°C / min, and holding it at that temperature for 30 minutes. During this stage, the polyimide matrix structure is initially stabilized through low-temperature annealing, while simultaneously promoting the formation of a preliminary dispersed network of silver nanoparticles in a low-energy state. Subsequently, the temperature is increased to 250°C at a rate of 15°C / min and held for 15 minutes. During this high-temperature stage, the grain boundary diffusion effect of the silver nanoparticles is utilized to form a denser thermally conductive pathway within the polyimide matrix. After annealing, the temperature is gradually reduced to room temperature at a rate of 5°C / min to prevent sudden cooling that could generate new stress within the buffer layer. The entire process requires strict control of the heating and cooling rates and holding time. The nitrogen environment prevents the oxidation of silver particles and ensures sufficient grain boundary diffusion. Ultimately, through stepped temperature control, the thermal conductivity of the buffer layer is significantly improved without damaging the polyimide properties, effectively suppressing thermal mismatch warpage between the printed circuit board and the substrate.

[0065] In some implementations, a stepped temperature control temperature compensation circuit is constructed under the following constraints. This temperature compensation circuit is directly integrated into the differential relaxation oscillator circuit, through the inductor branch ( , Series) and capacitor branch ( , , This application achieves temperature characteristic matching through the coordinated operation of parallel connections.

[0066] For example, temperature matching of the time constant of the inductor and capacitor branches is achieved by the following formula:

[0067] ;

[0068] in,

[0069] ,

[0070] ,

[0071] in, Indicates the coil inductance. This represents the rate of temperature change of the time constant of the inductor branch. This represents the rate of change of the time constant of the capacitor branch due to temperature. , These are the time constants of the inductor branch and the capacitor branch, respectively.

[0072] For example, the consistent temperature drift of the maximum value of the branch voltage U is achieved by the following formula:

[0073] ;

[0074] In the formula,

[0075] ,

[0076] ,

[0077] , These represent the maximum voltage values ​​of the inductor branch and the capacitor branch, respectively.

[0078] In some implementations, the dynamic delay compensation mechanism of the comparator unit is implemented based on the following method: through bias current. The temperature coefficient is inversely related to the temperature drift of the oscillation frequency, thus offsetting each other. and Uncovered linear temperature error; maximum absolute error ≤18μm at a target position of 450μm within the temperature range of −40℃ to 80℃.

[0079] This application's embodiments utilize multi-segment concentric ring-shaped copper coils formed through photolithography and electroplating processes. These coils can segmentally capture changes in the magnetic field gradient caused by target objects, significantly improving the positioning accuracy and distance resolution for metal objects and avoiding the detection blind spots caused by edge effects in traditional single coils. By using a combination of positive and negative temperature coefficient resistors (PTC+NTC) in series with an adjustable resistor-capacitor branch in the oscillator circuit, ambient temperature drift is dynamically offset, ensuring stable output frequency over a wide temperature range and solving the problem of false triggering caused by temperature fluctuations in industrial settings. The substrate and electronic chip employ a flip-chip interconnect structure, combined with gas-protected reflow soldering and composite adhesive buffer layer encapsulation processes, making them resistant to oil, dust, and high-pressure washing, with a lifespan of millions of operations, far exceeding that of mechanical contact sensors. By combining inductive and capacitive principles, they can simultaneously identify metal workpieces, plastic containers, or liquid levels, making them suitable for complex scenarios. High-precision mass production of the coils is achieved through photolithography and electroplating processes, avoiding yield losses associated with traditional wire-winding processes. Micromachining technology for solder pads improves packaging efficiency and significantly reduces unit costs.

[0080] This application provides a method for manufacturing a hybrid sensing proximity sensor based on semiconductor technology, used to manufacture the hybrid sensing proximity sensor based on semiconductor technology as described above. The method includes the following steps 210-250.

[0081] Step 210: Electroplating a copper layer on the substrate and forming a planar induction coil by photolithography, and integrating multiple concentric ring sub-coils.

[0082] In some embodiments, the copper layer is 25 μm thick; the planar induction coil is a planar helical coil. The terminals of the planar induction coil are interconnected with electronic chips via a substrate using a 100 μm thick printed circuit board.

[0083] In some implementations, the planar induction coil is formed on a silicon substrate by photolithography and electroplating processes, with an inductance of 4.3 μH and a resistance of 18 Ω. Its magnetic field distribution characteristics determine the effective measurement range of 100-1500 μm.

