Automatic regulating system for voltage fluctuation in electroplating process

Abstract

The application relates to the technical field of system for adjusting electric variables, and discloses an automatic adjusting system for voltage fluctuation of an electroplating process, which comprises a feature extraction module, a logic damping operation module and a pulse width modulation instruction reconstruction module.The feature extraction module is used for acquiring original sampling signals of transient changes of an end voltage of a controlled load and extracting voltage differential information.The logic damping operation module is used for generating a logic compensation instruction according to the voltage differential information and a preset logic damping operator.The pulse width modulation instruction reconstruction module is used for reconstructing a carrier phase and an on-off edge timing according to the logic compensation instruction to correct a duty ratio.The application implants a virtual damping vector with opposite polarity in a digital control loop, so that a damping response with exponential attenuation over time is generated for output voltage to suppress transient voltage oscillation.The application eliminates high-frequency ringing signals caused by coupling of parasitic parameters of a power transmission line and equivalent impedance of a controlled load, and guarantees voltage stability of a controlled load group during frequent switching of multi-task bits.

Description

Technical Field

[0001] This invention belongs to the field of system technology for regulating electrical variables, and particularly relates to an automatic regulation system for voltage fluctuations in the electroplating process. Background Technology

[0002] In high-precision electroplating and semiconductor wafer electrodeposition processes, the output voltage stability of the rectified power supply determines the crystal density and surface microstructure consistency of the plating layer. Existing technologies employ closed-loop feedback control systems based on proportional-integral-differential algorithms. These systems collect the voltage deviation signal at the output terminal and adjust the duty cycle of the pulse width modulation signal to maintain a constant cathode potential. However, as electroplating processes evolve towards more advanced technologies, the nonlinear double-layer charge characteristics exhibited at the cathode interface of the electrolytic cell become increasingly prominent. This causes the load impedance to fluctuate with the electrode reaction state. In this context, traditional feedback control mechanisms exhibit physical response delays, as their adjustment actions occur after the voltage deviation occurs, making it difficult to smooth transient fluctuations on the order of 1μs to 10μs. The industry typically uses the addition of large-capacity physical filter components to smooth the output voltage. This approach reduces the power density of the power supply system and is prone to coupling with parasitic inductance in transmission lines, inducing high-frequency ringing.

[0003] To address the regulation lag, approaches such as increasing the sampling frequency or optimizing the regulation gain often involve a trade-off between microprocessor processing power and overall system stability. Simply increasing the regulation intensity can easily trigger intermodulation oscillations between power supply units, leading to uncontrolled ripple amplification in the output voltage. This physical mismatch between the inherent lag property of the control law and the transient charge balance requirements of the load constitutes an inherent contradiction that existing regulation systems struggle to resolve. Existing control logic has bottlenecks in smoothing transient fluctuations. For example, Chinese invention patent CN116516456B discloses a method for automatic overvoltage protection in an intelligent electroplating line production heating system. By acquiring the temperature and voltage of the electroplating tank in real time, and using cloud feedback to reduce the power output to perform overvoltage protection when the monitored values ​​exceed a preset threshold, this is a post-event triggering mechanism based on a static threshold. The intervention of the cloud communication link amplifies the regulation lag effect. Faced with microsecond-level transient processes caused by electrochemical inertia, it lacks the ability to predict the cathode interface admittance trend in real time and cannot inject logic damping to absorb energy at the inflection point of ripple generation, making it difficult to solve the intermodulation oscillation problem of multi-machine parallel operation.

[0004] Therefore, how to identify the dynamic response characteristics of the load end through carrier phase micro-perturbation, and adjust the output characteristics of the regulation system in real time to smooth out transient fluctuations, is the technical problem to be solved by this invention. Summary of the Invention

[0005] This invention provides an automatic adjustment system for voltage fluctuations during electroplating. The system includes a feature extraction module, a logic damping calculation module, and a PWM modulation command reconstruction module.

[0006] The feature extraction module is used to acquire the original sampled signal that characterizes the transient changes in the voltage at the controlled load terminal, and extract the first-order voltage difference information and the second-order voltage difference information from it.

[0007] The logic damping operation module, whose input is connected to the feature extraction module, is used to perform weighted operations on the first-order voltage difference information and the second-order voltage difference information with the preset logic damping operator to generate a logic compensation instruction containing amplitude compensation component and phase lead component.

[0008] The PWM modulation instruction reconstruction module, whose input is connected to the logic damping operation module, is used to reconstruct the carrier phase and duty cycle timing of the pulse width modulation signal according to the logic compensation instruction, so as to correct the duty cycle of the digital control loop. In this module, the system implants a virtual damping vector with the opposite oscillation polarity to the original sampled signal into the digital control loop according to the logic compensation instruction. This makes the output voltage of the system exhibit a convergent waveform that decays exponentially with time and is accompanied by periodic oscillations when a sudden change occurs at the controlled load, so as to suppress the transient voltage oscillation at the controlled load.

