Double-second-order program-controlled filtering system for SERF atom magnetometer

Abstract

The dual-second-order program-controlled filtering system for the SERF atom magnetometer is designed and realized on the basis of a classical filtering theory, gain stability, adjustable cut-off frequency and quality factor Q and configurable filtering characteristics can be realized, an analog circuit structure is optimized, the optimal performance of a filter is realized in a simple network, and the filtering performance of the SERF atom magnetometer is improved. The method can be applied to a miniaturized single-beam SERF atom magnetometer prototype electronic measurement and control system for measuring weak biological magnetic signals of a human body, and a high-performance filter with low noise and controllable parameters can be formed. According to the invention, the power frequency noise, baseline drift and wake noise in the low-frequency weak magnetocardiogram signal can be effectively removed, and the broadband range capable of being flexibly configured and the stable 0dB in-band gain can be realized.

Description

Technical Field

[0001] The invention belongs to the technical field of SERF (Spin-Exchange Relaxation-Free) state atomic ultra-high sensitivity inertial measurement, and in particular relates to a dual second-order program-controlled filtering system for a SERF atomic magnetometer. Background Art

[0002] SERF atomic magnetometer can realize ultra-high sensitivity inertial measurement (SERF, Spin-Exchange Relaxation-Free, no spin exchange relaxation). The ultra-high sensitivity inertial measurement device is mainly composed of three parts: probe, light source and electronic measurement and control system. Figure 1 The structure of the single-beam SERF atomic magnetometer shown in the figure. The light source part uses a DFB (Distributed Feedback Laser) laser chip or a distributed Bragg reflection laser (Distributed Bragg Reflection, DBR) chip as the core seed light source, which then passes through two isolators in sequence and enters the beam shaping and splitting system, and enters the single-mode polarization-maintaining fiber through the fiber coupler to realize the input of the light source to the probe. The output Gaussian light is turned by the polarization combination mirror and converted into a circularly polarized light to enter the gas chamber. The photodetector detects the light intensity after passing through the gas chamber and converts it into a photocurrent. The photocurrent is connected to the electronic measurement and control system through a twisted pair shielded cable. The photocurrent signal connected to the electronic measurement and control system is first converted into a voltage by a transimpedance amplifier, and then the 1kHz signal carrying the measured signal is filtered out by the analog signal processing system and enters the phase-locked amplifier. The phase-locked amplifier performs data demodulation and inputs the demodulated signal into the data acquisition module to realize the measurement of weak magnetic field signals. Among them, the electronic measurement and control system mainly includes a signal generation module, a laser control module, a temperature measurement and heating module, an analog signal processing system, a data acquisition module, and a preamplifier circuit.

[0003] In the single-beam SERF atomic magnetometer that realizes extremely weak magnetic precision measurement, the current output from the photoelectric converter contains various high-frequency and large-amplitude noises in the signal after being converted into voltage by the transimpedance amplifier. The core of the analog signal processing system that extracts weak low-frequency signals from it is the filter. The voltage analog signal output by the transimpedance amplifier is filtered through the filter to accurately remove the high-frequency signal and completely retain the target signal without attenuation, which can avoid noise interference in the signal processing process, thereby further improving the sensitivity of the SERF atomic magnetometer system. Summary of the invention

[0004] The present invention is based on classical filtering theory, and designs and implements a dual second-order programmable filtering system for SERF atomic magnetometer, which is conducive to achieving gain stability, adjustable cutoff frequency and quality factor Q, configurable filtering characteristics, optimized analog circuit structure, and achieving the best performance of the filter in a concise network. It can be applied to the electronic measurement and control system of a miniaturized single-beam SERF atomic magnetometer prototype for measuring weak biomagnetic signals of the human body, and can form a high-performance filter with low noise and controllable parameters. The present invention can effectively remove power frequency noise, baseline drift and tail noise in low-frequency weak cardiomagnetic signals, and has a flexibly configurable wide-band range and a stable 0dB in-band gain.

[0005] The technical solution of the present invention is as follows:

[0006] A dual-second order program-controlled filtering system for a SERF atomic magnetometer, characterized in that it comprises a dual-second order filter, the dual-second order filter has a signal input interface, the dual-second order filter outputs a signal through a multiplexer, the dual-second order filter and the multiplexer are respectively connected to a control module, the control module is connected to a host computer, the dual-second order filter comprises a first operational amplifier, the negative input end of the first operational amplifier is connected to an input signal end through a first digital-to-analog converter, the output end of the first operational amplifier is connected to a second digital-to-analog converter and a first capacitor connected in parallel, the output end of the third operational amplifier is connected to a third digital-to-analog converter, the positive input end of the first operational amplifier is grounded, the output end of the first operational amplifier is connected to a negative input end of the second operational amplifier through a fourth digital-to-analog converter, and the negative input end of the second operational amplifier is connected to the output end of the second operational amplifier through a second capacitor. , the positive input terminal of the second operational amplifier is grounded, the output terminal of the second operational amplifier is connected to the negative input terminal of the third operational amplifier through the fifth resistor, the negative input terminal of the third operational amplifier is connected to the output terminal of the third operational amplifier through the sixth resistor, the output terminal of the third operational amplifier is connected to the negative input terminal of the fourth operational amplifier through the ninth resistor, the positive input terminal of the fourth operational amplifier is grounded, the negative input terminal of the fourth operational amplifier is connected to the output terminal of the fourth operational amplifier through the tenth resistor, the output terminal of the first operational amplifier is connected to the negative input terminal of the fourth operational amplifier through the eighth resistor, the input signal terminal is connected to the negative input terminal of the fourth operational amplifier through the seventh resistor, the output terminal of the fourth operational amplifier outputs a high-pass response HP, the output terminal of the third operational amplifier outputs an inverted low-pass response LP-, the output terminal of the second operational amplifier outputs a non-inverting low-pass response LP+, and the output terminal of the first operational amplifier outputs a band-pass response BP.

[0007] The host computer uses the frequency control word and the quality factor control word of the digital-to-analog converter DAC through the control module to achieve the filter cutoff frequency f cThe quality factor Q is precisely controlled by using the variable resistance characteristic of the DAC through the integrated feedback resistor R fb With f c The functional relationship between Q and filter cutoff frequency is used to realize the flexible adjustability of filter type parameters.

[0008] The invention comprises a first double-quad filter to a fourth double-quad filter, and a first-stage multiplexer to a fourth-stage multiplexer. The original signal Signal is respectively provided to the input end of the first double-quad filter and the respective input ends of the first-stage multiplexer to the fourth-stage multiplexer. The first double-quad filter provides a first-stage in-phase low-pass response and a first-stage high-pass response to the first-stage multiplexer. The first output signal of the first-stage multiplexer is transmitted to the input end of the second double-quad filter. The second double-quad filter provides a second-stage in-phase low-pass response and a second-stage high-pass response to the second-stage multiplexer. The second output signal of the second-stage multiplexer is transmitted to the input end of the third double-quad filter. The third double-quad filter provides a third-stage in-phase low-pass response and a third-stage high-pass response to the third-stage multiplexer. The third output signal of the third-stage multiplexer is transmitted to the input end of the fourth double-quad filter. The fourth double-quad filter provides a fourth-stage in-phase low-pass response and a fourth-stage high-pass response to the fourth-stage multiplexer. The fourth-stage multiplexer outputs a fourth output signal.

[0009] R fb With f c The functional relationship with Q is as follows:

[0010] D1=2 n *R fb *2π*C1*f c

[0011] D2=(2 n *R fb *2π*C1*f c ) / Q

[0012] Where D1 is the binary cutoff frequency control word of the DAC, n is the number of DAC bits, C1 is the first capacitor, and D2 is the binary quality factor control word of the DAC.

