Time-to-digital conversion assisted successive approximation analog-to-digital conversion system

By combining the quantization code weight relationship between SAR and TDC for internal calibration, the problem of accurate calibration of the TDC time quantization scale in the SAR and TDC hybrid analog-to-digital conversion system is solved, which improves conversion accuracy and stability, reduces circuit overhead and power consumption, and enhances environmental adaptability.

CN123122552APending Publication Date: 2026-11-17SHANGHAI FORMULA MICROELECTRONICS CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611431604.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-09-15
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

In a hybrid analog-to-digital converter system combining SAR and TDC, how can we achieve accurate calibration of the time quantization scale of TDC while reducing additional reference and matching circuits, in order to improve conversion accuracy and stability, especially under conditions of process, power supply voltage, and temperature variations?

Method used

By combining SAR successive approximation quantization with TDC time-domain quantization based on comparator decision time, and utilizing the equivalent quantization weight relationship between SAR quantization code and time quantization code, internal calibration is performed to adjust the delay parameters in the time-to-digital conversion circuit, thereby establishing a reliable time quantization reference.

Benefits of technology

It improves the conversion accuracy and stability of the SAR and TDC hybrid analog-to-digital conversion system, reduces additional circuit overhead and power consumption, lowers calibration errors, and enhances the system's environmental adaptability.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123122552A_ABST
    Figure CN123122552A_ABST
Patent Text Reader

Abstract

This application relates to the field of integrated circuit technology and discloses a time-to-digital conversion-assisted successive approximation analog-to-digital conversion system, comprising: a successive approximation analog-to-digital conversion circuit, including a capacitive digital-to-analog converter, a comparator, and successive approximation logic. The capacitive digital-to-analog converter is configured to perform successive approximation conversion on the sampled signal according to the control of the successive approximation logic to generate a successive approximation quantization result, and to form a residual signal at the input of the comparator; a time-to-digital conversion circuit, configured to quantize the decision time of the comparator in judging the residual signal to generate a time quantization result; and a calibration circuit, configured to adjust at least one delay parameter in the time-to-digital conversion circuit so that the corresponding time quantization code and the successive approximation quantization code satisfy a preset weight correspondence. This application can improve the conversion accuracy and stability of a hybrid SAR and TDC analog-to-digital conversion system.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This application relates to the field of integrated circuit technology, and in particular to a time-digital conversion-assisted successive approximation analog-to-digital conversion system. Background Technology

[0002] Successive Approximation Register Analog-to-Digital Converters (SAR ADCs) are widely used in high-speed, high-efficiency analog-to-digital conversion systems due to their relatively simple structure, high proportion of digital circuitry, and high energy efficiency. To achieve a performance trade-off between different application requirements, existing technologies have proposed SAR ADCs with dual-mode comparators, capable of switching between at least two operating modes, such as high-speed or dynamic mode and low-noise or calibrated mode. In high-speed mode, the comparator typically operates in enhanced regeneration mode to achieve minimum decision delay and maximum operating speed, but with little or no compensation for offset and noise. In high-precision mode, input equivalent noise and offset effects are reduced by changing bias conditions, introducing preamplifier structures, or enabling calibration paths, thereby improving decision reliability.

[0003] However, as the successive approximation process enters the low-bit conversion stage, the residual signal at the comparator input gradually decreases, and the impact of comparator noise, offset, and finite settling time on the low-bit quantization result increases accordingly. To balance high-speed conversion and low-bit quantization accuracy, some analog-to-digital converter (ADC) structures combine successive approximation quantization with time-domain quantization. In this type of structure, the correlation between the decision time of the dynamic comparator and its input residual amplitude can be utilized, allowing the comparator to be used simultaneously as a voltage-to-time converter (VTC). The decision time of the comparator is then quantized by a time-to-digital converter (TDC), thereby obtaining a time-quantization result related to the input residual amplitude. Thus, based on the high-bit conversion achieved using SAR quantization, low-bit quantization can be further assisted by time-domain information to improve the overall conversion efficiency of the ADC.

[0004] However, the delay characteristics of time-to-digital converters are often susceptible to variations in process, power supply voltage, and temperature (PVT). Therefore, in hybrid analog-to-digital converter architectures that combine SAR and TDC, accurately calibrating the time quantization scale of TDC is a crucial issue for ensuring analog-to-digital conversion performance.

