Time spectrometer system and method based on microchannel plate photomultiplier (MCP-PMT)

Abstract

The invention discloses a time spectrometer system and method based on a microchannel plate photomultiplier (MCP-PMT). The constant proportion discriminator is triggered by adopting an attenuation branch / delay branch differential zero crossing point, and performs latch width limiting and level conversion on a trigger pulse; fPGA-based multi-level carry chain time TDC is adopted, and parallel tap sampling, bubble suppression and code density calibration table look-up correction are carried out; a processing flow of time window histogram accumulation, a first effective hit strategy, a background adaptive threshold and multi-peak positioning and fitting is adopted. According to the invention, the picosecond-level timing precision is obtained at a low clock rate, the sensitivity and precision of weak light signal flight time distance measurement are significantly improved, and the method is suitable for laser radar imaging and high time sequence of time correlation single photon counting, can realize sub-hundred picosecond-level time resolution, and is suitable for flight time distance measurement under a weak light condition.

Description

Technical Field

[0001] This invention belongs to the field of electronic measurement technology, and further relates to a time spectrometer system and method based on microchannel plate photomultiplier tube (MCP-PMT) single-photon detection in the field of single-photon detection and high-precision time measurement. This invention can be used for precise time interval measurement in the fields of lidar imaging, quantum optics experiments such as time-correlated single-photon counting (TCSPC), and ultra-high time-resolution time-of-flight ranging. Background Technology

[0002] Precise time interval measurement has wide applications in scientific research experiments and practical life and production. In nuclear physics and particle physics experiments, it is necessary to measure the time of flight (TOF) of charged particles and the drift time of ionized electrons. Based on the measurement results, information such as particle momentum and particle trajectory can be obtained, the intrinsic mass of the particle can be deduced, and the type of charged particle can be identified. In high-energy physics experiments, high-precision time spectrometers are often used for particle lifetime detection. Precise time not only plays an important role in fundamental research fields such as nuclear physics, particle physics, geodynamics, relativity, pulsar period studies, and geodesy of artificial satellite dynamics, but also has widespread applications in fields such as aerospace, deep space communication, satellite launch and monitoring, geological mapping, navigation and communication, power transmission, and scientific metrology. In civilian applications, time measurement is widely used in automatic detection equipment, laser ranging, remote sensing imaging, medical image scanning, phase measurement, and frequency measurement. With the rapid development of science and technology, people have increasingly higher requirements for the reliability and time resolution of time measurement.

[0003] Cui Dajian, Ao Tianhong, Xi Shuiqing, et al. disclosed a high-sensitivity detector in their paper "Research Progress on InGaAs Single-Photon Avalanche Focal Plane (Invited)" (Infrared and Laser Engineering, 2023, 52(03): 9-19.). The paper summarizes recent advancements in single-photon detection technology, such as avalanche photodiodes (SPADs) or microchannel plate photomultiplier tubes, achieving sub-single-photon level detection sensitivity, enabling lidar to operate under extremely weak echo conditions. Microchannel plate photomultiplier tubes possess high quantum efficiency, low background noise, and sub-picosecond time broadening characteristics, making them particularly suitable for high time-resolution measurements. However, the detector still has shortcomings: under single-photon detection conditions, the average number of echo photons emitted per laser emission is typically much less than 1, the photon arrival time is randomly distributed, and the pulse amplitude output by the detector fluctuates significantly.

[0004] Suzhou University disclosed a photon counter in its patent application, "Coincidence Photon Counter" (application number 202410135669.5, publication number CN 118190183 A). This photon counter is based on photoelectric conversion and uses a field-programmable gate array (FPGA) to achieve high-speed real-time photon counting. The use of FPGA simplifies and flexibly designs the photon counter circuit, facilitating improvements and upgrades to its functionality. This photon counter adds a dual-channel photon coincidence measurement function (counting and time interval), effectively distinguishing between valid events and accidental coincidences caused by thermal motion, thereby increasing detection efficiency and measurement accuracy. The photon counter improves the light source to a TTL signal-modulated pulsed light source, offering the advantage of low cost and meeting the requirements of single-photon experiments. However, a remaining drawback of this photon counter is the significant time walk error that occurs when the amplitude changes, thus reducing the system's accuracy.

[0005] Huazhong Normal University disclosed a single-photon detector in its patent application, "A Low-Time-Jitter Single-Photon Detector" (Application No. 201710240460.5, Publication No. CN 107024289 B). This single-photon detector uses a limiting amplifier instead of a traditional fixed-threshold comparator to convert avalanche electrical signals with randomly fluctuating amplitudes into digital signals with consistent rise times and equal amplitudes, thereby significantly reducing time jitter. However, this single-photon detector still has shortcomings, including low requirements for time accuracy and limitations imposed by the device itself. Summary of the Invention

[0006] The purpose of this invention is to address the shortcomings of the prior art by providing a time spectrometer system and method based on microchannel plate photomultiplier tube single-photon detection (MCP-PMT). This aims to solve the technical problems in the prior art, such as insufficient ranging sensitivity, inadequate time resolution, and decreased accuracy due to timing jitter under weak light single-photon conditions.