[0084] For example, the fabrication process of a planar induction coil includes: firstly, a planar copper coil is fabricated on a silicon substrate using photolithography and electroplating. The core of this process is to form a high-precision microstructure through a combination of photolithography and electroplating. Specifically, a photoresist layer of approximately 25 μm thickness is spin-coated onto the silicon substrate surface, and ultraviolet photolithography is performed using a mask to form a groove structure corresponding to the coil pattern. Then, pure copper is electroplated within the groove, controlling the current density (approximately 2 A / dm²) and electroplating time (approximately 90 minutes) to ensure a uniform copper layer thickness of 25 μm. The coil in this application is a square spiral structure with a side length of 3 mm, a linewidth of 20 μm, and a 5 μm gap between adjacent coils, ultimately achieving an inductance of 4.3 μH and a resistance of 18 Ω. After electroplating, the coil needs to be removed, cleaned, and annealed (held at 300°C for 30 minutes) to eliminate internal stress in the copper layer and improve conductivity. This step must be completed in a Class 100 cleanroom to avoid short circuits or open circuits caused by micron-level linewidth contamination.

[0085] In some implementations, the planar induction coil integrates a multi-segment heterogeneous coil structure. By segmentally measuring the local magnetic field gradient changes caused by the target object, and combining this with digital signal processing algorithms, real-time identification of the target material is achieved. Specifically, this includes: dividing the original single coil into three groups of concentric ring-shaped sub-coils (inner ring, middle ring, and outer ring), with the linewidth and gap of each group of sub-coils increasing in a ratio of 1:1.5:2 to form differentiated magnetic field sensitive areas; and integrating a three-channel synchronous sampling circuit (corresponding to the inner, middle, and outer ring sub-coils) on the chip using 2M / 2P CMOS technology to collect the impedance changes of each sub-coil. Each channel includes: generating a 10MHz excitation signal through a switched capacitor network (SCM), which is then injected into the sub-coil after passing through a programmable gain amplifier (PGA) to detect the impedance. real part With the imaginary part ; and by calculating the magnetic field gradient ratio and comparing it with a pre-stored material feature library, a material classification signal is output.

[0086] Among them, the detection impedance real part With the imaginary part This can be expressed by the following formula:

[0087] ,

[0088] Where j represents the inner loop, middle loop, or outer loop, thus yielding the real part of the impedance. and the virtual part .

[0089] Subsequently, the impedance phase angle is calculated using the following arctangent function. ,

[0090] ,

[0091] Calculate the impedance phase angle This phase angle reflects the phase relationship between the inductive reactance and the resistance caused by the change in the magnetic field.

[0092] In some implementations, the magnetic field gradient ratio is calculated as follows:

[0093] Data normalization, that is, the impedance change of each sub-coil. Normalize according to the following formula:

[0094] ,

[0095] In the formula This is the reference impedance when there is no target.

[0096] Furthermore, the gradient ratio is calculated, that is, the magnetic field gradient ratio is defined according to the following formula. :

[0097] ,

[0098] By matching the pre-stored material feature library (G threshold ranges for metals such as aluminum, copper, and steel) using a lookup table method, a material classification signal is output.

[0099] Step 220: Prepare a positive temperature coefficient resistor in the metal layer of the electronic chip and a negative temperature coefficient resistor in the polycrystalline silicon layer of the electronic chip.

[0100] In some implementations, the fabrication process of the electronic chip includes: constructing a relaxation oscillator using a 2M / 2P / HR process; fabricating positive and negative temperature coefficient resistors on the metal layer (M1) and the polysilicon layer (P+), respectively; and laser trimming. , and parameter.

[0101] Step 230: Adjust the resistance parameters of the positive temperature coefficient resistor and / or negative temperature coefficient resistor to the target value using laser.

[0102] In some implementations, refer to the appendix. Figure 3 As shown, the electronic chip is a metal-oxide-semiconductor (MOS) chip. This MOS chip integrates a differential relaxation oscillator circuit, which includes two branches with temperature-matched characteristics: the inductor branch is the coil resistor. Series positive temperature coefficient resistor (Resistance 10Ω, TCR = 0.18% / ℃) and negative temperature coefficient resistor (The circuit consists of an 8Ω resistor with a TCR of -0.13% at ℃); the capacitor branch is composed of an NTC resistor. (Resistance value 5kΩ) and voltage divider resistor ( PTC and Composed of two branches (NTC) connected in parallel. The time constants of the two branches... and Temperature characteristics matching is achieved by adjusting the resistance parameters. At the same time, ensure the voltage division ratio , The temperature drift is consistent, thus eliminating the impact of temperature changes on measurement accuracy. The maximum voltage of the inductor branch (determined by the coil resistance) and , (Determined by the partial pressure); The maximum voltage of the capacitor branch (by...) and , (Determined by the partial pressure).

