Noise Reduction Topology and Method for Amplification Power Supply of Multi-Level Common-Mode Suppression Medical Precision Superconducting Quantum Interferometer

By employing a multi-level common-mode suppression structure, adaptive active common-mode compensation, and frequency domain segmentation noise reduction, the problems of insufficient common-mode noise suppression capability and incomplete frequency band of SQUID amplifier power supply are solved, achieving full-band system noise reduction and load adaptability, and improving noise suppression effect.

CN123122202APending Publication Date: 2026-11-17NANJING JIESI AUTOMOBILE TECH CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202611250777.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-09-02
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing SQUID amplifier power supply noise reduction technology lacks adaptive active common-mode compensation capability, has a single or incomplete noise reduction frequency band, fixed noise reduction parameters that cannot be dynamically adjusted according to load changes, and poor common-mode noise suppression effect.

Method used

A multi-level common-mode suppression structure is adopted, including an adaptive active common-mode compensation unit, a frequency domain segmentation multi-level noise reduction unit, a load-sensing feedforward prediction unit, and a temperature-adaptive bias compensation unit. Through active common-mode compensation, frequency domain segmentation noise reduction, and load prediction adjustment, noise reduction is achieved across the entire frequency band.

Benefits of technology

It improves common-mode noise suppression capability by more than 20dB, achieves full-band system noise reduction from DC to high frequency, adapts to scenarios with rapidly changing loads, and overcomes the shortcomings of traditional solutions.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123122202A_ABST
    Figure CN123122202A_ABST
Patent Text Reader

Abstract

This invention relates to the field of power supply noise reduction technology, specifically to a noise reduction topology and method for a multi-stage common-mode suppression power supply for a precision superconducting quantum interference device (QFID) for medical applications. The topology includes a power input terminal; an adaptive active common-mode compensation unit for detecting common-mode noise components in the power output and generating an active compensation signal with the opposite phase to the common-mode noise components to cancel the common-mode noise; and a frequency-domain segmentation multi-stage noise reduction unit for segmenting the power supply noise into at least two frequency bands and performing corresponding noise reduction processing on each band. By using adaptive active common-mode compensation instead of traditional passive common-mode suppression, the compensation coefficient can be adaptively adjusted according to load changes. By using a frequency-domain segmentation parallel noise reduction architecture instead of traditional serial multi-stage filtering, each noise reduction stage is optimized for a specific frequency band, achieving full-band system noise reduction from DC to high frequencies.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of power supply noise reduction technology, specifically to a noise reduction topology and method for amplification power supply in a multi-level common-mode suppression medical precision superconducting quantum interference instrument. Background Technology

[0002] Superconducting quantum interference devices (SQUIDs) are currently the most sensitive magnetic flux sensors known, and they have irreplaceable application value in fields such as biomagnetic imaging, geophysical exploration, and precision medical measurement. SQUID systems are extremely sensitive to power supply noise—any power fluctuation will couple into magnetic flux noise, directly affecting measurement accuracy and signal-to-noise ratio.

[0003] Several technical solutions for noise reduction in SQUID power supplies have been proposed in the prior art. Patent application CN104345758A discloses a bias amplifier circuit for a superconducting quantum interference device (SQUID). This circuit controls the connection mode of the preamplifier input by switching a switch, allowing the SQUID device to selectively operate in constant current bias mode or constant voltage bias mode, and forms a negative feedback amplification structure through a feedback resistor. However, this solution only achieves noise suppression through a single feedback resistor and does not involve active compensation for common-mode noise or multi-band segmentation processing.

[0004] Patent application CN102426343A discloses a readout circuit based on SQUID bias voltage reversal and a method for suppressing low-frequency noise. This scheme achieves bias reversal through a bias reversal circuit, thereby suppressing the generation of low-frequency noise. However, the above scheme mainly targets the suppression of low-frequency 1 / f noise and does not involve comprehensive noise reduction across the entire frequency band from DC to high frequencies, nor does it consider active compensation for common-mode noise.

