Ultrasonic transducer characteristic parameter testing system

Abstract

The invention discloses an ultrasonic transducer characteristic parameter test system, which belongs to the technical field of ultrasonic detection and comprises an excitation signal generation module, an impedance matching adjustment module, a resonance characteristic extraction module, a signal acquisition and processing module, a temperature control loading module, a physical vibration isolation module and a frequency spectrum calibration module. The excitation signal generation module is combined with an off-chip frequency locking feedback loop through an on-chip oscillator to generate a multi-frequency-point sweep frequency pulse array, the multi-frequency-point sweep frequency pulse array is connected to the resonance characteristic extraction module after passing through the impedance matching adjustment module, the multi-frequency-point sweep frequency pulse array and the frequency spectrum calibration module construct a bidirectional feedback path and share harmonic node synchronous calibration logic, and coupling offset compensation is achieved; the frequency spectrum calibration module performs bidirectional real-time error correction on the excitation output frequency and the resonance input phase; and the transducer is arranged in the thermal convection temperature equalizing cabin of the temperature control loading module and is overlapped with the physical vibration isolation module. The method has the advantages that high-precision, low-interference and full-frequency-domain dynamic testing of characteristic parameters of the ultrasonic transducer is achieved, and testing stability is improved.

Description

Technical Field

[0001] The present invention relates to the field of ultrasonic detection technology, and more particularly to a system for testing characteristic parameters of an ultrasonic transducer. Background Art

[0002] Traditional ultrasonic transducer characteristic parameter test systems are mostly based on impedance analysis or time-domain pulse reflection methods. Their development has gradually evolved towards higher precision, stronger stability, and wider adaptability. Early systems were limited by analog circuit drive and low-speed sampling, making it difficult to capture weak resonant responses and phase disturbances. In recent years, with the introduction of high-resolution ADCs, FPGAs, and digital filtering technologies, test systems have gradually acquired real-time spectrum scanning, dynamic temperature-controlled loading, and multi-channel parallel analysis capabilities, driving transducer performance evaluation towards intelligent and automated development.

[0003] However, existing technologies still face several key bottlenecks: the lack of a refined phase tracking mechanism for adjacent sampling points makes it difficult to identify perturbations at the resonant point; the gain control paths are mostly open-loop structures, making dynamic closed-loop correction impossible; frequency and phase corrections are typically unidirectional, which can easily lead to characteristic fitting deviations; and harmonic interference lacks a dedicated compensation channel in the frequency domain, affecting parameter extraction accuracy. Therefore, it is necessary to build an integrated test system capable of phase spectrum extraction, gain closed-loop control, bidirectional error correction, and coupling compensation to overcome existing limitations and meet the refined requirements for high-performance transducer characteristic evaluation. Summary of the Invention

[0004] The purpose of the present invention is to provide an ultrasonic transducer characteristic parameter testing system to solve the problems raised in the above-mentioned background technology: such as the lack of a refined phase tracking mechanism for adjacent sampling points, which makes it difficult to identify the perturbation of the resonance point; the gain control path is mostly an open-loop structure, which cannot achieve dynamic closed-loop correction; the frequency and phase correction are usually unidirectional adjustments, which easily lead to characteristic fitting deviations; harmonic interference fails to construct a special compensation channel in the frequency domain, affecting the accuracy of parameter extraction.

[0005] Technical solution: The ultrasonic transducer characteristic parameter testing system includes an excitation signal generation module, an impedance matching adjustment module, a resonance characteristic extraction module, a signal acquisition and processing module, a temperature control loading module, a physical vibration isolation module and a spectrum calibration module;

[0006] The excitation signal generation module forms a multi-frequency sweep pulse array by configuring an on-chip oscillator and an off-chip frequency-locked feedback loop, which is directly coupled to the impedance matching adjustment module. The impedance matching adjustment module includes a voltage-controlled matching network and a phase shifter array configured in parallel, which is used to implement amplitude adaptive matching within the range of 0.5Ω to 1000Ω; the matching output end is connected to the resonance characteristic extraction module, and the resonance characteristic extraction module forms a feedback linkage path with the spectrum calibration module through a bidirectional phase superposition path, and simultaneously integrates the amplitude voltage and current feedback signals into the signal acquisition and processing module. The resonance characteristic extraction module and the spectrum calibration module construct a coupling offset compensation channel through harmonic node synchronous calibration logic. The temperature control loading module and the physical vibration isolation module are arranged in a superimposed manner, and the transducer is embedded in the thermal convection temperature chamber of the temperature control loading module. The spectrum calibration module also performs bidirectional real-time error correction on the output frequency of the excitation signal generation module and the input phase of the resonance characteristic extraction module.

[0007] Preferably, the resonance characteristic extraction module is composed of a phase solution channel, a synchronous multi-phase acquisition array, an impedance vector conversion bridge and a harmonic residual control unit;

[0008] The phase resolution channel is arranged crosswise and synchronously with the voltage sampling path and the current sampling path. The impedance vector conversion bridge is composed of two sets of precision adjustable capacitors and a constant-amplitude impedance divider circuit. The phase resolution channel output and the harmonic residual control unit form a bidirectional closed loop. The synchronous multi-phase acquisition array uses an 8-channel parallel synchronization method to achieve a 100MSa / s sampling rate, so that the phase error is controlled within ±0.1°. The impedance vector conversion bridge maintains a vector angle stability error of less than 0.3% under signals below -10dB.

[0009] Preferably, the harmonic residual control unit has an embedded digital phase-locked loop array and a differential dynamic compensation chip to form a dual-channel frequency tracker. The phase-locked loop array adopts a three-level feedback sampling window structure to construct a dynamic phase-amplitude mapping matrix for the ±1kHz region around the excitation frequency, and uses the real-imaginary part fitting function to construct a residual decreasing trend line at the static equilibrium point, thereby gradually approaching the phase inflection point of the resonance point in multiple scans, so that the resonant frequency positioning accuracy of the equivalent inductor-capacitor network is higher than 10Hz.