[0013] The dual second-order filter circuit is connected to the control module via an SPI interface, and the selection circuit is connected to the control module via an I / O interface.

[0014] The control module is connected to the host computer via a UART interface.

[0015] The advantages of the present invention over the prior art are:

[0016] 1. Structural optimization of active filters. After investigation, comparison and simulation experiments on digital and analog filters, active and passive filters, and various advanced and mature systematic filter structures, it was finally determined that the filter circuit uses a dual-second order filter circuit composed of an operational amplifier, capacitors and resistors to achieve filter performance. It can output bandpass, low-pass and high-pass responses at the same time, has low sensitivity to components, and the components that determine the filter parameters (cutoff frequency, Q value and gain) can be adjusted independently, which fully meets the performance requirements and parameter indicators of the filter in the research objectives, and can adjust the parameters by programmable resistors on the basis of the dual-second order. It fundamentally solves the shortcomings of the active filter composed of traditional amplifiers in series, such as large and complex circuit structure and limited adjustment range when adjusting parameters, and has low self-noise level, low power consumption and low distortion rate signal output characteristics.

[0017] 2. The filtering parameters are independently adjustable. By modeling the dual-second-order circuit system function, the key parameter indicators of its filtering characteristics are obtained, and the functional relationship between the RC component values ​​or the conversion circuit structure is used to characterize the filtering performance parameters, such as cutoff frequency, quality factor, gain and roll-off order. After multiple scheme investigations and comparative experiments, it is determined to use the equivalent resistance of the digital-to-analog conversion chip as a programmable filter to flexibly adjust the filter's cutoff frequency and quality factor, and use a selection circuit composed of a multi-way switch to control the filter passband type and attenuation slope. Finally, the main control chip controls the DAC chip and the multi-way selector respectively to achieve flexible and adjustable parameters of the entire filtering system.

[0018] 3. Low system noise. Noise performance is one of the important indicators of the filter. Without considering non-ideal factors such as drift, the filtering accuracy of the system is determined by the equivalent input noise, that is, the equivalent input noise determines the output signal-to-noise ratio of the system. Therefore, it is of great significance to reduce the equivalent input noise of the system to a certain extent. The main source of filter noise is the analog part of the original signal filtering, including operational amplifiers, digital-to-analog converters, etc. In the actual selection process, attention should be paid to the noise level of the device, including self-noise and crosstalk between channels, so as to improve the circuit signal-to-noise ratio and minimize the distortion of the filter while ensuring the performance and stability of the filtering system. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 It is a schematic diagram of the structure of a single-beam SERF atomic magnetometer. Figure 1The electronic measurement and control system, a single-beam SERF atomic magnetometer probe, and a light source are included. The electronic measurement and control system (i.e., an analog signal processing system) includes a filter, the input end of the filter is connected to the photodetector outside the atomic gas chamber through a transimpedance amplifier, and the output end of the filter is connected to a data acquisition device through a phase-locked amplifier. A heating film is provided on the atomic gas chamber, and the heating film is connected to the temperature control system and the power amplifier in the electronic measurement and control system. The atomic gas chamber with the heating film is located in a coil, and the coil is connected to a signal generator in the electronic measurement and control system. The light source includes a laser, a beam shaping device, and a wavelength meter. The laser is connected to the photodetector through a beam shaping device, a polarization-maintaining fiber, a collimation module, a collimation lens, a reflector, and an atomic gas chamber in turn. SERF is spin-exchange relaxation-free.

[0020] Figure 2 yes Figure 1 Schematic diagram of the structure of China Electronics' measurement and control system. Figure 2 In the figure, A is a transimpedance amplifier, B is a filter, C is a phase-locked amplifier, D is a data acquisition module, E is a signal generator, and F is a magnetic compensation coil; a is the photocurrent output by the magnetometer gas chamber, b is the voltage signal output by the photocurrent after passing through the transimpedance amplifier, c is the output signal of the filtering system, which contains a 1kHz sinusoidal wave modulation signal, d is the signal demodulated by the phase-locked amplifier, e is the final output signal after being processed by the data acquisition module, f is the magnetic compensation signal generated by the signal generator, usually a 3Hz sinusoidal wave, and g is the 3Hz sinusoidal wave signal after filtering and purification by the filter.

[0021] Figure 3 It is a schematic diagram of a biquad filter circuit. Figure 3 U1~U4 are the first operational amplifier to the fourth operational amplifier, R1~R10 are the first resistor to the tenth resistor, and C1~C2 are the first capacitor to the second capacitor. The dotted box is the basic structure of the biquad filter. Vin represents the input signal, BP is the bandpass response output by U1, LP+ represents the in-phase low-pass response output by U2, LP- represents the inverting low-pass response output by U3, and HP represents the high-pass response output by U4.

[0022] Figure 4 yes Figure 3 The amplitude-frequency response and phase-frequency response curves of each output node. Figure 4 The middle horizontal axis is frequency (Hz), the left vertical axis is gain (dB), and the right vertical axis is phase (degrees). The output nodes include LP+, HP, and BP. Figure 4 The curves in the figure are simulations of a biquad filter with a cutoff frequency of 1kHz and a roll-off slope of 12dB / oct (decibels / octave).

[0023] Figure 5 The present invention is a schematic diagram of the structure of a dual second-order program-controlled filtering system for a SERF atomic magnetometer. Figure 5 The invention comprises a double second-order filter circuit or a double second-order filter (which has a DAC digital-to-analog converter, Digital Analog Converter), wherein the double second-order filter circuit has a signal input interface, the double second-order filter circuit outputs a signal through a selection circuit (which has a multiplexer), the double second-order filter circuit is connected to a control module through an SPI interface, the selection circuit is connected to the control module through an I / O interface, and the control module is connected to a host computer through a UART interface.

[0024] Figure 6 yes Figure 5 Schematic diagram of the biquad filter principle. Figure 6 The circuit includes a first integrator, a second integrator, an amplifier and an adder, the input signal Vin is connected to the first input terminal of the first integrator and the first input terminal of the adder respectively, the second input terminal of the first integrator is connected to the output terminal of the amplifier for low-pass feedback, and the output terminal V BP (Bandpass output signal) is respectively connected to the input end of the second integrator and the second input end of the adder, the third input end of the adder is connected to the output end of the amplifier, and the adder forms a high-pass output signal B HP , the output terminal V of the second integrator LP+ (V LP+ The amplifier forms an inverting low-pass output signal V LP- .

[0025] Figure 7 This is the schematic diagram of a current output DAC applied to a dual second-order filter circuit. Figure 7 The biquad filter circuit in Figure 3 Compared with the biquad filter circuit in Figure 7 DAC1 to DAC4 (the first digital-to-analog converter to the fourth digital-to-analog converter) in the Figure 3 The filter cutoff frequency f is controlled by using the DAC frequency control word and quality factor control word. c The variable resistance characteristic of DAC is used to precisely control the quality factor Q through the resistor R and f c The functional relationship between Q and filter cutoff frequency is used to realize the flexible adjustability of filter type parameters.

[0026] Figure 8 It is a schematic diagram of the filter attenuation and passband type selection circuit structure. Figure 8Among them, B1~B4 are the first double-quad filter to the fourth double-quad filter, K1~K4 are the first-stage multiplexer to the fourth-stage multiplexer, Signal is the original signal of the input filter, Output1~Output 4 are the first output signal to the fourth output signal of K1~K4, LP1~LP4 are the first-stage LP+ to the fourth-stage LP+, and HP1~HP4 are the first-stage HP to the fourth-stage HP.

[0027] Fig. 9 It is a schematic diagram of the amplitude-frequency characteristic curve of Butterworth low-pass filters of different orders when fc=50kHz.

[0028] Fig.10 It is a schematic diagram of the amplitude-frequency characteristic curve of Bessel low-pass filters of different orders when fc=50kHz.

