Magnetoelectric double-field non-contact vector phase discrimination and impedance estimation device and method

Abstract

The invention relates to the technical field of signal detection and processing, and provides a magnetoelectric double-field non-contact vector phase discrimination and impedance estimation device and method. The device comprises a magnetoelectric double-field non-contact sensing module, a reference phase-locked link module, an IQ demodulation and calibration module, a sampling and digital processing module, a calibration and self-calibration module and an impedance and phase calculation module. According to the device, the problems of high-voltage safety risk and interference resistance to an original circuit caused by traditional contact type sampling are avoided; through IQ channel mismatch correction, modulation degree desensitization processing and cross coupling elimination and parameter drift correction by using a corrected coupling matrix, the technical bottlenecks of probe drift, inter-channel parasitic coupling, demodulation imbalance, local oscillator drift, program modulation interference and the like in the non-contact technology are effectively solved; high-precision, high-stability and high-safety non-contact measurement of the high-voltage radio-frequency signal is realized.

Description

Technical Field

[0001] This application relates to the field of signal detection and processing technology, and in particular to a magnetoelectric dual-field non-contact vector phase detection and impedance estimation device and method. Background Technology

[0002] High-power shortwave transmitters (operating in the 3-30MHz frequency band) serve as key infrastructure in fields such as long-distance communication, emergency broadcasting, and military command communication. Their core function is to modulate baseband signals into high-frequency radio frequency signals and efficiently radiate them to the target area with the help of antennas. They play an irreplaceable role in scenarios such as disaster relief, transnational broadcasting, and communication in remote areas.

[0003] The performance and safety of this type of transmitter depend primarily on the impedance matching of the final stage circuit (such as the anode circuit of the final stage vacuum tube) and the output matching network (such as a π-type network or bandpass cavity). The anode of the final stage vacuum tube typically operates under conditions of 8-20kV DC high voltage and 5-30A RF high current. The output matching network must precisely match the output impedance of the vacuum tube (typically 50Ω) with the antenna impedance. If impedance mistuning occurs, it will cause RF energy reflection. Excessive reflected power will lead to anode overheating, vacuum tube filament burnout, and even breakdown of output components. At the same time, the transmission efficiency will drop sharply from over 85% to below 60%, seriously wasting energy and affecting communication quality.

[0004] To ensure the efficient and safe operation of the transmitter, it is necessary to acquire the voltage-current phase difference and complex impedance (including amplitude, resistance component, and reactance component) between the final stage circuit and the output matching network in real time. This information is then used as the core feedback signal to achieve automatic tuning (such as dynamically adjusting the capacitance / inductance parameters of the matching network). The phase detector, as the core component for phase difference measurement, directly determines the tuning effect based on its measurement accuracy. High-power shortwave transmitters have extremely stringent requirements for phase detection accuracy (phase error must be ≤ ±1°). If the phase detection error is too large, it will lead to tuning deviation, compromise impedance matching, and consequently cause equipment failure or communication interruption.

[0005] The core of a phase detector lies in "acquiring voltage and current signals and accurately calculating the phase difference", and the sampling method is a key factor that determines signal quality, system security and ease of maintenance.

[0006] Traditional phase detectors generally employ a "contact sampling" design, requiring sampling components (such as voltage divider capacitors and sampling resistors) to be directly cut-and-connected or paralleled to the transmitter's high-voltage RF nodes (such as the final-stage anode bus and output feeder). This presents the following problems: the sampling components must withstand tens of kilovolts of high voltage. If the voltage divider capacitor ages and breaks down, or the sampling resistor overheats and burns out, it can lead to a high-voltage ground short circuit, causing damage to the transmitter's final-stage vacuum tubes. Furthermore, installation and maintenance require contact with high-voltage components, and even after power is cut off, residual charge remains at the high-voltage nodes, posing a risk of injury from residual discharge. In addition, the voltage divider capacitor introduces additional parasitic capacitance, and the sampling resistor changes the loop impedance, both of which disrupt the original impedance matching between the transmitter's final-stage loop and the output matching network, resulting in RF signal distortion and increased harmonic components.

[0007] Some related technologies attempt to use non-contact sampling (such as single electric field coupling or magnetic field coupling), but due to technological limitations, key bottlenecks remain, failing to meet the accuracy requirements of high-power transmitters. First, non-contact voltage sampling relies on electric field coupling capacitance, which is easily affected by subtle changes in ambient temperature, humidity, and installation location, leading to fluctuations in signal amplitude and phase characteristics, with phase errors reaching ±5°. Furthermore, parasitic coupling easily occurs between the non-contact voltage probe (E probe) and the current probe (H probe), meaning the H probe couples the electric field component of the voltage signal, and the E probe couples the magnetic field component of the current signal, causing sampled signal distortion and consequently phase detection errors. In addition, during demodulation, the phase reference (local oscillator signal) is easily drifted by strong electromagnetic radiation interference from the transmitter, and there is a lack of effective re-acquisition mechanisms. Simultaneously, for modulation modes such as AM (amplitude modulation) and DRM (digital radio modulation), program content (such as audio signals and OFDM subcarrier data) can disturb the carrier frequency phase estimation, resulting in poor phase measurement stability. Summary of the Invention

[0008] In order to at least partially solve the above-mentioned technical problems in the related art, this application provides a magnetoelectric dual-field non-contact vector phase detection and impedance estimation device and method.

[0009] On the one hand, this application provides a magnetoelectric dual-field non-contact vector phase detection and impedance estimation device, which adopts the following technical solution: A magnetoelectric dual-field non-contact vector phase detector and impedance estimation device includes: The magnetoelectric dual-field non-contact sensing module includes an electric field-coupled voltage probe and a magnetic field-inducing current probe. The electric field-coupled voltage probe is used for non-contact induction sampling of radio frequency voltage signals, and the magnetic field-inducing current probe is used for non-contact induction sampling of radio frequency current signals. The reference phase-locked link module is used to extract the synchronous local oscillator signal from the signal under test; The IQ demodulation and calibration module is connected to the magnetoelectric dual-field non-contact sensing module and the reference phase-locked link module. It is used to perform phase-coordinated zero-IF demodulation of the radio frequency voltage signal and the radio frequency current signal using the synchronous local oscillator signal, and to perform I / Q channel mismatch correction to generate a baseband analog signal. The sampling and digital processing module, connected to the IQ demodulation and calibration module, is used to convert the baseband analog signal into a digital baseband signal and perform modulation desensitization processing on the digital baseband signal to generate a digital baseband component for phase detection; the calibration and self-calibration module is used to construct an initial coupling matrix and update the initial coupling matrix using a recursive least squares algorithm to generate a corrected coupling matrix for eliminating parameter drift. The impedance and phase calculation module is connected to the sampling and digital processing module and the calibration and self-calibration module. It is used to eliminate cross-coupling of the digital baseband components using the corrected coupling matrix, reconstruct the true voltage complex and the true current complex, and calculate impedance parameters and quality indicators based on the true voltage complex and the true current complex.

[0010] By employing the aforementioned technical solution, a complete non-contact vector phase detection and impedance estimation device is constructed through the coordinated operation of modules such as magnetoelectric dual-field non-contact sensing, reference phase-locked loop (PLL), IQ demodulation and calibration, digital processing and modulation desensitization, calibration and self-calibration, and impedance calculation. This solution fundamentally avoids the high-voltage safety risks and impedance interference problems of traditional contact sampling. Simultaneously, through IQ channel mismatch correction, modulation desensitization processing, and the use of a corrected coupling matrix for cross-coupling elimination and parameter drift correction, it effectively solves the technical bottlenecks existing in related non-contact technologies, such as probe drift, parasitic coupling between channels, demodulation imbalance, local oscillator drift, and program modulation interference. This achieves high-precision, high-stability, and high-safety non-contact measurement of high-voltage radio frequency signals.

[0011] Optionally, the magnetoelectric dual-field non-contact vector phase detector and impedance estimation device further includes: A front-end protection and band-limiting module is disposed between the magnetoelectric dual-field non-contact sensing module and the reference phase-locked link module, and is used to perform graded protection and selective filtering on the radio frequency voltage signal and the radio frequency current signal; The front-end protection and band-limiting module includes an electric field channel protection and band-limiting unit and a magnetic field channel protection and band-limiting unit. The electric field channel protection and band-limiting unit is disposed between the electric field coupled voltage probe and the reference phase-locked link module, and includes a high-resistance buffer element, a TVS clamping element and a π-type band-limiting element arranged in sequence. The magnetic field channel protection and band-limiting unit includes a switchable damping element and an RC band-limiting network.

[0012] By adopting the above technical solution, a front-end protection and band-limiting module is added between the sensing module and the subsequent links. This module achieves graded protection and selective filtering of radio frequency signals through high-impedance buffering, TVS clamping, and π-type band limiting in the electric field channel, as well as switchable damping and RC band-limiting networks in the magnetic field channel. This not only effectively protects the downstream circuitry from abnormal operating conditions such as high-voltage arcing and surges, but also deeply suppresses power frequency interference, high-frequency noise, and parasitic ringing. While ensuring the amplitude and phase integrity of signals in the 3-30MHz target frequency band, it significantly improves the device's anti-interference capability and operational reliability in strong electromagnetic environments.

[0013] Optionally, the reference phase-locked link module includes a near-field small loop pickup unit, a Schottky limiting unit, a programmable prescaler unit, and a digital PLL phase-locked unit connected in sequence; wherein, the digital PLL phase-locked unit adopts a dual-ring architecture of a capture ring and a tracking ring.

[0014] The above technical solution specifies the detailed configuration of the reference phase-locked link module. Non-contact parameter acquisition is achieved through near-field small-loop pickup, signal amplitude is stabilized using Schottky limiting, and the frequency adaptation range of the device is broadened through programmable prescaler. Its core employs a dual-loop digital PLL architecture of "capture loop + tracking loop," combining the fast locking characteristics of the capture loop with the narrow-band low-noise characteristics of the tracking loop. This ensures high stability and low phase jitter of the local oscillator signal, and provides rapid re-acquisition capability upon loss of lock, offering a precise and reliable phase reference for subsequent phase-coherent demodulation.

[0015] Optionally, the IQ demodulation and calibration module is used to perform coherent zero-IF demodulation on the RF voltage signal and the RF current signal using the synchronous local oscillator signal, and to perform I / Q channel mismatch correction to generate a baseband analog signal; wherein, the IQ demodulation and calibration module is configured to: mix the RF voltage signal with the in-phase component and the quadrature component of the synchronous local oscillator signal respectively to generate a baseband in-phase component and a baseband quadrature component of the voltage; and mix the RF current signal with the in-phase component and the quadrature component of the synchronous local oscillator signal respectively to generate a current... The baseband in-phase component and the baseband quadrature component of the current are used to generate the baseband analog signal. The baseband in-phase component, the baseband quadrature component, the baseband in-phase component, and the baseband quadrature component of the current are subjected to low-pass filtering. The filtered and denoised baseband in-phase component, the baseband quadrature component, the baseband in-phase component, and the baseband quadrature component of the current are then subjected to I / Q channel mismatch correction to generate the baseband analog signal, which includes the voltage in-phase component signal, the voltage quadrature component signal, the current in-phase component signal, and the current quadrature component signal.

[0016] The above technical solution specifically defines the detailed process of coherent zero-IF demodulation and I / Q mismatch correction in the IQ demodulation and calibration module. By mixing the RF voltage and current signals with the in-phase and quadrature components of the synchronous local oscillator, respectively, and then performing low-pass filtering, signal-to-baseband conversion is achieved. Crucially, this solution explicitly corrects the I / Q channel mismatch for the filtered baseband components. This effectively compensates for the amplitude and phase imbalance between the I and Q channels caused by hardware differences, significantly improves image rejection, reduces measurement errors introduced by demodulation distortion, and thus outputs a high-fidelity baseband analog signal.