[0007] The technical approach to achieving the objective of this invention is to introduce targeted key mechanisms in the three stages of front-end triggering, time measurement, and statistical calculation, forming a closed-loop synergy. Firstly, the system employs a constant ratio timing discriminator (CFD) with attenuation / delay differential zero-intersection triggering, and implements latching and standard level conversion on the trigger pulse. This ensures that the triggering time is largely unaffected by amplitude under different amplitude and rising edge conditions, thereby solving the problems of time slippage and instability under weak signals that are easily generated by traditional threshold triggering. Its outstanding substantive technical feature lies in: constructing... The amplitude factors at the zero intersection point cancel each other out, achieving first-order insensitivity to amplitude changes; and select and Make zero intersection near Maximum, the time jitter equivalent to noise The first aspect is the significant reduction in voltage. Simultaneously, latching and limiting the pulse width / amplitude of the analog comparator are used to cut off the secondary effects, and the pulse width / amplitude is stably injected into the TDC at LVDS / CMOS standard levels. Secondly, the time measurement stage employs a multi-stage carry chain TDC based on FPGA, combining parallel tap sampling, bubble suppression, and code density calibration lookup table correction. This achieves a significant improvement in equivalent time resolution and linearity without increasing the system clock frequency, avoiding the resolution bottleneck and temperature drift issues of high clock and analog TAC schemes. Its key technical features are: using a delay chain to subdivide a clock cycle into a large number of sub-units (gaining sub-clock interpolation capability), using bubble suppression to ensure the monotonicity of the thermometer code, and then using code density statistics to obtain the true bin width of each tap and mapping the non-uniform time scale to uniform time coordinates using a LUT, thereby simultaneously eliminating DNL / INL and environmental drift. Thirdly, the data processing module employs a time-window histogram accumulation, first-hit-spot strategy, background adaptive threshold, and multi-peak localization and fitting process. This significantly increases the number of effective samples participating in statistics and decision-making under the same acquisition duration and pulse repetition frequency, overcoming problems such as insufficient calculation points, pile-up, and large distance calculation errors caused by high false alarm rates. Its outstanding substantive technical feature lies in: aggregating discrete arrival events into sufficient statistics (histograms) so that the estimated variance increases with the number of samples. The invention proposes a high-precision time spectrometer method based on a microchannel plate photomultiplier tube (MCP-PMT) to address the problem that existing time measurement systems cannot accurately statistically analyze the photon arrival time distribution under extremely weak echo conditions. This invention utilizes a photon arrival time histogram and establishes a single-photon detection probability model and a time resolution error model based on an MCP-PMT single-photon detector, combined with a constant-ratio timing discriminator (CFD) and a high-precision time-to-digital converter (TDC). By constructing a photon arrival time histogram and establishing a single-photon detection probability model and a time resolution error model, it achieves highly sensitive and high-precision measurement of Time-of-Flight (ToF) signals. The method employs a background sliding window to achieve controlled false alarms similar to CFAR, and uses centroid / parabolic / Gaussian (or MLE) estimation and multi-peak separation in the peak region, significantly optimizing the distance and uncertainty in weak signal and multi-target scenarios.

[0008] The technical solution for achieving the objective of this invention is as follows.

[0009] The system of this invention includes a laser emission module, an MCP-PMT single-photon detector, a constant-ratio timing discriminator, a time-to-digital converter, and a data processing module; wherein:

[0010] The laser emitting module is used to generate ranging laser pulses and synchronously output a start reference signal as the Start input of the constant ratio timing discriminator; preferably, a low jitter synchronization path is generated by the laser driver or the emitted pulses are sampled by a high-speed photodiode and shaped to obtain a differential reference, with the start reference jitter not exceeding 10 ps, ​​which matches the input level of the time-to-digital converter;

[0011] The MCP-PMT single-photon detector is used to convert echo photons into nanosecond-level analog electrical signals for low-noise broadband front-end processing and signal splitting: the amplified detection pulse is split into two, one of which serves as the attenuation branch of the constant ratio timing discriminator, and the other serves as the delay reference branch; the passband is preferably ≥1 GHz, and the input and output are matched with 50 Ω to ensure pulse rise time and waveform fidelity;

[0012] The constant ratio timing discriminator includes an attenuation branch path and a delay branch path, which are used to achieve amplitude-independent constant time delay triggering under different amplitude and rising edge conditions, and output a digital trigger signal as the Stop input of the time-to-digital converter.

[0013] The time-to-digital converter is used to measure the time difference Δt between Start and Stop with picosecond resolution; it is preferably implemented based on a multi-level carry chain inside a field-programmable gate array (FPGA).

[0014] The data processing module is used to perform time window mapping and equal-width bin accumulation on the time difference Δt between multiple measurements Start / Stop to form a photon arrival time histogram, and to complete distance extraction and statistical evaluation: it performs peak detection (single / multiple peaks) on the histogram and obtains the echo center arrival time using centroid / parabolic / Gaussian or maximum likelihood fitting. Output distance With uncertainty It estimates the background / dark count rate and false alarm rate to support the "first effective hit" strategy to suppress pile-up, where c represents the speed of light in the propagation medium.