[0009] Preferably, when the system performs regulation, the output voltage satisfies the following attenuation constraint at the moment of load change: ,in, For real-time voltage output, The initial amplitude of the fluctuation. The attenuation coefficient is determined by a preset logic damping operator. The oscillation angular frequency, This is the initial phase.

[0010] Preferably, the feature extraction module includes a residual component identification unit, which is used to separate the high-frequency residual signal that characterizes high-frequency resonance from the original sampled signal; the logic damping operation module dynamically adjusts the gain weight of the preset logic damping operator according to the amplitude and phase characteristics of the high-frequency residual signal, so that the system exhibits high damping physical characteristics to cancel the high-frequency ringing signal caused by the coupling of the parasitic parameters of the transmission line and the equivalent impedance of the controlled load.

[0011] Preferably, the system also includes an asynchronous arbitration module, which is connected to the PWM modulation instruction reconstructing module. The asynchronous arbitration module is used to identify the characteristic ripple phase on the parallel bus under the condition of multiple machines in parallel operation. By fine-tuning the execution time of the logic compensation instructions of each power supply unit, the compensation pulses of multiple adjustment systems are staggered in the time domain.

[0012] Preferably, the system also includes a safety decision module, which is used to extract the low-frequency envelope characteristics of the output current and monitor the symbol switching frequency of the feedforward compensation gain in the logic compensation instruction. When the symbol switching frequency exceeds the preset safety threshold, the safety decision module outputs a gain correction instruction to the logic damping operation module to dynamically reduce the gain weight of the virtual inductor parameter.

[0013] Preferably, the safety decision module is also used to identify non-physical sampling noise caused by external electromagnetic interference, and to extract its early weak change characteristics by performing nonlinear transformation on the original sampling signal, thereby separating the non-physical sampling noise from the real load fluctuation signal.

[0014] Preferably, the sampling frequency of the feature extraction module is 100kHz to 500kHz to ensure that the first-order voltage difference information and the second-order voltage difference information cover the voltage transient evolution process of the dynamic response surface of the controlled load end within a time scale of 2μs to 10μs.

[0015] Preferably, when reconstructing the timing of the placeholder edge, the PWM modulation instruction reconstruction module shifts the carrier peak or trough position by 0.1μs to 0.5μs to make the release time of the compensation energy aligned with the rising edge of the endogenous current peak at the controlled load end.

[0016] Preferably, when fine-tuning the execution time, the asynchronous arbitration module sets the compensation pulse offset time between adjacent power supply units to be... The offset time satisfies the following proportional relationship: ,in, The system's switching cycle, This represents the total number of parallel power supply units, and It is an integer greater than or equal to 2; the system achieves global coordination of multi-machine group control by offsetting time.

[0017] Preferably, the safety decision module also includes a preset calibration unit. The preset calibration unit presets the static reference value of the logic damping operator based on the impedance spectrum data of the controlled load under different operating conditions, providing an initial physical anchor point for the generation of logic compensation instructions.

[0018] Compared with existing technologies, the automatic adjustment system for voltage fluctuations in the electroplating process of this invention has the following advantages:

[0019] 1. In the automatic regulation of voltage fluctuations, the edge characteristics of pulse modulation signals are used as active detection excitation. By capturing the real-time response of the output current to micro-perturbations in the phase, the charge saturation of the double-layer capacitor at the cathode interface can be directly inverted. This allows the system to capture the trend of interface admittance within a microsecond window before the overall voltage deviation occurs, thereby triggering the advance compensation of the virtual impedance model. This eliminates the response lag caused by error accumulation in traditional feedback regulation and ensures the transient stability of the output voltage under complex load jump conditions.

[0020] 2. By nonlinearly associating the extracted high-frequency residual signal with the logic damping operator, the control loop exhibits high damping characteristics for energy fluctuations at the logic level without changing the physical hardware topology. This effectively suppresses the high-frequency ringing phenomenon excited by the second-order resonant loop composed of the parasitic inductance of the power transmission cable and the equivalent impedance of the electroplating tank, eliminates the residual voltage oscillation at the far end of the tank, and achieves adaptive matching between the power supply regulation characteristics and the physical parameters of the long-distance transmission path.

[0021] 3. An asynchronous arbitration mechanism is adopted to identify the characteristic ripple phase on the parallel bus. By fine-tuning the execution time of the compensation command of each power supply unit, the compensation pulses of multiple regulation systems are staggered in the time domain. The intermodulation oscillation between parallel systems is canceled by the phase interference principle of physical signals. Global coordination of multi-machine group control is achieved without relying on external communication links, ensuring the voltage stability of large electroplating production lines when multiple task positions are frequently switched. Attached Figure Description

[0022] Figure 1 This is a flowchart of the signal feature extraction and logic compensation control process of the system of the present invention;

[0023] Figure 2 This is a timing diagram of PWM carrier phase reconstruction and load current response in this invention;

[0024] Figure 3 This is a diagram illustrating the hardware architecture and signal interaction of the DSP-based adjustment system of this invention. Detailed Implementation

[0025] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0026] It should be noted that all directional and positional terms used in this invention, such as: up, down, left, right, front, back, vertical, horizontal, inner, outer, top, bottom, transverse, longitudinal, center, etc., are only used to explain the relative positional relationship and connection between components in a specific state (as shown in the accompanying drawings). They are only for the convenience of describing this invention and do not require that this invention be constructed and operated in a specific orientation. Therefore, they should not be construed as limiting this invention. In addition, the descriptions of "first," "second," etc., in this invention are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated.