[0103] In one exemplary embodiment, an electronic interface circuit is constructed based on a 750nm complementary metal-oxide-semiconductor (2M / 2P / HR) process. First, the circuit is completed using electronic application automation (EDA) tools, including: laying out a differential relaxation oscillator comprising two branches, one of which is an inductive branch and the other a coil resistor. Series positive temperature coefficient resistor (Resistance 10Ω, TCR = 0.18% / ℃) and negative temperature coefficient resistor Composed of (8Ω, TCR=-0.1300% / ℃); the capacitor branch is an NTC resistor. (5kΩ) and voltage divider resistor (PTC) (NTC) parallel connection. Additionally, ensure the time constants of the two branches. , Temperature characteristics matching and voltage division ratio and The temperature drift is consistent. Chip fabrication can employ a standard tape-out process, including: first, depositing a metal layer 1 (M1, an aluminum-copper alloy layer (containing 0.5% Cu in Al), approximately 500 nm thick) and a polysilicon layer (P+) on a silicon wafer; defining the resistance pattern using photolithography; and then performing ion implantation and annealing to adjust the temperature coefficient of resistance of the thin film. Additionally, laser trimming equipment is used for precise adjustment of R... L1 R L2 R CThe resistance value was adjusted to the target parameter, with the error controlled within ±0.5%. Finally, the size of the electronic chip was optimized to 1.2mm × 1.2mm to match the coil area.

[0104] Step 240: The substrate is flip-chip interconnected with the electronic chip via a carrier plate.

[0105] In some embodiments, before flip-chip interconnecting the substrate with the electronic chip via the carrier, the process includes: fabricating microgroove structures in the pad area of ​​the carrier using a femtosecond laser to enhance solder wettability through capillary action; performing localized heating reflow soldering in the pad area under nitrogen atmosphere and initiating plasma cleaning; and printing and curing a polyimide composite adhesive between the carrier and the substrate to form a buffer layer.

[0106] In some embodiments, after printing polyimide composite adhesive between the carrier and the substrate and curing it to form a buffer layer, the process includes: performing a stepped temperature annealing on the buffer layer, wherein the stepped temperature annealing includes a low-temperature stable matrix structure stage and a high-temperature grain boundary diffusion stage, and is conducted under nitrogen protection throughout the stepped temperature annealing process.

[0107] For example, the substrate is flip-chip soldered onto the printed circuit board carrier; gold wire bonding connects the complementary metal-oxide-semiconductor electronic chip to the printed circuit board; and a 200μm epoxy resin protective layer is applied.

[0108] In some implementations, the hybrid inductive proximity sensor consists of a square planar copper coil (20 μm line width, 5 μm gap, and 25 μm copper layer thickness) with a side length of 3 mm vertically interconnected with a 750 nm metal-oxide-semiconductor electronic chip.

[0109] Step 250: Apply a protective layer to the interconnected structure surface.

[0110] In some implementations, a "sandwich" packaging structure is used to integrate the planar induction coil and the electronic chip. The specific process includes: flip-chip bonding the silicon-based planar induction coil to a 100μm thick flexible printed circuit board carrier; depositing tin-silver solder balls (50μm in diameter) on the pads using a ball-mounting machine; setting the reflow soldering temperature profile to a peak of 245℃ (holding for 30 seconds) to ensure a reliable electrical connection between the planar induction coil and the printed circuit board. Subsequently, the electronic chip, processed in the second step, is interconnected to the printed circuit board carrier using wire bonding technology. 25μm diameter gold wire is used for bonding, and the bonding pressure is controlled within the range of 80-100mN to avoid damaging the chip's passivation layer. During the packaging process, special care is taken to keep the top of the coil unobstructed to ensure maximum sensitivity of the sensing surface. Finally, epoxy resin is used for encapsulation, with the coating thickness controlled to 200μm using a dispensing machine, and the curing conditions are 120℃ for 60 minutes. After packaging, an airtightness test (helium mass spectrometer leak detector detects a leakage rate of <1×10-9Pa·m³ / s) and a mechanical shock test (performed in accordance with IEC60068-2-27 standard) are performed.

[0111] In some embodiments, to improve packaging quality, this application employs the following methods: flip-chip bonding instead of gold wire bonding, for example, using copper pillar bumps and anisotropic conductive adhesive (ACA) to achieve vertical interconnection between the electronic chip and the printed circuit board carrier, which can reduce parasitic inductance to below 0.1nH; local selective encapsulation, for example, defining the epoxy resin coverage area through photolithography to protect only the chip and solder joints, leaving a 0.5mm area at the top of the coil without a cover layer, thereby improving sensing sensitivity; laser-assisted reflow soldering, for example, using local laser heating (peak temperature 245°C) during the reflow of tin-silver solder balls to prevent the silicon-based coil from deforming due to overall high temperature.