[0015] Furthermore, the constant-ratio timing discriminator is used to achieve amplitude-independent constant-delay triggering under different amplitude and rising-edge conditions, and outputs a digital trigger signal as the Stop input of the time-to-digital converter; it includes an attenuation branch path and a delay branch path, respectively proportional to a certain factor. The input pulse is attenuated by a constant ratio and delayed by a fixed time. After a delay, the two signals form a differential signal in a high-speed differential comparator. When the difference zero intersection occurs and the input amplitude exceeds the threshold; A valid trigger pulse is output at the specified time; the trigger pulse is latched, limited, and level-converted to provide a standard logic level; through optimization... and make At its maximum, the discrimination jitter can be controlled to the order of 3 to 10 ps;

[0016] Furthermore, the input signal Hit of the time-to-digital converter is stably injected into the CARRY delay chain via an input chain composed of D flip-flops; the taps of each CARRY4 are sampled in parallel at the global clock edge and stored in the output registers respectively to form a thermometer code; a fine timing code is obtained through bubble suppression and encoding, and concatenated with a coarse counter to obtain a single Δt; a linearity correction lookup table is established in conjunction with code density testing to compensate for integral nonlinearity INL / differential nonlinearity DNL online, so as to improve the linearity and resolution across the entire temperature range.

[0017] The implementation steps of the method of the present invention include the following:

[0018] Step 1: Generate ranging laser pulses and establish an initial reference signal;

[0019] Step 2: Convert the collected echo photons into nanosecond-level analog electrical signals to establish two analog pulses;

[0020] Step 3: Amplitude-independent triggering of the two analog pulses;

[0021] Step 4: Measure the Start / Stop time difference;

[0022] Step 5: Through statistical analysis and calculation, a histogram of photon arrival times is generated;

[0023] Step 6: Output the distance and quality assessment results to the host computer or external system through the output interface of the data processing module, and feed back the optimized parameters to the laser emission module, MCP-PMT single-photon detector, constant ratio timing discriminator and time-to-digital converter for the next cycle measurement.

[0024] Furthermore, the steps of generating the ranging laser pulse and establishing the initial reference signal are as follows:

[0025] The first step is to select the measurement time window based on the target range. , ,in, This indicates the round-trip time corresponding to the pulse repetition frequency and the maximum range of the time window setting. Where c represents the speed of light in the propagation medium, Indicates the maximum measured distance. This represents the device delay and jitter margin, with a value ranging from 10 to 50 ns;

[0026] The second step is to select a value slightly lower than the upper limit of the cycle. To avoid cross-cycle aliasing;

[0027] The third step is to record the fixed delay of the reference path as a uniform correction value.

[0028] The fourth step is to extract the starting reference signal from the ranging laser pulse, and after limiting and shaping, use it as the Start input of the TDC, and verify that the jitter is ≤10 ps.

[0029] Furthermore, the steps for establishing two analog pulses are as follows:

[0030] The first step is to perform broadband low-noise amplification and 50 Ω matching on the output voltage pulse of the MCP-PMT single-photon detector to maintain consistency between the rising edge and amplitude.

[0031] The second step is to split the amplified pulse in two to form two signals: an attenuation branch for discrimination and a delay branch for time reference.

[0032] The third step involves setting a series current-limiting resistor, a fast clamping diode or TVS to ground / power supply, a small AC coupling capacitor, and a parallel adjustable RC discharge or active baseline recovery at the output of the MCP-PMT single-photon detector; and, after limiting, appropriately shaping the front-end ≥1 GHz to ensure that strong echoes are not saturated.

[0033] The fourth step is to send the two signals into a constant ratio timing discriminator (CFD) using equal-length / equivalent traces or time-adjustable coaxial / microstrip traces, with a group delay mismatch between channels ≤10 ps.

[0034] The fifth step is to send the detection gate width and trigger threshold parameters for this cycle to the CFD and TDC gating to ensure consistency with the measurement time window.

[0035] Furthermore, the step of amplitude-independent triggering of the two analog pulses is as follows:

[0036] Step 1, set the parameters:

[0037] Constant ratio decay ratio : 0.2~0.8, preferably 0.3~0.6;

[0038] Fixed time delay : Equivalent to or slightly smaller than the rising edge of the input pulse, 100–400 ps;

[0039] Minimum amplitude threshold Based on background and noise RMS settings The value of k ranges from 4 to 8 and is constrained by the first hit rate target;

[0040] The second step is to attenuate the input pulse by the attenuation branch according to the set constant ratio. Perform constant attenuation while maintaining waveform shape;

[0041] The third step is to apply a fixed time delay to the input pulse through the delay branch. The delayed signal is used as a time reference;

[0042] The fourth step involves a high-speed differential comparator performing zero-crossing detection on the two signals. If the signal exceeds the minimum amplitude threshold... When the trigger pulse is applied, a trigger pulse is output.