[0027] In the description of this invention, unless otherwise explicitly specified and limited, the terms installation, connection, and linking should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; they can refer to mechanical connections; they can refer to direct connections or indirect connections through an intermediate medium; they can refer to the internal connection of two components. For those skilled in the art, the specific meaning of the above terms in this invention can be understood according to the specific circumstances.

[0028] In the description of this specification, references to the terms "an embodiment," "some embodiments," "illustrative embodiments," "examples," "specific examples," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example, and the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0029] An automatic adjustment system for voltage fluctuations in an electroplating process includes a feature extraction module, a logic damping operation module, a PWM modulation command reconstruction module, an asynchronous arbitration module, and a safety decision module. The feature extraction module is used to acquire the raw sampled signal of voltage changes at the controlled load terminal. The system generates first-order and second-order voltage differential information. A logic damping operation module connected to the feature extraction module generates logic compensation instructions based on the voltage differential information and a preset logic damping operator. The logic compensation instructions are mapped to the controller's pulse width modulation duty cycle correction step. At the start of each 10.0μs switching cycle, the system reads the current value of the logic compensation instruction. If it is greater than 0.0, the current count value of the 12.0-bit pulse width modulation register is subtracted from the corresponding step increment value; if the value is less than 0.0, the count value is increased accordingly. Each unit of logic... The logic compensation value is forced to change the register count value corresponding to 2.0 system clock cycles, thereby directly converting the abstract virtual damping vector into a physical fine-tuning of the gate drive conduction time of the power switching device. This ensures that the compensation command can intervene in the digital control loop in real time. The PWM modulation command reconstruction module is connected to the logic damping operation module and is used to correct the duty cycle of the digital control loop according to the logic compensation command. The system implants a virtual damping vector in the digital control loop according to the logic compensation command, so that the output voltage generates a damping response that decays exponentially with time, thereby suppressing transient voltage oscillations at the controlled load.

[0030] In pulse electroplating, the output voltage stability of the rectifier system affects the density of the plating surface layer. Existing feedback regulation mechanisms arise after voltage deviation occurs, making it difficult to suppress transient fluctuations on the order of 1μs to 10μs. Therefore, this invention employs a feature extraction module to acquire the original sampling signal. The sampling frequency of the feature extraction module is set within the range of 100kHz to 500kHz to ensure that the differential information covers the voltage evolution process of the controlled load end within a time scale of 2μs to 10μs. The feature extraction module periodically reads the output voltage and uses a register to store the values ​​of three consecutive sampling points. The logic unit calculates the deviation between the current sampling point and the previous sampling point to obtain the first-order differential voltage information, and calculates the change in adjacent first-order differential values ​​to obtain the second-order differential voltage information. For example, when the system sets the reference voltage to 12V, if the sampling signal of three consecutive cycles... With the sequence 12V, 11.8V, and 11.5V, the calculated first-order difference is -0.2V and -0.3V, and the second-order difference is -0.1V. This numerical derivative characteristic characterizes the voltage drop trend, enabling the system to identify the starting point of the fluctuation before the overall voltage deviation occurs. Since the control loop cannot sense the load-side charge demand, it can cause high-frequency ringing signals on the transmission line. To address this challenge, the logic damping operation module performs a weighted operation on the received differential information and a preset logic damping operator to generate a logic compensation command that includes amplitude compensation components and phase lead components. When the system performs adjustment, the output voltage satisfies the following attenuation constraint at the moment of load abrupt change: ,in, For real-time voltage output, The initial amplitude of the fluctuation. The attenuation coefficient is determined by a preset logic damping operator. The oscillation angular frequency, As the initial phase, the logic damping operation module obtains the compensation gain based on the sign and amplitude mapping of the second-order voltage difference information. If a negative second-order difference is detected and its absolute value exceeds a preset threshold of 0.05V, the system increases the attenuation coefficient. The value of the value is used to implant a virtual damping vector with opposite polarity in the digital control loop, so that the output voltage presents a convergent waveform that decays exponentially with time.