[0112] This application embodiment integrates multiple concentric ring-shaped sub-coils using electroplated copper layers and photolithography, achieving high-precision coil patterning and avoiding the risk of inter-turn short circuits associated with traditional winding processes. The multi-segment coil structure can capture changes in magnetic field gradients segmentally, significantly improving the positioning accuracy of metallic targets while also supporting the detection of non-metallic targets, thus expanding application scenarios. By fabricating positive temperature coefficient (PTC) resistors on the metal layer of the electronic chip and negative temperature coefficient (NTC) resistors on the polycrystalline silicon layer, complementary resistance temperature drift is achieved using the intrinsic properties of the materials. This provides a physical basis for temperature compensation circuits, eliminating the need for external compensation components, simplifying the chip structure, and reducing thermal coupling interference. Furthermore, by dynamically adjusting the PTC / NTC resistance values ​​to the target parameters using laser, precise... The time constant temperature drift rate of the inductor and capacitor branches was matched, solving the problem of sensing distance drift caused by individual component differences in mass production and ensuring consistent detection across the entire temperature range. The substrate and electronic chip are interconnected via flip-chip bonding on a carrier board, replacing wire bonding, shortening the signal transmission path and reducing parasitic inductance. This structure enhances resistance to mechanical vibration, is compatible with surface mount technology, improves production line efficiency, and is suitable for high-vibration scenarios such as automotive welding. After interconnection, an epoxy resin or polyimide protective layer is applied, achieving sealing, moisture protection, chemical corrosion resistance, and mechanical protection. The composite adhesive layer buffers thermal stress, preventing solder joint cracking caused by temperature cycling and extending the sensor's service life in oily and high-pressure washing environments.

[0113] In some embodiments, the method further includes: synchronously acquiring the impedance signals of each sub-coil through a three-channel circuit; independently extracting the impedance phase angle of each impedance signal using a discrete Fourier transform (DFT) algorithm; calculating the magnetic field gradient ratio based on multiple impedance phase angles; comparing the magnetic field gradient ratio with a pre-stored feature library and outputting the target object material classification result for debugging the material recognition function of the hybrid inductive proximity sensor.

[0114] In one exemplary embodiment, the specific steps for extracting the impedance phase angle using the discrete Fourier transform algorithm include: First, a three-channel synchronous sampling circuit acquires the time-domain voltage and current signals of each sub-coil under 10MHz excitation, and converts them into digital signals via an analog-to-digital converter (ADC). Next, the digital signals are input into the Fourier transform algorithm module, which performs frequency domain transformation on the N sampling points within each cycle to calculate the complex amplitude components (real and imaginary parts) of the signal at a frequency of 10MHz.

[0115] In some embodiments, the three-channel synchronous sampling circuit is integrated based on a CMOS process with two layers of metal interconnection and / or two layers of polysilicon (2M / 2P). For each channel, a 10 MHz excitation signal is first generated by a switched-capacitor network, amplified by a programmable gain amplifier, and then injected into the corresponding sub-coil to detect the real and imaginary parts of the impedance. The collected time-domain electrical signal is converted into a digital signal by an analog-to-digital converter, and then the digital demodulation module uses the discrete Fourier transform algorithm to extract the impedance phase angle θ, calculate the magnetic field gradient ratio, and compare it with a pre-stored material feature library, and finally outputs a material classification signal to achieve material identification for the areas corresponding to different sub-coils.

[0116] In some embodiments, the method further includes: fixing the target position in a constant-temperature environment, measuring the oscillation frequency of the target fixed position, and dynamically adjusting the capacitance-branch voltage-dividing resistor ratio to make the voltage change rates of the inductor and the capacitance branch consistent; adjusting the temperature coefficient of the comparator bias current to cancel the residual linear drift and debug the temperature compensation function of the hybrid inductive proximity sensor.