[0010] Preferably, the differential dynamic compensation chip constructs a gain closed-loop path based on a CMOS ultra-low offset structure and a delay adjustable comparator, embeds an adjustable sampling window parameter unit, sets a sampling step of 5ns, constructs a phase drift spectrum of adjacent sampling points, and realizes accurate extraction of signal phase fluctuation trends with perturbations less than 0.1%. Finally, a nonlinear dual-parameter fitting plane is constructed at the data fitting end to invert the true reactance change trend at the resonant frequency.

[0011] Preferably, the temperature control loading module is composed of a Peltier heat pump array, a quasi-steady-state airflow convection unit and a nano-cavity thermal insulation stack assembly;

[0012] The heat pump array consists of four sets of reverse thermocouple elements, and achieves ±0.05°C precision control of the internal cavity temperature through a time-division switching mode. The airflow convection unit forms a closed-loop heat exchange circuit through the stepped distributed turbine ducts at both ends and the transducer mounting cavity. The nanocavity thermal insulation stack assembly adopts a graphene film-polyimide multilayer composite structure, achieving a thermal conductivity of less than 0.02W / m·K within a thickness of 3mm, shielding the heat flow outside the cavity.

[0013] Preferably, the transducer and the thermal convection cavity are installed in an oblique thermal coupling manner with a fixed angle of 85°±2°, the contact surface is coated with a phase change thermal conductive adhesive layer with a thickness of 0.2mm, and a high-conductivity micro-filler is used to increase the thermal conductivity to 8W / m·K. The internal convection flow channel is tuned through dynamic heat flow simulation to keep the surface temperature difference of the transducer within ±0.1°C during the steady-state test phase.

[0014] Preferably, the spectrum calibration module is composed of an on-chip frequency vector modulator, an off-chip variable storage array, and an asymmetric Gaussian filter group;

[0015] The frequency vector modulator adopts on-chip self-learning FPGA control logic to adjust the output frequency step interval and waveform factor in real time according to the external feedback error. The frequency accuracy adjustment resolution reaches 0.1Hz. The asymmetric Gaussian filter group uses a bidirectional window function to retain the harmonic envelope peak position while suppressing the high-frequency edge fluctuations of the spectrum, so that the signal energy main lobe position control error is less than ±0.2%.

[0016] Preferably, a multidimensional frequency-amplitude mapping table is set in the variable storage array, and the multidimensional frequency-amplitude mapping table updates the frequency error correction parameters 4 times per second through the on-chip EEPROM periodic write mechanism. The correction envelope injected into each group of frequency points includes at least three quantitative parameters: amplitude correction factor, phase correction factor, and harmonic content coefficient. The correction parameters are adaptively updated according to the Bayesian estimation model to achieve dynamic optimization based on the historical correction trajectory, thereby suppressing the step error caused by temperature drift and signal drift during the spectrum calibration process.

[0017] Preferably, the physical vibration isolation module is composed of a three-layer structure, including a carbon fiber support grid, a high-damping elastic support and a lower rigid aluminum alloy fixed base;

[0018] The high-damping elastic support adopts a butyl rubber-polyurethane two-component composite material with a hardness controlled at 60ShoreA and an embedded microporous vibration isolation net. The natural frequency of the entire physical vibration isolation module is controlled below 5Hz, isolating the measurement perturbations caused by external mechanical vibrations. A ceramic insulating bracket is set between the transducer and the test circuit and filled with polymer damping gel to avoid any low-frequency structural resonance affecting parameter fluctuations.

[0019] Compared with the prior art, the advantages of the present invention are:

[0020] (1) By constructing a high-density phase sampling network of components, the phase differences between adjacent points are dynamically captured to form a complete drift map, which is used to identify local structural resonance interference and improve the phase discrimination accuracy.

[0021] (2) Firmware gain closed-loop path design: A closed-loop gain control path is introduced in the signal acquisition and drive link, and the drive amplitude and receiving gain are adjusted through real-time feedback to achieve cross-band and consistent excitation and response amplitude control.

[0022] (3) A phase superposition feedback path is constructed between the resonance characteristic extraction module and the spectrum calibration module, which can superimpose and compare the original response phase and the expected phase in real time to form an active phase correction.

[0023] (4) Through the harmonic node synchronization calibration logic, the nonlinear harmonic excitation points are extracted and their frequency domain offset trends are identified, and a coupling offset compensation channel is constructed to effectively eliminate the characteristic distortion caused by harmonic disturbances.

[0024] (5) The spectrum calibration module not only corrects the excitation frequency, but also synchronously calibrates the phase of the response signal, realizing frequency-phase bidirectional linkage correction, significantly improving the stability and accuracy of the transducer characteristic spectrum.

