A white noise histogram-based TDC linearity compensation method

Abstract

The application discloses a TDC linearity compensation method based on white noise histogram, and the method can effectively eliminate the precision error between different phases caused by design, layout wiring and temperature by collecting the white noise histogram, extracting the offset error between different phases by software, and inversely compensating and calibrating the precision between different phases in the TDC. The method combines software and hardware, analyzes the obtained white noise data by software, extracts characteristic parameters, and realizes the flexibility of calibration. The hardware calibrates the ranging histogram data, reduces the application power consumption, and improves the operation speed. Without external complex equipment, the method can extract the characteristic value of the batch chips according to the test results, apply the results to all chips in the same batch, and thus obtain the batch compensation of the chips in the batch; or the method can calibrate and compensate the TDC linearity of single chips at different temperatures before ranging.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of integrated circuits, in particular to a white noise histogram-based TDC linearity compensation method for improving measurement accuracy. BACKGROUND

[0002] Direct time of flight (DTOF) ranging is a ranging method that uses a laser emitter, typically a vertical-cavity surface-emitting laser (VCSEL) to emit a laser pulse to a target object, the laser pulse is reflected back by the target object, and a single-photon-avalanche diode (SPAD) senses the reflected laser pulse and generates an electrical pulse. A time-digital-convertor (TDC) converts the time difference between the emitted and received electrical pulses into a digital signal, and a histogram statistical analysis is performed to obtain the distance of the target object. With the advancement of semiconductor technology and processes, DTOF technology has gained a large market application in industries, automobiles, consumer electronics, and the Internet of Things in recent years due to its strong anti-interference and fast dynamic response in long-distance and complex background light environments. Due to its use of the time-of-flight ranging principle, in high-precision applications, the subtle interference caused by the semiconductor manufacturing process, layout, design, and temperature effects leads to a ranging accuracy error in the speed of light coefficient. For example, a 0.1 ns time jitter corresponds to a 1.5 cm distance error. To address this time jitter, there are two approaches: one is to improve the precision from the process, design, etc., and the other is to compensate after the fact, i.e., to improve the precision based on the related calibration of the finished chip.

[0003] The present application is based on a finished chip, and is used for calibrating and compensating the uniformity of a TDC phase / delay stage. The basic principle of the TDC is to accumulate the count value of a high-precision periodic time slice, such as a PLL phase or delay line, to achieve the digitization of a target time period, thereby obtaining the conversion from the time period to the digital value. The high-precision minimum time slice represents the minimum precision of the TDC. Accordingly, one bin of the histogram corresponds to one high-precision time slice. In theory, the smaller the time slice, the higher the ranging accuracy. However, in practice, if the errors / noise caused by the design or process cannot be effectively controlled, the high precision is submerged in the noise error, and the theoretical high precision loses its significance. The selection of the specific time slice is not discussed in detail. The bin axis of the histogram is the periodic superposition of the periodic time slice, that is, the longer the time that can be measured by the histogram bin axis, the farther the target distance that can be measured, which also means a larger storage RAM. Due to the aforementioned process and design reasons, there is always a certain deviation between the periodic phase / delay units of the TDC, that is, the phase / delay units are not equal. This deviation causes the histogram bin value of some time period to be greater or less than that of the adjacent time period, thereby causing the peak value of the histogram to shift and causing the non-linear error of the ranging. Considering the horizontal characteristics of the ideal white noise histogram, the white noise histogram data of the actual chip is extracted and processed to make the processed histogram present the horizontal characteristics, that is, the compensation of the TDC phase / delay unit is realized. The compensation processing of the actual ranging histogram can realize the calibration and compensation of the actual ranging result, and obtain the expected ideal result. SUMMARY

[0004] The present application provides a TDC linearity compensation method based on a white noise histogram to solve the uniformity problem of the TDC count phase / delay unit.