[0031] The logic damping operation module retrieves the first-order voltage differential information stored in the register of the feature extraction module. With voltage second-order differential information A control law based on a complex impedance model is established, and a pre-defined logic damping operator converts the numerical derivative characteristics into a virtual damping vector through a mapping table. The gain command and control logic are self-consistent at the physical implementation level. The virtual inductance parameters involved in the safety decision module are adjusted collaboratively as the inductive components of the virtual damping vector in the frequency domain. The engineering debugging procedure is as follows: connect an equivalent impedance load and apply a step excitation in the system's static environment, monitor the output voltage convergence envelope using an oscilloscope, and adjust the first-order compensation coefficient. With second-order compensation coefficient This reduces the number of oscillation cycles of the output voltage after a sudden load change to less than two. The coefficient combination is stored in non-volatile memory as a physical anchor point for the generation of logic compensation instructions. The double-layer capacitor at the cathode interface of the electroplating tank has nonlinear energy storage characteristics, and the voltage change rate will induce an endogenous current spike. The PWM modulation instruction reconstruction module reconstructs the carrier phase and the timing of the occupancy edge of the pulse width modulation signal according to the logic compensation instructions. The PWM modulation instruction reconstruction module uses the pulse edge as a detection excitation and changes the on-time of the power switching device within a specific period in the modulation sequence to generate a phase jitter signal of 20ns to 50ns. The system synchronously captures the phase response of the output current to this tiny jitter and calculates the phase lead angle of the current response relative to the phase jitter signal. The phase lead angle is extracted based on a time correlation analysis process. Simultaneously with emitting a 50.0 ns phase jitter signal, the system activates an internal capture timer with a clock frequency of 200.0 MHz to monitor the zero-crossing moment of the rising edge of the output current after passing through the hysteresis comparator. The system compares this moment with the standard zero-crossing moment under jitter-free conditions, calculates the time difference between the two, and accumulates it in nanoseconds. The arithmetic mean of this time difference is calculated over 20.0 consecutive modulation cycles. This average difference is divided by the 10.0 μs switching cycle duration and multiplied by 360.0 to obtain the final phase lead angle value. If the measured average time difference is 100.0 ns, the mapped phase lead angle is 3.6 degrees. Used to characterize the charge saturation of the current electrolysis interface, when the phase lead angle When an offset occurs, the PWM modulation command reconstruction module performs a phase shift of 0.1μs to 0.5μs on the carrier position, so that the release time of the compensation energy is aligned with the rising edge of the current spike at the controlled load.

[0032] In multi-unit parallel operation, the sampling clock drift of each power supply unit is prone to intermodulation oscillation. To solve this problem, the system integrates an asynchronous arbitration module to identify the characteristic ripple phase on the parallel bus. The asynchronous arbitration module fine-tunes the execution time of the logic compensation instructions of each power supply unit, so that the compensation pulses of multiple regulation systems are staggered in the time domain. The physical phase coordination depends on the hardware capture of the zero-crossing point of the ripple signal below 300.0mV on the parallel bus. The asynchronous arbitration module of each power supply unit monitors the bus voltage fluctuation in real time through a high-speed comparator, and uses the zero-crossing point of the captured ripple rising edge as the zero-point reference for the synchronization time of the entire system. Once the synchronization signal is triggered, each power supply unit executes a specific compensation delay according to the preset hardware coding sequence number: the unit with sequence number 1.0 delays by 0.0μs, and the unit with sequence number 2.0 delays by 2.5μs. This staggered method based on the hardware trigger origin and fixed time delay steps effectively eliminates the risk of compensation pulse overlap between units due to the 5.0ppm clock frequency drift. The compensation pulse offset time between adjacent power supply units is set to... The offset time satisfies the following proportional relationship: ,in, The system's switching cycle, This represents the total number of parallel power supply units, and It is an integer greater than or equal to 2, for example, when the total number of parallel power sources When the value is 4 and the switching frequency is 100kHz, With a duration of 10μs, the asynchronous arbitration module staggers the trigger points of each unit by 2.5μs. This phase coordination mechanism uses the principle of physical signal interference to cancel intermodulation oscillations and ensures the voltage stability of large production lines when switching between multiple task positions.

[0033] When the PWM modulation instruction reconstructing module generates a 20ns to 50ns phase jitter signal, it uses a microprocessor timer comparator register to discretize and shift the carrier edge. The state sensing unit uses a high-order finite impulse response filter to extract the harmonic components in the output current corresponding to the phase jitter frequency, and calculates the phase lead angle based on the zero-crossing time difference of the harmonic components relative to the carrier's starting point. Phase lead angle Characterizing the charge saturation and equivalent double-layer capacitance at the electrolytic interface When the asynchronous arbitration module identifies the characteristic ripple phase of the parallel bus, it captures the composite ripple zero points generated by the superposition of the power supply unit switching frequencies in the bus voltage and calculates the time interval between two adjacent zero points. and the system switching cycle Comparison to determine that each power supply unit has a total of The time-domain position index in the parallel group is used to adjust the compensation pulse offset time. The logic input enables programmed control of physical signal phase interference in multi-machine parallel operation. In industrial electromagnetic environments, narrow pulse interference in the sampling circuit may be misinterpreted as voltage fluctuations, leading to system malfunctions. To address this, a safety decision module is implemented to extract the low-frequency envelope characteristics of the output current and monitor the sign switching frequency of the feedforward compensation gain in the logic compensation command. The safety decision module calculates the number of polarity reversals of the feedforward compensation gain within five consecutive modulation cycles. If the sign switching frequency exceeds a preset safety threshold of three times within a 10μs window, the safety decision module determines that the sampled signal contains non-physical sampling noise and immediately cuts off the output of the feedforward compensation gain. Simultaneously, the safety decision module uses a preset calibration unit to preset the static reference value of the logic damping operator based on the load impedance spectrum data, providing an initial physical anchor point for the logic compensation command. When the low-frequency envelope indicates an increase in load current density, the safety decision module dynamically adjusts the gain weight of the virtual inductor parameter to prevent overshoot under heavy load conditions and ensure the system's operational stability in extreme electromagnetic environments.