[0117] Refer to the appendix Figure 4-5 As shown, T1 and T2 represent the first temperature and the second temperature, where T1 < T2, and P(T1) and P(T2) are the intersection points of the voltage waveforms of the inductor branch and the capacitance branch corresponding to T1 and T2 respectively. The positions of the P(T1) point and the P(T2) point directly determine the oscillation period of the relaxation oscillator. During the operation of the relaxation oscillator, when the voltage of the capacitance branch reaches the same as that of the inductor branch, that is, the P(T1) point and the P(T2) point, the comparator flips, triggering the start or end of the oscillation period. Figure 4 Shows the situation where the intersection point P(T1) of the inductor branch and the capacitance branch during the charging cycle without compensation changes and shifts to P(T2) as the temperature changes (from T1 to T2): When the temperature rises, due to the temperature coefficient of the coil resistance (about 0.18% / °C), the time constant of the inductor branch changes, and the position of the intersection point changes significantly, causing the oscillation frequency to drift. This drift will cause a measurement error of up to the micron level in industrial applications. Figure 5 Shows the situation where the intersection point P(T1) of the inductor branch and the capacitance branch during the charging cycle after compensation based on the embodiment of the present present present application changes and shifts to P(T2) as the temperature changes (from T1 to T2): By connecting a positive / negative temperature coefficient resistor ( and <00) in series with the inductor branch and matching adjustable resistors ( 、 The coordinated design of the two branches ensures that their time constants are matched in temperature (this condition has been met in the embodiments of this application), while also ensuring consistent temperature drift of the voltage division ratio. Therefore, even with temperature changes, the crossover point position remains relatively stable, significantly reducing the impact of temperature on the oscillation frequency and achieving the technical specification of temperature drift ≤0.07% / ℃.

[0118] In one exemplary implementation, circuit calibration is performed based on the solution to a system of equations. This system of equations is established based on the following three constraints:

[0119] Constraint 1: Time constant matching

[0120] ,

[0121] in, The time constant of the inductor branch is... The time constant of the capacitor branch;

[0122] Constraint 2: Voltage Divider Matching

[0123] and The temperature coefficients are consistent;

[0124] Constraint 3, satisfy =10Ω =8Ω Boundary conditions with measured parameters such as 5kΩ.

[0125] In one exemplary implementation, the circuit calibration process includes: fixing the target at 450μm in a constant temperature chamber at -40℃, 25℃, and 80℃, and measuring the output frequency; fine-tuning the circuit using a digital potentiometer. and The resistance ratio makes and The temperature curves coincide; further, refer to the appendix. Figure 7 The temperature-frequency curve shown indicates that the temperature coefficient of the comparator bias current is adjusted to compensate for residual linearity error.

[0126] In one exemplary implementation, refer to the appendix Figure 7 As shown, the temperature compensation circuit debugging steps are as follows: First, set three temperature points in the constant temperature chamber: −40℃, 25℃, and 80℃. Fix the aluminum target at a distance of 450μm (this distance refers to the vertical distance between the aluminum target object and the sensing surface (coil plane) on the top of the sensor). Measure the output frequency change curve with temperature. Record the oscillation frequency (nominal value 10.5-30MHz) using an oscilloscope and calculate the temperature drift (formula: Δf / f·ΔT, where T is the temperature inside the chamber). If the measured drift exceeds 0.07% / ℃, fine-tuning is required. and The resistance ratio makes Temperature characteristics and Synchronous changes. Dynamic comparator delay compensation is achieved by adjusting the temperature coefficient of the bias current: a reference resistor is changed using a digital potentiometer, so that the temperature slope of the comparator propagation delay cancels out the residual linearity error. After debugging, it needs to be verified that the frequency fluctuation at the same target location does not exceed ±0.3% within the range of −40~80℃.

[0127] In some embodiments, the method further includes: replacing the reference voltage source with a dual-output comparator; monitoring the power consumption and waveform stability of the hybrid proximity sensor in operating mode and standby mode, for adjusting the power supply function of the hybrid proximity sensor.

[0128] In one exemplary embodiment, this application uses a dual-output comparator instead of a reference voltage source and utilizes complementary outputs to maintain circuit balance, resulting in a standby mode power consumption as low as 85μW and an operating mode power consumption of 100mW. Based on this, the application is compatible with industrial standard 5V control systems, solving the problem of poor interface compatibility between existing sensors and printed circuit board equipment. The packaging structure of this application adopts a "sandwich" stacked layout, that is, the coil substrate is flip-chip bonded to the electronic chip via a 100μm thick printed circuit board carrier, with the top sensing surface remaining unobstructed, and the overall size optimized to 3mm × 3mm × 1.2mm. Gold wire bonding achieves the electrical connection between the electronic chip and the printed circuit board, and a 200μm epoxy resin protective layer ensures environmental resistance; tests have shown that it maintains stable function under 80% humidity and 100,000 mechanical shocks.