[0005] In existing technologies, deviations in delay units within a time-delay control (TDC) caused by variations in process technology, voltage, and temperature can be estimated and compensated by replicating delay chains, delay matching circuits, or dedicated PVT tracking circuits. However, such calibration methods typically require additional reference delay paths, matching circuits, or other auxiliary calibration circuits, which can increase circuit area and power consumption. Furthermore, since the reference delay path is not entirely identical to the actual signal path formed by the comparator, multiplexer, and TDC during actual operation, even with matching methods, local mismatches may still exist, causing the PVT variations reflected by the reference path to not fully correspond to the time quantization deviations in the actual operating path.

[0006] Therefore, how to establish a reliable time quantization reference using the existing quantization information inside the analog-to-digital converter while reducing additional reference and matching circuits, and how to accurately calibrate the delay parameters in the TDC affected by PVT variations, so as to improve the conversion accuracy and stability of the SAR and TDC hybrid analog-to-digital conversion system, remains a technical problem that needs to be solved in this field. Summary of the Invention

[0007] The purpose of this application is to provide a successive approximation analog-to-digital conversion system assisted by time-to-digital conversion, which improves the conversion accuracy and stability of the SAR and TDC hybrid analog-to-digital conversion system.

[0008] In a first aspect, the present invention provides a time-digital conversion-assisted successive approximation analog-to-digital conversion system, comprising: A successive approximation analog-to-digital converter circuit includes a capacitive digital-to-analog converter, a comparator, and successive approximation logic. The capacitive digital-to-analog converter is configured to perform successive approximation conversion on a sampled signal according to the control of the successive approximation logic, so that the successive approximation logic generates a successive approximation quantization result, and the input terminal of the comparator forms a residual signal corresponding to the sampled signal. A time-to-digital converter circuit is configured to quantize the decision time of the comparator in judging the residual signal to generate a time quantization result, wherein at least a portion of the quantization code in the successive approximation quantization result has the same equivalent quantization weight as at least a portion of the time quantization code in the time quantization result. The calibration circuit is configured to adjust at least one delay parameter in the time-to-digital conversion circuit based on the equivalent quantization weight relationship between the successive approximation quantization result and the time quantization result, so that the corresponding time quantization code and the successive approximation quantization code satisfy a preset weight correspondence.

[0009] In a preferred embodiment, the time-to-digital conversion circuit includes a plurality of time shift units and a plurality of residual analytical delay units, and the calibration circuit is configured to adjust at least one of the plurality of time shift units and the plurality of residual analytical delay units so that the corresponding time quantization code and the successive approximation quantization code satisfy the equivalent quantization weight.

[0010] In a preferred embodiment, the time-to-digital conversion circuit includes n time shift units, n residual resolution delay units, n+1 flip-flops, and n+1 clock delay units. The inputs of the n time shift units and one of the clock delay units are coupled to the successive approximation logic via switches. The outputs of the n time shift units are simultaneously coupled to the input of one of the residual resolution delay units and the data input of one of the flip-flops. The remaining residual resolution delay units are sequentially coupled to the output of the remaining residual resolution delay unit. The data inputs of the remaining flip-flops are respectively coupled to the outputs of the n residual resolution delay units. The remaining clock delay units are sequentially coupled to the output of the remaining clock delay unit. The output of the remaining clock delay unit is coupled to the clock input of the remaining flip-flops. The outputs of the remaining clock delay units are sequentially coupled to the clock inputs of the remaining flip-flops. The n+1 flip-flops output n+1 time quantization codes for each transition in each of the multiple low-noise transition cycles in the lower bits.

[0011] In a preferred embodiment, the successive approximation conversion sequentially includes a plurality of high-speed conversion cycles for high-order bits and a plurality of low-noise conversion cycles for low-order bits. The time-to-digital converter is configured to generate a plurality of time quantization codes in each of the plurality of low-noise conversion cycles. The successive approximation quantization codes generated sequentially by the successive approximation logic in the plurality of low-noise conversion cycles have the same equivalent quantization weight as the highest-order time quantization codes generated sequentially by the time-to-digital converter in the plurality of low-noise conversion cycles.