[0043] Step 5, Limiting Width and Level Conversion: Use a monostable / programmable single-pulse generator to shape the trigger pulse into a fixed width, ranging from 5 to 20 nanoseconds; and convert it to a standard logic level that matches the TDC interface, with pulse jitter ≤10 ps;

[0044] The sixth step is to retain only the first valid trigger within the same measurement time window and suppress statistical bias caused by subsequent pile-up.

[0045] Furthermore, the step of measuring the Start / Stop time difference is as follows:

[0046] The first step is to connect Start and Stop to the start and hit channels of the time-to-digital converter, respectively.

[0047] The second step is to inject a multi-level carry chain delay structure into Stop after it is synchronized with the input trigger chain.

[0048] The third step is to sample each tap of the delay chain in parallel at the system clock edge to obtain the thermometer code.

[0049] The fourth step is to perform bubble suppression and encoding on the thermometer code to obtain the detailed timing value;

[0050] The fifth step is to combine the results with the coarse counter value to form a single time difference measurement result;

[0051] Step 6: Using a jitter source / asynchronous high-stability reference independent of the system clock, uniformly scan the delay chain with the Hit function and count the hits of each tap. And calculate the normalized bin width: By the accumulated boundary: Generate a LUT; process the original code. Corrected to: ;in, This represents the count of the i-th tap being hit. This represents the normalized bin width of the hit count for the i-th tap;

[0052] The seventh step involves dynamically adjusting the lookup table based on the temperature value detected by the temperature sensor and the power status detected by the power monitoring module, using a pre-stored temperature compensation coefficient, and real-time correcting the trigger threshold based on power voltage fluctuations to ensure consistent measurement across the entire temperature range.

[0053] Step 8: Detect and remove anomalies such as initial loss, double triggering, and illegal code type, and output the valid time difference sequence.

[0054] Furthermore, the solution steps are as follows:

[0055] The first step will be effective Mapping to equal-width bins within a preset time window and accumulating them forms a photon arrival time histogram; the effective... This refers to the time difference measurement value output after abnormalities are detected and eliminated in step 4.

[0056] The second step involves peak detection and peak region localization. Peak detection supports both single-peak and multi-peak modes, and sub-bin-level time estimation is performed in the peak region to obtain the echo arrival time. Peak detection and peak region localization first involve smoothing the histogram using 3- to 5-point median filtering or Gaussian smoothing, then combining this with an adaptive threshold to determine candidate peaks. The adaptive threshold is the background mean plus m times the background standard deviation. Where m ranges from 3 to 8, a window is taken within n bins before and after the candidate peak bin, where n ranges from 2 to 6; sub-bin estimation is performed using one of centroid, parabolic, or Gaussian fitting methods; when the peak spacing is less than several bins, EM / double Gaussian or constrained nonlinear fitting is used for separation; finally, quality evaluation is performed, and the outputs SNR, peak width FWHM, fitting residuals, etc. And the false alarm rate, and filter / rank multiple targets accordingly;

[0057] The third step is to convert the flight time into the target distance: It also outputs the following quality metrics: peak height, integral intensity, SNR, and distance uncertainty. False alarm rate; where represents the speed of light in the propagation medium. This is expressed as an estimate of the echo arrival time; the fourth step is to estimate the background / darkness count rate based on the tail window or sliding window, and adaptively update the threshold / gate width / integration count;

[0058] The fifth step is to perform peak separation and result sorting under multi-target echo conditions, and output the multi-target distance sequence.

[0059] Furthermore, the optimized parameters include, but are not limited to:

[0060] Parameters related to the constant ratio timing discriminator CFD: constant ratio coefficient Fixed delay Trigger threshold Latch pulse width and input equivalent delay balancing;

[0061] Parameters related to the time-to-digital converter (TDC): code density calibration lookup table (LUT), bubble suppression parameters, coarse and fine timing stitching bit width, temperature compensation coefficient, and recalibration cycle;

[0062] Time window and gating related parameters: Measurement time window First effective hit switch / strategy, detection gate width, integration count and PRF upper limit settings;

[0063] Background adaptation-related parameters: background estimation window width, threshold coefficient Dynamic integration strategy;

[0064] Output and host computer interface: multi-peak sorting threshold, quality index threshold and anomaly removal rules;

[0065] The above parameters are updated based on the statistical results and quality indicators from step 5, and take effect in the next cycle to achieve closed-loop optimization.

[0066] Compared with the prior art, the invention has the following advantages:

[0067] First, because the constant-ratio timing discriminator in the system of this invention uses attenuation branch / delay branch differential zero-intersection triggering and performs latching and level conversion on the trigger pulse, the triggering time is basically unaffected by the amplitude under echo conditions with different amplitudes and different rising edges. This overcomes the shortcomings of existing technologies, such as threshold triggering easily causing time slippage and instability under weak signals. The system of this invention can maintain stable triggering accuracy under conditions of weak echo, strong background, and limited front-end bandwidth, and achieve highly reliable time-of-flight measurement under low signal-to-noise ratio, thus improving ranging stability and repeatability.