[0005] In summary, the existing SQUID amplifier power supply noise reduction technology has the following shortcomings: (1) Common mode suppression mainly relies on passive means such as differential structure and common mode choke, and lacks adaptive active common mode compensation capability; (2) The noise reduction frequency band is single or incomplete, and lacks a systematic noise reduction architecture for the entire frequency band; (3) The noise reduction parameters are fixed and do not have the ability to dynamically adjust according to load changes. Summary of the Invention

[0006] In order to overcome the deficiencies in the prior art, the present invention aims to provide a noise reduction topology and method for amplification power supply of a multi-level common-mode suppression medical precision superconducting quantum interference instrument, so as to solve the problems mentioned in the background art.

[0007] To achieve the above objectives, on the one hand, the present invention provides a noise reduction topology for a multi-level common-mode suppression amplification power supply for a medical precision superconducting quantum interference instrument, including a power input terminal for connecting to an external DC power supply; An adaptive active common-mode compensation unit, whose input terminal is coupled to the power input terminal, is used to detect the common-mode noise component in the power output and generate an active compensation signal that is opposite in phase to the common-mode noise component in order to cancel the common-mode noise. A frequency domain segmentation multi-level noise reduction unit, whose input is coupled to the output of the adaptive active common-mode compensation unit, is used to segment power supply noise into at least two frequency bands in the frequency domain and perform corresponding noise reduction processing on each frequency band. The load-sensing feedforward prediction unit has its input end coupled to the load end. It is used to collect load current change signals in real time, predict the noise spectrum characteristics that will be generated based on the load current change signals, and adjust the working parameters of the frequency domain segmentation multi-stage noise reduction unit in advance. The temperature adaptive bias compensation unit has its input terminal coupled to a temperature sensor to monitor the ambient temperature in real time and dynamically adjust the power supply bias parameters according to temperature changes. And a power output terminal, coupled to the output terminal of the frequency domain segmentation multi-stage noise reduction unit, for outputting a stable power supply after noise reduction to the superconducting quantum interference instrument amplifier circuit.

[0008] As a further improvement to this technical solution, the adaptive active common-mode compensation unit includes a common-mode noise detection subunit, used to extract the common-mode noise component in the power supply output; A high-speed error amplifier is used to generate a gain-adjustable amplified signal based on the common-mode noise component. And a compensation signal injection subunit for injecting the amplified signal into the power supply circuit in a manner opposite to the phase of the common-mode noise.

[0009] As a further improvement to this technical solution, the frequency domain segmentation multi-level noise reduction unit includes an ultra-low frequency noise reduction subunit for processing noise in the DC to 1Hz frequency band. The ultra-low frequency noise reduction subunit includes an ultra-low noise linear regulator and an active feedback circuit. A low-frequency noise reduction subunit is used to process noise in the 1Hz to 1kHz frequency band. The low-frequency noise reduction subunit includes an active common-mode compensation circuit and a differential drive circuit. An intermediate frequency noise reduction subunit is used to process noise in the frequency band from 1 kHz to 100 kHz. The intermediate frequency noise reduction subunit includes a multi-stage passive LC notch filter and an electromagnetic shielding layer. A high-frequency noise reduction subunit is used to process noise in the frequency band greater than 100kHz. The high-frequency noise reduction subunit includes a distributed bypass capacitor array and a ferrite absorption element.

[0010] As a further improvement to this technical solution, a frequency domain crossover optimization module is provided between the ultra-low frequency noise reduction subunit, the low frequency noise reduction subunit, the mid frequency noise reduction subunit, and the high frequency noise reduction subunit. The frequency domain crossover optimization module is used to dynamically adjust the frequency band boundary points between each subunit according to the real-time noise spectrum distribution, so as to achieve seamless connection of noise suppression.

[0011] As a further improvement to this technical solution, the load sensing feedforward prediction unit includes a load current sampling subunit for real-time acquisition of load current. The digital signal processor has a built-in digital twin model of the load characteristics of the superconducting quantum interference device amplifier circuit, which is used to predict the noise spectrum characteristics based on the load current variation trend. And a feedforward parameter adjustment subunit, used to adjust the gain and filtering parameters of each subunit in the frequency domain segmentation multi-level noise reduction unit in advance according to the prediction results.

[0012] As a further improvement to this technical solution, the temperature adaptive bias compensation unit includes a temperature sensor interface for receiving temperature signals from a temperature sensor. A high-order temperature compensation network, comprising at least three compensation branches with different temperature compensation coefficients, with each compensation branch connected in parallel; And a bias parameter adjustment subunit, used to dynamically adjust the bias voltage and bias current of the power reference source according to the temperature signal and the output of the higher-order temperature compensation network.