[0005] To solve the above technical problems, the present application is realized by the following technical scheme:

[0006] This invention discloses a TDC linearity compensation method based on white noise histograms, implemented using a proximity / range sensor chip. The proximity / range sensor data link includes TDC, histograms, and software / hardware-implemented data processing functions. The TDC linearity compensation method comprises two parts: calibration and application. Calibration is used to extract offset parameters from the TDC phase data. The calibration stage involves acquiring and processing the white noise histogram to extract calibration compensation coefficients. Application involves using the parameters obtained before ranging to compensate the ranging histogram data. The application stage is the actual ranging stage, where the compensation system obtained in the calibration stage is used to adjust / compensate the ranging histogram to obtain TDC-homogenized data, which is then used as input for subsequent filtering, peak search, and other digital signal processing (DSP) operations.

[0007] Calibration includes:

[0008] S1. Start calibration;

[0009] S2. White noise acquisition: To maximize data acquisition while ensuring the histogram is not saturated, obtain a white noise histogram;

[0010] S3. White noise histogram analysis: Phase processing is performed on the white noise histogram;

[0011] S4. Extract calibration coefficients: Save the calibration coefficients obtained in the above histogram analysis steps for use in distance measurement compensation during the application phase.

[0012] Application phase:

[0013] P1. Configure calibration coefficients: Configure the calibration coefficients obtained in the calibration phase into the corresponding registers for subsequent application;

[0014] P2. Target ranging: Collect the histogram of the target to be measured;

[0015] P3, Histogram Compensation: Read the histogram data, multiply it sequentially with the corresponding calibration coefficients, shift the product 10 bits to the right to compensate for the amplification done during parameter extraction, and use the result as the calibrated histogram data and as the input bin data for subsequent score search;

[0016] P4. Obtain a frame of ranging results: After the above histogram compensation results are processed by DSP, including smoothing filtering and peak search, the target distance information is obtained.

[0017] If more distance frames are needed, repeat the application phases P1-P4.

[0018] Furthermore, in step S2, bin data of the same phase / delay level are extracted from the white noise histogram according to the actual phase / delay level of TDC, classified and summarized, and the amplitude information of each phase / delay level is obtained; and the amplitude difference parameters of each phase / delay level are extracted according to the uniformity characteristics of white noise, the purpose of which is to achieve uniform normalization of the histogram data after compensation.

[0019] Furthermore, in actual ranging, the bin data of the phase delay level corresponding to the ranging histogram is applied to compensate for the TDC phase / delay level deviation.

[0020] The present invention has the following advantages over the prior art:

[0021] 1. This invention is a calibration algorithm that, through on-chip compensation after chip design, can eliminate the TDC linearity problem caused by chip design, process performance, differences between chips and temperature changes, thereby improving ranging accuracy;

[0022] 2. This invention can adopt a batch compensation mode, that is, by extracting a certain number of sample chips / modules from a large batch of chips produced in the same process and batch, the calibration coefficients of each sample are obtained. Then, data analysis (such as mean method or minimum error method) is performed on the calibration system of these samples to obtain the mean value of calibration coefficients with universal statistical characteristics. This mean value is applied to all chips / modules in the batch to obtain a general performance improvement, thereby reducing costs for mass production.

[0023] 3. This invention is designed for special applications requiring high precision. It can also perform individual calibration on each chip / module to obtain specific calibration coefficients for each chip / module, thereby maximizing performance improvement and increasing the profit margin of a single product.

[0024] Of course, any product implementing this invention does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0025] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 This is a 128-bin, 16-phase white noise histogram collected during the calibration phase of a specific embodiment;

[0027] Figure 2 for Figure 1 The phase histogram is obtained by decomposing the white noise histogram into 16 phases and summing the phases in phase.

[0028] Figure 3 Flowchart of the method of the present invention;

[0029] Figure 4 This is a schematic diagram showing two calibration modes: batch mode and single-chip mode. Detailed Implementation

[0030] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0031] like Figure 3 The diagram shows the flowchart of each step of the specific operation process. The example histogram is 128 bins deep and the TDC is 16 phases. The embodiments described are merely a few specific examples for illustrating the present invention, and the present invention includes but is not limited to these specific examples.

[0032] The present invention discloses a TDC linearity compensation method based on white noise histogram, which is implemented based on a proximity / range sensor chip. The proximity / range sensor data link contains TDC, histogram, and software / hardware implemented data processing functions. The TDC linearity compensation method includes two parts: calibration and application. Calibration is used to extract the offset parameters of TDC phase data, and application is to compensate the range histogram data with the parameters obtained before application in the range measurement.