[0029] Fig.11 It is a schematic diagram of the amplitude-frequency characteristic curve of Butterworth high-pass filters of different orders when fc=50kHz.

[0030] Fig.12 It is a schematic diagram of the amplitude-frequency characteristic curve of Bessel high-pass filters of different orders when fc=50kHz.

[0031] Fig.13 It is a schematic diagram of the amplitude-frequency characteristic curve of Butterworth low-pass filters of different orders when fc=100kHz.

[0032] Fig.14 It is a schematic diagram of the amplitude-frequency characteristic curve of Butterworth low-pass filters of different orders when fc=250kHz.

[0033] Fig.15 It is a schematic diagram of the amplitude-frequency characteristic curve of Butterworth low-pass filters of different orders when fc=500kHz.

[0034] Fig.16 This is the FFT analysis of the filter output signal. FFT is the Fast Fourier Transform.

[0035] Fig.17 This is a comparison chart of the FFT analysis of the filter input and output signals.

[0036] Fig.18 This is the FFT analysis diagram of the sensitivity of the single-beam SERF atomic magnetometer. DETAILED DESCRIPTION

[0037] Below is the attached figure ( Figure 1-Figure 18 ) and Examples illustrate the present invention.

[0038] Figure 1 It is a schematic diagram of the structure of a single-beam SERF atomic magnetometer. Figure 2 yes Figure 1 Schematic diagram of the structure of China Electronics' measurement and control system. Figure 3 It is a schematic diagram of a biquad filter circuit. Figure 4 yes Figure 3 Simulation curves of amplitude-frequency response and phase-frequency response of each output node. Figure 5 The present invention is a schematic diagram of the structure of a dual second-order program-controlled filtering system for a SERF atomic magnetometer. Figure 6 yes Figure 5 Schematic diagram of the biquad filter principle. Figure 7 This is the schematic diagram of a current output DAC applied to a dual second-order filter circuit. Figure 8 It is a schematic diagram of the filter attenuation and passband type selection circuit structure. Fig. 9 It is a schematic diagram of the amplitude-frequency characteristic curve of Butterworth low-pass filters of different orders when fc=50kHz. Fig.10 It is a schematic diagram of the amplitude-frequency characteristic curve of Bessel low-pass filters of different orders when fc=50kHz. Fig.11 It is a schematic diagram of the amplitude-frequency characteristic curve of Butterworth high-pass filters of different orders when fc=50kHz. Fig.12 It is a schematic diagram of the amplitude-frequency characteristic curve of Bessel high-pass filters of different orders when fc=50kHz. Fig.13 It is a schematic diagram of the amplitude-frequency characteristic curve of Butterworth low-pass filters of different orders when fc=100kHz. Fig.14 It is a schematic diagram of the amplitude-frequency characteristic curve of Butterworth low-pass filters of different orders when fc=250kHz. Fig.15 It is a schematic diagram of the amplitude-frequency characteristic curve of Butterworth low-pass filters of different orders when fc=500kHz. Fig.16 It is the FFT analysis diagram of the filter output signal. Fig.17 This is a comparison chart of the FFT analysis of the filter input and output signals. Fig.18 This is the FFT analysis of the sensitivity of the single-beam SERF atomic magnetometer. Figures 1 to 18As shown, a dual-second order program-controlled filtering system for a SERF atomic magnetometer includes a dual-second order filter, the dual-second order filter has a signal input interface (for receiving an input signal Vin), the dual-second order filter outputs a signal through a multiplexer, the dual-second order filter and the multiplexer are respectively connected to a control module, the control module is connected to a host computer, the dual-second order filter includes a first operational amplifier U1, the negative input terminal (-) of the first operational amplifier U1 is connected to the input signal Vin terminal through a first digital-to-analog converter DAC1, the output terminal of the first operational amplifier U1 is connected through a second digital-to-analog converter DAC2 and a first capacitor C1 connected in parallel, the output terminal of the third operational amplifier U3 is connected through a third digital-to-analog converter DAC3, the positive input terminal (+) of the first operational amplifier U1 is grounded, the output terminal of the first operational amplifier U1 is connected to the negative input terminal (-) of the second operational amplifier U2 through a fourth digital-to-analog converter DAC4, the negative input terminal (-) of the second operational amplifier U2 is connected to the output terminal of the second operational amplifier U2 through a second capacitor C2, and the second operational amplifier U2 is connected to the output terminal of the second operational amplifier U2 through a fourth digital-to-analog converter DAC4. The positive input terminal (+) of the amplifier U2 is grounded, the output terminal of the second operational amplifier U2 is connected to the negative input terminal (-) of the third operational amplifier U3 through the fifth resistor R5, the negative input terminal (-) of the third operational amplifier U3 is connected to the output terminal of the third operational amplifier U3 through the sixth resistor R6, the output terminal of the third operational amplifier U3 is connected to the negative input terminal (-) of the fourth operational amplifier U4 through the ninth resistor R9, the positive input terminal (+) of the fourth operational amplifier U4 is grounded, the negative input terminal (-) of the fourth operational amplifier U4 is connected to the output terminal of the fourth operational amplifier U4 through the tenth resistor R10, the output terminal of the first operational amplifier U1 is connected to the negative input terminal (-) of the fourth operational amplifier U4 through the eighth resistor R8, the input signal Vin terminal is connected to the negative input terminal (-) of the fourth operational amplifier U4 through the seventh resistor R7, the output terminal of the fourth operational amplifier U4 outputs a high-pass response HP, the output terminal of the third operational amplifier U3 outputs an inverted low-pass response LP-, the output terminal of the second operational amplifier U2 outputs a non-inverting low-pass response LP+, and the output terminal of the first operational amplifier U1 outputs a bandpass response BP. The host computer uses the frequency control word and the quality factor control word of the digital-to-analog converter DAC through the control module to achieve the filter cutoff frequency f c The quality factor Q is precisely controlled by using the variable resistance characteristic of the DAC through the integrated feedback resistor R fb With f c The functional relationship between Q and filter cutoff frequency is used to realize the flexible adjustability of filter type parameters.

[0039] The invention comprises a first double-quad filter B1 to a fourth double-quad filter B4, and a first-stage multiplexer K1 to a fourth-stage multiplexer K4, wherein the original signal Signal is respectively provided to the input end of the first double-quad filter B1 and the respective input ends of the first-stage multiplexer K1 to the fourth-stage multiplexer K4, the first double-quad filter B1 provides the first-stage in-phase low-pass response LP1 and the first-stage high-pass response HP1 to the first-stage multiplexer K1, the first output signal Output1 of the first-stage multiplexer K1 is transmitted to the input end of the second double-quad filter B2, and the second double-quad filter B2 provides the second-stage in-phase high-pass response HP1 to the second-stage multiplexer K2. The second output signal Output2 of the second-stage multiplexer K2 is transmitted to the input end of the third double-second-order filter B3, the third double-second-order filter B3 provides the third-stage in-phase low-pass response LP3 and the third-stage high-pass response HP3 to the third-stage multiplexer K3, the third output signal Output3 of the third-stage multiplexer K3 is transmitted to the input end of the fourth double-second-order filter B4, the fourth double-second-order filter B4 provides the fourth-stage in-phase low-pass response LP4 and the fourth-stage high-pass response HP4 to the fourth-stage multiplexer K4, and the fourth-stage multiplexer K4 outputs the fourth output signal Output4.

[0040] R fb With f c The functional relationship with Q is as follows:

[0041] D1=2 n *R fb *2π*C1*f c

[0042] D2=(2 n *R fb *2π*C1*f c ) / Q

[0043] Where D1 is the binary cutoff frequency control word of the DAC, n is the number of DAC bits, C1 is the first capacitor, and D2 is the binary quality factor control word of the DAC.