[0017] Optionally, the sampling and digital processing module is used to convert the baseband analog signal into a digital baseband signal, and to perform modulation desensitization processing on the digital baseband signal to generate a digital baseband component for phase detection; wherein, the sampling and digital processing module is configured to: use a co-source clock control to perform synchronous analog-to-digital conversion on the voltage in-phase component signal, the voltage quadrature component signal, the current in-phase component signal, and the current quadrature component signal to generate a digital baseband signal; set the baseband equivalent noise bandwidth and the digital integration window width according to the transmitter's operating mode; and process the digital baseband signal through a long integration or digital averaging algorithm to reduce the disturbance of the program modulation component on the carrier frequency phase estimation to generate a digital baseband component for phase detection.

[0018] The above technical solution specifically defines the functionality of the sampling and digital processing modules. By employing synchronous analog-to-digital conversion with a shared clock, strict consistency in the sampling timing of the four I / Q signals (voltage and current) is ensured, avoiding phase errors introduced by asynchronous clocks between channels. Furthermore, by setting the bandwidth and integration window according to the operating mode and using long integral or digital averaging algorithms for processing, modulation desensitization is achieved, effectively reducing the disturbance of program content components on carrier frequency phase estimation under AM, DRM, and other modulation modes, and significantly improving the stability of phase measurement under modulated signals.

[0019] Optionally, the calibration and self-calibration module is used to construct an initial coupling matrix and update the initial coupling matrix using a recursive least squares algorithm to generate a corrected coupling matrix for eliminating parameter drift; wherein, the calibration and self-calibration module is configured to: construct the initial coupling matrix during transmitter offline; introduce imperceptible perturbation excitation and collect perturbation data during transmitter operation; and use a recursive least squares algorithm to iteratively update the initial coupling matrix using the perturbation data to generate the corrected coupling matrix.

[0020] The above technical solution specifically defines the operating mode of the calibration and self-calibration modules. This solution employs an "initial calibration + dynamic correction" strategy: an initial coupling matrix is ​​constructed during transmitter offline periods to correct inherent hardware cross-coupling; during transmitter operation, imperceptible perturbation excitation is introduced, and the matrix is ​​iteratively updated using recursive least squares (RLS) algorithm based on the perturbation data. This design achieves "non-cast" self-calibration, dynamically compensating for parameter drift caused by changes in environmental temperature and humidity or device aging, ensuring the long-term stability and accuracy of the device measurements, while reducing the maintenance costs of manual calibration.

[0021] On the other hand, this application also provides a non-contact vector phase detection and impedance estimation method for magnetoelectric dual-field, which adopts the following technical solution: A non-contact vector phase detection and impedance estimation method for magnetoelectric dual-field systems includes the following steps: S1. Extract radio frequency voltage and radio frequency current signals through non-contact induction sampling.

[0022] S2. Extract the synchronous local oscillator signal from the measured signal.

[0023] S3. The synchronous local oscillator signal is used to perform coherent zero-IF demodulation on the radio frequency voltage signal and the radio frequency current signal, and I / Q channel mismatch correction is performed to generate a baseband analog signal.

[0024] S4. Convert the baseband analog signal into a digital baseband signal, and perform modulation and desensitization processing on the digital baseband signal to generate a digital baseband component for phase detection.

[0025] S5. Construct an initial coupling matrix and update the initial coupling matrix using a recursive least squares algorithm to generate a corrected coupling matrix for eliminating parameter drift.

[0026] S6. Use the modified coupling matrix to eliminate cross-coupling of the digital baseband components and reconstruct the true voltage complex and the true current complex.

[0027] S7. Calculate the impedance parameters and quality indicators based on the actual voltage complex and the actual current complex.

[0028] The above technical solution provides a complete method for non-contact vector phase detection and impedance estimation in a dual-field magnetoelectric system. This method fundamentally avoids high-voltage safety risks and circuit interference problems through non-contact inductive sampling; it ensures the stability of the phase reference and high fidelity of demodulation by extracting the synchronous local oscillator, phase-coherent demodulation, and I / Q correction; it solves the problem of interference from program content on phase measurement through modulation desensitization processing; and it overcomes the difficulties of parasitic coupling and parameter drift in non-contact measurement by constructing and recursively updating the coupling matrix and using this matrix to eliminate cross-coupling; ultimately achieving high-precision impedance and phase estimation.

[0029] Optionally, step S3 includes: mixing the radio frequency voltage signal with the in-phase component and quadrature component of the synchronous local oscillator signal respectively to generate a baseband in-phase component and a baseband quadrature component of the voltage; mixing the radio frequency current signal with the in-phase component and quadrature component of the synchronous local oscillator signal respectively to generate a baseband in-phase component and a baseband quadrature component of the current; performing low-pass filtering on the baseband in-phase component, the baseband quadrature component, the baseband in-phase component, and the baseband quadrature component of the voltage; and performing I / Q channel mismatch correction on the filtered and denoised baseband in-phase component, the baseband quadrature component, the baseband in-phase component, and the baseband quadrature component of the current to generate the baseband analog signal, wherein the baseband analog signal includes a voltage in-phase component signal, a voltage quadrature component signal, a current in-phase component signal, and a current quadrature component signal.

[0030] The above technical solution specifically defines the IQ demodulation and calibration process. This step, through a detailed process of mixing, low-pass filtering, and I / Q channel mismatch correction, effectively compensates for the amplitude and phase imbalance between hardware channels, improves image rejection, reduces measurement errors introduced by demodulation distortion, and provides a high-fidelity baseband signal for subsequent digital processing.

[0031] Optionally, step S4 includes: using a co-source clock control to perform synchronous analog-to-digital conversion on the voltage in-phase component signal, the voltage quadrature component signal, the current in-phase component signal, and the current quadrature component signal to generate a digital baseband signal; setting the baseband equivalent noise bandwidth and the digital integration window width according to the transmitter's operating mode; and processing the digital baseband signal using a long integration or digital averaging algorithm to reduce the disturbance of the program modulation component on the carrier frequency phase estimation to generate a digital baseband component for phase detection.

[0032] The above technical solution specifically defines the sampling and digital processing procedures. This step ensures the synchronicity of multi-channel sampling by using a synchronous analog-to-digital converter with a shared clock source, avoiding phase errors introduced by asynchronous clocks. Simultaneously, through long integral or digital averaging algorithms, the disturbance of program modulation components on carrier frequency phase estimation is effectively reduced, achieving modulation desensitization and ensuring the stability of phase measurement under modulation mode.

[0033] Optionally, step S5 includes: constructing an initial coupling matrix during transmitter offline operation; introducing imperceptible perturbation excitation and collecting perturbation data during transmitter operation; and using a recursive least squares algorithm to iteratively update the initial coupling matrix with the perturbation data to generate the corrected coupling matrix.

[0034] The above technical solution specifically defines the calibration and self-calibration process. This step achieves "non-cast" dynamic self-calibration by constructing an initial matrix offline and introducing imperceptible perturbations online, then updating the matrix using the RLS algorithm. This enables the method to effectively compensate for parameter drift caused by environmental changes and device aging, ensuring the long-term stability and high accuracy of the measurement results.

[0035] In summary, this application includes at least one of the following beneficial technical effects: 1. A complete non-contact vector phase detection and impedance estimation device was constructed through the collaborative work of modules such as magnetoelectric dual-field non-contact sensing, reference phase-locked loop (PLL), IQ demodulation and calibration, digital processing and modulation desensitization, calibration and self-calibration, and impedance calculation. This scheme fundamentally avoids the high-voltage safety risks and impedance interference problems of traditional contact sampling. Simultaneously, through IQ channel mismatch correction, modulation desensitization processing, and the use of the corrected coupling matrix for cross-coupling elimination and parameter drift correction, it effectively solves the technical bottlenecks existing in related non-contact technologies, such as probe drift, parasitic coupling between channels, demodulation imbalance, local oscillator drift, and program modulation interference. This achieves high-precision, high-stability, and high-safety non-contact measurement of high-voltage radio frequency signals.

[0036] 2. By mixing the RF voltage and current signals with the in-phase and quadrature components of the synchronous local oscillator, respectively, and then performing low-pass filtering, signal-to-baseband conversion is achieved. Crucially, this scheme explicitly corrects for I / Q channel mismatch in the filtered baseband components. This effectively compensates for amplitude and phase imbalances between the I and Q channels caused by hardware differences, significantly improves image rejection, reduces measurement errors introduced by demodulation distortion, and thus outputs a high-fidelity baseband analog signal.

[0037] 3. By employing synchronous analog-to-digital conversion with co-current clock control, strict consistency in the sampling times of the four I / Q signals (voltage and current) is ensured, avoiding phase errors introduced by asynchronous clocks between channels. Furthermore, by setting the bandwidth and integration window according to the operating mode and using long integral or digital averaging algorithms, modulation desensitization is achieved, effectively reducing the disturbance of program content components on carrier frequency phase estimation under AM, DRM, and other modulation modes, significantly improving the stability of phase measurement under modulated signals.

[0038] 4. A "initial calibration + dynamic correction" strategy is adopted: An initial coupling matrix is ​​constructed during transmitter offline operation to correct inherent hardware cross-coupling; during transmitter operation, imperceptible perturbation excitation is introduced, and the matrix is ​​iteratively updated using the recursive least squares (RLS) algorithm based on the perturbation data. This design achieves "non-cast" self-calibration, dynamically compensating for parameter drift caused by changes in environmental temperature and humidity or device aging, ensuring the long-term stability and accuracy of the device measurements, while reducing the maintenance costs of manual calibration. Attached Figure Description

[0039] Figure 1 This paper shows a schematic diagram of a non-contact vector phase detection and impedance estimation device for a dual-field magnetoelectric system according to an embodiment of this application. Figure 2 A schematic diagram of the capacitively coupled voltage probe in an embodiment of this application is shown; Figure 3 A schematic diagram of the structure of the magnetic induction current probe in an embodiment of this application is shown; Figure 4 A schematic diagram of a reference phase-locked link module according to an embodiment of this application is shown; Figure 5 A schematic diagram of the IQ demodulation and calibration module according to an embodiment of this application is shown; Figure 6 This illustration shows a matrix decoupling self-calibration process diagram of the calibration and self-calibration module in an embodiment of this application; Figure 7 A schematic diagram of the calculation logic of the impedance and phase calculation module in an embodiment of this application is shown; Figure 8 The diagram shows a flowchart of a non-contact vector phase detection and impedance estimation method for a dual-field magnetoelectric field according to an embodiment of this application.

[0040] Figure labeling: 10. Magnetoelectric dual-field non-contact sensing module; 11. Electric field coupled voltage probe; 12. Magnetic field induced current probe; 20. Front-end protection and band-limiting module; 21. Electric field channel protection and band-limiting unit; 22. Magnetic field channel protection and band-limiting unit; 30. Reference phase-locked link module; 31. Near-field small loop pickup unit; 32. Schottky limiting unit; 33. Programmable prescaler unit; 34. Digital PLL phase-locked unit; 40. IQ demodulation and calibration module; 41. Phase coherent demodulation unit; 42. Low-pass filter unit; 43. Channel mismatch calibration unit; 50. Sampling and digital processing module; 51. ZIF sampling unit; 52. Same-source clock control unit; 53. Modulation desensitization unit; 60. Calibration and self-calibration module; 61. Initial calibration unit; 62. Self-calibration unit; 70. Impedance and phase calculation module. Detailed Implementation

[0041] The following combination Figures 1-8 This application will be described in further detail.

[0042] This application discloses a magnetoelectric dual-field non-contact vector phase detection and impedance estimation device.

[0043] Figure 1 This diagram illustrates a module schematic of a magnetoelectric dual-field non-contact vector phase detection and impedance estimation device according to an embodiment of this application. Figure 1 As shown, the magnetoelectric dual-field non-contact vector phase detection and impedance estimation device includes a magnetoelectric dual-field non-contact sensing module 10, a front-end protection and band-limiting module 20, a reference phase-locked link module 30, an IQ demodulation and calibration module 40, a sampling and digital processing module 50, a calibration and self-calibration module 60, and an impedance and phase calculation module 70.