[0034] Example 1: In the electroplating line for producing high aspect ratio blind vias on printed circuit boards, the system faces a distributed power supply environment consisting of four high-frequency rectifier power supplies connected in parallel. The controlled load exhibits complex impedance characteristics due to the nonlinear energy storage features of the double-layer capacitance at the cathode interface. Furthermore, the long-distance power transmission cables introduce parasitic inductances of 5μH to 10μH. Under dynamic conditions of frequent switching between multiple workstations, the rectifier system output generates a high-frequency ringing signal caused by the coupling between the parasitic parameters of the transmission lines and the equivalent impedance of the controlled load. This causes voltage fluctuations to enter a microsecond-level response dead zone, directly resulting in uneven plating thickness inside the blind vias. When a change in ion concentration at the electrode interface causes an impedance drop, the feature extraction module acquires the original sampling signal at a sampling frequency of 500kHz. The system continuously extracts first-order and second-order voltage difference information by calculating the deviation between adjacent sampling periods. The logic damping operation module receives the numerical derivative features and performs weighted operations with preset logic damping operators, dynamically adjusting the attenuation coefficient. The compensation strength is locked at the 10V dropout inflection point, and the output voltage satisfies the constraint relationship at the moment of load change. ,in, For real-time voltage output, The initial amplitude of the fluctuation. The attenuation coefficient is determined by a preset logic damping operator. The oscillation angular frequency, As the initial phase, the PWM modulation instruction reconstruction module performs a 0.1μs phase shift on the carrier phase according to the generated logic compensation instruction. Simultaneously, it uses the pulse edge as an active detection excitation to change the conduction time of a specific period, generating a 50ns phase jitter signal. The state sensing unit synchronously captures the phase response of the output current to this jitter and calculates the phase lead angle. Phase lead angle By characterizing the charge saturation of the electrolytic interface in real time, the release of compensation energy is precisely aligned with the rising edge of the current spike generated at the controlled load end, thereby achieving phase alignment between the feature prediction drive and the physical damping vector, reducing the convergence time of the system during load abrupt changes from 5ms to the 50μs level.

[0035] The asynchronous arbitration module extracts the zero-point phase of the bus ripple under multi-machine parallel operation, based on the total number of parallel power supply units. The system switching cycle is 4. Execute offset time with an initial specification of 10 μs. The allocation logic is used to calculate the offset time of the compensation pulse for each unit. It is 2.5 μs, where, This refers to the compensation pulse offset time between adjacent power supply units. The system's switching cycle, To determine the total number of parallel power supply units, the regulation pulses of the four rectifier power supplies are uniformly and interleaved in the time domain. This utilizes the principle of phase interference of physical signals to cancel intermodulation oscillations between the parallel systems. The safety decision module calculates the symbol switching frequency of the feedforward compensation gain within a 10μs window. The system determines that the current number of polarity reversals is less than the safety threshold and maintains the execution of the adjustment command, so that the residual ripple amplitude of the bus voltage is maintained within the range of 0.05V, ensuring the global consistency of the coating growth in the high aspect ratio blind hole.

[0036] Example 2: To verify the transient suppression capability and engineering feasibility of the adjustment system of the present invention under dynamic load conditions, a physical test platform containing four parallel rectified power supplies was built. The test platform was equipped with a simulated electroplating tank load unit and a digital oscilloscope with a sampling frequency of 1 GS / s and a bandwidth of 200 MHz. The data used in the experiment came from the real-time acquisition of bus electrical variables by the platform under different operating conditions. The core parameter sampling frequency was set. The technical consideration is to capture the microsecond-level potential drop inflection point at the controlled load end. This requires balancing the interrupt handling load of the controller with the resolution of the transient signal. When the spectral characteristic bandwidth of the controlled load end is in the range of 50kHz to 100kHz, the sampling frequency is determined to satisfy the Nyquist sampling theorem and allow for phase margin. It is 500kHz, where, The sampling frequency is expressed in kHz. This value serves as the sampling reference for this experiment, covering the evolution of both first-order and second-order voltage differential information.