[0033] like Figure 1 The diagram shows a 128-bin, 16-phase white noise histogram acquired during the calibration phase. The horizontal axis represents the bin value, and the vertical axis represents the SPAD count at the corresponding bin position. Figure 2 for Figure 1 The histogram of neutral white noise is decomposed into 16 phases, and the phase histogram is obtained by summing the in-phase values; for example... Figure 3 This is a flowchart of the system's operation, including the calibration phase and the application phase specifically for distance compensation; such as Figure 4 Two calibration modes are available: batch mode and single-chip mode;

[0034] Calibration includes:

[0035] S1. Start calibration, corresponding to... Figure 3 The first step in;

[0036] S2. White Noise Acquisition: While ensuring the histogram remains unsaturated, maximize data acquisition to obtain a white noise histogram. Specifically, as follows... Figure 1The noise histogram shown is 128 bins deep.

[0037] S3. White noise histogram analysis: Phase processing is performed on the white noise histogram;

[0038] 1. Categorize the bin values ​​from 0 to 127 according to their phase values ​​from 0 to 15.

[0039] bin0->phase0,

[0040] bin1->phase1,

[0041] …

[0042] bin15->phase15,

[0043] bin16->phase0,

[0044] bin17->phase1,

[0045] …

[0046] bin127->phase15

[0047] 2. Sum (or average) the count values ​​for the same phase to obtain... Figure 2 The histogram shown has a total of 16 phases on the horizontal axis. As can be seen from the figure, there is a relatively obvious phase nonlinearity between the phases.

[0048] Dividing the highest peak count value (in this special case, bin2) by the count values ​​of each phase, i.e., highest count value / count value of each phase, yields the calibration coefficients (characteristic parameters) for each phase. Multiplying this set of system values ​​by the corresponding phase count values ​​gives the normalized phase histogram (linearization). Considering hardware implementation, the coefficients need to be integerized. Combined with the maximum count value (e.g., 1024), the coefficients can be amplified, for example, to 1024 (2^32) / 2^32. 10 Rounding up to the nearest integer () times the original value is used to improve accuracy.

[0049] S4. Extract calibration coefficients: Save the calibration coefficients obtained in the above histogram analysis steps. Specifically, save the set of calibration coefficients (defined as coef0~15) obtained in the above histogram analysis for use in distance compensation during the application phase.

[0050] Application phase:

[0051] P1. Configure calibration coefficients: Configure the calibration coefficients obtained in the calibration phase into the corresponding registers for subsequent application;

[0052] P2. Target ranging: Collect the histogram of the target to be measured;

[0053] P3, Histogram Compensation: Read the histogram data, multiply it sequentially with the corresponding calibration coefficients, shift the product 10 bits to the right to compensate for the amplification (1024) made during parameter extraction, and use the result as the calibrated histogram data and as the input bin data for subsequent score search;

[0054] Right now:

[0055] bin0*coef0 / 1024->BIN0

[0056] bin1*coef1 / 1024->BIN1

[0057] …

[0058] bin15*coef15 / 1024->BIN15

[0059] bin16*coef0 / 1024->BIN16

[0060] …

[0061] bin127*coef15 / 1024->BIN127

[0062] P4. Obtain a frame of ranging results: After the above histogram compensation results are processed by DSP, including smoothing filtering and peak search, the target distance information is obtained.

[0063] If more distance frames are needed, repeat the application phases P1-P4.

[0064] In step S2, bin data of the same phase / delay level are extracted from the white noise histogram based on the actual phase / delay level of TDC, classified and summarized, and the amplitude information of each phase / delay level is obtained; and the amplitude difference parameters of each phase / delay level are extracted based on the uniformity characteristics of white noise, with the aim of achieving uniform normalization of the histogram data after compensation.

[0065] In actual ranging, the bin data of the phase delay level corresponding to the ranging histogram is used to compensate for the TDC phase / delay level deviation. In the specific implementation of this invention, hardware implementation is adopted, but software implementation of compensation is also within the scope of protection of this invention.