[0013] As a further improvement to this technical solution, each compensation branch of the high-order temperature compensation network sets independent compensation coefficients for three temperature points: 4.2K in the liquid helium temperature range, 77K in the liquid nitrogen temperature range, and 300K in the room temperature range. Continuous temperature compensation across the entire temperature range is achieved through piecewise linear interpolation.

[0014] As a further improvement to this technical solution, a common-mode feedback control loop is also included. The input terminal of the common-mode feedback control loop is coupled to the power supply output terminal, and the output terminal of the common-mode feedback control loop is coupled to the adaptive active common-mode compensation unit. This loop is used to feed back the residual common-mode noise from the power supply output terminal to the adaptive active common-mode compensation unit to achieve closed-loop control.

[0015] On the other hand, the present invention provides a noise reduction method for a noise reduction topology of a multi-level common-mode suppression amplification power supply for a medical precision superconducting quantum interference device. Using the aforementioned noise reduction topology of the multi-level common-mode suppression amplification power supply for a medical precision superconducting quantum interference device, the method includes the following steps: S1. Connect an external DC power supply, detect the common-mode noise component in the power supply output through an adaptive active common-mode compensation unit, and generate an active compensation signal with the opposite phase to the common-mode noise component and inject it into the power supply circuit to cancel the common-mode noise. S2. Divide the power supply noise after common-mode compensation into at least two frequency bands according to the frequency domain, and perform corresponding noise reduction processing on each frequency band, including active feedback noise reduction in the ultra-low frequency band, active common-mode compensation noise reduction in the low frequency band, passive LC notch noise reduction in the mid-frequency band, and distributed bypass noise reduction in the high frequency band. S3. Real-time acquisition of load current change signals, prediction of the upcoming noise spectrum characteristics based on the load current change signals using a digital twin model, and adjustment of the noise reduction processing parameters for each frequency band in advance. S4. Real-time monitoring of ambient temperature and dynamic adjustment of power supply bias parameters through a high-order temperature compensation network based on temperature changes, so as to optimize temperature compensation and noise suppression in a coordinated manner. S5 outputs a stable power supply after noise reduction to the superconducting quantum interference instrument amplifier circuit.

[0016] As a further improvement to this technical solution, step S2 also includes a dynamic optimization step for frequency domain crossover points: real-time acquisition of the noise spectrum distribution at the power output terminal, and dynamic adjustment of the boundary frequency between each frequency band according to the noise spectrum distribution, so that the noise reduction processing of each frequency band can be seamlessly connected. Step S3 further includes a feedforward-feedback composite control step: weighting and fusing the feedforward prediction result with the real-time feedback signal at the output end to generate the final control parameters.

[0017] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. The noise reduction topology and method of the multi-level common-mode suppression medical precision superconducting quantum interference instrument amplification power supply adopts adaptive active common-mode compensation instead of traditional passive common-mode suppression. The active compensation improves the common-mode noise suppression capability by more than 20dB compared with passive devices, and can adaptively adjust the compensation coefficient according to the load change, overcoming the defect of the passive solution where the common-mode rejection ratio decreases when the load changes.

[0018] 2. The noise reduction topology and method of the multi-stage common-mode suppression medical precision superconducting quantum interference instrument amplification power supply adopts a frequency domain segmentation parallel noise reduction architecture instead of the traditional series multi-stage filtering. Each noise reduction stage is optimized for a specific frequency band, avoiding the problem of overlapping or omission of frequency bands in series filtering, and realizing full-band system noise reduction from DC to high frequency.

[0019] 3. The noise reduction topology and method of the multi-level common-mode suppression medical precision superconducting quantum interference instrument amplification power supply, by introducing a load-sensing feedforward prediction mechanism to replace simple feedback noise reduction, and by predicting noise in advance and adjusting parameters in advance through a digital twin model, overcomes the inherent response delay of feedback control, and is particularly suitable for scenarios with rapidly changing loads in SQUID systems. Attached Figure Description

[0020] The accompanying drawings described herein are for illustrative purposes only.

[0021] Figure 1 This is a schematic diagram of the overall control principle of the noise reduction topology of the present invention.

[0022] Figure 2 This is the overall control flowchart of the noise reduction method for the noise reduction topology of the present invention.