[0044] The dual-field non-contact sensing module 10 includes an electric field-coupled voltage probe 11 and a magnetic field-inducing current probe 12. The electric field-coupled voltage probe 11 is used for non-contact induction sampling of the radio frequency voltage signal, and the magnetic field-inducing current probe 12 is used for non-contact induction sampling of the radio frequency current signal. A front-end protection and band-limiting module 20 is located between the dual-field non-contact sensing module 10 and the reference phase-locked link module 30, and is used for graded protection and selective filtering of the radio frequency voltage and current signals. The reference phase-locked link module 30 is used to extract the synchronous local oscillator signal from the measured signal. The IQ demodulation and calibration module 40 is connected to the dual-field non-contact sensing module 10 and the reference phase-locked link module 30, and is used to perform coherent zero-IF demodulation of the radio frequency voltage and current signals using the synchronous local oscillator signal, and to perform I / Q channel mismatch correction to generate a baseband analog signal. The sampling and digital processing module 50, connected to the IQ demodulation and calibration module 40, is used to convert the baseband analog signal into a digital baseband signal and perform modulation and desensitization processing on the digital baseband signal to generate digital baseband components for phase detection. The calibration and self-calibration module 60 is used to construct an initial coupling matrix and update the initial coupling matrix using a recursive least squares algorithm to generate a corrected coupling matrix for eliminating parameter drift. The impedance and phase calculation module 70, connected to the sampling and digital processing module 50 and the calibration and self-calibration module 60, is used to perform cross-coupling elimination on the digital baseband components using the corrected coupling matrix, reconstruct the true voltage complex and true current complex, and calculate impedance parameters and quality indicators based on the true voltage complex and true current complex.

[0045] The following is a detailed description of each module of the device with reference to the accompanying drawings.

[0046] The magnetoelectric dual-field non-contact sensing module 10 serves as the signal acquisition front end of the device. It breaks through the technical limitations of traditional contact sampling and adopts a dual-probe collaborative design of capacitively coupled voltage probe 11 and magnetically inductive current probe 12. Based on the core logic of "near-field induction + no electrical connection", it achieves "circuit-free and highly stable coupling" sampling of high-voltage radio frequency signals, thereby avoiding high-voltage safety risks and interference problems of the original circuit from the root, and providing a pure original signal for subsequent high-precision measurement.

[0047] Figure 2 A schematic diagram of the capacitively coupled voltage probe in an embodiment of this application is shown. Figure 2As shown, the capacitively coupled voltage probe 11, as an electric field (E) sampling unit, adopts an integrated design of "main electrode - guard electrode - insulating support". Its core principle is to pick up voltage signals through near-field coupling of the electric field. The main electrode is made of a metal material with excellent conductivity and maintains a safe distance from the high voltage being measured. It utilizes the small coupling capacitance naturally formed between the two to induce near-field displacement current. The guard electrode surrounding the main electrode uses bootstrap driving technology to "track" the induced potential of the main electrode in real time, completely eliminating the problem of electric field distortion at the edge of the main electrode, significantly improving the stability of the coupling capacitance, and effectively reducing interference caused by environmental humidity and slight changes in installation position. To avoid the capacitively coupled voltage probe 11 affecting the original circuit of the transmitter, a high input impedance buffer stage is specially designed at the signal output end. With its extremely low static equivalent input capacitance, it ensures that the impedance matching state of the original output network will not be changed after connection, achieving the goal of "sampling without interference".

[0048] Figure 3 A schematic diagram of the structure of the magnetic induction current probe in an embodiment of this application is shown. Figure 3 As shown, the magnetic induction current probe 12 serves as a magnetic field (H) sampling unit. Based on the core principle of electromagnetic induction, it employs an open-type structure for easy on-site installation, allowing for quick fixation without cutting the transmitter feeder. The core magnetic core is made of materials with excellent high-frequency characteristics, and combined with a carefully designed secondary winding, it can efficiently sense the magnetic field signal generated by the current in the feeder. Equipped with a switchable load and band-limited network, it can flexibly adapt to the feeder characteristics of different transmitters, while filtering out high-frequency noise and parasitic ringing interference. This unit is equivalent to a broadband current transformer, possessing a stable amplitude and phase response within the target operating frequency band. It can accurately capture current signal characteristics without introducing additional impedance due to series connection, perfectly meeting the "disturbance-free sampling" requirement under high-voltage, high-current conditions.

[0049] The front-end protection and band-limiting module 20 serves as a crucial intermediate link connecting the magnetoelectric dual-field non-contact sensing module 10 with the subsequent reference phase-locked link module 30 and IQ demodulation and calibration module 40. Its core function is to "preserve the amplitude and phase integrity of the E / H signal within the target frequency band of 3–30MHz to the greatest extent possible while ensuring the safety of the subsequent circuitry." Through a design scheme of "graded protection + precise filtering," it achieves three core functions: "resistance to strong surges, deep filtering of clutter, and stable amplitude and phase response," providing a high-purity, low-distortion signal foundation for subsequent phase-coherent demodulation and impedance / phase calculation. The front-end protection and band-limiting module 20 includes an electric field channel protection and band-limiting unit 21 and a magnetic field channel protection and band-limiting unit 22.

[0050] The electric field channel protection and band-limiting unit 21 adopts a three-level collaborative architecture of high-impedance buffer element, TVS clamping element and π-type band-limiting element for the weak electric field coupling signal picked up by the capacitively coupled voltage probe 11. This avoids the impact of high voltage surge, power frequency interference and UHF noise on the subsequent circuits, and ensures low-loss transmission of the target frequency band signal.

[0051] The high-impedance buffer element is selected from a wideband amplifier circuit with a JFET / CMOS architecture, and its equivalent input impedance is ≥10Ω. 12 Ω, static equivalent input capacitance ≤3pF, which can match the high impedance output characteristics of the E probe and stabilize the equivalent value of the coupling capacitor Cc, while avoiding signal attenuation and phase shift, and the signal transmission loss ≤0.05dB.

[0052] The TVS clamping element uses a 2kV (1.2 / 50μs) surge-tolerant bidirectional TVS diode, which is combined with a Schottky diode to form a cascaded clamping circuit. For sudden surges such as high-voltage discharge and lightning strikes, it can clamp abnormal voltages to a safe threshold (≤5V) within nanoseconds, while being compatible with the peak range of normal signals (±1.8V) without introducing additional distortion.

[0053] The π-type band-limiting element consists of a 100pF high-frequency ceramic capacitor and a 1kΩ metal film resistor forming a symmetrical π-type network, specifically designed for deep suppression of 50 / 60Hz power frequency interference, grid harmonics, and >30MHz VHF clutter. Specifically, it attenuates power frequency interference by ≥60dB, >30MHz clutter by ≥40dB, and strictly controls signal attenuation within the 3–30MHz target frequency band to ≤0.1dB, with an amplitude-frequency response flatness of ≤±0.05dB.

[0054] The magnetic field channel protection and band-limiting unit 22, based on the secondary side signal characteristics of the magnetic induction current probe 12, adopts a coordinated design of switchable damping elements and RC band-limiting network to address issues such as high-frequency parasitic ringing and phase nonlinearity, ensuring the consistency of phase response within the target frequency band.

[0055] The switchable damping element employs a 100Ω / 200Ω high-precision alloy resistor on the secondary side, supporting flexible switching based on the measured signal strength, frequency band, and load requirements. Low impedance damping (100Ω) is suitable for strong signal scenarios, suppressing overload distortion; high impedance damping (200Ω) is suitable for weak signal scenarios, improving signal amplitude stability. In both modes, the signal distortion is ≤0.01%.

[0056] The RC band-limited network is a low-pass filter structure composed of high-frequency capacitors and resistors. Working in conjunction with switchable damping elements, it provides a dual guarantee of "damping to suppress ringing + RC filtering to remove noise." It can attenuate parasitic ringing at frequencies >30MHz by ≥40dB, preventing modulation distortion of the baseband phase, while ensuring a phase response linearity error ≤0.5° and a phase delay consistency ≤1ns / MHz within the 3–30MHz frequency band, providing a precise phase foundation for subsequent phase synchronization and phase detection.

[0057] The reference phase-locked link module 30 is the core reference module for coherent demodulation in this device. Its core function is to provide a low-jitter, highly stable, and wide-adaptability local oscillator signal. By accurately extracting the phase reference of the measured signal, it ensures the phase synchronization of E / H dual-channel demodulation, directly determining the upper limit of impedance and phase measurement accuracy. The reference phase-locked link module 30 adopts a full-link optimized design of "signal pickup - amplitude stabilization and shaping - frequency division and adaptation - dual-loop phase-locking - loss-of-lock redundancy", which balances wide-band adaptability and phase stability in complex electromagnetic environments.

[0058] Figure 4 A schematic diagram of a reference phase-locked link module according to an embodiment of this application is shown. Figure 4 As shown, the reference phase-locked link module 30 includes a near-field small loop pickup unit 31, a Schottky limiting unit 32, a programmable prescaler unit 33, and a digital PLL phase-locked link unit 34 connected in sequence. The parameters of each unit are adapted according to actual engineering requirements, taking into account both performance and reliability.

[0059] The near-field small loop pickup unit 31 uses a high-frequency oxygen-free copper small loop antenna and employs a shielded package (single-point grounding of the shielding shell) to suppress environmental electromagnetic interference. This unit picks up the carrier component of the measured signal through near-field coupling, eliminating the need for electrical connection with high-voltage conductors. Furthermore, the coupling gain has a fine-tuning function, ensuring stable reference phase extraction even in complex environments.

[0060] The Schottky limiting unit 32 employs a dual Schottky diode symmetrical limiting circuit, coupled with an impedance matching network, to stabilize the reference signal amplitude within a certain range. Its limiting response time is short, and the phase disturbance for signals in the 1.6-60MHz target frequency band does not exceed 0.08°. This not only avoids overload distortion in subsequent circuits but also ensures phase-locked loop stability in weak signal scenarios.

[0061] The programmable prescaler unit 33 employs a low-noise wideband divider chip, supporting DSP programming to adjust the division factor (range 1-16, step size 1). It can divide input signals from 1.6-60MHz to a suitable range of 100kHz-60MHz to meet the PLL input frequency requirements. The phase noise transfer factor of this divider is no higher than -18dB / decade, avoiding phase jitter accumulation during the division process.

[0062] The digital PLL phase-locked unit 34 adopts a dual-loop architecture of "capture loop + tracking loop", and implements loop control through FPGA. Its parameter configuration balances locking speed and stability. The capture loop bandwidth is 1-3kHz, and it can achieve fast locking within 30ms thanks to its wideband search characteristics. The tracking loop bandwidth is 50-200Hz, and it automatically switches after locking. It uses narrowband characteristics to suppress phase noise, strictly controlling the phase jitter of the local oscillator signal to no more than 0.1° (RMS value), and the output phase noise is no higher than -105dBc / Hz@1kHz offset.

[0063] To ensure demodulation continuity, the reference phase-locked link module 30 incorporates highly reliable loss-of-lock logic, with parameter settings tailored to actual application scenarios.

[0064] The reference phase-locked link module 30 monitors the power amplitude and phase stability of the reference signal in real time. When the power of the reference signal changes suddenly by ±3dB, or the phase change is not less than 5°, it is immediately determined to be in a state of loss of lock. At the same time, a hardware interrupt is triggered synchronously to record the timestamp of the loss of lock (accuracy not exceeding 1μs), providing a basis for subsequent data cleaning and fault analysis.

[0065] Within 10ms of losing lock-in, the reacquisition procedure is automatically initiated. The digital PLL phase-locked unit 34 switches from the tracking loop back to the acquisition loop and quickly locates the target carrier using an optimized frequency search algorithm. During reacquisition, the reference phase-locked link module 30 outputs the temporary value of the LO signal from the previous moment to avoid demodulation data loss and ensure demodulation continuity.