[0129] In one exemplary embodiment, to address the industrial demand for 0-5V unipolar power supply, the function of a dual-output comparator as a substitute for a reference voltage source is verified. Specifically, in operating mode, the chip current (nominal 20mA) is monitored using a power load tester to calculate power consumption; in standby mode, a 0V logic signal is applied to shut down the oscillator, and leakage current is measured (<17μA, corresponding to 85μW power consumption). The duty cycle stability of the output square wave is observed using an oscilloscope; the frequency offset should be less than ±0.1% when the power supply voltage fluctuates by ±5%. Furthermore, it is necessary to ensure that the circuit can maintain oscillator oscillation even without a reference voltage. This is achieved by adjusting the comparator threshold voltage (VTH) and hysteresis width (>10mV) to avoid false triggering due to noise interference.

[0130] In some embodiments, the method further includes: testing the minimum discernible displacement of the hybrid inductive proximity sensor using a micro-displacement driving device; measuring the frequency offset of the hybrid inductive proximity sensor at a fixed position during temperature cycling; and conducting long-term operational tests of the hybrid inductive proximity sensor under mechanical shock and humidity conditions.

[0131] In some implementations, refer to the appendix. Figure 6 As shown, the performance verification data of this application show that the sensor achieves a longitudinal resolution of 120nm (<500μm range) for aluminum targets, a full-scale nonlinear error of ≤0.25%, and a response frequency coverage of 10.5-30MHz (corresponding to changes in target distance). Compared to existing technologies, this application reduces parasitic inductance by 60% and mass production costs by 40% through 3D integration, and can meet the needs of complex industrial environments without external calibration. This technology has been verified through prototype testing and is applicable to scenarios such as precision positioning, robot joint monitoring, and semiconductor equipment clearance control in the field of intelligent manufacturing, demonstrating significant technological advantages and industrialization prospects.

[0132] In one exemplary embodiment, the resolution test includes: using piezoelectric ceramic to drive an aluminum target (1 mm thick), moving it in 5 μm steps to a range of 500 μm, recording the frequency-displacement curve, and calculating the minimum resolvable displacement (120 nm).

[0133] In one exemplary embodiment, reference is made to the appendix. Figure 7 As shown, the temperature drift test includes: fixing the target at two positions, 450μm and infinity, in a temperature chamber of −40~80℃, measuring the change in output frequency, and verifying that the drift is ≤0.07% / ℃.

[0134] In one exemplary implementation, the long-term stability test includes: continuous operation for 1000 hours, recording the zero-point drift (frequency offset when the target is removed) every 24 hours, with a requirement of <±0.5MHz.

[0135] In one exemplary embodiment, the environmental tolerance test includes: 85% humidity storage (240 hours) and vibration test (10-2000Hz sweep frequency, 15g acceleration).

[0136] In some embodiments, the method further includes: using copper pillar bumps and anisotropic conductive adhesive to achieve vertical interconnection between the carrier and the substrate; defining the epoxy resin covered area through a mask while retaining the coil induction surface; and completing solder reflow under local laser heating.

[0137] This application addresses the technical bottleneck of directly achieving micron-level negative temperature coefficient resistors in complementary metal-oxide-semiconductor (CMOS) processes by employing a combined positive and negative temperature coefficient resistor compensation strategy. By establishing a five-variable equation system (including two time constant matching equations, two voltage divider matching equations, and three boundary condition equations), the following solution is obtained: , , , and The optimal parameter combination was determined. Experimental verification shows that within the temperature range of −40℃ to 80℃, this compensation application controls the temperature drift to within 0.07% / ℃, improving stability by more than 60% compared to traditional external negative temperature coefficient resistor compensation. To eliminate residual linear temperature error, a dynamic comparator delay compensation mechanism is further introduced, that is, by adjusting the temperature characteristics of the comparator bias current, it is made to act inversely to the temperature drift of the oscillation frequency, ultimately achieving stable output across the entire temperature range. In terms of power supply architecture, this application abandons the traditional three-wire power supply mode (−2.5V / GND / +2.5V) and adopts a 0-5V unipolar power supply.