[0023] Figure 3 This is a control flowchart of the adaptive active common-mode compensation unit in this invention.

[0024] Figure 4 This is a control flowchart of the frequency domain segmentation multi-stage noise reduction unit in this invention.

[0025] Figure 5 This is a control flowchart of the load sensing feedforward prediction unit in this invention. Detailed Implementation

[0026] Under the guidance of this invention, any possible variations of this invention should be considered within the scope of this invention.

[0027] Example 1: Overall Topology like Figure 1 As shown, the noise reduction topology of the multi-level common-mode suppression medical precision superconducting quantum interference instrument amplification power supply in this embodiment includes: The power input terminal V_IN is used to connect an external DC power supply, with an input voltage range of 5V to 24V.

[0028] An adaptive active common-mode compensation unit has its input coupled to the power input terminal V_IN. This unit includes a common-mode noise detection subunit, a high-speed error amplifier, and a compensation signal injection subunit. The common-mode noise detection subunit extracts the common-mode noise component from the power output through a common-mode choke and a differential amplifier; the high-speed error amplifier generates an adjustable-gain amplified signal based on the common-mode noise component, and its gain coefficient is dynamically adjusted by the adaptive control module based on real-time detection results; the compensation signal injection subunit injects the amplified signal into the power circuit in a manner opposite in phase to the common-mode noise through a compensation transformer or compensation capacitor, achieving active common-mode cancellation. This active common-mode compensation scheme is based on the current detection-current compensation principle, but this invention further introduces adaptive gain control, enabling it to automatically adjust the compensation intensity according to load and temperature changes.

[0029] A frequency-domain segmentation multi-stage noise reduction unit has its input coupled to the output of the adaptive active common-mode compensation unit. This unit includes four parallel branches: The ultra-low frequency noise reduction subunit uses the LT3042 series ultra-low noise linear regulator, combined with an active feedback circuit, to specifically handle noise in the DC to 1Hz frequency band, with an output noise density as low as 0.8nV / √Hz.

[0030] The low-frequency noise reduction subunit employs an active common-mode compensation circuit and a differential drive circuit, specifically designed to handle noise in the 1Hz to 1kHz frequency band. This subunit includes an active filter composed of a high-speed operational amplifier, with a common-mode rejection ratio (CMRR) of no less than 120dB at 1kHz.

[0031] The intermediate frequency noise reduction subunit employs a multi-stage passive LC notch filter in conjunction with an electromagnetic shielding layer, specifically designed to handle noise in the 1kHz to 100kHz frequency band. The center frequency of the notch filter can be dynamically adjusted based on the actual noise spectrum using a frequency domain crossover optimization module.

[0032] The high-frequency noise reduction subunit employs a distributed bypass capacitor array and ferrite absorption elements to specifically handle noise in the frequency band above 100kHz. The bypass capacitor array contains multiple ceramic capacitors of different capacitance values, such as 100pF, 1nF, 10nF, and 100nF connected in parallel, to cover different high-frequency points.

[0033] The frequency domain crossover optimization module collects the noise spectrum distribution at the power output terminal in real time, analyzes the noise spectrum characteristics through Fast Fourier Transform (FFT), and dynamically adjusts the frequency band boundary frequency between each sub-unit based on the analysis results. For example, the boundary points of 1Hz / 1kHz / 100kHz are finely adjusted according to the actual noise distribution to achieve seamless connection of noise reduction processing for each frequency band.

[0034] The load-sensing feedforward prediction unit has its input coupled to the load, i.e., the SQUID amplifier circuit. This unit includes a load current sampling subunit, a digital signal processor, and a feedforward parameter adjustment subunit. The load current sampling subunit acquires the load current signal in real time using a high-precision current sensor, such as a Hall effect sensor or a sampling resistor + differential amplifier. The digital signal processor has a built-in digital twin model of the load characteristics of the SQUID amplifier circuit. This model is pre-established through system identification and can predict the upcoming noise spectrum characteristics based on the load current change trend. The feedforward parameter adjustment subunit sends parameter adjustment commands to the frequency domain segmentation multi-stage noise reduction unit in advance based on the prediction results, adjusting the gain and filtering parameters of each subunit in advance.