[0044] The dual second-order filter circuit is connected to the control module via an SPI interface, the selection circuit is connected to the control module via an I / O interface, and the control module is connected to a host computer via a UART interface.

[0045] The present invention starts from the high-performance filter of SERF (Spin-Exchange Relaxation-Free) atomic magnetometer analog signal processing system, focuses on the structural optimization and performance improvement of analog circuits, carries out research and design of high signal-to-noise ratio parameter-adjustable filters, deeply analyzes key theoretical and application issues such as system function model, parameter control adjustment, performance performance, etc. of dual-second-order state variable filters in active filters, develops a filtering system with stable performance and relatively independent and controllable parameters, and is applied to the electronic measurement and control system of a miniaturized single-beam SERF atomic magnetometer prototype for measuring weak biomagnetic signals of the human body, and at the same time lays a solid foundation for the further miniaturization of subsequent SERF atomic magnetometers and the realization of ultra-high sensitivity magnetic field measurement accuracy.

[0046] Figure 1 The application scenarios of filters are included in this paper, which are used in the electronic measurement and control system of a miniaturized single-beam SERF atomic magnetometer prototype. The working process of the filter in the atomic magnetometer is as follows Figure 2 As shown. Wherein A is a transimpedance amplifier, B is the filtering system in the present invention, C is a phase-locked amplifier, D is a data acquisition module, E is a signal generator, and F is a magnetic compensation coil; a is the photocurrent output by the magnetometer gas chamber, b is the voltage signal output by the photocurrent after passing through the transimpedance amplifier, c is the output signal of the filtering system, which contains a 1kHz sine wave modulation signal, d is the signal demodulated by the phase-locked amplifier, e is the final output signal after processing by the data acquisition module, f is the magnetic compensation signal generated by the signal generator, usually a 3Hz sine wave, and g is a 3Hz sine wave signal after filtering and purification by a filter. The photocurrent signal entering the electronic measurement and control system is first converted into a voltage by the transimpedance amplifier, enters the analog signal processing system, and is filtered out of 3Hz and 1kHz sine waves respectively through a bandpass filter for noise reduction, and then input into the phase-locked amplifier and demodulated with the synchronous trigger signal of the signal generator as a reference. Finally, the demodulated signal is transmitted to the data acquisition module to achieve the measurement of weak magnetic field signals.

[0047] In classical filtering theory, filters can be divided into analog filters and digital filters according to the different signals they process. According to the different components, analog filters can be divided into passive filters and active filters.

[0048] Active filter circuits are widely used, mainly for low-frequency (1Hz~1MHz) small signal processing. They can be used to suppress interference and noise, attenuate useless frequency signals, and highlight useful frequency signals to achieve the purpose of improving signal-to-noise ratio or frequency selection. Active filters have many unique advantages, such as standardized design, modularization, and convenient debugging. In addition, due to the high gain and input resistance of the operational amplifier and the low output resistance, they can provide a certain signal gain and buffering effect, and a high-order filter can be obtained by simple cascading. It has a certain system function and cutoff frequency analytical expression. In practical applications, active filters are generally selected based on comprehensive consideration of circuit filtering characteristics and signal gain. The main indicators for selecting filters are ripple amplitude, cutoff frequency, quality factor, stability, delay, etc.

[0049] The present invention provides a dual second-order program-controlled filtering system for a SERF atomic magnetometer, and its specific performance indicators are as follows:

[0050] (1) Filter mode: low pass / high pass optional;

[0051] (2) Filter approximation type: Butterworth / Bessel optional;

[0052] (3) Filter frequency band: adjustable within 1Hz~500kHz, with adjustment resolution of 3 bits (±1% accuracy);

[0053] (4) Transition band attenuation (roll-off): 12 / 24 / 36 / 48dB / oct, i.e. 2 / 4 / 6 / 8 order optional;

[0054] (6) Gain: 0dB.

[0055] The main work of the present invention is as follows:

[0056] (1) Conduct an in-depth study on the working characteristics of the filter, and formulate corresponding technical indicators in combination with the actual application scenarios of the magnetometer. Through research and summary of different schemes, simulation and comparison of performance, a dual second-order programmable filter scheme is selected as the invention content. Based on the basic principle of the dual second-order filter, a model simulation is established and the system transfer function is calculated. The functional relationship between parameters such as cutoff frequency, Q value, passband type, order and circuit elements is summarized to achieve precise program control. Combined with the working principle and deduction model of the current output digital-to-analog converter, a set of programmable filter design methods for dual second-order digital-to-analog converters is proposed based on the filter normalization design model.

[0057] Solution 1: RC active filter cascade. When the cutoff frequency or filter type (quality factor Q) is changed, the values ​​of many components in the circuit will change. When using programmable filter switching, not only is the algorithm complex, the control difficulty increases, and too many programmable components will also cause circuit redundancy.

[0058] Solution 2: State variable filter. The state variable filter is composed of three operational amplifiers and resistors. It is one of the most commonly used circuit structures in today's filter systems. It has three outputs: low-pass, high-pass and band-pass. The cutoff frequency and Q value can be changed by changing the resistor. However, through simulation, it was found that the passband flatness of the bandpass and high-pass of the state variable filter was poor, and the waveform distortion was more obvious.

[0059] Solution 3: Double second-order filter. The circuit diagram of double second-order filter is as follows Figure 3 As shown, U1, U2, U3, and U4 are the first, second, third, and fourth operational amplifiers, respectively, R1 to R9 represent the first to ninth resistors, respectively, and C1 and C2 represent the first capacitor and the second capacitor, respectively. The basic structure of the biquad filter is composed of three operational amplifiers U1, U2, and U3 in the dotted box, two capacitors C1 and C2, and six resistors R1 to R6. A biquad filter circuit can simultaneously output bandpass and lowpass responses from different operational amplifier output terminals. When a fourth operational amplifier and more components are added, highpass, bandstop, and all-pass filters can even be obtained. Figure 3 The fourth-stage operational amplifier U4 and the seventh to tenth resistors R7 to R10 are connected, Vin represents the input signal of the biquad circuit, BP represents the bandpass response output by the first-stage amplifier U1, LP+ represents the in-phase low-pass response output by the second-stage amplifier, LP- represents the inverting low-pass response output by the third-stage amplifier, and HP represents the high-pass response output by the fourth-stage amplifier.

[0060] The biquad filter with a cutoff frequency of 1kHz and a roll-off slope of 12dB / oct is simulated. The simulation diagrams of the bandpass, lowpass and highpass response frequency characteristics are as follows: Figure 4 As shown, the solid line represents the filter amplitude-frequency response, and the dotted line represents the filter phase-frequency response. From the simulation diagram, it can be seen that the biquad filter can not only output multiple types of filter signals at the same time like the state variable filter, but its amplitude-frequency characteristic curve also has a flat passband, stable transition band attenuation, and a smaller distortion rate. Whether from the perspective of circuit structure or filtering performance, the biquad filter network structure is the most suitable filtering solution for the SERF atomic magnetometer signal processing system.

[0061] Biquad filters have absolute advantages that other circuits do not have. The main advantage of biquad filters is that voltage gain, bandwidth and center frequency are all independently adjustable. Different operational amplifiers in the circuit can simultaneously output three types of filter responses: low-pass, high-pass and band-pass. Thanks to the stable and diverse performance, compact and efficient circuit structure and easy development of biquad filters, it has been widely used in scientific research and the commercial development of electronic products.

[0062] The filtering characteristics of the biquad are closely related to the resistor and capacitor. Changing the value of the resistor or capacitor can change the filter's cutoff frequency and other parameters. In the present invention, the purpose of accurately controlling the filter's cutoff frequency, Q value, gain and other parameters is achieved by using programmable resistors to change the resistance value.