[0138] The hybrid proximity sensor proposed in this application achieves the following significant technical effects: Experimental verification shows that the sensor achieves a longitudinal resolution of 120nm (<500μm range) for aluminum targets, with a full-scale nonlinearity error ≤0.25%, effectively solving the accuracy bottleneck problem caused by temperature drift in traditional sensors; within a temperature range of −40~80℃, a dual-branch temperature compensation circuit (R... L1 =10Ω, R L2 =8Ω, R C The synergistic effect of the 5kΩ wire bond and dynamic comparator delay adjustment controls the temperature drift to within 0.07% / ℃, improving stability by more than 60% compared to traditional external NTC compensation. This performance meets the requirements of high-precision industrial scenarios such as precision positioning of semiconductor equipment (±50nm level) and robot joint monitoring (<1μm displacement detection). In terms of integration, a three-dimensional stacked packaging structure is used to vertically interconnect the planar copper coil (square side length 3mm) with the CMOS chip. Parasitic effects are eliminated by a 100μm thick PCB substrate, optimizing the overall size to 3mm×3mm×1.2mm, which is 75% smaller than the traditional split type. This structure achieves environmental resistance through gold wire bonding and an epoxy resin protective layer (200μm), maintaining stable function in 80% humidity and 100,000 mechanical shock tests. The single-supply (0-5V) architecture eliminates reference voltage dependence through dual-output comparators, achieving a standby power consumption as low as 85μW and a working mode power consumption of 100mW. This represents a 40% reduction in energy consumption compared to traditional three-wire power supplies, and it is compatible with industrial standard PLC control systems. This application reduces mass production costs by 40% through the hybrid integration of planar induction coils using photolithography and electroplating processes with complementary metal-oxide-semiconductor (CMOS) technology, and can meet the demands of complex industrial environments without external calibration. Real-world testing in applications such as printing press roller gap monitoring and semiconductor wafer handling robots has demonstrated that the sensor maintains a repeatability accuracy of ±0.3μm after 1000 hours of continuous operation, indicating its potential for large-scale application.

[0139] This application aims to address the technical challenges of miniaturization, temperature stability, and integration deficiencies in traditional industrial sensors, particularly for non-contact displacement detection in industrial automation. It employs a three-dimensional stacked packaging structure, achieving high-precision displacement detection through the collaborative design of a planar electroplated coil and a complementary metal-oxide-semiconductor (CMOS) electronic chip. By integrating CMOS technology with planar induction coils using etching and electroplating processes, it overcomes the miniaturization bottleneck and temperature drift issues inherent in discrete components in traditional sensors, making it suitable for high-precision, low-power, single-supply intelligent manufacturing scenarios. The three-dimensional stacked structure integrates a planar copper coil and a CMOS electronic chip. A dual-branch temperature compensation circuit is used, with the inductor branch connected in series with positive / negative temperature coefficient resistors and the capacitor branch matched with an adjustable resistor. Through synergistic compensation of the time constant and the temperature characteristics of voltage division, a temperature drift of ≤0.07% / ℃ is achieved within the range of −40~80℃. Combined with dynamic comparator delay adjustment and a single-supply architecture, it is suitable for high-precision industrial displacement detection.

[0140] It is understood that the above embodiments are merely exemplary implementations used to illustrate the principles of this application, and this application is not limited thereto. For those skilled in the art, various modifications and improvements can be made without departing from the spirit and substance of this application, and these modifications and improvements are also considered to be within the scope of protection of this application.

Claims

1. A hybrid proximity sensor based on semiconductor technology, characterized in that, include: Planar induction coils and electronic chips, The planar induction coil is a copper coil structure formed on a substrate through photolithography and electroplating processes. The planar induction coil integrates multiple concentric ring sub-coils, which are used to measure the magnetic field gradient changes caused by the target object in segments. The planar induction coil has a linewidth of 20 μm, a gap of 5 μm, and a thickness of 25 μm. The multiple concentric ring sub-coils include an inner ring, a middle ring, and an outer ring sub-coils. The linewidth and gap between the inner ring, the middle ring, and the outer ring increase in a ratio of 1:1.5:2 to form differentiated magnetic field sensitive areas. The electronic chip integrates an oscillator circuit, which includes an inductor branch with a series combination of positive and negative temperature coefficient resistors, a capacitor branch with a parallel adjustable resistor, and a comparator unit for dynamically adjusting the delay. The inductor branch consists of a coil resistor. Positive temperature coefficient resistor With negative temperature coefficient resistors Series connection; capacitor branch is an adjustable resistor With two-component voltage resistors and The parallel configuration has a time constant calculated based on the following formula: in, Represents the time constant. This indicates the capacitance value in the capacitor branch; the positive temperature coefficient resistor The resistance is 10Ω and the temperature coefficient of resistance is 0.18% / ℃. This negative temperature coefficient resistor... The resistance is 8Ω and the temperature coefficient of resistance is -0.13% / ℃. The adjustable resistor... The resistance is 5kΩ, and the oscillator circuit achieves temperature compensation through time constant temperature characteristic matching of the inductor branch and the capacitor branch, satisfying the following constraints: in, , ;in, This represents the inductance of the coil, where, This represents the rate of temperature change of the time constant of the inductor branch. This represents the rate of change of the time constant of the capacitor branch due to temperature. , These represent the time constants of the inductor branch and the capacitor branch, respectively; the consistent temperature drift of the maximum value of the branch voltage U is achieved by the following formula: In the formula, , ,in, , These represent the maximum values ​​of the inductor branch voltage and the capacitor branch voltage, respectively; the comparator unit dynamically compensates for the residual temperature drift of the oscillation frequency by adjusting the temperature characteristics of the bias current. The hybrid proximity sensor's packaging structure includes a substrate that is flip-chip interconnected with the electronic chip via a carrier board, and the surface of the hybrid proximity sensor's packaging structure is covered with a protective layer; the pre-packaging treatment of the hybrid proximity sensor includes pad micromachining, gas-shielded reflow soldering, and a composite adhesive buffer layer.