[0037] To simulate the real electromagnetic environment of an industrial site, Gaussian white noise with a signal-to-noise ratio of 20dB and 50Hz power frequency interference harmonics were superimposed on the original voltage sampling circuit. The experimental group used the automatic adjustment system of this invention, while the control group used proportional, integral, and derivative adjustment. When a step current change from 10A to 50A was simulated at the controlled load, the output voltage of the control group showed a deviation of 1.32V and a convergence time of 8.4ms. In contrast, the experimental group of this invention obtained the original sampling signal within 2μs of detecting the voltage change. The first-order voltage difference is -0.24V, and the second-order voltage difference is -0.07V. The original sampled signal is in volts (V). The logic damping module adjusts the attenuation coefficient based on the second-order voltage difference information. This suppresses output voltage fluctuations to within 0.138V, and the processed ripple signal converges within 47.6μs, filtering out actively superimposed background noise. As the attenuation coefficient, when verifying the detection logic of the PWM modulation command reconstructing module, the system introduces a 40ns phase jitter signal into the modulation sequence. The state sensing unit synchronously captures the response component of the output current to this jitter and calculates the phase lead angle. It is 14.5°, where, The phase lead angle, measured in degrees, characterizes the charge saturation state at the cathode interface of the electroplating tank. The PWM modulation command reconstruction module uses this phase lead angle as a reference. By shifting the carrier phase by 0.22 μs, the release time of the compensation energy is aligned with the rising edge of the current spike generated at the controlled load. The measured data shows that, under the adjustment of the sample group of this invention, due to the phase alignment between the physical damping vector and the interface charge requirement, the voltage drop slope at the output of the rectified power supply is reduced from 0.15 V / μs in the original state to 0.02 V / μs, which confirms the compensation efficiency of the carrier phase reconstruction mechanism for electrochemical inertia.

[0038] To demonstrate the rationality of the technical parameter range, an out-of-range control group was set up, and the sampling frequency was adjusted. When the frequency was reduced to 80kHz, the experimental results showed that, due to the sampling frequency being below 100kHz, the system could not obtain effective second-order differential characteristics before the overall voltage deviation occurred, resulting in a logic lag of 120.5μs in the adjustment command and an oscillation of 0.85V in the output voltage. This was further investigated when examining the feedforward compensation gain. When the absolute value of the second-order voltage difference exceeds 0.15V, the performance improvement curve shows a saturation trend, and the growth rate of key performance indicators decreases from 1.25 to 0.18. This confirms that the 0.05V fluctuation threshold and related parameter weights defined in this invention are working windows after engineering trade-offs. By setting a partially missing control group, while maintaining logic damping operations, the asynchronous arbitration module is removed. Under the condition of four rectifier power supplies operating in parallel and each unit clock having a 5ppm drift, a low-frequency intermodulation oscillation with a peak-to-peak value of 0.68V is generated on the parallel bus. However, in the sample group of this invention, after the asynchronous arbitration module intervenes, the performance improvement curve shows a saturation trend, and the growth rate of key performance indicators decreases from 1.25 to 0.18. This confirms that the 0.05V fluctuation threshold and related parameter weights defined in this invention are working windows after engineering trade-offs. The parameter is set to 4, which sets the compensation pulse offset time for each unit. The system switching cycle is set to 2.5 μs. With a time interval of 10μs, the command trigger points of each power supply unit are evenly and alternately distributed in the time domain. The intermodulation oscillation between parallel systems is canceled by the phase interference principle of physical signals. The measured residual ripple amplitude of the bus voltage drops to 0.042V. This experimental conclusion confirms that the present invention achieves voltage stability under multi-machine group control through the coordinated operation of various functional modules.

[0039] Example 3: In a new generation of vertical continuous electroplating production line for printed circuit boards, the power supply cluster faces high-frequency pulse group interference caused by the switching of adjacent high-power frequency converters. This interference acts on the sampling circuit of the feature extraction module and generates random voltage spikes with pulse widths ranging from 100ns to 500ns. If the system logic misinterprets the signal as load voltage fluctuations, it will drive the PWM modulation instruction reconstruction module to generate an incorrect phase translation operation, causing the output voltage to oscillate near the equilibrium point. The system uses a digital signal processor with a main frequency of not less than 100MHz and equipped with a single-cycle multiplication instruction. The feature extraction module performs multiplication in each switching cycle. The analog-to-digital conversion sampling is triggered 2μs after the start to obtain the current sample value. ,in, The switching period of the system is expressed in μs. The original sampled signal, in volts (V), is used by the logic unit to calculate the difference between the current sampled value and the sampled value of the previous cycle to obtain the first-order differential voltage information. The second-order voltage difference information is obtained by calculating the difference between adjacent first-order difference values. The logic damping operation module performs weighted compensation operations and generates logic compensation instructions. The specific calculation formula is as follows: ,in, This is a logical compensation instruction. The first-order compensation coefficient is... The second-order compensation coefficient is used in this embodiment. Set to 1.25 and Set to 0.45, if If the absolute value exceeds 0.05V and the duration reaches 2 sampling cycles, the system determines that the controlled load has entered a transient fluctuation state and increases the feedforward compensation gain. The weight.