[0063] The Digital Analog Converter (DAC) can convert the binary value into voltage or current, and the voltage or current is proportional to the input binary value, so it can accurately be equivalent to resistance. Finally, the main control chip is used to control the current output DAC as the programmable resistor of the filter circuit. The variable resistor value is accurately controlled by the program, the circuit structure is optimized, the noise error is reduced, the hardware design and debugging complexity is reduced, and the simplest circuit design is constructed with the least number of devices to achieve the best circuit performance.

[0064] (2) The hardware and software functional architecture of the dual second-order filter system was designed and implemented. The main tasks of the hardware design include chip selection, circuit construction and simulation, schematic design, PCB design, etc. Based on the system solution, the stacking scheme design of power supply, filter circuit, selection circuit, control module, etc. was completed in accordance with the modular thinking, and the basic performance of the hardware circuit was tested. The main tasks of the software design include the design and implementation of the DAC control word algorithm, the selection module and the logic flow and embedded code of the serial communication, and finally the upper computer interface was built based on the MATLAB GUI. Through the above work, the visual human-computer interaction of the adjustable parameter filter was realized, and a high-performance, parameter-adjustable filter that can be used in the electronic measurement and control system of the single-beam SERF atomic magnetometer was developed.

[0065] (3) The system performance was tested and analyzed, including independent and joint testing. The frequency characteristics of the filter were analyzed by controlling the variables, verifying the effectiveness and accuracy of the host computer and instructions in controlling the filter system. A filter with a bandwidth of 1MHz, a cutoff frequency of 1 to 500kHz, a fixed gain, a flat passband, and adjustable parameters such as order and type was developed. The filter was tested in a joint test of the single-beam SERF atomic magnetometer system, and the time and frequency domain characteristics of the filter input and output signals were collected and analyzed. The overall sensitivity of the SERF atomic magnetometer system with the filter of the present invention was measured to be 30fT / Hz. 1 / 2 @30.5Hz, which is comparable to existing commercial instruments, verifying the performance of the filter system of the present invention.

[0066] A dual second-order program-controlled filtering system for SERF atomic magnetometer, the block diagram of which is shown in Figure 5 In the process of designing the system, a modular design concept is adopted to achieve the simplification and efficiency of the circuit structure to the greatest extent on the basis of achieving the optimal performance of the filtering system. The system consists of four parts: a coupling host computer, a dual second-order circuit, a selection circuit and a control module.

[0067] The original signal is input into the dual second-order filter circuit. The module is composed of four dual second-order circuits in cascade, forming an eight-order analog filter. Each dual second-order circuit is connected through a multiplexer. These interfaces derived from the dual second-order circuit and multiple multiplexers together constitute the selection circuit. The control module composed of the STM32 single-chip microcomputer and its peripheral circuits inputs the binary control word D to the current output DAC in the dual second-order filter circuit through the SPI interface to realize programmable resistance to accurately change the cutoff frequency and quality factor (Butterworth / Bessel filter) in the filter system; the control module also controls the multiplexer switch in the selection circuit through the I / O interface, and controls the on and off of different combination switches to achieve the passband type (high pass / low pass) and filter order (2 / 4 / 6 / 8 order), that is, the roll-off slope (12 / 24 / 36 / 48dB / oct) of the filter, and finally outputs the filtered signal. Due to the optimal low-power structure and low-noise characteristics of the dual-quadratic filter circuit itself, combined with the improved high-precision and high-reliability program control method, the entire filtering system can perfectly achieve the design goals of high performance, low noise, stable and controllable performance, flexible and adjustable parameters, and easy operation.

[0068] A design of a dual second-order program-controlled filter system for a SERF atomic magnetometer includes the following steps:

[0069] The first step is to model the system function according to the principle of the double-quadratic filter circuit, and compare it with the normalized standard formula of the filter design to obtain the functional relationship between the filter parameters (cutoff frequency, quality factor and gain) and the capacitor and resistor. The program-controlled variable resistor scheme is created using this functional relationship to achieve precise control of the cutoff frequency, quality factor and gain of the double-quadratic filter system.

[0070] The second step is to convert the digital-to-analog converter into a controllable resistor. Combining the functional relationship between the filter parameters and the capacitor and resistor in the first step, the binary numerical word of the control parameter is calculated and summarized, so that the DAC equivalent resistor can be accurately controlled proportionally through the main control chip, making the filter parameters (cutoff frequency, quality factor and gain) adjustable.

[0071] The third step is to grasp the characteristics of the biquad circuit in that different amplifiers in a group of circuits can output high-pass, low-pass and band-pass responses, and combine the characteristics of the active filter stage cascade to change the slope of the transition band to design and select the circuit to achieve the switching function of high-pass, low-pass and filter order;

[0072] The fourth step is to design and implement the control module program in C language on the Keil uVision5 platform based on the hardware architecture and circuit connection of the filtering system, which mainly includes the control program of the DAC equivalent variable resistor according to the cutoff frequency and quality factor control word algorithm, the program for writing the true value of each channel of the selection circuit according to the preset mode, the definition and control program of the JSON format command through the serial port, and the program for the microcontroller to communicate with the host computer through the RS485 protocol; the host computer control program is designed based on the GUI on the Matlab platform, and the filter parameters are adjusted through the UI interface displayed by the computer. Only by achieving precise control of key parameters in the program can the excellent characteristics of low noise, low power consumption, compact structure and high device utilization of this analog filtering circuit be maximized.

[0073] A design of a dual second-order program-controlled filter system for a SERF atomic magnetometer includes the following steps:

[0074] The first step is to solve the normalized system function of the biquad filter circuit.

[0075] like Figure 3 As shown in the figure, the circuit structure in the dotted box is the basic unit of the biquad filter. The biquad filter circuit consists of an integrator and an amplifier. Its circuit model is shown in the figure. Figure 5 As shown. Operational amplifier U1 has local feedback R2, C1 and input resistor R1, forming an inverting lossy integrator structure; amplifier U2 and its associated input resistor R4 and feedback capacitor C2 are an inverting integrator; with input resistor R5 and feedback resistor R6 is an inverting amplifier, which forms the main feedback loop of the circuit from U3 through R3 to U1. If a high-pass output is required, an adder needs to be added. The adder consists of the fourth-stage operational amplifier U4 and resistors R7, R8, R9, R 10 Composition. in represents the input signal of the biquad circuit, V BP represents the bandpass response of the first stage amplifier U1 output, V LP+ represents the in-phase low-pass response of the second-stage amplifier output, V LP- represents the inverting low-pass response of the third-stage amplifier output, HP represents the high-pass response of the fourth-stage amplifier output, and Figure 3 The output in corresponds to .

[0076] The LT1361CS8 is selected as the operational amplifier chip in the present invention. It is a dual-channel precision operational amplifier with rail-to-rail output, a maximum offset voltage of 1mV and a maximum input bias current of 1μA, and a 9nV / Hz 1 / 2The input noise voltage is 0.08V / μs, and it has a fast response speed of 800V / μs. It has a gain-bandwidth product of up to 50MHz at a ±15V power supply and a 200kHz operating frequency, which provides a guarantee for the realization of a wide range of 1-500kHz cut-off frequency of the filter in this paper. Even the dual-second-order filter circuit composed of the amplifier has ample bandwidth when it is cascaded in multiple stages.

[0077] From the amplifier "virtual short" and "virtual break" principle, we can know

[0078]

[0079] Among them, V in is the filter input signal, V BP is the bandpass output signal of the first-stage amplifier U1 of the biquad circuit, V LP1 is the in-phase low-pass output of the second-stage amplifier U2, V LP2 is the inverting low-pass output of the third-stage amplifier U3, V HP is the high-pass output of the fourth-stage amplifier U4. R1-R6 represent the first to sixth resistors, respectively, and C1 and C2 represent the first and second capacitors, respectively. s is the complex frequency in the Laplace transform.