2. The hybrid proximity sensor based on semiconductor technology according to claim 1, characterized in that, The electronic chip is manufactured using complementary metal-oxide-semiconductor (CMOS) technology.

3. A manufacturing method for manufacturing a hybrid sensing proximity sensor based on semiconductor technology as described in claim 1 or 2, characterized in that, include: A copper layer is electroplated on the substrate and a planar induction coil is formed by photolithography, and multiple concentric ring sub-coils are integrated. A positive temperature coefficient resistor is fabricated in the metal layer of an electronic chip, and a negative temperature coefficient resistor is fabricated in the polycrystalline silicon layer of an electronic chip. The resistance parameters of positive temperature coefficient resistors and / or negative temperature coefficient resistors are adjusted to the target value using laser technology. The substrate is flip-chip interconnected with the electronic chip via a carrier plate; A protective layer is applied to the surface of the interconnected structure.

4. The manufacturing method according to claim 3, characterized in that, Before flip-chip interconnecting the substrate with the electronic chip via the carrier plate, the following steps are included: A femtosecond laser is used to process microgroove structures in the pad area of ​​the carrier board to enhance solder wettability through capillary action; Localized heating reflow soldering of the pad area is performed in a nitrogen atmosphere, and plasma cleaning is initiated. A polyimide composite adhesive is printed between the carrier plate and the substrate and cured to form a buffer layer.

5. The manufacturing method according to claim 3, characterized in that, After printing polyimide composite adhesive between the carrier plate and the substrate and curing it to form a buffer layer, the process includes: The buffer layer is subjected to stepped heating annealing, wherein the stepped heating annealing includes a low-temperature stable matrix structure stage and a high-temperature grain boundary diffusion stage, and nitrogen protection is applied throughout the stepped heating annealing process.

6. The manufacturing method according to claim 3, characterized in that, The method further includes: The impedance signals of each sub-coil are acquired synchronously through a three-channel circuit; The impedance phase angle of each impedance signal is extracted independently using the discrete Fourier transform algorithm; Calculate the magnetic field gradient ratio based on the multiple impedance phase angles; The magnetic field gradient ratio is compared with the pre-stored feature library and the target object material classification result is output for debugging the material recognition function of the hybrid induction proximity sensor.

7. The manufacturing method according to claim 3, characterized in that, The method further includes: The target position is fixed in a constant temperature environment. The oscillation frequency at the fixed target position is measured. The voltage divider resistor ratio of the capacitor branch is dynamically adjusted to make the inductance and the voltage change rate of the capacitor branch consistent. Adjust the temperature coefficient of the comparator bias current to counteract residual linear drift and to adjust the temperature compensation function of the hybrid inductive proximity sensor.

8. The manufacturing method according to claim 3, characterized in that, The method further includes: A dual-output comparator is used to replace the reference voltage source; The power consumption and waveform stability of the hybrid proximity sensor are monitored in both working and standby modes to debug the power supply function of the hybrid proximity sensor.

9. The manufacturing method according to claim 3, characterized in that, The method further includes: The minimum discernible displacement of the hybrid inductive proximity sensor was tested using a micro-displacement driving device. The frequency offset of the hybrid inductive proximity sensor at a fixed location is measured during temperature cycling. Long-term operation tests were conducted on the hybrid proximity sensor under mechanical shock and humidity conditions.

10. The manufacturing method according to claim 4, characterized in that, The method further includes: The vertical interconnection between the carrier plate and the substrate is achieved by using copper pillar bumps and anisotropic conductive adhesive. The epoxy resin coverage area is defined by a mask, while the coil induction surface is preserved; Solder reflow is achieved under localized laser heating.

Images