[0068] Secondly, because the method of this invention employs a multi-stage carry chain time (TDC) based on FPGA and combines parallel tap sampling, bubble suppression, and code density calibration lookup table correction, the equivalent time resolution and linearity are significantly improved without increasing the system clock frequency. This overcomes the problems of limited resolution and large temperature drift caused by existing technologies that rely on high clock speeds or analog TAC. This allows the invention to achieve picosecond-level timing accuracy at lower clock rates, reducing the requirements for hardware operating frequencies and dedicated components, and significantly improving system consistency and engineering feasibility.

[0069] Third, because the method of this invention employs time window histogram accumulation, a first-hit-success strategy, background adaptive thresholding, and multi-peak localization and fitting processing, the number of effective samples participating in statistics and decision-making is significantly increased under the same acquisition duration and pulse repetition frequency. This overcomes the problems of insufficient calculation points, pile-up bias, and large distance calculation errors when the false alarm rate is high in existing technologies. This allows the invention to achieve sub-picosecond time resolution and millimeter-level distance resolution even under extremely low signal-to-noise ratio and multi-echo conditions, significantly improving the measurement capabilities of high-temporal-resolution applications such as lidar imaging and time-correlated single-photon counting. Attached Figure Description

[0070] Figure 1 This is a functional flowchart of a time spectrometer provided in an embodiment of the present invention;

[0071] Figure 2 This is a block diagram of the TCSPC system based on a multi-level carry chain in an FPGA according to the present invention;

[0072] Figure 3 This is a flowchart of the constant fractional timing discriminator architecture described in the method of the present invention. Detailed Implementation

[0073] The present invention will now be further described with reference to the accompanying drawings and embodiments.

[0074] Figure 1 This is a flowchart illustrating the time spectrometer function provided in an embodiment of the present invention. Figure 1The modules shown from top to bottom are: laser emission module → MCP-PMT single-photon detector → constant ratio timing discriminator → time-to-digital converter (TDC) → data processing module. Figure 1 The signal chain is also marked as follows: optical signal, analog electrical signal, digital signal, and time data / count flow.

[0075] The working principle of the system link is as follows Figure 1 As shown, the laser emission module outputs a synchronization start reference signal as the Start of the TDC. The echo photons are detected by the MCP-PMT and output nanosecond-level analog electrical signals; the CFD performs constant proportional threshold triggering on the analog pulses and outputs digital signals; the TDC measures the time difference Δt using Start / Stop; the data processing module maps multiple Δt values ​​to time windows and accumulates them by bin to form a photon arrival time histogram, realizing distance extraction and statistical evaluation.

[0076] Let N be the average number of photons returned per pulse. s If the system's equivalent detection efficiency is η, then the probability of "detecting at least one signal photon" is:

[0077] ;

[0078] Probability of background / dark counting First valid hit:

[0079] ;

[0080] The input pulse is divided into an attenuation branch and a delay branch to form a differential. Zero intersection point Corresponding to a fixed proportion of trigger times It is not sensitive to amplitude. The CFD residual jitter approximation is:

[0081] ;

[0082] By optimizing α, Maximize |dv / dt|. Accuracy can be controlled within 3~10 ps. The entire simulation process parameters include light pulse generation:

[0083] ;

[0084] in, Indicates the center time of the Gaussian signal. This represents the time standard deviation of a Gaussian signal.

[0085] Delay and channel response will Delay The electrical signal is obtained by convolving it with h(t) according to the attenuation of reflectivity:

[0086]

[0087] CFD triggering:

[0088] ;

[0089] When s(t) > A digital edge is generated at a time, and the trigger time is recorded. .

[0090] Figure 2 This is a block diagram of the TCSPC system based on a multi-level carry chain in an FPGA, which can greatly improve the linearity of the system compared to a single carry chain method.

[0091] See Figure 2 The Time-to-Digital Converter (TDC) measures Δt using Start / Stop. This invention employs multi-level carry chain technology within an FPGA, effectively improving the linearity of the TDC. In the multi-level carry chain time-to-digital converter circuit, the differential signal output from the comparator is processed through the main carry chain (4 levels) and four sub-carry chains within the FPGA to achieve a TDC system with high linearity and high time resolution.

[0092] In the multi-level carry-to-digital converter circuit, a portion of the input signal (Hit) passes through an input chain composed of multiple D flip-flops and enters the multi-level carry chain (CARRY4). Each CARRY4 unit contains four D flip-flops, and the input signal is passed sequentially through these flip-flops to form the carry chain. The clock signal (CLOCK) serves as the clock signal for the D flip-flops. The clock signal (CLOCK) is used to synchronize the operation of all D flip-flops. Each register obtains data from different D flip-flops, and the output register collects and stores this data for subsequent processing after the time-to-digital conversion.

[0093] The output registers include output register 1, output register 2, output register 3, and output register 4. A delay chain is constructed using cascaded delay units (Carry) within the FPGA, with each Carry4 containing four delay units. The COUT of each Carry4 is connected to the CIN of the next Carry4, forming a delay chain through cascading. Each COUT acts as a tap output to a flip-flop, and data is sampled using an external clock to obtain sampling results Q1, Q2, Q3, and Q4. Compared to traditional tapped delay chains that store all results in a single memory, this topology-structured delay chain stores the four results generated in a single Carry4 separately, further reducing the error caused by "bubble" data. In the design, identical bits in each 4-bit carry chain (e.g., C0 uses a new carry chain, C1 uses a new carry chain) form a main chain, and there is also a 4-level sub-carry chain structure to improve the linearity of the time-delayed conversion (TDC).