[0066] The reference phase-locked link module 30 adjusts the prescaler coefficient through the programmable prescaler unit 33. Combined with the wideband voltage-controlled oscillator (VCO) inside the digital PLL phase-locked unit 34, it can adapt to the 1.6-60MHz extended frequency band without hardware replacement, meeting the carrier frequency requirements of different topology RF systems. The digital PLL phase-locked unit 34 uses temperature-compensated RC components, ensuring that the LO signal frequency drift does not exceed ±15ppm and the phase drift does not exceed ±0.06° / ℃ within an operating temperature range of -40 to 85℃. The reference phase-locked link module 30 incorporates multi-stage power supply filtering and a metal shielding cover, which can resist power grid fluctuations and external electromagnetic interference, further improving the stability of the phase reference. After passing through a low-noise buffer amplifier, the LO signal is distributed to the E / H dual-channel IQ demodulation module via an equal-length transmission line, ensuring that the phase consistency of the two LO signals does not exceed 0.05°, providing precise synchronization for dual-channel coherent demodulation.

[0067] Figure 5 A schematic diagram of the IQ demodulation and calibration module according to an embodiment of this application is shown. Figure 5 As shown, the IQ demodulation and calibration module 40, as the core component for converting RF signals into valid baseband information, performs coherent demodulation, noise suppression, and channel mismatch calibration on the E / H RF signals output from the front end, thereby outputting high-fidelity baseband I / Q signals. This provides a precise amplitude and phase basis for subsequent impedance and phase calculations. The design of this module focuses on three core elements: demodulation linearity, noise suppression, and mismatch calibration. The parameter settings are simplified and meet actual engineering needs. The IQ demodulation and calibration module 40 includes a coherent demodulation unit 41, a low-pass filter unit 42, and a channel mismatch calibration unit 43.

[0068] In the coherent demodulation unit 41, the E / H dual-channel signals are respectively connected to independent double-balanced mixers, and coherently mixed with the quadrature local oscillator (LO) signals (0° / 90°) provided by the reference phase-locked link to complete the down-conversion of the RF signal to the baseband I / Q signal. A mixer with high isolation is selected to avoid interference caused by LO signal leakage, ensure that the conversion loss is controlled within a reasonable range, and guarantee the effective transfer of signal energy. The quadrature LO signal is implemented through power-divided phase shifting, and the quadrature error is strictly controlled, laying the foundation for reducing the original channel mismatch. An impedance matching network is configured at the mixer input to avoid amplitude and phase distortion caused by signal reflection.

[0069] The low-pass filter unit 42 employs a two-stage filtering design of "analog anti-aliasing + digital low-pass," focusing on suppressing image frequency components, high-frequency noise, and out-of-band interference. Its baseband equivalent noise bandwidth (ENBW) can flexibly adapt to different measurement scenarios. The first-stage analog anti-aliasing filter element filters out high-frequency components generated by mixing, preventing aliasing during subsequent sampling and ensuring the linearity of the signal phase. The second-stage digital low-pass filter element further suppresses noise; the ENBW can be adjusted within the range of 10-200Hz, balancing high accuracy in static measurements with fast response characteristics in dynamic tuning.

[0070] Due to differences in hardware components, amplitude mismatch (ε) is prone to occur in the I / Q channels. g ) and phase mismatch (ε φ In cases where ε exists, this will directly lead to an increase in the image frequency component, thereby affecting the accuracy of phase measurement. Ideally, the amplitudes of the I / Q channels are consistent and their phases are orthogonal (90°), allowing for accurate reconstruction of the signal's amplitude and phase after demodulation; however, when ε exists... g (I / Q amplitude ratio deviates from 1) and ε φ When the phase deviates by 90°, image frequency interference is introduced, causing the phase measurement error Δφ to differ from ε. g ε φ They show a positive correlation, and the magnitude of the error increases with the degree of mismatch.

[0071] The channel mismatch calibration unit 43 eliminates the effects of mismatch through digital domain calibration. Its core lies in constructing a mismatch model based on a known reference signal and compensating for amplitude differences and phase deviations through linear correction. After calibration, the image frequency suppression can reach ≥40dB, significantly reducing the interference of mismatch on measurement accuracy and ensuring that demodulation phase error is within a minimal range.

[0072] The sampling and digital processing module 50, as a key link between IQ demodulation and matrix decoupling, focuses on achieving high-precision digital conversion of signals. Simultaneously, through clock synchronization control and modulation desensitization design, it ensures the stability of carrier frequency phase estimation, providing a low-jitter, anti-interference digital signal foundation for subsequent impedance / phase calculations. The module's design combines simplicity and practicality, emphasizing core logic and mathematical analysis. The sampling and digital processing module 50 comprises a ZIF sampling unit 51, a co-current clock control unit 52, and a modulation desensitization unit 53.

[0073] The ZIF sampling unit 51 supports both analog ZIF (Zero Intermediate Frequency) and digital ZIF architectures, adapting to the cost and performance requirements of different systems. Its core logic focuses on "near-DC signal extraction," eliminating the need for complex parameter configuration. The analog ZIF architecture directly low-pass filters the near-DC signal after IQ demodulation, followed by low-speed sampling, thus simplifying analog link design and reducing hardware complexity. The digital ZIF architecture performs high-speed sampling of E / H RF signals, achieving zero-IF conversion in the digital domain through frequency synthesis and mixing, flexibly adapting to wide-band requirements.

[0074] The co-current clock control unit 52 uses a shared co-current clock for the I / Q channels, with the core objective of suppressing the impact of phase jitter on measurement accuracy. This unit samples the jitter σ. t (Sampling time deviation caused by clock jitter) is converted into phase error σ φ It can be expressed as the following formula: σ φ ≈ω0·σ t (Where ω0 is the angular frequency of the measured signal, ω0 = 2πf0, and f0 is the carrier frequency). By using a common clock source, the sampling times of the I / Q channels are strictly synchronized, and σ t Controlled within a very small range, thus making σ φ To meet measurement accuracy requirements and avoid introducing additional phase errors due to asynchronous clocks between channels.

[0075] The modulation desensitization unit 53, targeting AM / DRM and other modulated signals, optimizes the baseband integration window and equivalent noise bandwidth (ENBW) to achieve modulation desensitization, preventing program modulation from disturbing the carrier phase estimation. The estimated carrier phase φ0 is obtained from the baseband signal within the integration window T. avg The integral within the range is calculated, and the formula is: φ0=arg[∫0 Tavg [s(t)dt], where s(t) is the demodulated baseband signal, containing carrier phase information and program modulation components (AM for amplitude modulation, DRM for orthogonal frequency division multiplexing). The frequency range of the program modulation components is typically 0–10 kHz (AM) or higher (DRM subcarriers), determined by selecting an appropriate integration window T. avg With ENBW (ENBW and T) avg The condition approximately satisfies ENBW≈1 / (2T_avg), which suppresses high-frequency modulation components during the integration process. When ENBW is controlled within 10–200Hz, T... avg For 5–50ms, the cumulative effect of the modulation component within the integration window is significantly weakened, and φ0 mainly reflects the true phase of the carrier frequency, unaffected by program content disturbances, ensuring the stability of phase estimation.

[0076] Figure 6 A schematic diagram of the matrix decoupling self-calibration process of the calibration and self-calibration module in an embodiment of this application is shown. Figure 6 As shown, the calibration and self-calibration module 60 is the core mechanism for eliminating hardware coupling, parameter drift, and environmental interference. It adopts a two-layer design scheme of "initial benchmark construction + dynamic real-time adjustment" to ensure the long-term accuracy of non-contact measurements. The module has a simple logical structure, focusing on two core objectives: "precise calibration" and "stable updates." The calibration and self-calibration module 60 includes an initial calibration unit 61 and a self-calibration unit 62.

[0077] During the initial installation phase, the initial calibration unit 61 provides initial calibration parameters for the system through data acquisition and algorithm solving under typical operating conditions, thereby eliminating inherent hardware errors (such as E / H channel cross-coupling and I / Q imbalance). Specifically, it utilizes two typical states of the system under test ("approximate matching" and "slight detuning") to cover conventional operating scenarios, acquiring E / H channel output signals and solving for calibration parameters through mathematical modeling. "Approximate matching" refers to the measured impedance being close to the system's characteristic impedance, with minimal reflection, suitable for calibration of small-signal linear relationships; "slight detuning" refers to the measured impedance deviating from the characteristic impedance, with moderate reflection, suitable for calibration of medium-to-strong signal coupling characteristics.

[0078] Assume the true voltage / current vector is x = [V, I] T (where V represents voltage and I represents current), the output signal vector of the E / H sensor is y = [y E ,y H ] T Where y E The electric field sample value, y H The magnetic field sampling values ​​and the two satisfy a linear relationship: y = K·x + n, where K is the coupling matrix (containing main coupling and cross-coupling terms), and n is the measurement noise. The residual is minimized using the least squares method. Solving the initial calibration matrix Simultaneously acquire amplitude / phase imbalance correction parameters (such as amplitude ratio and phase deviation) of the I / Q channels to provide a reference for subsequent measurements.

[0079] The self-calibration unit 62 is used to dynamically compensate for parameter drift. It operates without system downtime, updating calibration parameters in real time through perturbation excitation and confidence assessment to offset accuracy degradation caused by factors such as hardware aging and temperature changes. The self-calibration unit 62 applies a small excitation of 1-2 LSB to the system under test (its amplitude is much smaller than a normal signal and will not cause perceptual interference), acquires E / H channel response data, and updates the calibration matrix based on a recursive algorithm. in Let be the calibration matrix at time t, α be the correction coefficient (used to adjust the update step size), and Δy and Δx be the output and input changes caused by perturbations, respectively, to ensure a smooth update process without affecting normal operation. Real-time monitoring of the signal-to-noise ratio (SNR) and residuals is also performed. When the SNR is too low (e.g., <20dB) or the residual exceeds the threshold (e.g., >5% of full scale), parameter updates are automatically frozen to prevent abnormal data from contaminating the calibration results; all calibration parameters ( The I / Q imbalance parameters are stored in a versioned manner, allowing for rollback to the most recent stable version, ensuring that the calibration process is traceable and recoverable.

[0080] Figure 7 A schematic diagram illustrating the calculation logic of the impedance and phase calculation module in an embodiment of this application is shown. Figure 7 As shown, the impedance and phase calculation module 70, as the core conversion unit of the device from "signal processing to practical output," plays a crucial role in connecting matrix decoupling calibration and external linkage control. Its core function lies in calculating the reconstructed true voltage complex... With current complex It accurately calculates core parameters such as impedance and phase, and outputs multi-dimensional quality indicators, ultimately generating targeted alarms and automatic tuning suggestions, providing a basis for decision-making for the efficient operation and safety protection of the transmitter.

[0081] The core calculation parameters of the impedance and phase calculation module 70 cover reconstruction based on matrix decoupling. (including amplitude) Phase θ V )and (including amplitude) Phase θ I Complex impedance (Comprehensive characterization of the amplitude and phase relationship between voltage and current); impedance amplitude (Reflects the overall magnitude of the circuit impedance); resistance component ( The real part corresponds to the active power loss in the circuit; reactance component ( The imaginary part corresponds to the energy storage characteristics of the circuit (positive for inductive, negative for capacitive); phase difference: (The phase shift between voltage and current visually reflects the tuning state.)

[0082] The impedance and phase calculation module 70 simultaneously calculates multiple key quality indicators to comprehensively evaluate the reliability of measurement results and system operating status, providing a reference for subsequent decision-making. Output quality indicators include signal-to-noise ratio (SNR), which reflects... and Signal purity; a higher value indicates less noise interference and a more reliable measurement basis. Image rejection reflects the effectiveness of IQ demodulation and calibration; a value ≥40dB is acceptable, ensuring that channel mismatch does not introduce significant errors. Residual error... Characterization calibration matrix The accuracy of the fit to the actual signal; the smaller the residual, the more thorough the elimination of cross-coupling; condition number. reflect Reversible stability, The smaller, and The smaller the reconstruction error, the higher the confidence level. Based on the above indicators, it is divided into three levels: "high / medium / low", which directly guides the tuning suggestions and alarm logic triggering.