[0035] The temperature adaptive bias compensation unit, with its input coupled to a temperature sensor, is configured within the SQUID operating environment. This unit includes a temperature sensor interface, a high-order temperature compensation network, and a bias parameter adjustment subunit. The high-order temperature compensation network comprises three compensation branches with different temperature compensation coefficients, connected in parallel. These branches set independent compensation coefficients for three temperature points: 4.2K in the liquid helium range, 77K in the liquid nitrogen range, and 300K at room temperature. Continuous temperature compensation across the entire temperature range is achieved through piecewise linear interpolation. The bias parameter adjustment subunit dynamically adjusts the bias voltage and bias current of the power reference source based on the temperature signal and the output of the high-order temperature compensation network, ensuring that the SQUID achieves optimal bias conditions at different operating temperatures.

[0036] The common-mode feedback control loop has its input coupled to the power supply output V_OUT and its output coupled to the adaptive active common-mode compensation unit. This loop feeds back the residual common-mode noise from the power supply output to the adaptive active common-mode compensation unit, forming a closed-loop control with the feedforward compensation, further improving the common-mode rejection accuracy.

[0037] The power output terminal V_OUT is coupled to the output terminal of the frequency domain segmentation multi-stage noise reduction unit, used to output a stable, noise-reduced power supply to the SQUID amplifier circuit. An output filter capacitor, such as a 10μF tantalum capacitor in parallel with a 0.1μF ceramic capacitor, is provided at the output terminal to further smooth the output voltage.

[0038] In this embodiment, differential transmission is used for signal transmission between units to minimize common-mode noise introduced during signal transmission.

[0039] Example 2: Noise Reduction Method Flowchart like Figures 2-5 As shown, the noise reduction method of the multi-level common-mode suppression medical precision superconducting quantum interference device amplification power supply in this embodiment uses the aforementioned multi-level common-mode suppression medical precision superconducting quantum interference device amplification power supply noise reduction topology and includes the following steps: Step S1: Adaptive Active Common Mode Compensation After connecting an external DC power supply, the common-mode noise component in the power supply output is first detected by an adaptive active common-mode compensation unit. This step specifically includes: S1-1: Extract the common-mode noise signal V_cm from the power supply output using a common-mode choke and a differential amplifier; S1-2: Input V_cm into the high-speed error amplifier. The error amplifier generates an amplified signal V_comp=G×V_cm based on the preset gain coefficient G, with the initial value determined by the system calibration. S1-3: V_comp is injected into the power supply circuit in reverse mode through a compensation transformer or compensation capacitor to achieve active cancellation of common-mode noise; S1-4: Monitor the residual common-mode noise at the power supply output in real time, and adjust the gain coefficient G using an adaptive algorithm to minimize the residual common-mode noise. The adaptive algorithm can be either the Least Mean Square (LMS) algorithm or the Recursive Least Squares (RLS) algorithm.

[0040] Step S2: Frequency Domain Segmentation and Multi-Stage Noise Reduction The power supply noise after common-mode compensation is divided into four frequency bands, and corresponding noise reduction processing is performed on each frequency band. This step specifically includes: S2-1: The power output signal is subjected to FFT analysis by the frequency domain crossover optimization module to obtain the real-time noise spectrum distribution; S2-2: Based on the noise spectrum distribution, determine the boundary frequencies of each frequency band: f1 is between DC and 1Hz, f2 is between 1Hz and 1kHz, and f3 is between 1kHz and 100kHz. S2-3: The ultra-low frequency noise reduction subunit performs ultra-low noise linear regulation and active feedback noise reduction on noise in the DC to f1 frequency band; S2-4: The low-frequency noise reduction subunit performs active common-mode compensation and differential drive noise reduction on the noise in the f1 to f2 frequency band; S2-5: The intermediate frequency noise reduction subunit performs multi-stage passive LC notch filtering and electromagnetic shielding noise reduction on the noise in the f2 to f3 frequency band; S2-6: The high-frequency noise reduction subunit performs distributed bypass and ferrite absorption noise reduction for noise in the frequency band greater than f3; S2-7: Real-time monitoring of residual noise after noise reduction in each frequency band, dynamic adjustment of the boundary point frequencies f1, f2, f3 and the parameters of each sub-unit to achieve seamless connection of noise reduction processing in each frequency band.