[0080] Combining the above formulas, we can get the system transfer function of the positive phase low-pass output of the second stage amplifier U2:

[0081]

[0082] Normalizing (1.4) yields the system transfer function of the in-phase low-pass response of the biquad filter:

[0083]

[0084] The fourth stage amplifier U4 satisfies

[0085]

[0086] Simplified, the system transfer function of the high-pass filter response of the fourth-stage amplifier U4 is obtained:

[0087]

[0088] Normalization gives the system transfer function of the high-pass response of the biquad filter:

[0089]

[0090] Any biquad voltage transfer function has the same topology. The transfer function coefficients are determined by the resistor and capacitor values. By changing the component values ​​(preferably only changing the resistors), different transfer functions can be generated.

[0091]

[0092] Comparing the above system function with the normalized filter function expression, we can get the functional relationship expressions of the different parameters of the low-pass and high-pass filter systems and the capacitance and resistance. The denominator of the low-pass and high-pass filter response system functions is the same, so the cutoff frequency f c The relationship is the same as that of the Q value, that is,

[0093]

[0094] where ω c is the intermediate variable, f c is the filter cutoff frequency, indicating the frequency value corresponding to the filter gain of -3dB. Q is the filter quality factor. For bandpass filters, Q is the bandwidth ratio between the center frequency f0 and the -3dB points on both sides; for low-pass and high-pass filters, Q represents the ratio of the gain at -3dB to the bandwidth gain. Q represents the quality of the filter pole and also reflects the cutoff frequency characteristics. The Q value can be used to distinguish the three functions of Butterworth, Chebyshev and Bessel. Their second-order low-pass filter Q values ​​are 0.707, 0.96 and 0.56 respectively.

[0095] ω c =2πf c (1.12)

[0096] Usually, R5=R6, R7=R8=R9, and This minimizes the sensitivity of Q. For ease of design, C1 = C2, then R3 = R4. Equations (1.11) and (1.12) can be simplified to

[0097]

[0098]

[0099] The amplifier gain can be expressed as

[0100]

[0101] Combined with the above formula, the resistance value can be determined according to the cut-off frequency, quality factor, and gain setting value. The setting steps are as follows:

[0102] a. Select the nominal values ​​of R3, R4, C1, and C2 according to the target cutoff frequency fc;

[0103]

[0104] b. Determine the Q value by filter approximation (Butterworth, Bessel), where

[0105] Q Bessel =0.56, then select the value of R2

[0106] R2=QR3 (1.18)

[0107] c. Determine R1 and R by low-pass and high-pass gains respectively 10 Value

[0108]

[0109] R 10 =G HP R9 (1.20)

[0110] The second step is to convert the digital-to-analog converter into a controllable resistor. Combining the functional relationship between the filter parameters and the capacitor and resistor in the first step, the binary numerical word of the control parameter is calculated and summarized, so that the DAC equivalent resistor can be accurately controlled proportionally through the main control chip, making the filter parameters (cutoff frequency, quality factor and gain) adjustable.

[0111] A digital analog converter (DAC) can convert a binary value into a voltage or current that is proportional to the input binary value D. Select a current output DAC that meets the bit accuracy requirements. According to Ohm's law, it can be regarded as a resistor with voltage input and current output. When a DAC with a higher bit number is selected, a larger resistance dynamic range can be achieved, so that the equivalent resistance value can be accurately adjusted by controlling the binary control word D, thereby adjusting the filter parameters.

[0112] DACs are usually binary-weighted and R / 2R ladder-type. Since the number of resistors inside a binary-weighted DAC increases with the number of input bits, the absolute accuracy of the resistor value will be affected by temperature changes, resulting in an increase in the difference between resistor values, making it difficult to achieve the target accuracy, and affecting the accuracy and stability of the DAC. Therefore, binary-weighted DACs are only used in situations where the number of input bits is low and the accuracy requirements are not high. The R / 2R ladder-type digital-to-analog converter overcomes the limitations of the binary-weighted D / A converter and is the most commonly used structure in integrated DACs.

[0113] In practical applications, the reference voltage V is provided externally. ref The main control chip (single-chip microcomputer) controls the pin analog switch to input the data bit, and the output voltage is

[0114]

[0115] Where n is the number of DAC bits, V ref is the external reference input voltage of DAC, and D is defined as the input word of DAC, then D=2 n- 1 dn-1 +2 n-2 d n-2 +...+d0, where d x (x=0…n-1) represents the xth bit of the binary number D, which is 1 or 0, so the output voltage can be expressed as

[0116]

[0117] The output current can be obtained from Ohm's law

[0118]

[0119] Where R fb is the integrated feedback resistor inside the DAC. The equivalent resistor value of the DAC is

[0120]

[0121] Therefore, the range of the DAC equivalent resistance is R fb ~2 n ·R fb .

[0122] The related resistors R1, R2, R3, and R4 that affect the parameters in the biquad circuit architecture are replaced by current output DACs, and the remaining components are all fixed-value capacitors or resistors. Figure 7 DAC1, DAC2, DAC3, and DAC4 are the replaced components. Combining the relationship between circuit components and parameters and the DAC equivalent resistance formula in the above text, the frequency control word and quality factor control word of the DAC are calculated to achieve the filter cutoff frequency f c With precise control of the quality factor Q.

[0123] The filter gain in the present invention is set to 1, then R1 = R3, and then the combination of equation (1.18) and equation (1.25) can be obtained that R1, R3, and R4 are related to the binary cutoff frequency control word D1 of the DAC

[0124] D1=2 n R fb 2πC1f c (1.25)

[0125] This formula corresponds to the binary control word written by DAC1, DAC3, and DAC4 replacing R1, R3, and R4, which can be used to adjust the cutoff frequency f of the filter. c .

[0126] From equation (1.19), we can get the expression of the resistor R2 involved in adjusting the quality factor Q. Similarly, combined with (1.25), we can get the binary quality factor control word D2 of R2 with respect to the DAC:

[0127]

[0128] This formula corresponds to the binary control word written by DAC2 replacing R2, which can be used to adjust the quality factor Q of the filter.

[0129] The digital-to-analog converter DAC is the key to the adjustable filter of the present invention. The number of bits n of the digital-to-analog converter is the basis for achieving the adjustment accuracy. The present invention selects a 14-bit AD5453 current output digital-to-analog converter. According to the design indicators that the filter frequency band can be adjusted within 1Hz to 500kHz and the adjustment resolution is 3 bits (±1% accuracy) in this study, a DAC with a suitable number of bits can be selected according to the above design steps.

[0130] The third step is to select the cutoff frequency, Q value and passband gain, and adopt the method of connecting multiple stages of double second-order filters in series. The output of the low-pass or high-pass filter of each double second-order filter circuit unit is connected to the next stage unit respectively, so as to achieve a higher order and better cutoff characteristic filtering effect.