[0066] The technical implementation of this invention is generally achieved through two stages:

[0067] The system consists of two phases: calibration and application (ranging). The calibration phase involves acquiring and processing white noise histograms to extract calibration compensation coefficients. The application phase is the actual ranging phase, where the compensation system obtained in the calibration phase is used to adjust / compensate the ranging histogram, obtaining TDC-homogenized data, which is then used for subsequent filtering, peak search, and other digital signal processing.

[0068] Digital signal processing (DSP) input.

[0069] This invention acquires white noise histograms, extracts offset errors between different phases using software, and performs reverse compensation calibration on the accuracy of different phases in the TDC (Time-Digital Displacement Control). This effectively eliminates accuracy errors between different phases caused by design, layout, routing, and temperature. It employs a combined hardware and software approach: software analyzes the acquired white noise data and extracts feature parameters, enabling flexible calibration; hardware calibrates the ranging histogram data, reducing power consumption and improving operating speed. Furthermore, without the need for complex external equipment, it can extract feature values ​​from batches of chips based on test results and apply the results to all chips in the same batch, thus achieving batch compensation for that batch. It can also perform TDC linearity calibration and compensation for individual chips at different temperatures before ranging. This achieves improved mass production performance or higher-precision dynamic performance compensation for individual chips.

[0070] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A TDC linearity compensation method based on white noise histogram, characterized in that, Based on a proximity / range sensor chip, the proximity / range sensor data link contains TDC, histogram, and software / hardware implemented data processing functions; the TDC linearity compensation method includes two parts: calibration and application. Calibration is used to extract the offset parameters of the TDC phase data, and application is to compensate the ranging histogram data with the parameters obtained before application in ranging. Calibration includes: S1. Start calibration; S2. White noise acquisition: To maximize data acquisition while ensuring the histogram is not saturated, obtain a white noise histogram; S3. White noise histogram analysis: Phase processing is performed on the white noise histogram; First, the bin values ​​from 0 to 127 are categorized sequentially according to phase from 0 to 15. bin0->phase0, bin1->phase1, … bin15->phase15, bin16->phase0, bin17->phase1, … bin127->phase15 Then, the count data of the same phase are summed or averaged to obtain a histogram. The horizontal axis corresponds to a total of 16 phases, and there is a relatively obvious phase nonlinearity between each phase. Divide the highest peak count value by the count value of each phase, i.e., highest count value / count value of each phase, to obtain the calibration coefficient of each phase; multiply this set of coefficients by the corresponding phase count value to obtain the normalized phase histogram; round the coefficients to integers, combine them with the maximum count value, amplify the coefficients and then round them to improve accuracy; S4. Extract calibration coefficients: Save the calibration coefficients obtained in the above histogram analysis steps. Specifically, save the set of calibration coefficients obtained in the above histogram analysis for use in distance compensation during the application phase. Application phase: P1. Configure calibration coefficients: Configure the calibration coefficients obtained in the calibration phase into the corresponding registers for subsequent application; P2. Target ranging: Collect the histogram of the target to be measured; P3, Histogram Compensation: Read the histogram data, multiply it sequentially with the corresponding calibration coefficients, shift the product 10 bits to the right to compensate for the amplification done during parameter extraction, and use the result as the calibrated histogram data and as the input bin data for subsequent score search; P4. Obtain a frame of ranging results: After the above histogram compensation results are processed by DSP, including smoothing filtering and peak search, the target distance information is obtained. If more distance frames are needed, repeat the application phases P1-P4.

2. The TDC linearity compensation method based on white noise histogram according to claim 1, characterized in that, In step S2, bin data of the same phase / delay level are extracted from the white noise histogram according to the actual phase / delay level of TDC, classified and summarized, and the amplitude information of each phase / delay level is obtained; and the amplitude difference parameters of each phase / delay level are extracted according to the uniformity characteristics of white noise, the purpose of which is to achieve uniform normalization of the histogram data after compensation.

3. The TDC linearity compensation method based on white noise histogram according to claim 2, characterized in that, In actual ranging, the bin data of the phase delay level corresponding to the ranging histogram is used to compensate for the TDC phase / delay level deviation.

Images