[0025] (6) In the process of characteristic parameter extraction, multi-dimensional feedback signals such as frequency response, amplitude stability, and phase change rate are introduced to dynamically optimize the sampling path and test rhythm to improve the accuracy and efficiency of parameter analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 This is a schematic diagram of the overall system of an ultrasonic transducer characteristic parameter testing system of the present invention; DETAILED DESCRIPTION

[0027] For examples, see Figure 1 ,An ultrasonic transducer characteristic parameter testing system includes an excitation signal generation module, an impedance matching adjustment module, a resonance characteristic extraction module, a signal acquisition and processing module, a temperature control loading module, a physical vibration isolation module and a spectrum calibration module;

[0028] The excitation signal generation module forms a multi-frequency sweep pulse array by configuring an on-chip oscillator and an off-chip frequency-locked feedback loop, which is directly coupled to the impedance matching adjustment module. The impedance matching adjustment module includes a voltage-controlled matching network and a phase shifter array configured in parallel, which is used to implement amplitude adaptive matching in the range of 0.5Ω to 1000Ω; the matching output end is connected to the resonance characteristic extraction module, which forms a feedback linkage path with the spectrum calibration module through a bidirectional phase superposition path, and at the same time integrates the amplitude voltage and current feedback signals into the signal acquisition and processing module. The resonance characteristic extraction module and the spectrum calibration module construct a coupling offset compensation channel through the harmonic node synchronous calibration logic. The temperature control loading module and the physical vibration isolation module are arranged in superposition with each other, and the transducer is embedded in the thermal convection equalization chamber of the temperature control loading module. The spectrum calibration module also performs bidirectional real-time error correction on the output frequency of the excitation signal generation module and the input phase of the resonance characteristic extraction module.

[0029] Specifically, the feedback linkage path structure and data path are as follows:

[0030] The feedback linkage path includes the following core units:

[0031] ResonantPhaseSynthesizer

[0032] Overlay Phase Buffer (PhaseBufferRegisterFile)

[0033] SpectrumFeedbackInterfaceController

[0034] Cooperative Tuning Scheduler

[0035] These modules collaborate to build a programmable phase coupling path using on-chip FPGA logic resources. Their functions are as follows:

[0036] Linkage process description:

[0037] Step 1: Initial resonance point extraction

[0038] The system first samples the amplitude and phase of the transducer echo signal within the sweep frequency range, identifying possible resonance points (where the amplitude is minimal or the phase jumps). The phase information of these points is stored in a superposition buffer, forming the first frame of the phase spectrum.

[0039] Step 2: Bidirectional phase superposition construction

[0040] The system performs bidirectional phase difference vector superposition processing on two frames of sampling data:

[0041]

[0042] in, Represents the phase of the kth sampling point at frequency f:

[0043] w k The weighting factor is set based on experience (the default is triangular window distribution);

[0044] N is the sampling window length.

[0045] Through this superposition, the system can amplify the intensity of the phase mutation near the resonance point and suppress the background phase noise.

[0046] Step 3: Reverse input of spectrum calibration parameters

[0047] The multi-dimensional frequency-amplitude mapping table included in the spectrum calibration module provides the following feedback after each scan:

[0048] Current frequency deviation correction value Δf i ;

[0049] Amplitude gain correction factor G i ;

[0050] Phase correction offset δφ i .

[0051] This information is read by the spectrum feedback interface controller and updated to the resonance point phase builder in real time to correct the phase reference during phase superposition:

[0052]

[0053] This ensures that the phase mutation calculation results remain physically consistent despite temperature drift or waveform distortion.

[0054] Step 4: Co-modulation Optimization

[0055] If there is misidentification in the superposition curve or the suspected resonance point is not obvious, the coordinated modulation scheduling unit will start the fine-tuning mechanism:

[0056] Adjust the frequency step interval Δf to 0.1Hz accuracy;

[0057] Perform secondary sampling at the candidate resonance point;

[0058] Asymmetric Gaussian filter groups are used to further suppress edge noise;

[0059] A Savitzky-Golay third-order fitting is performed on the phase drift trend line to extract the inflection point curvature.

[0060] Finally, the system uses the phase transition slope at the resonance point fitting peak as the final criterion to determine the true resonance point.

[0061] Specifically, the coupling offset compensation channel consists of the following units:

[0062] Harmonic Spectrum Identifier

[0063] Synchronous Calibration Comparator Unit (SynchronousCalibrationComparatorUnit)

[0064] Coupling Offset Matrix Generator

[0065] Band Reconstruction Modulator (BandReallocationModulator)

[0066] The above units are integrated and constructed within an on-chip FPGA or SoC platform, which can dynamically identify harmonic nodes and implement compensation control.

[0067] Workflow and compensation mechanism:

[0068] Step 1: Harmonic node identification and synchronization lock

[0069] The harmonic spectrum identifier constructs a main frequency-harmonic index table based on the fast Fourier transform (FFT) result of the real-time sampling signal. After each frame of data is sampled, the system automatically extracts the main frequency f0 and its certificate multiple frequency f n =n·f0 (where n=2, 3, 4...) frequency amplitude and phase.

[0070] For harmonic frequencies with amplitudes greater than the noise floor, the system marks them as "active harmonic nodes" and records their instantaneous phase information φ n (t), used for subsequent compensation matrix construction.

[0071] Step 2: Create a synchronous calibration logic reference table

[0072] After each round of harmonic identification, the system will calculate the phase drift Δφ between the main frequency point and the active harmonic node n =φ n (t)-n·φ0(t) is recorded and added to the synchronization phase offset table:

[0073] SyncTable={(f n ,Δφ n ,A n )}

[0074] Among them, A n is the amplitude of the harmonic component, which is used for subsequent weighting.

[0075] Step 3: Construct the coupling offset compensation matrix

[0076] The coupling compensation matrix generator constructs the following offset compensation coefficient matrix C based on SyncTable ij

[0077]

[0078] Where: i, j are the numbers of adjacent measurement channels;

[0079] α ij is the inter-channel coupling coefficient (obtained through static calibration);

[0080]

[0081] N is the number of currently active harmonic nodes.

[0082] This matrix is used to represent the transverse coupling phase error caused by harmonic interference.

[0083] Step 4: Compensation Signal and Band Modulation

[0084] Finally, the band reconstruction modulator applies the compensation matrix to the sampled data channel, introducing an inverse compensation phase at each measurement point:

[0085] φ′ i =φ i -∑ j C ij

[0086] This enables real-time compensation of harmonic coupling offset errors in the measurement data, effectively improving the accuracy of determining the position of the transducer resonance point.