[0041] Step S3: Load-aware feedforward prediction The system acquires load current change signals in real time, predicts the upcoming noise spectrum characteristics using a digital twin model, and adjusts the noise reduction parameters for each frequency band in advance. This step specifically includes: S3-1: The load current signal I_load(t) of the SQUID amplifier circuit is acquired in real time by a high-precision current sensor; S3-2: Input I_load(t) into the digital signal processor. The SQUID amplifier circuit digital twin model built into the processor predicts the noise spectrum characteristics in the future time interval Δt based on the current value and the current change rate dI_load / dt. Δt = 10μs to 100μs, which can be set according to the system response speed requirements. S3-3: Based on the prediction results, send parameter adjustment instructions to the frequency domain segmentation multi-stage noise reduction unit in advance, including: adjustment of the feedback coefficient of the ultra-low frequency noise reduction sub-unit, adjustment of the common mode compensation gain of the low frequency noise reduction sub-unit, adjustment of the notch frequency of the intermediate frequency noise reduction sub-unit, and switching of the bypass capacitor of the high frequency noise reduction sub-unit. S3-4: The feedforward prediction results are weighted and fused with the real-time feedback signal from the power supply output to generate the final control parameters. The weighting coefficients can be dynamically adjusted according to the actual operating conditions.

[0042] Step S4: Temperature Adaptive Bias Compensation The system monitors the ambient temperature in real time and dynamically adjusts the power supply bias parameters based on temperature changes using a high-order temperature compensation network. This step specifically includes: S4-1: Real-time acquisition of the temperature signal T of the SQUID's working environment via a temperature sensor, covering the entire temperature range from 4.2K to 300K; S4-2: Input T into the high-order temperature compensation network, which contains three compensation branches that set independent compensation coefficients α1, α2, and α3 for the three reference temperature points of 4.2K, 77K, and 300K respectively; S4-3: Based on the relationship between the current temperature T and the three reference temperature points, calculate the comprehensive compensation coefficient α(T) at the current temperature through piecewise linear interpolation; S4-4: Dynamically adjust the bias voltage V_bias and bias current I_bias of the power supply reference source according to α(T) so that SQUID can obtain the optimal bias state at different operating temperatures. S4-5: Coordinate the temperature compensation results with the parameters of each noise reduction unit; for example, when temperature changes cause the SQUID operating point to shift, synchronously adjust the operating parameters of the noise reduction unit to match the new operating point.

[0043] Step S5: Power Output After the above four steps, a stable, noise-reduced power supply is output to the SQUID amplifier circuit through the power output terminal. An output filter capacitor is provided at the output terminal to further smooth the output voltage waveform.

[0044] Example 3: Selection of Key Parameters In a preferred embodiment of the present invention, the key parameters of each unit are as follows: Adaptive active common-mode compensation unit: The high-speed error amplifier is an operational amplifier with a bandwidth of not less than 10MHz and a slew rate of not less than 20V / μs; the compensation transformer adopts a manganese-zinc ferrite high-frequency magnetic core with a turns ratio of 1:1 and an operating frequency covering 10Hz to 1MHz.

[0045] Ultra-low frequency noise reduction subunit: Uses ultra-low noise linear regulator chips, such as ADI's LT3042 or TI's TPS7A47, with an output voltage noise density of less than 0.8nV / √Hz, such as 10Hz to 100kHz; the loop bandwidth of the active feedback circuit is designed to be 10Hz.

[0046] Low-frequency noise reduction subunit: The core operational amplifier of the active common-mode compensation circuit has a CMRR of not less than 120dB (at 1kHz); the output impedance of the differential drive circuit is not higher than 0.1Ω.

[0047] Intermediate frequency noise reduction subunit: The LC notch filter adopts a third-order Butterworth structure with a notch depth of not less than 40dB; the electromagnetic shielding layer is made of permalloy (μ-metal) material with a thickness of not less than 0.5mm.

[0048] High-frequency noise reduction subunit: The distributed bypass capacitor array includes four capacitance levels: 100pF, 1nF, 10nF, and 100nF. The level can be switched via MOSFET switching according to the actual noise spectrum. The ferrite absorption element uses nickel-zinc ferrite beads with an impedance of not less than 600Ω (at 100MHz).

[0049] Digital Signal Processor: Select a DSP chip with floating-point operation capability, such as TI TMS320F28379D, with an ADC sampling rate of not less than 1MSPS and an FFT operation point count of not less than 1024 points.