[0083] The impedance and phase calculation module 70, relying on core parameters and quality indicators, conforms to the transmitter's operational requirements, generating precise tuning guidance and risk alarms to achieve a closed loop of "measurement-decision-control". Specifically, for near-matching states (|X| << |Z|, φ ≈ 0°), it outputs a "tuning met" message, maintaining the current operating parameters with a "high" confidence level. For slightly detuned states (|X| is small, φ absolute value ≤ 10°), it provides tuning direction suggestions based on the sign of X—when X > 0 (inductive detuning), it suggests reducing the inductor or increasing the capacitor; when X < 0 (capacitive detuning), it suggests increasing the inductor or decreasing the capacitor, while simultaneously outputting a "slight detuning" warning message. For severe detuning / abnormal conditions, when φ changes abruptly (e.g., |Δφ / Δt|>10° / ms), the residual exceeds the threshold, or the confidence level is "low", a "severe detuning" alarm is triggered; if it is accompanied by an increase in harmonic amplitude, it is directly determined that there is a risk of resonance or arcing, an "emergency backoff" alarm is output, and the transmitter protection logic is activated.

[0084] In some embodiments, the magnetoelectric dual-field non-contact vector phase detector and impedance estimation device is configured with Ethernet, serial, and dry contact interfaces to simplify parameters and focus on functional adaptation, addressing the needs of centralized transmitter monitoring, local control, and emergency linkage. The Ethernet interface uses an industrial-grade interface with integrated electrical isolation to resist interference; it supports the Modbus-TCP protocol (for interfacing with transmitter monitoring systems) and custom protocols (for transmitting impedance / phase parameters and calibration logs), meeting the needs of long-distance, high-volume centralized monitoring and remote configuration. The serial interface uses an industrial-grade RS485 differential interface, supports the Modbus-RTU protocol, and is compatible with field PLCs, control cabinets, and other equipment; it transmits simplified measurement data and basic alarm signals, suitable for short-distance local control. The dry contact interface uses a programmable hard-wired interface; the input side receives transmitter operating status signals for synchronization, and the output side outputs statuses such as normal measurement and detuning alarms; it directly links to the transmitter hardware protection circuit, making emergency response more reliable and avoiding software delay risks.

[0085] Furthermore, the magnetoelectric dual-field non-contact vector phase detection and impedance estimation device, combined with the transmitter's high-voltage, high-current operating characteristics, achieves closed-loop protection through an "anomaly detection-graded action" mechanism, with its core logic focusing on key risk points. Specifically, it monitors abnormal changes in phase difference φ (such as resonance offset, load abrupt changes), and upon detection of an anomaly, outputs a detuning alarm; if normal operation is not restored within a short time, it automatically triggers power reduction and freezes the tuning actuator to prevent further detuning. It analyzes the voltage / current spectrum, and if harmonic content exceeds limits or signal distortion intensifies, it outputs a harmonic alarm; if the abnormality persists, it limits transmitter power and prompts a check of the load matching network. It identifies typical high-voltage discharge signals (such as sudden narrow pulses, sudden rises in noise floor), and upon detection, immediately triggers an emergency backoff operation—outputting a severe alarm, cutting off the high-voltage power supply, and recording fault data for subsequent analysis; after fault resolution, manual reset is required to restore operation. All anomalies and protection actions are stored locally, supporting data export and traceability, facilitating fault analysis and tuning optimization.

[0086] In some embodiments, the E / H probe is fixed to an insulating bracket adapted to the high-voltage environment and has no electrical connection with the high-voltage conductors (anode busbar, feeder) to avoid the risk of high-voltage conduction; its installation position must ensure a stable electric field distribution and reduce coupling drift. The creepage distance and clearance between the probe and the high-voltage conductor should meet the safety standards for high-voltage equipment and allow for environmental margins; an insulating baffle is installed between them to prevent insulation failure caused by contact with foreign objects.

[0087] To address the strong electromagnetic radiation environment of the transmitter, interference is suppressed through structural design: the probe and core module (front-end protection, reference phase-locked loop) are encapsulated with metal shielding, ensuring the direction of the shielding shell gaps is orthogonal to the transmitter's main current direction, thus cutting off parasitic coupling paths; the probe cable uses double-shielded wire to reduce signal interference. All grounding parts of the system (shielding shell, signal ground, power ground) are connected to the same grounding electrode (grounding resistance ≤1Ω), eliminating ground loop interference caused by potential differences and ensuring signal purity. The E-probe, insulating bracket, and cable insulation layers are all made of high-voltage and high-temperature resistant special materials (such as PTFE, FR4, and silicone rubber) to adapt to the high-temperature and high-humidity environment of the transmitter room. High-voltage contact components must undergo withstand voltage and insulation resistance tests before leaving the factory to ensure no risk of breakdown or flashover; their surfaces are coated with anti-fouling coatings, and internal moisture-proof measures are taken to avoid environmental factors affecting insulation performance. During actual installation, it is necessary to ensure that the gaps are standardized, the grounding is reliable, and the shielding direction is correct, constructing a three-dimensional protection system of "insulation-shielding-grounding" to adapt to high-voltage, high-current transmitter scenarios.

[0088] The implementation principle of the magnetoelectric dual-field non-contact vector phase detection and impedance estimation device in this application is as follows: A complete non-contact vector phase detection and impedance estimation device is constructed through the coordinated operation of modules such as magnetoelectric dual-field non-contact sensing, reference phase-locked loop, IQ demodulation and calibration, digital processing and modulation desensitization, calibration and self-calibration, and impedance calculation. This scheme fundamentally avoids the high-voltage safety risks and impedance interference problems of traditional contact sampling. Simultaneously, through IQ channel mismatch correction, modulation desensitization processing, and the use of the corrected coupling matrix for cross-coupling elimination and parameter drift correction, it effectively solves the technical bottlenecks existing in related non-contact technologies, such as probe drift, parasitic coupling between channels, demodulation imbalance, local oscillator drift, and program modulation interference, achieving high-precision, high-stability, and high-safety non-contact measurement of high-voltage radio frequency signals.

[0089] This application also discloses a non-contact vector phase detection and impedance estimation method for a magnetoelectric dual-field system.

[0090] Figure 8 This diagram illustrates a flowchart of a non-contact vector phase detection and impedance estimation method using a dual-field magnetoelectric system according to an embodiment of this application. Figure 8 As shown, the method includes the following steps: S1. Extract radio frequency voltage and radio frequency current signals through non-contact induction sampling.

[0091] S2. Extract the synchronous local oscillator signal from the measured signal.

[0092] S3. The synchronous local oscillator signal is used to perform coherent zero-IF demodulation on the radio frequency voltage signal and the radio frequency current signal, and I / Q channel mismatch correction is performed to generate a baseband analog signal.

[0093] S4. Convert the baseband analog signal into a digital baseband signal, and perform modulation and desensitization processing on the digital baseband signal to generate a digital baseband component for phase detection.

[0094] S5. Construct an initial coupling matrix and update the initial coupling matrix using a recursive least squares algorithm to generate a corrected coupling matrix for eliminating parameter drift.

[0095] S6. Use the modified coupling matrix to eliminate cross-coupling of the digital baseband components and reconstruct the true voltage complex and the true current complex.

[0096] S7. Calculate the impedance parameters and quality indicators based on the actual voltage complex and the actual current complex.

[0097] In some embodiments, to address the core requirement of non-contact measurement in high-voltage, high-current shortwave transmitters (e.g., the anode circuit of the final stage electron tube and the output matching network), an E / H dual-probe sampling modeling and cross-coupling decoupling system with "algorithm-driven" as its core is constructed. Through mathematical modeling, the signal transmission law is quantified, and parasitic interference is eliminated with precise algorithms, achieving high-precision reconstruction of the true voltage and current amplitude and phase information, providing core algorithmic support for subsequent impedance / phase calculations.

[0098] The core objective of E / H probe sampling modeling is to provide an input signal with "amplitude and phase characteristics standardization" for the decoupling algorithm. Through the collaborative design of probe structure and algorithm, it is ensured that there is a linear mapping relationship between the sampled signal and the real electrical parameters.

[0099] The goal of the E-probe capacitive coupling algorithm is to achieve non-contact linear sampling of line voltage and output a signal with a fixed phase relationship to the voltage. The displacement current coupling algorithm, based on the near-field coupling effect of the electric field, constructs a displacement current model I. c =jωC c V (ω=2πf, f=3~30MHz, C c =1~10pF is the coupling capacitor). The high-impedance approximation algorithm uses the JFET / CMOS architecture to realize the equivalent input impedance Z. in ≥10 12 Ω, full (ωC) c Z in Under condition <<1, the output voltage model is simplified to V s ≈jωC c Z in ·V, ensure V s It has a fixed phase of +90° with V and a linear amplitude mapping; the coupling stabilization algorithm uses a guard bootstrap circuit to ensure that the potential difference between the main / guard electrode is ≤0.1V, suppressing edge electric field distortion and reducing C at the algorithm level. c Drift error caused by installation deviations and changes in temperature and humidity.

[0100] The goal of the H-probe mutual inductance coupling modeling algorithm is to achieve non-contact linear sampling of the line current and output a flat amplitude-frequency characteristic signal that is approximately in phase with the current. The mutual inductance calculation algorithm constructs the mutual inductance model. (k≥0.8 is the coupling coefficient, N=5~20 turns is the number of turns in the winding, μ is the magnetic core permeability); Amplitude-phase calibration algorithm is used to construct the output voltage model. Through load selection (R) L =50 / 100Ω makes ωL leak <<R LAn RC band-limited network (33pF / / 1kΩ, cutoff frequency 48MHz) was designed to ensure that H(ω) has a flat amplitude and frequency response and a phase shift ≤0.5° in the 3-30MHz frequency band, ultimately achieving V s A linear, distortion-free mapping to I.

[0101] The cross-coupling matrix decoupling core algorithm addresses the signal distortion problem caused by parasitic coupling between E / H probes, specifically the cross-coupling terms of voltage in the H channel and current in the E channel. It designs a full-link decoupling algorithm system to achieve accurate reconstruction of the real signal.

[0102] Specifically, a linear coupling model between the E / H channel output and the actual electrical parameters is established: in, Let k be the coupling matrix. v =jωC c Z in The main voltage coupling coefficient. (main current coupling coefficient) This represents the noise vector (including electromagnetic interference and thermal noise). The core of the algorithm is to separate the main coupling and cross coupling through matrix operations, eliminating the influence of α, β, and n.

[0103] To ensure that the matrix is ​​invertible, det(K) = k must be satisfied. v k i -αβ≠0, guaranteed through a two-stage algorithm. The cross-coupling suppression algorithm uses the E-probe's bootstrap circuit to suppress β (leakage magnetic interference), and the H-probe's shielding seam is orthogonal to the main current direction to suppress α (parasitic capacitance coupling), reducing the weight of the cross-coupling term by ≥60% at the algorithm level. The main coupling enhancement algorithm increases Z... in Increase k v Optimize core and winding design to improve k i , ensure k v k i >>αβ, numerically guarantees that the determinant of the matrix is ​​nonzero.

[0104] Then, the coupling matrix is ​​implemented using the least squares algorithm. Accurate estimation: Under two typical operating conditions—"approximate matching" (|X|<<|Z|) and "slight mistuning" (|X| is at its minimum)—multiple sets of {V} were collected. s ,I s Sample and corresponding reference electrical parameters {V ref ,I ref}; Objective function construction and solution: Constructing the least squares objective function min∑[V s I s ]-K·[V ref I ref Solving the problem through iterative calculation. Ensure calibration error is ≤3%.

[0105] Algorithms for real signal reconstruction and impedance calculation. Reconstruction algorithm: based on calibrated signals. Real signal reconstruction is achieved by matrix inversion: in, Directly eliminates cross-coupling and noise interference; impedance parameter calculation algorithm: based on reconstructed complex quantities. and We obtain this through complex number operations: φ=θ V -θ I , R=|Z|cosφ, X=|Z|sinφ.