[0087] Implementation effects and advantages:

[0088] This compensation channel effectively reduces signal phase drift and frequency shifts caused by parasitic coupling between multiple channels and non-ideal filter tail harmonic leakage. In a typical high-frequency resonance (5MHz) test scenario, the compensation reduced transducer impedance curve fluctuations by over 40%, and the error in phase characteristic curvature extraction was reduced to ±0.1°.

[0089] In addition, the logic channel has dynamic adaptability, can automatically identify the number of active harmonic nodes and adjust the size of the compensation matrix, and is suitable for the precise resonance characteristic extraction of various types of transducers (piezoelectric, magnetostrictive, CMUT, etc.).

[0090] Specifically, two-way real-time correction path, synchronous correction:

[0091] The output frequency of the excitation signal generator (ESG);

[0092] Input phase data for the Resonant Feature Extraction Module (RFEM).

[0093] The core of this mechanism is a dual-path frequency-phase error tracking and compensation loop (Dual-Path Frequency-Phase Error Compensation Loop), which includes the following key logic units:

[0094] 1. System composition structure:

[0095] Frequency Drift Monitor (FDM)

[0096] Reference Phase-Locked Loop (R-PLL)

[0097] Dynamic Frequency Adjuster (DFA)

[0098] Input Phase Corrector (IPC)

[0099] 2. Working principle and process:

[0100] Step 1: Real-time identification and correction of excitation signal frequency error

[0101] The spectrum calibration module continuously monitors the output frequency of the excitation signal generation module ESG and compares it with the ideal reference frequency source (such as crystal oscillator, GPS or on-chip constant temperature oscillator) through the frequency stability monitoring unit FDM. If a deviation (Δf = f actual -f ref ), a frequency correction instruction is generated.

[0102] This instruction acts on the ESG through the output frequency dynamic adjuster DFA and performs fine-tuning according to the following modulation method:

[0103] f out ′=f out -k f Δf

[0104] where k f The modulation coefficient is automatically adjusted to maintain system stability, with a typical accuracy of ±10Hz.

[0105] Step 2: Real-time calibration of the resonance characteristic input phase

[0106] Before the transducer response signal is input to RFEM, the system uses the reference phase-locked loop (R-PLL) to detect the phase offset of the sampled signal, using the excitation frequency as the phase reference:

[0107] Δφ(t)=φ in (t)-φ ref (t)

[0108] This phase shift may be due to signal chain delays, analog circuit drift, or harmonic interference.

[0109] The input phase accuracy corrector IPC dynamically adjusts the input signal according to Δφ(t) to align it with the current excitation frequency reference phase:

[0110] φ in ′(t)=φ in (t)-k φ Δφ(t)

[0111] where k φ It is the phase correction coefficient, which is generally less than 1 to ensure smooth phase tracking without oscillation.

[0112] Step 3: Bidirectional feedback control logic

[0113] The frequency and phase error detection and correction mechanisms work in tandem. Specifically, when the frequency offset Δf is corrected, the system automatically refreshes the R-PLL reference phase, ensuring that the phase calibration logic remains consistent with the latest excitation frequency, preventing phase misalignment caused by frequency drift. The system forms the following bidirectional linkage path:

[0114] ESG Excitation Transducer Response RFEM Deviation Feedback SCM Correction ESG / RFEM

[0115] Specifically, the excitation signal generation module includes a high-precision on-chip clock oscillator, a digitally controlled frequency synthesizer (DDS), a frequency phase-locked feedback control unit, a signal shaping output stage, and a power amplifier module. The on-chip clock oscillator uses a temperature-compensated crystal oscillator (TCXO) with a frequency stability better than ±0.1ppm, providing a stable reference signal for the DDS.

[0116] The DDS module is based on the AD9959 four-channel digital frequency synthesis chip, which can output any frequency signal in the range of 10kHz to 30MHz, with a minimum frequency step of 0.01Hz. The output signal can be selected as sine wave, square wave or pulse sequence.

[0117] The frequency phase-locked feedback control unit adopts a phase-locked loop (PLL) circuit, which realizes closed-loop correction control by sampling the frequency offset and feeding it back to the DDS control port. The frequency error is less than ±0.05%.

[0118] The signal shaping output stage includes an adjustable boost filter module and a multi-channel signal multiplexing channel, which is used to adjust the output voltage amplitude (0.1Vpp~10Vpp) and select different coupling modes (series / parallel) according to the test object.

[0119] The power amplifier module uses a linear broadband amplifier (such as OPA2673) with an output impedance matched to 50Ω to drive the impedance matching adjustment module.

[0120] Specifically, the impedance matching adjustment module consists of a voltage-controlled variable impedance network, a digital phase shifter array, a power bidirectional coupler and an impedance adjustment control unit, with the purpose of achieving dynamic adaptive matching of different transducer impedance states.

[0121] The voltage-controlled variable impedance network consists of a capacitor (C)-inductor (L) array that can be switched through a multi-bit digital potentiometer or PIN diode network to form a π-type or T-type matching structure, covering an impedance range of 0.5 to 1000Ω.

[0122] The digital phase shifter array is based on a four-quadrant vector modulator to achieve 360° full-angle phase adjustment with an accuracy of ±1°, which is used to calibrate the phase relationship between the excitation signal and the transducer input signal.

[0123] The power bidirectional coupler monitors the input power and reflected power in real time, and generates forward / reverse power vector data for reference by the control unit.

[0124] The impedance adjustment control unit integrates a fuzzy control algorithm to adjust the L, C combination and phase shift angle in real time according to the voltage standing wave ratio (VSWR) or S11 parameters to minimize reflections.