[0050] Temperature sensor: Select a low-temperature suitable silicon diode temperature sensor, such as the LakeShore DT-670 series, with an accuracy of no less than ±0.5K in the temperature range of 4.2K to 300K.

[0051] Industrial applicability

[0052] This invention can be widely applied in SQUID applications with extremely high requirements for power supply noise, such as medical precision superconducting quantum interference devices, biomagnetic imaging systems, ultra-low field nuclear magnetic resonance systems, and geophysical exploration. Each unit of this invention can be implemented using existing mature analog / mixed-signal integrated circuit technology, exhibiting good industrial applicability and manufacturability.

[0053] The above embodiments are only for illustrating the technical concept and features of the present invention, and are intended to enable those skilled in the art to understand the content of the present invention and implement it accordingly. They should not be construed as limiting the scope of protection of the present invention. All equivalent changes or modifications made in accordance with the spirit and essence of the present invention should be covered within the scope of protection of the present invention.

Claims

1. A noise reduction topology for a multi-level common-mode suppression amplification power supply in a medical precision superconducting quantum interference instrument, characterized in that: Includes a power input terminal for connecting to an external DC power supply; An adaptive active common-mode compensation unit, whose input terminal is coupled to the power input terminal, is used to detect the common-mode noise component in the power output and generate an active compensation signal that is opposite in phase to the common-mode noise component in order to cancel the common-mode noise. A frequency domain segmentation multi-level noise reduction unit, whose input is coupled to the output of the adaptive active common-mode compensation unit, is used to segment power supply noise into at least two frequency bands in the frequency domain and perform corresponding noise reduction processing on each frequency band. The load-sensing feedforward prediction unit has its input end coupled to the load end. It is used to collect load current change signals in real time, predict the noise spectrum characteristics that will be generated based on the load current change signals, and adjust the working parameters of the frequency domain segmentation multi-stage noise reduction unit in advance. The temperature adaptive bias compensation unit has its input terminal coupled to a temperature sensor to monitor the ambient temperature in real time and dynamically adjust the power supply bias parameters according to temperature changes. And a power output terminal, coupled to the output terminal of the frequency domain segmentation multi-stage noise reduction unit, for outputting a stable power supply after noise reduction to the superconducting quantum interference instrument amplifier circuit.

2. The noise reduction topology of the multi-level common-mode suppression amplification power supply for a medical precision superconducting quantum interference instrument according to claim 1, characterized in that: The adaptive active common-mode compensation unit includes a common-mode noise detection subunit, used to extract the common-mode noise component in the power supply output; A high-speed error amplifier is used to generate a gain-adjustable amplified signal based on the common-mode noise component. And a compensation signal injection subunit for injecting the amplified signal into the power supply circuit in a manner opposite to the phase of the common-mode noise.

3. The noise reduction topology of the multi-level common-mode suppression amplification power supply for a medical precision superconducting quantum interference device according to claim 2, characterized in that: The frequency domain segmentation multi-stage noise reduction unit includes an ultra-low frequency noise reduction subunit for processing noise in the DC to 1Hz frequency band. The ultra-low frequency noise reduction subunit includes an ultra-low noise linear regulator and an active feedback circuit. A low-frequency noise reduction subunit is used to process noise in the 1Hz to 1kHz frequency band. The low-frequency noise reduction subunit includes an active common-mode compensation circuit and a differential drive circuit. An intermediate frequency noise reduction subunit is used to process noise in the frequency band from 1 kHz to 100 kHz. The intermediate frequency noise reduction subunit includes a multi-stage passive LC notch filter and an electromagnetic shielding layer. A high-frequency noise reduction subunit is used to process noise in the frequency band greater than 100kHz. The high-frequency noise reduction subunit includes a distributed bypass capacitor array and a ferrite absorption element.

4. The noise reduction topology of the multi-level common-mode suppression amplification power supply for a medical precision superconducting quantum interference device according to claim 3, characterized in that: A frequency domain crossover optimization module is provided between the ultra-low frequency noise reduction subunit, the low frequency noise reduction subunit, the mid frequency noise reduction subunit, and the high frequency noise reduction subunit. The frequency domain crossover optimization module is used to dynamically adjust the frequency band boundary points between each subunit according to the real-time noise spectrum distribution, so as to achieve seamless connection of noise suppression.