[0106] Error suppression and adaptive update algorithm. This addresses the issues of reconstruction error and matrix condition number. Based on the positive correlation characteristics, a three-level error suppression algorithm is designed; the signal-to-noise ratio (SNR) enhancement algorithm involves front-end amplitude limiting to stabilize the signal amplitude, band-limited filtering to remove out-of-band clutter, and baseband integration (ENBW = 10–200 Hz) to suppress noise, resulting in an SNR improvement of ≥35 dB; the dynamic correction algorithm is based on perturbation identification technology, applying a small excitation of 1–2 LSB (without program awareness) to the tuning actuator, and updating in real time using a recursive least squares algorithm. Compensate for parameter changes caused by hardware aging and temperature drift to ensure Structure-assisted algorithm: The E / H probe installation spacing is ≥100mm, and double-shielded wires are used for transmission to reduce error sources introduced by parasitic coupling.

[0107] By constructing a full-link algorithm system of "sampling modeling - calibration estimation - reconstruction correction", three core breakthroughs were achieved: First, the amplitude and phase characteristics of the E / H signal were fixed through the probe modeling algorithm to form a standardized sampling input, providing a unified processing basis for the decoupling algorithm; Second, with the innovative design of the cross-coupling matrix decoupling algorithm, parasitic interference was accurately quantified and eliminated, reducing the signal reconstruction error from ≥20% to ≤3%, achieving high-precision interference suppression; At the same time, relying on the synergistic effect of least squares calibration and perturbation recursive update algorithm, the model parameters were ensured to dynamically adapt to the complex operating conditions of high-voltage, high-current transmitters, providing continuous and stable high-precision data support for subsequent impedance assessment, automatic tuning, and safety linkage.

[0108] In some embodiments, "high-precision amplitude and phase extraction" is achieved through coherent zero-IF demodulation. For non-contact measurement of 3-30MHz high-voltage, high-current transmitters, a full-link design of "phase reference locking - amplitude and phase error correction - signal purity enhancement - error budget control" is used to convert the RF signal acquired by the E / H probe into a pure baseband complex signal that can be directly used for impedance calculation. This overcomes the technical challenges of "phase reference drift, I / Q imbalance distortion, and modulation disturbance interference" in traditional demodulation. Specific solutions include reference phase-locked loop and local oscillator allocation, non-contact acquisition of reference signals, and pre-division frequency reduction processing.

[0109] The reference phase-locked loop (PLL) and local oscillator (LO) allocation steps are designed in a coordinated manner through "near-field non-contact parameter acquisition - signal shaping - dual-loop PLL - loss-of-lock redundancy processing" to obtain a stable LO that is in phase with the transmitter carrier, providing a high-precision phase reference for phase-coherent demodulation.

[0110] The non-contact acquisition step of the reference signal uses a miniature near-field loop antenna (diameter ≤20mm) to non-contactly couple a very weak carrier signal around the high-voltage conductor under test (such as the anode bus or output feeder) (avoiding access to high-voltage nodes); after the signal is shaped by a 2-3 stage Schottky limiter (such as 1N5711), a square wave reference signal with stable amplitude is output (amplitude fluctuation ≤±5%), completely eliminating the influence of transmitter carrier power fluctuation (such as ±10dB) on phase-locked loop stability.

[0111] The pre-division down-division process uses a high-speed frequency divider (such as 74HC4060) to divide the limited reference signal to 1-10MHz. This frequency range is compatible with the stable locking range of a digital PLL (phase-locked loop). The frequency division design can reduce the phase noise transmission coefficient of the PLL and significantly improve the phase purity of the final local oscillator.

[0112] Furthermore, a dual-ring digital PLL architecture, employing a "capture ring + tracking ring" dual-ring design, balances locking speed and long-term stability. The capture ring bandwidth is set to 1-3kHz, enabling rapid carrier locking via a wide-range frequency sweep (locking time ≤100ms), adapting to fine-tuning of the carrier frequency within ±500Hz during daily transmitter operation. The tracking ring bandwidth is set to 50-200Hz, using narrow-band tracking characteristics to suppress interference from environmental electromagnetic interference (such as clutter within the 30MHz band) on the local oscillator phase. After locking, the local oscillator phase jitter is ≤0.1° / ms, meeting the accuracy requirements of impedance calculation for the phase reference.

[0113] When the reference signal is lost (e.g., the near-field loop coupling is interrupted) or the carrier power changes abruptly, the PLL immediately initiates the "fast reacquisition" process and records the moment of loss of lock through the timestamp module. In the subsequent digital processing stage, the demodulated data of that period is marked and cleaned based on the timestamp of loss of lock to avoid invalid data from participating in impedance calculation and to ensure the reliability of the results.

[0114] Due to the non-ideal characteristics of hardware circuits (mixers, amplifiers), amplitude mismatch (ε) is very likely to occur in the I / Q demodulation channel. g ) and phase mismatch (ε φ This leads to demodulation complex envelope distortion. In some embodiments, a linear imbalance model is constructed and a least-squares correction algorithm is used to achieve high-precision elimination of amplitude and phase errors, as detailed below: First, a linear imbalance model is constructed, assuming the ideal complex envelope is z = Re{z} + jIm{z} and the uncorrected complex envelope is zm. The two are correlated through a linear matrix: Where: ε g ε is the amplitude difference coefficient for the I / Q channels (typically ≤5%, caused by inconsistent amplifier gain); φ d is the phase deviation of the I / Q channel (typically ≤3°, caused by the phase difference of the mixer local oscillator); d is the DC offset vector (generated by the drift of the circuit's static operating point, typically ≤5mV).

[0115] Then, the correction matrix solving algorithm is executed, employing two complementary methods to solve the correction matrix. To ensure calibration accuracy under different operating conditions: Then, image suppression is minimized through iterative adjustments. The elements (amplitude correction coefficient, phase compensation angle) are used to minimize the power of the image frequency component in the demodulated output, ultimately achieving an image suppression degree of ≥40dB (meeting the clutter suppression requirements in the 3-30MHz frequency band). Then, reference phase sequence fitting is performed. A standard sinusoidal signal with known phase (such as a 10MHz sinusoidal signal generated by the internal DDS, with a phase accuracy ≤0.01°) is input into the demodulation channel, and 20-50 sets of {z} are collected. m The samples in the ,z} are solved using the least squares algorithm. : Finally, real-time correction is applied, and the inverse of the correction matrix is ​​used. The embedded baseband digital signal processing flow performs real-time correction on the demodulated I / Q data; the amplitude and phase consistency error of the I / Q channels after correction is controlled within ≤0.5% (amplitude) and ≤0.1° (phase), providing high-precision complex input for subsequent impedance calculation.

[0116] The baseband filtering, integration, and modulation desensitization steps address the issue of program content (audio / OFDM data) disturbing carrier frequency phase estimation in modulation modes such as AM (amplitude modulation) and DRM (digital radio). Modulation desensitization is achieved through "narrowband filtering + mode-specific integration + frequency offset control" to ensure long-term stability of phase measurements. Specifically, dual-integral / moving average filtering is performed. In the baseband processing stage, a "first-stage RC analog filter + second-stage digital moving average" structure is adopted, setting the baseband equivalent noise bandwidth (ENBW) to 10-200Hz. High-frequency noise (such as electromagnetic interference) within the 30MHz band is suppressed through time-domain averaging, while simultaneously reducing the disturbance amplitude of the program modulation signal to carrier frequency phase estimation by ≥20dB.

[0117] The mode-specific desensitization architecture is as follows: For AM mode (amplitude modulation broadcasting), a narrow ENBW (10-50Hz) is selected to address the amplitude fluctuations of the 0-10kHz audio envelope, ensuring that the sensitivity of phase estimation to amplitude changes is ≤0.1° / dB, thus avoiding phase measurement deviations caused by alternating audio strength. For DRM mode (digital broadcasting), the characteristic of OFDM subcarriers being symmetrically distributed around the carrier frequency is utilized, combined with a digital frequency-locked loop (DFLL) to achieve carrier frequency tracking (residual frequency offset ≤0.1Hz), ensuring that phase estimation depends only on the carrier frequency component and is independent of the OFDM data content of the subcarriers, thereby completely eliminating interference from program data to the measurement.

[0118] Then, precise control of DC and frequency offset is performed, including: DC bias removal, using an 8th-order digital high-pass filter (cutoff frequency 0.1Hz) to control the DC component in the baseband signal to ≤1mV, avoiding phase accumulation error caused by DC drift; residual frequency offset control, using a DFLL to limit the residual carrier frequency offset Δf to |Δf|<0.1Hz, ensuring that the cumulative phase error within the 1s integration window is <0.1°, meeting the measurement accuracy requirements of phase difference φ in impedance calculation (≤±1°).

[0119] In addition, sampling clock jitter and ADC quantization error introduce phase noise. By optimizing the quantization error budget and hardware selection, the impact on phase measurement can be controlled within an acceptable range (≤0.1°).

[0120] First, a phase noise budget model is constructed. The phase error variance is determined by both sampling jitter and quantization error. The specific budget formula is as follows: Where f0 is the transmitter carrier frequency (3-30MHz), and when the maximum value of 30MHz is taken, (2πf0)≈1.88×10 8 rad / s;σ t Substituting the sampling clock jitter (controlled to ≤10ps), we get (2πf0)σ.t ≈0.0188°; σ q For ADC quantization phase error (≤0.05° for 14-16 bit ADC); final total phase error It meets the accuracy requirements.

[0121] Then, the ADC quantization bit width selection is determined. A 14-16 bit resolution baseband ADC (such as ADS127L01) can be used, where a 14-bit ADC can meet the quantization error σ. q For requirements of ≤0.05°, a 16-bit ADC can further reduce this to σ. q ≤0.01°, while avoiding the increase in hardware cost and power consumption caused by excessively high bit width (such as 18 bits).

[0122] Then, an anti-saturation protection design can be implemented by connecting a Schottky limiter in series at the front end of the ADC and setting the limiting level to 90% of the ADC's full scale. When abnormal situations such as high-voltage arcing or resonant spikes occur in the transmitter, the limiter can quickly clamp the signal amplitude of the input ADC to a safe range, avoiding long signal recovery due to ADC saturation and ensuring the continuity of the demodulation process.

[0123] In some embodiments, calibration and non-stop self-calibration are key guarantees for "long-term high-precision measurement." A full-cycle approach—including initial calibration to establish a baseline, operational perturbation correction for drift, confidence management to ensure reliability, and aging records to support maintenance—ensures the coupling matrix... With I / Q correction matrix Long-term stability is required to meet the transmitter's "uninterrupted broadcasting and high reliability" operational requirements.

[0124] First, initial value calibration (baseline parameter establishment) is performed by obtaining the initial value of the coupling matrix through "dual typical working condition sampling + least squares fitting". Initial values ​​of the I / Q imbalance correction matrix This provides a benchmark for subsequent measurements. For the selection of calibration conditions, two conditions covering common transmitter operating states are chosen to ensure the universality of calibration parameters. For the approximately matched condition, the output matching network is manually tuned so that the reactance component |X| << |Z| (e.g., |X| ≤ 0.1|Z|), at which point the phase difference φ ≈ 0°, simulating the normal operating state of the transmitter. For the slightly detuned condition, the capacitance / inductance of the matching network is finely adjusted so that |X| ≈ 0.3|Z| (either inductive or capacitive detuning is acceptable), at which point φ ≈ ±15°, simulating a slightly detuned state of the transmitter.

[0125] Then, data acquisition and fitting calculations are performed, including: data acquisition, collecting 20-50 sets of E / H probe outputs {V} under two working conditions. s ,Is} (i.e., sample {y) i Each sampling session lasts 1 second to ensure the validity of the statistical data. Reference values ​​are obtained by measuring the transmitter output impedance using a standard impedance analyzer (e.g., Agilent E4990A, accuracy ±0.1%), and the true voltage {V} and current {I} (i.e., the sample {x}) are derived. i Perform least-squares fitting and solve using the following formula. and : After calibration, The fitting error is ≤3%. The corresponding image suppression is ≥40dB, which meets the initial measurement accuracy requirements.