[0125] Specifically, the signal acquisition and processing module consists of a differential sampling and amplification unit, a high-speed ADC array, a synchronous sampling controller, a digital signal processor (DSP), and a host computer communication interface. It is primarily responsible for sampling, processing, and analyzing the stimulus response signals. The differential sampling and amplification unit uses an ultra-low-noise instrumentation amplifier (such as the INA828) to amplify the voltage and current feedback signals and convert them into differential signals, improving their resistance to common-mode interference.

[0126] The high-speed ADC array uses a synchronous sampling analog-to-digital converter (such as AD7606) with a sampling rate of 1MSa / s to 10MSa / s, a resolution of 16 to 18 bits, and 4 to 8 channels to collect signals such as voltage, current, and harmonic response at both ends of the transducer.

[0127] The synchronous sampling controller is based on an FPGA logic control module (such as Xilinx Artix-7). It ensures that the synchronous sampling error of all channels is less than ±5ns through a clock distribution network and supports external trigger synchronization (Trigger-In) mode.

[0128] The digital signal processor integrates FFT spectrum analysis, phase difference calculation, impedance inverse calculation and resonance feature extraction functions, and can output indicators such as resonance frequency, quality factor, series and parallel electrical parameters in real time.

[0129] The communication interface supports USB3.0, high-speed UART and LAN interfaces, and the collected data is uploaded to the host computer GUI interface for visual display and storage analysis.

[0130] The resonance characteristic extraction module consists of a phase solution channel, a synchronous multi-phase acquisition array, an impedance vector conversion bridge and a harmonic residual control unit;

[0131] The phase resolution channel is cross-synchronously arranged with the voltage sampling path and the current sampling path. The impedance vector conversion bridge consists of two sets of precision adjustable capacitors and equal-amplitude impedance divider circuits. The phase resolution channel output and the harmonic residual control unit form a bidirectional closed loop. The synchronous multi-phase acquisition array uses 8-channel parallel synchronization to achieve a 100MSa / s sampling rate, so that the phase error is controlled within ±0.1°. The impedance vector conversion bridge maintains a vector angle stability error of less than 0.3% under signals below -10dB.

[0132] Specifically, the phase resolution channel of the resonance characteristic extraction module achieves precise signal phase resolution through cross-synchronization with the voltage and current sampling paths. This channel digitizes the sampled signal using a high-speed analog-to-digital converter (ADC) and calculates the phase difference between the voltage and current signals using a digital signal processing algorithm. The synchronous multiphase acquisition array utilizes eight parallel channels to achieve a high sampling rate of 100MSa / s, ensuring phase error within ±0.1°.

[0133] The impedance vector conversion bridge consists of two sets of precision adjustable capacitors and a constant-amplitude impedance divider circuit. By adjusting the capacitance, the bridge adapts to varying impedances, ensuring maximum signal power transmission. At signals below -10dB, the bridge maintains vector amplitude stability error below 0.3%, meeting the requirements of high-precision testing.

[0134] The phase solver output forms a bidirectional closed loop with the harmonic residual control unit. This closed loop structure uses a real-time feedback mechanism to adjust system parameters and optimize the accuracy and stability of resonant characteristic extraction.

[0135] The harmonic residual control unit embeds a digital phase-locked loop array and a differential dynamic compensation chip to form a dual-channel frequency tracker. The phase-locked loop array adopts a three-level feedback sampling window structure to construct a dynamic phase-amplitude mapping matrix for the ±1kHz region around the excitation frequency. The real-imaginary part fitting function is used to construct a residual decreasing trend line at the static equilibrium point, thereby gradually approaching the phase inflection point of the resonant point in multiple scans, so that the resonant frequency positioning accuracy of the equivalent inductor-capacitor network is higher than 10Hz.

[0136] The digital phase-locked loop (PLL) array embedded in the harmonic residual control unit utilizes a three-level feedback sampling window structure. This structure uses a multi-level feedback mechanism to finely sample the ±1kHz region around the excitation frequency, constructing a dynamic phase-amplitude mapping matrix. This matrix records the phase and amplitude relationships at different frequencies, providing data support for subsequent fitting analysis.

[0137] Using a real-imaginary fitting function, a decreasing residual trendline is constructed at the static equilibrium point. This trendline analyzes the real and imaginary changes in phase and amplitude to predict the system's residual trend and guide the direction and amplitude of frequency adjustment. Over multiple scans, the system gradually approaches the phase inflection point of the resonant point, achieving high-precision positioning of the resonant frequency of the equivalent inductor-capacitor network, with a positioning accuracy exceeding 10Hz.

[0138] The differential dynamic compensation chip constructs a gain closed-loop path based on a CMOS ultra-low offset structure and a delay-adjustable comparator, embeds an adjustable sampling window parameter unit, sets a sampling step size of 5ns, and constructs a phase drift spectrum of adjacent sampling points to achieve accurate extraction of signal phase fluctuation trends with perturbations less than 0.1%. Finally, a nonlinear dual-parameter fitting plane is constructed at the data fitting end to invert the true reactance change trend at the resonant frequency.

[0139] Specifically, the gain closed-loop path contains three submodules:

[0140] Dynamic sampling controller: Based on the PLL phase-locked mechanism, the echo time window is repositioned after each pulse transmission to adjust the sampling starting point;

[0141] Phase offset calculation unit: using the reference sampling point S ref The phase offset angle is calculated from the complex amplitude of the current sampling point Si using the following formula:

[0142]

[0143] Gain adjustment unit: According to Calculate the current "response instability index" U of the system i =α|Δφ i |+β.