5. The noise reduction topology of the multi-level common-mode suppression amplification power supply for a medical precision superconducting quantum interference device according to claim 4, characterized in that: The load sensing feedforward prediction unit includes a load current sampling subunit for real-time acquisition of load current. The digital signal processor has a built-in digital twin model of the load characteristics of the superconducting quantum interference device amplifier circuit, which is used to predict the noise spectrum characteristics based on the load current variation trend. And a feedforward parameter adjustment subunit, used to adjust the gain and filtering parameters of each subunit in the frequency domain segmentation multi-level noise reduction unit in advance according to the prediction results.

6. The noise reduction topology of the multi-level common-mode suppression amplification power supply for a medical precision superconducting quantum interference device according to claim 5, characterized in that: The temperature adaptive bias compensation unit includes a temperature sensor interface for receiving temperature signals from a temperature sensor. A high-order temperature compensation network, comprising at least three compensation branches with different temperature compensation coefficients, with each compensation branch connected in parallel; And a bias parameter adjustment subunit, used to dynamically adjust the bias voltage and bias current of the power reference source according to the temperature signal and the output of the higher-order temperature compensation network.

7. The noise reduction topology of the multi-level common-mode suppression amplification power supply for a medical precision superconducting quantum interference device according to claim 6, characterized in that: Each compensation branch of the high-order temperature compensation network sets independent compensation coefficients for three temperature points: 4.2K in the liquid helium temperature range, 77K in the liquid nitrogen temperature range, and 300K in the room temperature range. Continuous temperature compensation across the entire temperature range is achieved through piecewise linear interpolation.

8. The noise reduction topology of the multi-level common-mode suppression amplification power supply for a medical precision superconducting quantum interference device according to claim 7, characterized in that: It also includes a common-mode feedback control loop, the input of which is coupled to the power supply output and the output of which is coupled to the adaptive active common-mode compensation unit, for feeding back the residual common-mode noise from the power supply output to the adaptive active common-mode compensation unit to achieve closed-loop control.

9. A noise reduction method for a multi-level common-mode suppression amplification power supply for a medical precision superconducting quantum interference device, using the noise reduction topology for a medical precision superconducting quantum interference device as described in claim 8, characterized in that... Includes the following steps: S1. Connect an external DC power supply, detect the common-mode noise component in the power supply output through an adaptive active common-mode compensation unit, and generate an active compensation signal with the opposite phase to the common-mode noise component and inject it into the power supply circuit to cancel the common-mode noise. S2. Divide the power supply noise after common-mode compensation into at least two frequency bands according to the frequency domain, and perform corresponding noise reduction processing on each frequency band, including active feedback noise reduction in the ultra-low frequency band, active common-mode compensation noise reduction in the low frequency band, passive LC notch noise reduction in the mid-frequency band, and distributed bypass noise reduction in the high frequency band. S3. Real-time acquisition of load current change signals, prediction of the upcoming noise spectrum characteristics based on the load current change signals using a digital twin model, and adjustment of the noise reduction processing parameters for each frequency band in advance. S4. Real-time monitoring of ambient temperature and dynamic adjustment of power supply bias parameters through a high-order temperature compensation network based on temperature changes, so as to optimize temperature compensation and noise suppression in a coordinated manner. S5 outputs a stable power supply after noise reduction to the superconducting quantum interference instrument amplifier circuit.

10. The noise reduction method for the noise reduction topology of the multi-level common-mode suppression medical precision superconducting quantum interference instrument amplification power supply according to claim 9, characterized in that, Step S2 also includes a frequency domain crossover point dynamic optimization step: real-time acquisition of the noise spectrum distribution at the power output terminal, and dynamic adjustment of the boundary frequency between each frequency band according to the noise spectrum distribution, so that the noise reduction processing of each frequency band can be seamlessly connected. Step S3 further includes a feedforward-feedback composite control step: weighting and fusing the feedforward prediction result with the real-time feedback signal at the output end to generate the final control parameters.

Citation Information

Patent Citations

  • Readout circuit based on SQUID (Superconducting Quantum Interference Device) offset voltage reversal and method for inhibiting low-frequency noises

    CN102426343A

  • Biased amplifier circuit of superconducting quantum interference device

    CN104345758A