[0126] In some embodiments, runtime perturbation identification (dynamic parameter correction) is performed. This is to compensate for hardware aging (such as E-probe coupling capacitor C). c Drift, decrease in the permeability μ of the H-probe core, and changes in ambient temperature and humidity can all contribute to this. The offset is achieved by the system through "perturbation injection - response observation - recursive update" without interrupting broadcasting. The real-time correction is as follows: First, a non-perceptible micro-perturbation injection is performed to apply a minimum step perturbation to the capacitor box / short circuit of the transmitter matching network; the perturbation amplitude is far below the program index requirements to ensure that the audience has no program perception and does not affect normal broadcasting.

[0127] Then, response observation and incremental modeling are performed, including: conducting response observation and collecting the change in E / H probe output Δy={ΔV} before and after the perturbation. s ,ΔI s Simultaneously record the known impedance change Δx = {ΔV, ΔI} corresponding to the perturbation (obtained through theoretical calculation or pre-calibration). Establish A linear incremental model is used to isolate the drift effects of main coupling and cross coupling. The recursive least squares (RLS) algorithm is employed, updating in real time based on {Δy, Δx}. The update cycle is set to 10-30 minutes (configurable based on environmental stability); after each update, The deviation from the true value is ≤2%, ensuring long-term measurement accuracy.

[0128] In some embodiments, to avoid abnormal operating conditions such as strong electromagnetic interference and signal loss... Error updates employ a confidence management and rollback mechanism (parameter reliability assurance). The confidence assessment metrics for ensuring parameter reliability include signal-to-noise ratio (SNR) and residuals. Specifically, the baseband SNR of the E / H channel is monitored in real time; when the SNR < 25dB (insufficient signal quality, measurement error easily exceeds limits), the system is immediately frozen. Update. And calculate the deviation between the model's predicted values ​​and the actual output. (L2 norm), when the residual is >5% of full scale ( (Drift exceeds limit), triggering a low confidence alarm.

[0129] The parameter rollback mechanism includes: establishing a parameter version library and retaining the 10 most recent stable versions. (Each version includes labels for SNR, residual, ambient temperature, etc.) Freeze immediately when a low confidence condition is triggered. Update and automatically revert to the previous stable version (SNR≥30dB, residual ≤3% of full scale). Simultaneously record anomaly timestamps and operating conditions (e.g., carrier frequency, power) for subsequent maintenance personnel to analyze the cause of the anomaly (e.g., probe loosening, interference source intrusion).

[0130] In some embodiments, a parameter aging profile is established to record... The drift curves over time and with ambient temperature provide data for preventative maintenance, preventing measurement errors from exceeding limits due to parameter aging. Specifically, the following parameters can be automatically recorded hourly: Key elements, including the main coupling coefficient k v k i Cross-coupling terms α and β; environmental parameters, including probe-near temperature (-40–85°C) and relative humidity (10%–90% RH); health indicators, including condition number The corresponding mirror suppression degree.

[0131] Furthermore, drift analysis and maintenance early warning are performed. The monthly drift rate of parameters (such as k) is derived through linear fitting. v A monthly drift rate ≤ 0.5% is considered normal. When the parameter drift rate exceeds the preset threshold (e.g., monthly drift rate > 1%), or When matrix stability decreases, a maintenance warning message is generated to prompt the user to check the probe installation status (such as whether the E probe spacing is offset) or replace aging parts (such as the H probe magnetic core) to avoid the risk of decreased measurement accuracy in advance.

[0132] The implementation principle of the non-contact vector phase detection and impedance estimation method in the dual-field magnetoelectric system of this application is as follows: A complete method for non-contact vector phase detection and impedance estimation in the dual-field magnetoelectric system is provided. This method fundamentally avoids high-voltage safety risks and circuit interference problems through non-contact inductive sampling; by extracting the synchronous local oscillator and phase-coherent demodulation and I / Q correction, the stability of the phase reference and the high fidelity of demodulation are ensured; through modulation desensitization processing, the interference of program content on phase measurement is solved; by constructing and recursively updating the coupling matrix and using this matrix to eliminate cross-coupling, the problems of parasitic coupling and parameter drift in non-contact measurement are overcome; ultimately, high-precision impedance and phase estimation is achieved.

[0133] In some embodiments, the device and / or method can be applied to high-voltage final-stage vacuum tubes (suitable for kilovolt-level voltage / ampere-level current scenarios). Addressing the measurement requirements of the anode circuit of a high-voltage, high-current transmitter final-stage vacuum tube, focusing on the stringent operating conditions of high anode voltage and high current, the specific design is as follows: It strictly matches the operating parameters of the anode circuit of the final stage vacuum tube, and the operating frequency band covers the 3-30MHz shortwave band. It is compatible with the final stage circuit of push-pull / parallel architecture of quadrupole / pentode vacuum tubes.

[0134] The main sampling electrode of the E-probe (capacitive voltage sampling) is made of a 12×8mm aluminum sheet (balancing conductivity and lightweight). During installation, it maintains a distance of 8-12mm from the measured anode bus / feeder (balancing coupling efficiency and high voltage insulation). It adopts a three-sided surrounding guard electrode, combined with a bootstrap circuit driven by a non-inverting follower, to suppress the electric field distortion at the edge of the main electrode. The buffer stage is a JFET architecture high-impedance amplifier circuit with an equivalent input capacitance ≤3pF (avoiding interference with the original impedance characteristics of the anode circuit, meeting the "near-zero insertion effect" requirement in the disclosure document).

[0135] The H probe (magnetic ring current sampling) uses NiZn ferrite material for its core (which has low magnetic loss in the HF band and is suitable for 3-30MHz) and adopts an open-and-close snap-fit ​​structure; the secondary winding has N=10 turns, is equipped with a 50Ω / 100Ω switchable alloy resistor load, and is connected in parallel with a 33pF / / 1kΩ RC band-limited network.

[0136] The reference phase acquisition uses a near-field small loop antenna with a diameter ≤20mm for non-contact coupling of the carrier signal → the amplitude is shaped and stabilized by a 2-stage Schottky limiter → the frequency is predivided to 1-5MHz (adapting to the input range of the digital PLL) → the digital PLL (dual-loop structure: acquisition bandwidth 1-3kHz, tracking bandwidth 50-200Hz) outputs a stable local oscillator (LO).

[0137] The I / Q demodulation channel employs a dual-channel symmetrical hardware structure. In the digital domain, a least-squares algorithm corrects amplitude and phase imbalance, ensuring an image rejection ratio ≥40dB. The baseband equivalent noise bandwidth (ENBW) is set to 20-100Hz. Under a data averaging duration of 200ms, the phase measurement error is ≤0.5°, and the complex impedance amplitude error is ≤3%.

[0138] In addition, the front end is equipped with a multi-stage limiting circuit consisting of a gas discharge tube, a TVS diode, and a clamping diode to resist high-voltage arcing surges; the minimum creepage distance between the probe and the high-voltage conductor is configured according to the station standard; the probe bracket and the system housing adopt a single-point star grounding (grounding resistance ≤ 1Ω to avoid parasitic loops).

[0139] In some embodiments, the device and / or method can be applied to general non-contact measurement scenarios, adapting to non-contact measurements using most shortwave transmitter output matching networks (such as π-type networks, bandpass cavities), as specifically designed below: The main electrode of the capacitance sampler (E probe) uses a 12×8mm aluminum / copper sheet (the copper sheet is suitable for high-current scenarios), and is installed with a 6–10mm gap from the busbar under test. Guard electrodes surround three sides of the main electrode, with their slots orthogonal to the direction of the busbar current (aimed at reducing parasitic coupling, aligning with the "optimization of shielding slot direction" design concept in the specification). The guard circuit employs in-phase bootstrap technology, following the main electrode potential to reduce the coupling capacitance C. c Sensitivity to installation distance; the buffer stage adopts a high input impedance broadband amplification architecture (JFET input stage), and the equivalent input capacitance is no more than 3pF (to avoid disturbing the network under test).

[0140] The magnetic core of the magnetic ring sampler (H probe) is made of openable NiZn ferrite (which has low loss characteristics in the HF band and a permeability μ≈10). 4 Secondary side turns N = 10 turns; configured with a 50Ω / 100Ω switchable load (switched via relay to adapt to different current amplitudes), and a 33pF / / 1kΩ RC band-limited network in parallel (used to suppress high-frequency parasitic signals with frequencies greater than 50MHz).

[0141] The reference phase and demodulation are achieved by picking up the reference phase with a near-field small antenna with a length not exceeding 50mm, stabilizing the amplitude and controlling the power through a "Schottky limiting pair + attenuation network", dividing the frequency to 5MHz (this is the optimal locking frequency of the PLL), and then outputting the LO through a digital PLL phase-locked loop; the I / Q demodulation channels share the LO to ensure phase synchronization, and imbalance correction is achieved in the digital domain by estimating the amplitude mismatch coefficient and phase offset, with an image rejection of not less than 40dB.

[0142] During initial calibration, 20–50 frames were acquired under both "approximate matching" and "slight mistuning" conditions. s ,I s The data, combined with the reference impedance measured by a standard impedance analyzer, is used to extrapolate the true {V,I}, and the coupling matrix is ​​solved using the least squares algorithm. The initial value; during operation, a perturbation is triggered every 10–30 minutes (applying a minimum step of 1–2 LSB to the tuner actuator, which will not cause program perception), and the observed output change Δy = {ΔV} s ,ΔI s} and the known impedance increment Δx={ΔV,ΔI}, through Iterative update

[0143] Output complex impedance amplitude |Z|, resistance component R, reactance component X, phase difference φ, and quality indicators such as SNR, model residual, and calibration confidence; generate tuning suggestions when |X| increases or |φ|>10°; trigger protection alarm when the residual is greater than 5% of full scale and the harmonic amplitude rises more than 10%.

[0144] The front end is equipped with multi-stage protection consisting of a gas discharge tube, a TVS diode, and a clamping diode, enabling it to withstand surges of at least 2kV (1.2 / 50μs) without damage; it maintains a creepage distance of at least 8mm from high-voltage components, and the insulation components withstand temperatures no lower than 105℃; the outer casing has a cross-sectional area of ​​at least 2.5mm². 2 A single-point star grounding is achieved using copper strips. Under conditions of 10kHz resolution bandwidth (RBW) and 200ms data averaging, the standard deviation of phase measurement is no greater than 0.3–0.5°, and the complex impedance amplitude error after calibration does not exceed 3%.

[0145] In some embodiments, the apparatus and / or method can be applied to static tuning and acceptance testing scenarios, adapting to the need for "high-precision measurement," and suitable for transmitter static tuning (impedance calibration when no program is being broadcast) and new equipment acceptance testing (verifying the compliance of output network impedance). The specific design is as follows: The baseband equivalent noise bandwidth (ENBW) is reduced to 10–20Hz (narrowband filtering enhances noise suppression, consistent with the "baseband integration to suppress noise" approach), and the integration time is extended to 1–2s (further improving accuracy through time-domain averaging). The digital PLL tracking bandwidth is reduced to below 50Hz to reduce environmental noise disturbances to the local oscillator phase, ensuring LO phase jitter ≤0.05° / ms. Phase measurement accuracy is better than ±0.5°, and complex impedance amplitude error ≤2%; however, dynamic response is reduced, with a step response time ≥1s, making it unsuitable for rapid detuning scenarios.

[0146] In some embodiments, the device and / or method can be applied to fast tuning scenarios in DRM mode, adapting to real-time impedance tracking of OFDM signals, as specifically designed below: The baseband equivalent noise bandwidth (ENBW) is increased to 200–300Hz (widening the bandwidth to accelerate response), coupled with a wavelet threshold adaptive denoising algorithm (suppressing noise over a wide bandwidth); the ADC sampling rate is increased to 500kS / s (I / Q imbalance correction uses a real-time recursive algorithm to avoid static correction lag). Performance and scenario adaptability: dynamic response time ≤100ms, meeting the fast tuning closed-loop requirements of DRM mode; phase jitter is slightly increased, but still within the allowable range, enabling real-time impedance tracking under dynamic changes of OFDM subcarriers.