[0040] Safety decision module monitors feedforward compensation gain To identify high-frequency burst interference, the module uses a 5-inch circular buffer in its storage unit and records the polarity changes of the most recent 5 sampling periods. The sign of the positive or negative data is used by the security decision module to calculate the number of times adjacent data in the buffer have inconsistent polarities, and the number of times the polarity is reversed within a 10μs time window. After four iterations, the system determines that the input signal contains sampling noise with a frequency higher than the load response bandwidth. The safety decision module then outputs a logic status bit to update the logic compensation command. The output value is corrected to 0, and the phase shift function of the PWM modulation instruction reconstruction module is blocked. When the sampling circuit is disturbed and produces a jump of 12.5V, 11.5V, and 12.6V, the calculated differential sequence drives the gain polarity switching. The safety decision module triggers protection and stops the phase shift operation at the fourth polarity switching point to maintain the reference stability of the bus voltage under the interference environment.

[0041] Example 4: In the initial deployment scenario for electroplating tanks of different specifications, the system executes standardized engineering calibration procedures to establish feedforward compensation gain. With voltage second-order differential information Based on the corresponding relationship, the rectified power supply is connected to a test load with a known equivalent impedance. An external signal generator is used to generate a step current excitation at different attenuation coefficients. The response waveform of the output voltage is recorded at the set value, and the logic unit analyzes the obtained first-order differential voltage information. With voltage second-order differential information The transient amplitude is selected, and the coefficient combination that minimizes the output voltage convergence time and keeps the overshoot within 0.1V is stored in a lookup table in non-volatile memory. By performing cyclic tests at 5 discrete load points, the system obtains a set of compensation weight reference values ​​covering light load to full load conditions, which serve as the basis for logic compensation instructions. It provides data input for real-time computing.

[0042] When the system is applied to field conditions with different transmission cable lengths, the system executes the field deployment pre-deployment commissioning procedure to calibrate the initial physical reference of the PWM modulation command reconstructing module. In the static standby state, the PWM modulation command reconstructing module generates a reference signal with a fixed frequency and a pulse width of 50%, and superimposes a phase jitter signal with an amplitude of 50ns as a detection vector. The state sensing unit captures the current feedback at the controlled load end and calculates the initial phase lead angle caused by the parasitic inductance of the line. The logic unit will advance the initial phase angle. As the zero-point reference for the compensation algorithm, and to correct the initial weight of the carrier phase shift, the adjustment system completes the initial setting of the distributed inductance parameters for a specific production line.

[0043] Example 5: In the system calibration procedure for acidic copper plating bath environments with different ion concentrations, the system executes a parameter determination process based on frequency response analysis. A small sinusoidal excitation signal with an amplitude of 10mV and a frequency continuously varying from 10Hz to 100kHz is applied across the electrodes. The state sensing unit simultaneously records the current response and generates an equivalent impedance spectrum of the bath solution. The logic operation unit determines the double-layer capacitance of the controlled load by identifying the semi-circular characteristics in the spectrum. With solution resistance And according to the formula Calculate the characteristic time constant The system selects an attenuation coefficient. The time constant is 1.5 to 2.5 times the reciprocal of the time constant, locking the operational benchmark of the logic damping module within the physical response range of the tank solution. It is a double-layer capacitor, with the unit being F. The resistance of the solution is expressed in units of Ω. , The characteristic time constant is expressed in seconds. This is the attenuation coefficient.

[0044] When a change in the electrode spacing causes a drift of more than 1 μH in the distributed inductance parameter, the system executes an online phase alignment calibration procedure to determine the phase shift step size of the PWM modulation command reconstruction module. The state sensing unit extracts the third harmonic component of the output current and calculates the phase deviation during the phase jitter period. The logic unit performs a linear mapping and sets the phase shift step size. Phase deviation Satisfy proportional relationship ,in, This represents the phase shift step size, in μs. This is the phase deviation, expressed in degrees (°). and To obtain a fixed coefficient determined through pre-calibration, the system adjusts the carrier phase to align the peak energy of the compensation pulse with the peak value of the current spike within a 100ns error range, thereby eliminating dynamic mismatch caused by differences in physical wiring.

[0045] The embodiments of this application have been described above with reference to the accompanying drawings. Unless otherwise specified, the embodiments and features in the embodiments of this application can be combined with each other. This application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit of this application and the scope of protection of this invention, and all of these forms are within the protection scope of this application.