[0094] Figure 3 This is a flowchart of the constant-ratio timing discriminator architecture described in this invention. The nanosecond-level weak electrical signal output from the MCP-PMT is amplified and divided into a discrimination path and a delay path. The differential trigger and delay chain module uses a precision delay line and a proportional attenuator to form two inputs, one of which goes directly to a high-speed comparator. The other input is delayed and then input to the comparator, forming a fixed-delay differential trigger to achieve constant-ratio timing discrimination. In this embodiment of the invention, the digital conversion and logic output module includes a high-speed comparator (MAX9691), a latch (MC10EP51), and a level conversion circuit. It is responsible for shaping the differential trigger output into a standard TTL logic pulse and using it as the "stop" signal for the TDC input.

[0095] See Figure 3The circuit architecture of this time-spectrum analyzer system mainly consists of an analog front-end amplification and splitting module, a differential triggering and delay chain module, and a digital conversion and logic output module. The analog front-end amplification and splitting structure uses a high-speed, low-noise OPA847 operational amplifier to amplify the weak nanosecond-level pulse signal output from the MCP-PMT detector. A splitting and biasing network using MCH4021 RF transistors divides the amplified signal into two branches: one branch serves as the constant-ratio timing discrimination (CFD) signal path, and the other branch serves as the delay reference path. The differential processing and delay chain are implemented by dividing the input pulse into two paths and then performing differential triggering processing on each: one path is attenuated by a constant ratio and directly fed to a high-speed comparator as a reference; the other path is fed to the other input of the comparator after passing through a precisely matched delay line, forming a delay chain with a fixed time difference. The two signals form a differential input in the comparator. When the pulse amplitude rises to a preset ratio, the comparator is triggered to flip, achieving constant-ratio timing discrimination. The high-speed comparator preferably uses the MAX9691, whose ECL differential input structure ensures a constant-delay trigger signal output under pulses of varying amplitudes and rising edges, making it insensitive to time deviations caused by amplitude variations. The narrow pulse output by the comparator enters the digital conversion and logic output module for logic shaping and level conversion: the comparator output pulse is latched by a high-speed ECLD flip-flop to limit the output pulse width and prevent glitch triggering, and then the signal level is converted to a standard TTL output by a TTL output driver circuit. Finally, this module outputs a standard digital pulse corresponding to the photon arrival time, which can be directly used as the "stop" signal input for the TDC or for subsequent digital processing units to acquire and process. In the signal processing flow, after the single-photon echo arrives at the MCP-PMT, it is converted into an analog electrical signal. This pulse is amplified by the OPA847 front end and simultaneously enters the discrimination branch and the delay branch. The comparator performs constant-ratio threshold differential triggering on the two inputs. When the input pulse reaches the set ratio threshold, it outputs a digital trigger pulse signal corresponding to the photon arrival time. The digital pulse, after being shaped by subsequent logic gates and latches, is converted into a standard TTL logic level output trigger signal, thus providing a precise and stable timing pulse. This trigger pulse serves as the stop signal input to the time measurement circuit (TDC), and together with the laser emission start reference, enables high-precision measurement of the time of flight Δt.

[0096] The system module design of this invention has significant advantages: First, the analog path has extremely high bandwidth, enabling distortion-free transmission and amplification of picosecond-level broadened detection signals, ensuring the fidelity of the time signal and rapid response; second, the use of constant ratio timing technology combined with fully differential discrimination triggering significantly improves triggering time accuracy and reduces time travel errors caused by changes in input pulse amplitude, making the output trigger delay essentially unaffected by signal amplitude; furthermore, the output stage simultaneously provides both TTL and other logic level signals, offering strong compatibility and facilitating interface with subsequent high-precision TDC timing modules or other digital processing circuits to achieve seamless signal acquisition and processing.

[0097] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the technical concept of the present invention should be included within the protection scope of the present invention.