[0144] |Ai -A ref |, and adjust the gain G according to the following feedback function i :

[0145] G i =G i-1 ·(1-λU i )

[0146] Where: λ is the feedback sensitivity coefficient, and the empirical value is set to 0.01~0.05.

[0147] All calculations are completed in parallel in the firmware through a pipeline structure. Data calculation and gain update can be completed within one sampling cycle, ensuring that the system response delay is controlled within 50ns.

[0148] 2. Construction process of phase drift spectrum of adjacent sampling points:

[0149] After sampling is completed, the main control chip extracts the complex signal of each point in the sampling window And perform the following processing steps:

[0150] Step 1: Phase difference calculation

[0151] Traverse all adjacent sampling point pairs and calculate:

[0152] δφ i =φ i+1 -φ i

[0153] And for all δφ i Perform phase unwrapping processing to avoid misjudgment due to -π / π periodic jumps.

[0154] Step 2: Construct the drift matrix

[0155] Generate phase shift matrix M = [δφ1,δφ2,…,δφ n ]], and perform sliding window filtering on it with the window size set to 3 to 5 points to improve the smoothness of the phase change curve.

[0156] Step 3: 2D graph visualization

[0157] The horizontal axis is the sampling point position i and the vertical axis is δφ n i. Build a two-dimensional heat map or line graph of phase drift. To enhance trend identification, introduce color mapping:

[0158] Blue indicates phase stability (change < 0.01 rad);

[0159] Yellow indicates slight drift;

[0160] Red indicates the phase mutation region.

[0161] Step 4: Identify the slight change in the resonance point

[0162] Based on abnormal peaks or sudden changes in the phase drift curve, the system uses the Savitzky-Golay smooth differential method to extract the phase change slope and mark the local maximum and minimum points. These points represent the drift behavior of the transducer's resonant frequency under different temperatures or mechanical loads.

[0163] Step 5: Fitting and Prediction

[0164] The system inputs the spectrum data into the least squares fitting model to construct the phase drift surface function:

[0165] φ(i,T)=a·i+b·T+c·iT+d

[0166] Where i is the sampling point number and T is the temperature factor. This model can be used to predict the phase response behavior of the transducer at different temperatures and sampling times.

[0167] 3. Actual application effect:

[0168] In testing a typical transducer (center frequency of 5MHz), the system's phase drift map can identify a minimum phase change of 0.003 rad between adjacent sampling points. Its sensitivity to detecting minute temperature drifts and slight resonant shifts is approximately 6.3 times higher than traditional FFT analysis methods. Furthermore, thanks to the firmware's closed-loop gain, which adjusts the sampling amplitude in real time, the system's output waveform's overall signal-to-noise ratio is improved by approximately 20%, demonstrating excellent adaptive testing capabilities and anti-interference performance.

[0169] The temperature-controlled loading module consists of a Peltier heat pump array, a quasi-steady-state airflow convection unit, and a nano-cavity thermal insulation stack assembly;

[0170] The heat pump array consists of four sets of reverse-direction thermocouple elements, which achieve ±0.05°C temperature control within the cavity through a time-sharing switching mode. The airflow convection unit forms a closed-loop heat exchange circuit with the transducer mounting cavity through the stepped distributed turbine ducts at both ends. The nanocavity thermal insulation stack assembly adopts a graphene film-polyimide multilayer composite structure, achieving a thermal conductivity of less than 0.02W / m·K within a thickness of 3mm, shielding the heat flow outside the cavity.

[0171] The transducer and the thermal convection cavity are installed in an oblique thermal coupling manner with a fixed angle of 85°±2°. The contact surface is coated with a 0.2mm thick phase-change thermal conductive adhesive layer, and high-conductivity micro-fillers are used to increase the thermal conductivity to 8W / m·K. The internal convection flow channel is tuned through dynamic heat flow simulation to keep the transducer surface temperature difference within ±0.1°C during the steady-state test phase.

[0172] Specifically, the temperature-controlled loading module utilizes a compact Peltier heat pump array as a controllable heat source. Four alternating groups of Bi2Te3-based thermoelectric elements operating in reverse mode are used. Time-division drive current switching enables the heat pump array to achieve high-precision temperature control within ±0.05°C. The module's housing is insulated with a graphene film-polyimide composite insulation barrier, with a total thickness of less than 3mm, effectively isolating the module from external thermal radiation and convection interference, ensuring thermal field stability within the cavity.

[0173] The quasi-steady-state airflow convection unit uses micro-turbo fans to form a closed return flow path in the stepped air ducts at both ends. The wind speed and direction are adjustable, with a maximum wind speed of 1.2m / s. The transducer installation cavity maintains uniform heat convection distribution in this airflow loop. During the transducer assembly process, an 85°±2° inclined plug-in thermal coupling structure is adopted. The plug-in angle is precisely positioned by the guide rail, and a 0.2mm thick phase change thermal conductive adhesive is applied to the interface surface (the thermal conductivity coefficient is initially 4W / m·K, enhanced to 8W / m·K with high-conductivity fillers) to ensure that thermal resistance is minimized and the heat exchange response rate is improved.

[0174] The simulation uses CFD software (such as Ansys Fluent) to model and optimize the internal thermal flow field, adjusting the cross-sectional structure of the convection channel to achieve a temperature difference of ≤±0.1°C on the transducer surface during the steady-state phase. This structure ensures that the temperature-controlled loading module can achieve stable control under different ambient temperatures, improving the repeatability and environmental adaptability of the transducer characteristic test.

[0175] The spectrum calibration module consists of an on-chip frequency vector modulator, an off-chip variable storage array, and an asymmetric Gaussian filter group;

[0176] The frequency vector modulator uses on-chip self-learning FPGA control logic to adjust the output frequency step interval and waveform factor in real time based on external feedback error. The frequency accuracy adjustment resolution reaches 0.1Hz. The asymmetric Gaussian filter group uses a bidirectional window function to retain the harmonic envelope peak position while suppressing high-frequency edge fluctuations of the spectrum, so that the signal energy main lobe position control error is less than ±0.2%.