[0131] When using a biquad filter circuit to form a filter system, if you want to achieve richer functions, you need to add a selection circuit to achieve adjustable roll-off frequency and passband type. One of the biggest features of a biquad filter is that the high-pass, low-pass and band-pass filter output interfaces are independent of each other. The so-called cascade of biquad filters is to connect the low-pass and high-pass outputs of a biquad circuit to the input of the next biquad circuit, so as to achieve a higher order and a steeper roll-off frequency cutoff characteristic. The logic diagram of the selection circuit is as follows: Figure 8 As shown, B1, B2, B3, and B4 in the figure represent four identical dual-quadratic filter circuits, and their internal dual-quadratic structures are Figure 7 As shown, K1, K2, K3, and K4 represent four multiplexers respectively. The multiplexer chip used in the present invention is ADG5412BF from ADI. The multiplexer contains a four-channel single-pole / single-throw (SPST) switch, which can be controlled by the microcontroller through the I / O port to connect any one of the output and input signals. LP1 and HP1 are respectively the positive phase low-pass response outputs of the first-stage biquad circuit B1 (i.e. Figure 3 LP+ in the circuit and high-pass response output; LP2 and HP2 are respectively the positive phase low-pass response outputs of the second-stage biquad circuit B2 (i.e. Figure 3 LP+) and high-pass response output; LP3 and HP3 are respectively the positive phase low-pass response outputs of the third-stage biquad circuit B3 (i.e. Figure 3 LP+) and high-pass response output; LP4 and HP4 are respectively the positive phase low-pass response output of the fourth-stage biquad circuit B4 (i.e. Figure 3LP+) and high-pass response output; Output1~4 are the output signals of the first-stage K1~fourth-stage K4 multiplexers respectively; Signal is the original signal of the input filter, which is short-circuited and connected to the input end of the multiplexers K1~K4 to achieve the purpose of adjusting the order.

[0132] The low-pass and high-pass outputs in B1, B2, B3 and B4 are connected to selectors K1, K2, K3 and K4 respectively. The program sent by the master control controls the output filter type and filter order. Each dual second-order circuit can achieve a second-order filter circuit. Connecting two dual second-order circuits in series can achieve a fourth-order filter, and so on. At the same time, the original signal is connected to the four selectors. The original signal is connected to achieve other situations except the eighth-order filter. In conjunction with the multi-way selector, the unnecessary dual second-order circuits are short-circuited when different signals are turned on and off. It should be noted that when multiple dual second-order circuits are cascaded, if you want to achieve a high-order low-pass filtering effect, you need to connect the low-pass output of the previous stage to the signal input of the next stage, and the high-pass output pin can be floating; if you want to achieve a high-order low-pass and high-pass filtering effect, you need to connect the high-pass output of the previous stage to the signal input of the next stage, and the low-pass output pin is floating.

[0133] For example, when the 6th-order high-pass mode is selected, the multiplexer K1 disconnects the LP1 and HP1 signal channels output by B1, turns on the channel where the input signal Signal is located and outputs Output1 to the input end of B2; after filtering by the double second-order filter circuit B2, the HP2 loop of the multiplexer K2 is turned on, the other channels are turned off, and the signal is output to Output2 to the input end of the double second-order filter circuit B3; after filtering by the double second-order filter circuit B3, the HP3 loop of the multiplexer K3 is turned on, the other channels are turned off, and the signal is output to Output3 to the input end of the double second-order filter circuit B4; after filtering by the double second-order filter circuit B4, the HP4 loop of the multiplexer K4 is turned on, the other channels are turned off, and the filtered signal Output4 is finally output. By analogy, the system can generate a total of eight output modes with a combination of orders and channel types, including second-order low-pass, second-order high-pass, fourth-order low-pass, fourth-order high-pass, sixth-order low-pass, sixth-order high-pass, eighth-order low-pass, and eighth-order high-pass.

[0134] The fourth step is to design and implement the control module program in C language on the Keil uVision5 platform based on the hardware architecture and circuit connection of the filtering system, which mainly includes the control program of the DAC equivalent variable resistor according to the cutoff frequency and quality factor control word algorithm, the program for writing the true value of each channel of the selection circuit according to the preset mode, the definition and control program of the JSON format command through the serial port, and the program for the microcontroller to communicate with the host computer through the RS485 protocol; the host computer control program is designed based on the GUI on the Matlab platform, and the filter parameters are adjusted through the UI interface displayed by the computer. Only by achieving precise control of key parameters in the program can the excellent characteristics of low noise, low power consumption, compact structure and high device utilization of this analog filtering circuit be maximized.

[0135] The control module in this study is mainly used to realize the functional configuration and working status indication of the analog circuit part. The control module uses the STM32F407VET6 microcontroller as the core processor, receives host computer instructions through the RS485 serial interface, and can also use the serial port debugging assistant to input instructions to drive the microcontroller for instruction transmission. The microcontroller realizes all-round control and adjustment of the filter system through the following methods:

[0136] (1) Communicate with the host computer through the RS485 interface to realize the control and adjustment of the filter system through the MATLAB GUI interface; you can also use the ST-Link downloader and the virtual serial port to perform SWD download and data exchange, and use the serial port debugging assistant to complete the preliminary debugging with the help of the JSON format commands set in the program. In the control panel area of ​​the host computer, enter the cutoff frequency (unit: Hz), select the filter type (Butterworth / Bessel), passband type (LowPass / Highpass), and roll-off frequency (12 / 24 / 36 / 48dB / oct corresponding to 2 / 4 / 6 / 8 orders respectively), switch the cutoff frequency, and click "SET" to complete the setting. The above settings can be read back in real time in the command window at intervals of 1s, and displayed in real time in the real-time display area "Current configuration" at the bottom of the window.

[0137] (2) Input binary values ​​to the digital-to-analog conversion chip DAC in the dual-quadratic filter through SPI communication to generate a suitable output current, so that the DAC can be equivalent to a variable resistor, and its resistance value is finely divided according to the different binary values. Under the premise of a given cutoff frequency and filter type (Butterworth / Bessel), the functional relationship between the cutoff frequency and Q value of the dual-quadratic filter and the specific resistor is reversely deduced based on the previous model database, and then the binary D of the precise input DAC chip is deduced, achieving flexible adjustability of the filter cutoff frequency and filter type parameters;

[0138] (3) The multiplexer in the programmable filter module is controlled through the I / O interface. After receiving the instructions about the passband type (high pass / low pass) and roll-off slope (12 / 24 / 36 / 48dB / oct) from the host computer or the key, the decoding calculation is performed to control the binary value of the multiplexer input to determine the shutdown of a specific path, and its shutdown state is determined by controlling the switch level. Specifically, one of the three outputs of low pass, high pass, and band pass can be selected by controlling the four-choice switch, or one of the switches connected to the input signal can be turned on so that the corresponding part of the filter is bypassed and grounded, so that the part is switched out of the complete cascade filter to achieve the required stopband attenuation slope and passband type; to achieve flexible adjustability of the filter passband type and attenuation slope parameters.

[0139] In order to reduce the size of the system and improve the stability and integration of the system, this system integrates the STM32F407VET6 chip and its minimum system circuit with the remaining modules in the filtering system and their peripheral circuits on a PCB board.

[0140] Experimental verification and performance testing are carried out for the filter under the scenario of adjusting parameters. First, the power module, dual second-order filter circuit, selection circuit combined with the control module and the host computer of the filter system are tested for software and hardware performance and the scheme is verified to evaluate the system circuit performance; then the filter system is connected to the single-beam SERF magnetometer system for overall performance joint debugging test, the sensitivity of the magnetometer prototype when the self-developed filter is used is calibrated, and the actual performance of the filter system in the target application scenario is evaluated.

[0141] (1) The performance of the designed filtering system was tested. The download circuit interface of the filtering system circuit board was connected to the computer using the SWD downloader to ensure the download and synchronization of the control program. The RS485 interface was connected to the computer using a USB to RS485 serial port cable to ensure real-time communication between the host computer and the slave computer. The input and output BNC interfaces were connected to the network analyzer through connecting cables, and the amplitude-frequency characteristic curve was viewed through spectrum analysis.

[0142] By summarizing the data displayed by the network analyzer under different filter types and orders, the following test results can be obtained:

[0143] (1.1) The amplitude-frequency characteristics of the 2nd, 4th, 6th, and 8th order (12, 24, 36, and 48 dB / oct) Butterworth and Bessel low-pass filters at a cutoff frequency of 50 kHz are shown in the figure below: Fig. 9 and Fig.10 shown.