Claims

1. A time-spectrum analyzer system based on a microchannel plate photomultiplier tube (MCP-PMT), characterized in that, The system includes a laser emission module, an MCP-PMT single-photon detector, a constant-ratio timing discriminator, a time-to-digital converter, and a data processing module; among which: The laser emitting module is used to generate ranging laser pulses and synchronously output a start reference signal as the Start input of the constant ratio timing discriminator; preferably, a low jitter synchronization path is generated by the laser driver or the emitted pulses are sampled by a high-speed photodiode and shaped to obtain a differential reference, with the start reference jitter not exceeding 10 ps, ​​which matches the input level of the time-to-digital converter; The MCP-PMT single-photon detector is used to convert echo photons into nanosecond-level analog electrical signals for low-noise broadband front-end processing and signal splitting: the amplified detection pulse is split into two, one of which serves as the attenuation branch of the constant ratio timing discriminator, and the other serves as the delay reference branch; the passband is preferably ≥1 GHz, and the input and output are matched with 50 Ω to ensure pulse rise time and waveform fidelity; The constant ratio timing discriminator includes an attenuation branch path and a delay branch path, which are used to achieve amplitude-independent constant time delay triggering under different amplitude and rising edge conditions, and output a digital trigger signal as the Stop input of the time-to-digital converter. The time-to-digital converter is used to measure the time difference Δt between Start and Stop with picosecond resolution; it is preferably implemented based on a multi-level carry chain inside a field-programmable gate array (FPGA). The data processing module is used to perform time window mapping and equal-width bin accumulation on the time difference Δt between multiple Start / Stop measurements to form a photon arrival time histogram, and to complete distance extraction and statistical evaluation: it performs peak detection (single / multiple peaks) on the histogram and obtains the echo center arrival time using centroid / parabolic / Gaussian or maximum likelihood fitting. Output distance: With uncertainty It estimates the background / dark count rate and false alarm rate to support the "first effective hit" strategy to suppress pile-up, where c represents the speed of light in the propagation medium.

2. The time spectrometer system according to claim 1, characterized in that, The constant ratio timing discriminator uses a constant ratio attenuation coefficient respectively. For input pulse Perform constant-proportional decay with a fixed time delay After a delay, the two signals form a differential signal in a high-speed differential comparator: When the difference zero intersection occurs and the input amplitude exceeds the threshold; A valid trigger pulse is output at the specified time; the trigger pulse is latched, limited, and level-converted to provide a standard logic level; through optimization... and ,make At its maximum, the discrimination jitter can be controlled to the order of 3-10 ps. This indicates a modulo operation.

3. The time spectrometer system according to claim 1, characterized in that, The input signal Hit of the time-to-digital converter is stably injected into the CARRY delay chain through an input chain composed of D flip-flops; the taps of each CARRY4 are sampled in parallel at the global clock edge and stored in the output register respectively to form a thermometer code; after bubble suppression and encoding, a fine timing code is obtained, which is concatenated with a coarse counter to obtain a single Δt. A linearity correction lookup table is established in conjunction with code density testing to compensate for integral nonlinearity (INL) and differential nonlinearity (DNL) online, thereby improving linearity and resolution across the entire temperature range.

4. A time spectrometer method based on a microchannel plate photomultiplier tube (MCP-PMT) according to claim 1, characterized in that, The implementation steps of this method include the following: Step 1: Generate ranging laser pulses and establish an initial reference signal; Step 2: Convert the collected echo photons into nanosecond-level analog electrical signals to establish two analog pulses; Step 3: Amplitude-independent triggering of the two analog pulses; Step 4: Measure the Start / Stop time difference; Step 5: Through statistical analysis and calculation, a histogram of photon arrival times is generated; Step 6: Output the distance and quality assessment results to the host computer or external system through the output interface of the data processing module, and feed back the optimized parameters to the laser emission module, MCP-PMT single-photon detector, constant ratio timing discriminator and time-to-digital converter for the next cycle measurement.

5. The time-spectrum method according to claim 4, characterized in that, The steps in step 1 for generating the ranging laser pulse and establishing the initial reference signal are as follows: The first step is to select the measurement time window based on the target range. , ,in, This indicates the round-trip time corresponding to the pulse repetition frequency and the maximum range of the time window setting. Where c represents the speed of light in the propagation medium, Indicates the maximum measured distance. This represents the device delay and jitter margin, with a value ranging from 10 to 50 ns; The second step is to select a value slightly lower than the upper limit of the cycle. To avoid cross-cycle aliasing; The third step is to record the fixed delay of the reference path as a uniform correction value. The fourth step is to extract the starting reference signal from the ranging laser pulse, and after limiting and shaping, use it as the Start input of the TDC, and verify that the jitter is ≤10 ps.

6. The time-spectrum method according to claim 4, characterized in that, The steps for establishing two analog pulses in step 2 are as follows: The first step is to perform broadband low-noise amplification and 50 Ω matching on the output voltage pulse of the MCP-PMT single-photon detector to maintain consistency between the rising edge and amplitude. The second step is to split the amplified pulse in two to form two signals: an attenuation branch for discrimination and a delay branch for time reference. The third step involves setting a series current-limiting resistor, a fast clamping diode or TVS to ground / power supply, a small AC coupling capacitor, and a parallel adjustable RC discharge or active baseline recovery at the output of the MCP-PMT single-photon detector; and, after limiting, appropriately shaping the front-end ≥1 GHz to ensure that strong echoes are not saturated. The fourth step is to send the two signals into a constant ratio timing discriminator (CFD) using equal-length / equivalent traces or time-adjustable coaxial / microstrip traces, with a group delay mismatch between channels ≤10 ps. The fifth step is to send the detection gate width and trigger threshold parameters for this cycle to the CFD and TDC gating to ensure consistency with the measurement time window.