[0177] A multidimensional frequency-amplitude mapping table is set in the variable storage array. The multidimensional frequency-amplitude mapping table updates the frequency error correction parameters four times per second through the on-chip EEPROM periodic write mechanism. The correction envelope injected into each set of frequency points includes at least three quantitative parameters: amplitude correction factor, phase correction factor, and harmonic content coefficient. The correction parameters are adaptively updated according to the Bayesian estimation model to achieve dynamic optimization based on the historical correction trajectory, thereby suppressing the step error caused by temperature drift and signal drift during the spectrum calibration process.

[0178] Specifically, the temperature-controlled loading module utilizes a compact Peltier heat pump array as a controllable heat source. Four alternating groups of Bi2Te3-based thermoelectric elements operating in reverse mode are used. Time-division drive current switching enables the heat pump array to achieve high-precision temperature control within ±0.05°C. The module's housing is insulated with a graphene film-polyimide composite insulation barrier, with a total thickness of less than 3mm, effectively isolating the module from external thermal radiation and convection interference, ensuring thermal field stability within the cavity.

[0179] The quasi-steady-state airflow convection unit uses micro-turbo fans to form a closed return flow path in the stepped air ducts at both ends. The wind speed and direction are adjustable, with a maximum wind speed of 1.2m / s. The transducer installation cavity maintains uniform heat convection distribution in this airflow loop. During the transducer assembly process, an 85°±2° inclined plug-in thermal coupling structure is adopted. The plug-in angle is precisely positioned by the guide rail, and a 0.2mm thick phase change thermal conductive adhesive is applied to the interface surface (the thermal conductivity coefficient is initially 4W / m·K, enhanced to 8W / m·K with high-conductivity fillers) to ensure that thermal resistance is minimized and the heat exchange response rate is improved.

[0180] The simulation uses CFD software (such as Ansys Fluent) to model and optimize the internal thermal flow field, adjusting the cross-sectional structure of the convection channel to achieve a temperature difference of ≤±0.1°C on the transducer surface during the steady-state phase. This structure ensures that the temperature-controlled loading module can achieve stable control under different ambient temperatures, improving the repeatability and environmental adaptability of the transducer characteristic test.

[0181] The physical vibration isolation module consists of a three-layer structure, including a carbon fiber support grid, a high-damping elastic support, and a lower rigid aluminum alloy fixed base;

[0182] The high-damping elastic support is made of a butyl rubber-polyurethane two-component composite material with a hardness controlled at 60ShoreA. It has an embedded microporous vibration isolation net. The natural frequency of the entire physical vibration isolation module is controlled below 5Hz, isolating the measurement perturbations caused by external mechanical vibrations. A ceramic insulating bracket is set between the transducer and the test circuit and filled with polymer damping gel to avoid any low-frequency structural resonance affecting parameter fluctuations.

[0183] Specifically, the physical vibration isolation module adopts a three-layer structure: the upper layer is a carbon fiber woven support grid to ensure the installation rigidity and position stability of the transducer; the middle layer is a high-damping elastic support, which is composed of a composite of butyl rubber and polyurethane, with a hardness controlled at 60ShoreA, and a microporous mesh vibration isolation layer embedded inside, which has good low-frequency vibration absorption performance; the lower layer is a high-strength aluminum alloy base, which is used for overall fixation and docking with the external platform.

[0184] The module's natural frequency is controlled below 5Hz, ensuring it's out of the frequency range of common mechanical vibration sources and effectively isolating it from external mechanical disturbances. To prevent minor mechanical coupling from interfering with the transducer output, a ceramic insulating bracket filled with polymer damping gel (damping factor > 0.9) is placed between the transducer and the signal transmission circuit to further suppress low-frequency structural resonance and improve system testing stability.

[0185] In actual tests, under typical laboratory background vibration conditions (<0.02gRMS), the system measured the signal phase jitter amplitude to drop from 0.8° to less than 0.1°, and the signal amplitude fluctuation suppression rate increased to more than 92%, verifying the significant improvement effect of the physical vibration isolation module on the high-sensitivity test accuracy of ultrasonic transducers.

[0186] The above shows and describes the basic principles, main features and advantages of the present invention; those skilled in the art should understand that the present invention is not limited to the above embodiments, and the above embodiments and descriptions are only preferred examples of the present invention and are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, and these changes and improvements all fall within the scope of the present invention to be protected; the scope of protection claimed in the present invention is defined by the attached claims and their equivalents.

Claims

1. An ultrasonic transducer characteristic parameter testing system, characterized in that: The ultrasonic transducer characteristic parameter testing system includes an excitation signal generation module, an impedance matching adjustment module, a resonance characteristic extraction module, a signal acquisition and processing module, a temperature control loading module, a physical vibration isolation module and a spectrum calibration module; The excitation signal generation module forms a multi-frequency sweep pulse array by configuring an on-chip oscillator and an off-chip frequency-locked feedback loop, which is directly coupled to the impedance matching adjustment module. The impedance matching adjustment module includes a voltage-controlled matching network and a phase shifter array configured in parallel, which is used to implement amplitude adaptive matching within the range of 0.5Ω to 1000Ω; the matching output end is connected to the resonance characteristic extraction module, and the resonance characteristic extraction module forms a feedback linkage path with the spectrum calibration module through a bidirectional phase superposition path, and simultaneously integrates the amplitude voltage and current feedback signals into the signal acquisition and processing module. The resonance characteristic extraction module and the spectrum calibration module construct a coupling offset compensation channel through harmonic node synchronous calibration logic. The temperature control loading module and the physical vibration isolation module are arranged in a superimposed manner, and the transducer is embedded in the thermal convection temperature chamber of the temperature control loading module. The spectrum calibration module also performs bidirectional real-time error correction on the output frequency of the excitation signal generation module and the input phase of the resonance characteristic extraction module.