[0144] (1.2) The amplitude-frequency characteristic curves of 2nd, 4th, 6th, and 8th order Butterworth and Bessel high-pass filters at a cutoff frequency of 50kHz are as follows: Fig.11 and Fig.12 shown.

[0145] (1.3) The amplitude-frequency characteristic curve of the Butterworth low-pass filter at cut-off frequencies of 100kHz, 250kHz, and 500kHz is as follows: Fig.13 , Fig.14 and Fig.15 shown.

[0146] The test results show that the filtering system can achieve the filtering effect under the adjustment of the host computer. The Bessel filter attenuates faster at the cutoff frequency than the Butterworth filter, which is consistent with the theoretical analysis. The filter can change the corresponding cutoff frequency and quality factor according to the different binary control words written in the control program. Under each filter type and cutoff frequency, whether it is a high-pass or low-pass filter, the 2nd, 4th, 6th, and 8th order transition band attenuation slopes are significantly different, and all meet the set 12, 24, 36, and 48dB / oct. Regardless of the setting, the filter gain can be maintained at 0dB, and the bandwidth of the filter can reach 1MHz.

[0147] (2) In order to verify the actual application performance of the self-developed filtering system, the filter was connected to the electronic measurement and control system in the single-beam SERF atomic magnetometer test platform for verification. Before the formal measurement, the magnetometer needs to be subjected to 30.5Hz three-axis magnetic compensation. After the gas chamber of the atomic magnetometer of the system is heated and stabilized at 140°C, the photocurrent signal output by the photodetector is converted into a voltage signal by the transimpedance amplifier. The voltage signal enters the filtering system as the input of the filter, and after filtering, it is input to the subsequent phase-locked amplifier for further processing.

[0148] Perform FFT on the collected filter output signal, the result is as follows Fig.16As shown in the FFT results, it can be seen that the filter retains the 1kHz signal containing weak cardio-cerebral magnetic signals while filtering out high-frequency signals. At the same time, there are even harmonics such as 2kHz, 4kHz, and 6kHz, and odd harmonics such as 970Hz, 1030Hz, 2970Hz, 3030Hz, 6970Hz, and 7030Hz after the 30Hz supplementary magnetic field, and the sum and difference frequency signals.

[0149] Perform FFT on the signals before and after filtering, and the results are as follows Fig.17 As shown, by comparison, it can be seen that the filtering system has an absolute inhibitory effect on the high-frequency interference of the magnetometer signal, while retaining the important signal containing weak biomagnetic information in the low-frequency band, and can ensure that the amplitude of the key signal before and after filtering is not distorted.

[0150] Finally, the output signal of the magnetometer connected to the self-developed filtering system was collected to obtain the overall sensitivity of the single-beam SERF atomic magnetometer under the joint test. The results are as follows: Fig.18 As shown, the sensitivity is 30fT / Hz 1 / 2 @30.5Hz, meeting the sensitivity requirements of the magnetometer system.

[0151] The contents not described in detail in the specification of the present invention belong to the prior art known to the professional and technical personnel in the field. It is pointed out here that the above description helps those skilled in the art to understand the invention, but does not limit the protection scope of the invention. Any equivalent replacement, modification and / or simplification of the above description without departing from the essence of the invention falls within the protection scope of the invention.

Claims

1. A dual second-order program-controlled filtering system for a SERF atomic magnetometer, characterized in that: The invention comprises a double second-order filter, wherein the double second-order filter has a signal input interface, the double second-order filter outputs a signal through a multiplexer, the double second-order filter and the multiplexer are respectively connected to a control module, the control module is connected to a host computer, the double second-order filter comprises a first operational amplifier, the negative input end of the first operational amplifier is connected to the input signal end through a first digital-to-analog converter, the output end of the first operational amplifier is connected to the output end of the first operational amplifier through a second digital-to-analog converter and a first capacitor connected in parallel, the output end of the third operational amplifier is connected to the output end of the third operational amplifier through a third digital-to-analog converter, the positive input end of the first operational amplifier is grounded, the output end of the first operational amplifier is connected to the negative input end of the second operational amplifier through a fourth digital-to-analog converter, the negative input end of the second operational amplifier is connected to the output end of the second operational amplifier through a second capacitor, and the positive input end of the second operational amplifier is grounded The output terminal of the second operational amplifier is connected to the negative input terminal of the third operational amplifier through the fifth resistor, the negative input terminal of the third operational amplifier is connected to the output terminal of the third operational amplifier through the sixth resistor, the output terminal of the third operational amplifier is connected to the negative input terminal of the fourth operational amplifier through the ninth resistor, the positive input terminal of the fourth operational amplifier is grounded, the negative input terminal of the fourth operational amplifier is connected to the output terminal of the fourth operational amplifier through the tenth resistor, the output terminal of the first operational amplifier is connected to the negative input terminal of the fourth operational amplifier through the eighth resistor, the input signal terminal is connected to the negative input terminal of the fourth operational amplifier through the seventh resistor, the output terminal of the fourth operational amplifier outputs a high-pass response HP, the output terminal of the third operational amplifier outputs an inverted low-pass response LP-, the output terminal of the second operational amplifier outputs a non-inverting low-pass response LP+, and the output terminal of the first operational amplifier outputs a band-pass response BP.

2. The dual second-order program-controlled filtering system for SERF atomic magnetometer according to claim 1, characterized in that: The host computer uses the frequency control word and the quality factor control word of the digital-to-analog converter DAC through the control module to achieve the filter cutoff frequency f c The quality factor Q is precisely controlled by using the variable resistance characteristic of the DAC through the integrated feedback resistor R fb With f c The functional relationship between Q and filter cutoff frequency is used to realize the flexible adjustability of filter type parameters.

3. The dual second-order program-controlled filtering system for SERF atomic magnetometer according to claim 1, characterized in that: The invention comprises a first biquad filter to a fourth biquad filter, and a first-stage multiplexer to a fourth-stage multiplexer. The original signal Signal is respectively provided to the input end of the first biquad filter and the respective input ends of the first-stage multiplexer to the fourth-stage multiplexer. The first biquad filter provides a first-stage in-phase low-pass response and a first-stage high-pass response to the first-stage multiplexer. The first output signal of the first-stage multiplexer is transmitted to the input end of the second biquad filter. The second biquad filter provides a second-stage in-phase low-pass response and a second-stage high-pass response to the second-stage multiplexer. The second output signal of the second-stage multiplexer is transmitted to the input end of the third biquad filter. The third biquad filter provides a third-stage in-phase low-pass response and a third-stage high-pass response to the third-stage multiplexer. The third output signal of the third-stage multiplexer is transmitted to the input end of the fourth biquad filter. The fourth biquad filter provides a fourth-stage in-phase low-pass response and a fourth-stage high-pass response to the fourth-stage multiplexer. The fourth-stage multiplexer outputs a fourth output signal.

4. The dual second-order program-controlled filtering system for SERF atomic magnetometer according to claim 2, characterized in that: R fb With f c The functional relationship with Q is as follows: D1=2 n *R fb *2π*C1*f c D2=(2 n *R fb *2π*C1*f c ) / Q Where D1 is the binary cutoff frequency control word of the DAC, n is the number of DAC bits, C1 is the first capacitor, and D2 is the binary quality factor control word of the DAC.

5. The dual second-order program-controlled filtering system for SERF atomic magnetometer according to claim 1, characterized in that: The dual second-order filter circuit is connected to the control module via an SPI interface, and the selection circuit is connected to the control module via an I / O interface.

6. The dual second-order program-controlled filtering system for SERF atomic magnetometer according to claim 1, characterized in that: The control module is connected to the host computer via a UART interface.

Images