7. The time-spectrum method according to claim 4, characterized in that, The steps for amplitude-independent triggering of the two analog pulses described in step 3 are as follows: Step 1, set the parameters: Constant ratio decay ratio : 0.2~0.8, preferably 0.3~0.6; Fixed time delay : Equivalent to or slightly smaller than the rising edge of the input pulse, 100–400 ps; Minimum amplitude threshold Based on background and noise RMS settings The value of k ranges from 4 to 8 and is constrained by the first hit rate target; The second step is to attenuate the input pulse by the attenuation branch according to the set constant ratio. Perform constant attenuation while maintaining waveform shape; The third step is to apply a fixed time delay to the input pulse through the delay branch. The delayed signal is used as a time reference; The fourth step involves a high-speed differential comparator performing zero-crossing detection on the two signals. If the signal exceeds the minimum amplitude threshold... When the trigger pulse is applied, a trigger pulse is output. Step 5, Limiting Width and Level Conversion: Use a monostable / programmable single pulse generator to shape the trigger pulse into a fixed width, ranging from 5 to 20 nanoseconds; and convert it to a standard logic level that matches the TDC interface, with pulse jitter ≤10ps; The sixth step is to retain only the first valid trigger within the same measurement time window and suppress statistical bias caused by subsequent pile-up.

8. The time-spectrum method according to claim 4, characterized in that, The steps for measuring the Start / Stop time difference described in step 4 are as follows: The first step is to connect Start and Stop to the start and hit channels of the time-to-digital converter, respectively. The second step is to inject a multi-level carry chain delay structure into Stop after it is synchronized with the input trigger chain. The third step is to sample each tap of the delay chain in parallel at the system clock edge to obtain the thermometer code. The fourth step is to perform bubble suppression and encoding on the thermometer code to obtain the detailed timing value; The fifth step is to combine the results with the coarse counter value to form a single time difference measurement result; Step 6: Using a jitter source / asynchronous high-stability reference independent of the system clock, uniformly scan the delay chain with the Hit function and count the hits of each tap. And calculate the normalized bin width: By the accumulated boundary: Generate a LUT; process the original code. Corrected to: ;in, This represents the count of the i-th tap being hit. This represents the normalized bin width of the hit count for the i-th tap; The seventh step involves dynamically adjusting the lookup table based on the temperature value detected by the temperature sensor and the power status detected by the power monitoring module, using a pre-stored temperature compensation coefficient, and real-time correcting the trigger threshold based on power voltage fluctuations to ensure consistent measurement across the entire temperature range. Step 8: Detect and remove anomalies such as initial loss, double triggering, and illegal code type, and output the valid time difference sequence.

9. The time-spectrum method according to claim 4, characterized in that, The solution steps described in step 5 are as follows: The first step is to calculate the time difference between Start and Stop. Mapping to equal-width bins within a preset time window and accumulating them forms a photon arrival time histogram; the effective... This refers to the time difference measurement value output after abnormalities are detected and eliminated in step 4. The second step involves peak detection and peak region localization. Peak detection supports both single-peak and multi-peak modes, and sub-bin-level time estimation is performed in the peak region to obtain the echo arrival time. Peak detection and peak region localization first involve smoothing the histogram using 3- to 5-point median filtering or Gaussian smoothing, then combining this with an adaptive threshold to determine candidate peaks. The adaptive threshold is the background mean plus m times the background standard deviation. Where m ranges from 3 to 8, a window is taken within n bins before and after the candidate peak bin, where n ranges from 2 to 6; sub-bin estimation is performed using one of centroid, parabolic, or Gaussian fitting methods; when the peak spacing is less than several bins, EM / double Gaussian or constrained nonlinear fitting is used for separation; finally, quality evaluation is performed, and the outputs SNR, peak width FWHM, fitting residuals, etc. And the false alarm rate, and filter / rank multiple targets accordingly; The third step is to convert the flight time into the target distance: It also outputs the following quality metrics: peak height, integral intensity, SNR, and distance uncertainty. False alarm rate; where represents the speed of light in the propagation medium. This is expressed as an estimate of the echo arrival time; The fourth step is to estimate the background / darkness count rate based on the tail window or sliding window, and adaptively update the threshold / door width / integration count. The fifth step is to perform peak separation and result sorting under multi-target echo conditions, and output the multi-target distance sequence.

10. The time-spectrum method according to claim 4, characterized in that, The optimized parameters mentioned in step 6 include, but are not limited to: Parameters related to the constant ratio timing discriminator CFD: constant ratio coefficient Fixed delay Trigger threshold Latch pulse width and input equivalent delay balancing; Parameters related to the time-to-digital converter (TDC): code density calibration lookup table (LUT), bubble suppression parameters, coarse and fine timing stitching bit width, temperature compensation coefficient, and recalibration cycle; Time window and gating related parameters: Measurement time window First effective hit switch / strategy, detection gate width, integration count and PRF upper limit settings; Background adaptation-related parameters: background estimation window width, threshold coefficient Dynamic integration strategy; Output and host computer interface: multi-peak sorting threshold, quality index threshold and anomaly removal rules; The above parameters are updated based on the statistical results and quality indicators from step 5, and take effect in the next cycle to achieve closed-loop optimization.

Images