[0147] In summary, this application has at least the following advantages over related technologies: 1. Significantly reduced high-voltage safety risks: Related technologies require the voltage divider chain and sampling resistor to be directly connected to the high-voltage radio frequency node of the transmitter (such as the anode circuit of the final stage electron tube), which poses potential risks such as device breakdown; This application uses a dual-probe non-contact architecture of "E-field capacitive coupling + H-field electromagnetic induction", which does not form an electrical connection with the high-voltage node throughout the process, avoiding high-voltage safety hazards from the sampling source, and is fully applicable to the high-voltage and high-current working conditions of high-power shortwave transmitters.

[0148] 2. Measurement accuracy meets high-precision tuning requirements: Related technologies suffer from poor phase measurement accuracy due to indirect calculation of the final stage current, failure to handle E / H channel cross-coupling, and I / Q demodulation amplitude-phase imbalance. This application eliminates channel parasitic interference by using a matrix decoupling algorithm, calibrates I / Q amplitude-phase imbalance in the digital domain, and directly obtains the vector relationship between voltage and current, significantly improving phase measurement accuracy and meeting the stringent phase accuracy requirements of high-power shortwave transmitter tuning.

[0149] 3. No interference from the network under test, ensuring transmission performance: Related technologies require the insertion of voltage divider capacitors or sampling resistors in the transmitter circuit, introducing parasitic parameters and insertion loss, which disrupts the original impedance matching and resonance characteristics; This application optimizes the high-impedance buffer design (low static equivalent input capacitance) of the E probe and the open-close structure of the H probe to achieve "near-zero insertion effect", without changing the characteristics of the transmitter's final stage circuit, and without affecting the RF signal transmission and transmission efficiency.

[0150] 4. Enhanced reliability under abnormal operating conditions: Related technologies lack effective protection under abnormal operating conditions such as high-voltage arcing and high-power pulses, which can easily lead to the failure of phase detection circuits. This application designs a multi-level surge protection structure at the front end, and monitors abnormal states such as phase change and harmonic exceedance, triggering graded protection actions such as alarms, power reduction, and emergency backoff, so that it can still operate stably under complex operating conditions.

[0151] 5. Good long-term stability and reduced operation and maintenance costs: Related technologies rely on high-Q capacitors, which are easily affected by temperature, humidity and device aging, and require frequent manual calibration. This application uses temperature-compensated phase-locked loop elements and a non-stop broadcast perturbation self-calibration mechanism to automatically offset parameter drift, reduce manual intervention, maintain measurement accuracy for a long time, and reduce operation and maintenance costs and broadcast downtime risks.

[0152] 6. Improved anti-modulation interference capability of phase measurement: Related technologies are subject to interference from program content under modulation modes such as AM / DRM, resulting in large fluctuations in phase measurement. This application achieves modulation desensitization through adjustable equivalent noise bandwidth and mode-specific integration window design, which reduces the impact of program content on carrier frequency phase estimation, reduces phase measurement fluctuations, avoids malfunction of the tuning control unit, and ensures stable operation of the transmitter.

[0153] 7. Stable phase reference and rapid recovery after lock-up: Related technologies use a single-loop phase-locked architecture, which is prone to phase reference drift and requires manual intervention after lock-up. This application adopts a dual-loop phase-locked architecture of "capture loop + tracking loop" to reduce local oscillator phase jitter, and designs a lock-up detection and rapid re-acquisition mechanism to ensure the continuous and reliable phase reference, avoid demodulation data loss, and ensure measurement continuity.

[0154] The above are all preferred embodiments of this application, and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the structure, shape and principle of this application should be covered within the scope of protection of this application.

Claims

1. A magnetoelectric dual-field non-contact vector phase detector and impedance estimation device, characterized in that, include: The magnetoelectric dual-field non-contact sensing module (10) includes an electric field coupled voltage probe (11) and a magnetic field induced current probe (12). The electric field coupled voltage probe (11) is used for non-contact induction sampling of radio frequency voltage signals, and the magnetic field induced current probe (12) is used for non-contact induction sampling of radio frequency current signals. The reference phase-locked link module (30) is used to extract the synchronous local oscillator signal from the signal under test; The IQ demodulation and calibration module (40) is connected to the magnetoelectric dual-field non-contact sensing module (10) and the reference phase-locked link module (30) to perform phase-coordinate zero-IF demodulation of the radio frequency voltage signal and the radio frequency current signal using the synchronous local oscillator signal, and to perform I / Q channel mismatch correction to generate a baseband analog signal. The sampling and digital processing module (50) is connected to the IQ demodulation and calibration module (40) and is used to convert the baseband analog signal into a digital baseband signal and perform modulation and desensitization processing on the digital baseband signal to generate a digital baseband component for phase detection. The calibration and self-calibration module (60) is used to construct an initial coupling matrix and update the initial coupling matrix using a recursive least squares algorithm to generate a corrected coupling matrix for eliminating parameter drift. Impedance and phase calculation module (70) is connected to the sampling and digital processing module (50) and the calibration and self-calibration module (60). It is used to eliminate cross-coupling of the digital baseband components using the modified coupling matrix, and reconstruct the real voltage complex and the real current complex. It also calculates impedance parameters and quality indicators based on the real voltage complex and the real current complex.

2. The magnetoelectric dual-field non-contact vector phase detector and impedance estimation device according to claim 1, characterized in that, Also includes: The front-end protection and band-limiting module (20) is located between the magnetoelectric dual-field non-contact sensing module (10) and the reference phase-locked link module (30) for graded protection and selective filtering of the radio frequency voltage signal and the radio frequency current signal; The front-end protection and band-limiting module (20) includes an electric field channel protection and band-limiting unit (21) and a magnetic field channel protection and band-limiting unit (22). The electric field channel protection and band-limiting unit (21) is disposed between the electric field coupled voltage probe (11) and the reference phase-locked link module (30), and includes a high-resistance buffer element, a TVS clamping element and a π-type band-limiting element arranged in sequence. The magnetic field channel protection and band-limiting unit (22) includes a switchable damping element and an RC band-limiting network.

3. The magnetoelectric dual-field non-contact vector phase detector and impedance estimation device according to claim 1, characterized in that, The reference phase-locked link module (30) includes a near-field small loop pickup unit (31), a Schottky limiting unit (32), a programmable prescaler unit (33), and a digital PLL phase-locked unit (34) connected in sequence; wherein, the digital PLL phase-locked unit (34) adopts a dual-ring architecture of a capture ring and a tracking ring.

4. The magnetoelectric dual-field non-contact vector phase detector and impedance estimation device according to any one of claims 1-3, characterized in that, The IQ demodulation and calibration module (40) is used to perform coherent zero-IF demodulation on the RF voltage signal and the RF current signal using the synchronous local oscillator signal, and to perform I / Q channel mismatch correction to generate a baseband analog signal; wherein, the IQ demodulation and calibration module (40) is configured to: mix the RF voltage signal with the in-phase component and the quadrature component of the synchronous local oscillator signal respectively to generate a baseband in-phase component and a baseband quadrature component of the voltage; and mix the RF current signal with the in-phase component and the quadrature component of the synchronous local oscillator signal respectively to generate an RF voltage signal. The baseband in-phase component of the voltage and the baseband quadrature component of the current are generated. Low-pass filtering is applied to the baseband in-phase component of the voltage, the baseband quadrature component of the voltage, the baseband in-phase component of the current, and the baseband quadrature component of the current. I / Q channel mismatch correction is then applied to the filtered and denoised baseband in-phase component of the voltage, the baseband quadrature component of the voltage, the baseband in-phase component of the current, and the baseband quadrature component of the current to generate the baseband analog signal. The baseband analog signal includes the voltage in-phase component signal, the voltage quadrature component signal, the current in-phase component signal, and the current quadrature component signal.

5. The magnetoelectric dual-field non-contact vector phase detector and impedance estimation device according to claim 4, characterized in that, The sampling and digital processing module (50) is used to convert the baseband analog signal into a digital baseband signal and perform modulation desensitization processing on the digital baseband signal to generate a digital baseband component for phase detection; wherein, the sampling and digital processing module (50) is configured to: use a co-source clock control to perform synchronous analog-to-digital conversion on the voltage in-phase component signal, the voltage quadrature component signal, the current in-phase component signal and the current quadrature component signal to generate a digital baseband signal; set the baseband equivalent noise bandwidth and digital integration window width according to the transmitter's operating mode; and process the digital baseband signal through a long integration or digital averaging algorithm to reduce the disturbance of the program modulation component on the carrier frequency phase estimation to generate a digital baseband component for phase detection.

6. The magnetoelectric dual-field non-contact vector phase detector and impedance estimation device according to claim 5, characterized in that, The calibration and self-calibration module (60) is used to construct an initial coupling matrix and update the initial coupling matrix using a recursive least squares algorithm to generate a corrected coupling matrix for eliminating parameter drift; wherein, the calibration and self-calibration module (60) is configured to: construct an initial coupling matrix during transmitter offline; introduce imperceptible perturbation excitation and collect perturbation data during transmitter operation; and use a recursive least squares algorithm to iteratively update the initial coupling matrix using the perturbation data to generate the corrected coupling matrix.

7. A non-contact vector phase detection and impedance estimation method for a dual-field magnetoelectric system, characterized in that, Includes the following steps: S1. Radio frequency voltage and radio frequency current signals are extracted through non-contact induction sampling; S2. Extract the synchronous local oscillator signal from the measured signal; S3. The synchronous local oscillator signal is used to perform coherent zero-IF demodulation on the radio frequency voltage signal and the radio frequency current signal, and I / Q channel mismatch correction is performed to generate a baseband analog signal. S4. Convert the baseband analog signal into a digital baseband signal, and perform modulation and desensitization processing on the digital baseband signal to generate a digital baseband component for phase detection. S5. Construct an initial coupling matrix and update the initial coupling matrix using a recursive least squares algorithm to generate a corrected coupling matrix for eliminating parameter drift. S6. Use the modified coupling matrix to eliminate cross-coupling of the digital baseband components and reconstruct the true voltage complex and the true current complex. S7. Calculate the impedance parameters and quality indicators based on the actual voltage complex and the actual current complex.

8. The magnetoelectric dual-field non-contact vector phase detection and impedance estimation method according to claim 7, characterized in that, Step S3 includes: mixing the radio frequency voltage signal with the in-phase component and quadrature component of the synchronous local oscillator signal respectively to generate a baseband in-phase component and a baseband quadrature component of the voltage; mixing the radio frequency current signal with the in-phase component and quadrature component of the synchronous local oscillator signal respectively to generate a baseband in-phase component and a baseband quadrature component of the current; performing low-pass filtering on the baseband in-phase component, the baseband quadrature component, the baseband in-phase component, and the baseband quadrature component of the voltage; and performing I / Q channel mismatch correction on the filtered and denoised baseband in-phase component, the baseband quadrature component, the baseband in-phase component, and the baseband quadrature component of the current to generate the baseband analog signal, wherein the baseband analog signal includes a voltage in-phase component signal, a voltage quadrature component signal, a current in-phase component signal, and a current quadrature component signal.

9. The magnetoelectric dual-field non-contact vector phase detection and impedance estimation method according to claim 8, characterized in that, Step S4 includes: using a co-source clock control to perform synchronous analog-to-digital conversion on the voltage in-phase component signal, the voltage quadrature component signal, the current in-phase component signal, and the current quadrature component signal to generate a digital baseband signal; setting the baseband equivalent noise bandwidth and the digital integration window width according to the transmitter's operating mode; and processing the digital baseband signal through a long integration or digital averaging algorithm to reduce the disturbance of the program modulation component on the carrier frequency phase estimation to generate a digital baseband component for phase detection.

10. The magnetoelectric dual-field non-contact vector phase detection and impedance estimation method according to claim 9, characterized in that, Step S5 includes: constructing an initial coupling matrix during transmitter offline operation; introducing imperceptible perturbation excitation and collecting perturbation data during transmitter operation; and using a recursive least squares algorithm to iteratively update the initial coupling matrix using the perturbation data to generate the corrected coupling matrix.

Images