Claims

1. An automatic adjustment system for voltage fluctuations during electroplating, characterized in that, The system includes a feature extraction module, a logic damping operation module, and a PWM modulation instruction reconstructing module. The feature extraction module is used to acquire the original sampled signal that characterizes the transient changes in the voltage at the controlled load terminal, and extract the first-order voltage difference information and the second-order voltage difference information from it. The logic damping operation module, whose input is connected to the feature extraction module, is used to perform weighted operations on the first-order voltage difference information and the second-order voltage difference information with the preset logic damping operator to generate a logic compensation instruction containing amplitude compensation component and phase lead component. The PWM modulation instruction reconstruction module, whose input is connected to the logic damping operation module, is used to reconstruct the carrier phase and duty cycle timing of the pulse width modulation signal according to the logic compensation instruction, so as to correct the duty cycle of the digital control loop. In this module, the system implants a virtual damping vector with the opposite oscillation polarity to the original sampled signal into the digital control loop according to the logic compensation instruction, so that when the output voltage of the system changes abruptly at the controlled load, it presents a convergent waveform that decays exponentially with time and is accompanied by periodic oscillations, so as to suppress the transient voltage oscillation at the controlled load. In addition, the sampling frequency of the feature extraction module is 100kHz to 500kHz to ensure that the first-order voltage difference information and the second-order voltage difference information cover the voltage transient evolution process of the dynamic response surface of the controlled load end within the time scale of 2μs to 10μs. When reconstructing the timing of the placeholder edge, the PWM modulation instruction reconstruction module shifts the carrier peak or trough position by 0.1μs to 0.5μs to make the release time of the compensation energy aligned with the rising edge of the endogenous current peak at the controlled load end. The safety decision module also includes a preset calibration unit. Based on the impedance spectrum data of the controlled load under different operating conditions, the preset calibration unit presets the static reference value of the logic damping operator, providing an initial physical anchor point for the generation of logic compensation instructions. The PWM modulation instruction reconstruction module reconstructs the carrier phase and positional edge timing of the pulse width modulation signal based on the logic compensation instruction. Using the pulse edge as the detection excitation, the module changes the on-time of the power switching device within a specific period in the modulation sequence, generating a phase jitter signal of 20ns to 50ns. The system synchronously captures the phase response of the output current to this minute jitter and calculates the phase lead angle of the current response relative to the phase jitter signal. ; When the phase lead angle When an offset occurs, the PWM modulation command reconstruction module performs a phase shift of 0.1μs to 0.5μs on the carrier position, so that the release time of the compensation energy is aligned with the rising edge of the current spike at the controlled load.

2. The automatic adjustment system for voltage fluctuations in the electroplating process according to claim 1, characterized in that, When the system performs regulation, the output voltage satisfies the following attenuation constraint at the moment of load change: ,in, For real-time voltage output, The initial amplitude of the fluctuation. The attenuation coefficient is determined by a preset logic damping operator. The oscillation angular frequency, This is the initial phase.

3. The automatic adjustment system for voltage fluctuations in the electroplating process according to claim 1, characterized in that, The feature extraction module includes a residual component identification unit, which is used to separate the high-frequency residual signal that characterizes high-frequency resonance from the original sampled signal; The logic damping operation module dynamically adjusts the gain weight of the preset logic damping operator based on the amplitude and phase characteristics of the high-frequency residual signal, so that the system exhibits high damping physical characteristics.

4. The automatic adjustment system for voltage fluctuations in the electroplating process according to claim 1, characterized in that, The system also includes an asynchronous arbitration module, which is connected to the PWM modulation instruction reconstruction module. It is used to identify the characteristic ripple phase on the parallel bus under multi-machine parallel operation. By fine-tuning the execution time of the logic compensation instructions of each power supply unit, the compensation pulses of multiple regulation systems are staggered in the time domain.

5. The automatic adjustment system for voltage fluctuations in the electroplating process according to claim 1, characterized in that, The system also includes a safety decision module, which is used to extract the low-frequency envelope characteristics of the output current and monitor the symbol switching frequency of the feedforward compensation gain in the logic compensation instruction. When the symbol switching frequency exceeds the preset safety threshold, the safety decision module outputs a gain correction instruction to the logic damping operation module to dynamically reduce the gain weight of the virtual inductor parameter.

6. The automatic adjustment system for voltage fluctuations in the electroplating process according to claim 5, characterized in that, The safety decision module is also used to identify non-physical sampling noise caused by external electromagnetic interference. By performing a nonlinear transformation on the original sampling signal to extract its early weak change characteristics, the non-physical sampling noise is separated from the real load fluctuation signal.

7. An automatic adjustment system for voltage fluctuations in the electroplating process according to claim 4, characterized in that, When fine-tuning the execution time, the asynchronous arbitration module sets the compensation pulse offset time between adjacent power supply units to be... The offset time satisfies the following proportional relationship: ,in, The system's switching cycle, This represents the total number of parallel power supply units, and It is an integer greater than or equal to 2; the system achieves global coordination of multi-machine group control by offsetting time.

Images