2. The ultrasonic transducer characteristic parameter testing system according to claim 1, characterized in that: The resonance characteristic extraction module is composed of a phase solution channel, a synchronous multi-phase acquisition array, an impedance vector conversion bridge and a harmonic residual control unit; The phase resolution channel is arranged crosswise and synchronously with the voltage sampling path and the current sampling path. The impedance vector conversion bridge is composed of two sets of precision adjustable capacitors and a constant-amplitude impedance divider circuit. The phase resolution channel output and the harmonic residual control unit form a bidirectional closed loop. The synchronous multi-phase acquisition array uses an 8-channel parallel synchronization method to achieve a 100MSa / s sampling rate, so that the phase error is controlled within ±0.1°. The impedance vector conversion bridge maintains a vector angle stability error of less than 0.3% when the signal is below -10dB.

3. The ultrasonic transducer characteristic parameter testing system according to claim 2, characterized in that: The harmonic residual control unit embeds a digital phase-locked loop array and a differential dynamic compensation chip to form a dual-channel frequency tracker. The phase-locked loop array adopts a three-level feedback sampling window structure to construct a dynamic phase-amplitude mapping matrix for the ±1kHz region around the excitation frequency, and uses a real-imaginary part fitting function to construct a residual decreasing trend line at the static equilibrium point, thereby gradually approaching the phase inflection point of the resonance point in multiple scans, so that the resonant frequency positioning accuracy of the equivalent inductor-capacitor network is higher than 10Hz.

4. The ultrasonic transducer characteristic parameter testing system according to claim 3, characterized in that: The differential dynamic compensation chip constructs a gain closed-loop path based on a CMOS ultra-low offset structure and a delay-adjustable comparator, embeds an adjustable sampling window parameter unit, sets a sampling step size of 5ns, and constructs a phase drift spectrum of adjacent sampling points to achieve accurate extraction of signal phase fluctuation trends with perturbations less than 0.1%. Finally, a nonlinear dual-parameter fitting plane is constructed at the data fitting end to invert the true reactance change trend at the resonant frequency.

5. The ultrasonic transducer characteristic parameter testing system according to claim 1, characterized in that: The temperature control loading module is composed of a Peltier heat pump array, a quasi-steady airflow convection unit and a nano-cavity thermal insulation stack assembly; The heat pump array consists of four sets of reverse thermocouple elements, and achieves ±0.05°C precision control of the internal cavity temperature through a time-division switching mode. The airflow convection unit forms a closed-loop heat exchange circuit through the stepped distributed turbine ducts at both ends and the transducer mounting cavity. The nanocavity thermal insulation stack assembly adopts a graphene film-polyimide multilayer composite structure, achieving a thermal conductivity of less than 0.02W / m·K within a thickness of 3mm, shielding the heat flow outside the cavity.

6. The ultrasonic transducer characteristic parameter testing system according to claim 5, characterized in that: The transducer and the thermal convection cavity are installed in an oblique insertion thermal coupling manner with a fixed angle of 85°±2°. The contact surface is coated with a phase change thermal conductive adhesive layer with a thickness of 0.2mm, and high-conductivity micro-fillers are used to increase the thermal conductivity coefficient to 8W / m·K. The internal convection flow channel is tuned through dynamic heat flow simulation to keep the surface temperature difference of the transducer within ±0.1°C during the steady-state test phase.

7. The ultrasonic transducer characteristic parameter testing system according to claim 1, characterized in that: The spectrum calibration module is composed of an on-chip frequency vector modulator, an off-chip variable storage array, and an asymmetric Gaussian filter group; The frequency vector modulator adopts on-chip self-learning FPGA control logic to adjust the output frequency step interval and waveform factor in real time according to the external feedback error. The frequency accuracy adjustment resolution reaches 0.1Hz. The asymmetric Gaussian filter group uses a bidirectional window function to retain the harmonic envelope peak position while suppressing the high-frequency edge fluctuations of the spectrum, so that the signal energy main lobe position control error is less than ±0.2%.

8. The ultrasonic transducer characteristic parameter testing system according to claim 7, characterized in that: A multidimensional frequency-amplitude mapping table is set in the variable storage array. The multidimensional frequency-amplitude mapping table updates the frequency error correction parameters four times per second through an on-chip EEPROM periodic write mechanism. The correction envelope injected into each set of frequency points includes at least three quantitative parameters: an amplitude correction factor, a phase correction factor, and a harmonic content coefficient. The correction parameters are adaptively updated according to a Bayesian estimation model to achieve dynamic optimization based on historical correction trajectories, thereby suppressing step errors caused by temperature drift and signal drift during spectrum calibration.

9. The ultrasonic transducer characteristic parameter testing system according to claim 1, characterized in that: The physical vibration isolation module consists of a three-layer structure, including a carbon fiber support grid, a high-damping elastic support, and a lower rigid aluminum alloy fixed base; The high-damping elastic support adopts a butyl rubber-polyurethane two-component composite material with a hardness controlled at 60ShoreA and an embedded microporous vibration isolation net. The natural frequency of the entire physical vibration isolation module is controlled below 5Hz, isolating the measurement perturbations caused by external mechanical vibrations. A ceramic insulating bracket is set between the transducer and the test circuit and filled with polymer damping gel to avoid any low-frequency structural resonance affecting parameter fluctuations.

Images