[0012] In a preferred embodiment, the low-noise conversion cycle includes three conversion cycles, and the equivalent quantization weights of the successive approximation quantization codes generated by the successive approximation logic in the three conversion cycles are 16, 8, and 4, respectively. The time-to-digital conversion circuit generates four time quantization codes in each of the three conversion cycles, and the equivalent quantization weights among the four time quantization codes are 4, 3, 2, and 1, respectively.

[0013] In a preferred embodiment, the time-to-digital conversion circuit includes first to third time shift units, first to third residual analysis delay units, first to fourth flip-flops, and first to fourth clock delay units. The inputs of the first to third time shift units and the input of the fourth clock delay unit are respectively coupled to the successive approximation logic via switches. The outputs of the first to third time shift units are simultaneously coupled to the input of the third residual analysis delay unit and the data input of the fourth flip-flop. The second and first residual analysis delay units are sequentially coupled to the output of the third residual analysis delay unit. The data inputs of the third to first flip-flops are respectively coupled to the outputs of the third to first residual analysis delay units. The third to first clock delay units are sequentially coupled to the output of the fourth clock delay unit. The output of the fourth to first clock delay unit is coupled to the clock input of the fourth to first flip-flops. The fourth to the first triggers output the fourth to the first time quantization codes in each of the three conversion cycles.

[0014] In a preferred embodiment, the successive approximation logic outputs a comparator clock to the comparator and the time-to-digital converter circuit. The falling edge of the transition phase of the comparator clock determines the triggering time of the fourth flip-flop for time quantization. The rising edge of the transition phase of the comparator clock determines the timing at which the output signals of the first to third time shift units are sent to the fourth flip-flop.

[0015] In a preferred embodiment, the calibration circuit adjusting at least one delay parameter in the time-to-digital conversion circuit based on the equivalent quantization weight relationship between the successive approximation quantization result and the time quantization result further includes: The third time shift unit is adjusted so that the fourth time quantization code generated in the first low-noise conversion cycle corresponds to and is equal to the successive approximation quantization code generated in the second low-noise conversion cycle. Adjust the second and third residual analysis delay units so that the second time quantization code generated in the first low-noise conversion cycle corresponds to the successive approximation quantization code generated in the third low-noise conversion cycle; The third time shift unit is adjusted so that the second time quantization code generated in the first low-noise conversion cycle corresponds to and is equal to the successive approximation quantization code generated in the second low-noise conversion cycle. Adjust the third residual analysis delay unit so that the third time quantization code generated in the first low-noise conversion cycle is equal to the sum of the successive approximation quantization codes generated in the second and third low-noise conversion cycles; Adjust the second residual analysis delay unit so that the second time quantization code generated in the first low-noise conversion cycle corresponds to the successive approximation quantization code generated in the second low-noise conversion cycle; Adjust the first residual analysis delay unit so that the first time quantization code generated in the first low-noise conversion cycle corresponds to the successive approximation quantization code generated in the third low-noise conversion cycle. Adjust the second time shift unit so that the second time quantization code generated in the second low-noise transition cycle corresponds to and is equal to the successive approximation quantization code generated in the third low-noise transition cycle; Adjust the first residual resolution delay unit so that the second time quantization code generated in the third low-noise conversion cycle is equal to the first time quantization code generated in the second low-noise conversion cycle.

[0016] In a preferred embodiment, after the calibration circuit adjusts the first to third residual analysis delay units, the delay of the third residual analysis delay unit is... The delay of the second residual analysis delay unit is The delay of the first residual analysis delay unit is ,in is the time constant.

[0017] It should be understood that, within the scope of this invention, the above-described technical features of this invention and the technical features specifically described below (such as in the embodiments) can be combined with each other to form new or preferred technical solutions. Due to space limitations, they will not be described in detail here. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. It should be understood that the accompanying drawings described below are merely some implementation examples of the present invention, and those skilled in the art can obtain other implementation examples based on these drawings without creative effort.

[0019] Figure 1 A schematic diagram of a sub-analog-to-digital converter with one channel is shown.

[0020] Figure 2 A schematic diagram of the time-to-digital conversion circuit 20 is shown.

[0021] Figure 3 A schematic diagram of the internal timing of a sub-analog-to-digital converter for one channel is shown.

[0022] Figure 4 The comparator decision time characteristics and the code weight relationship between SAR code and TDC code are shown.

[0023] Figure 5 The step-by-step calibration and code weight alignment relationship between the time translation unit and the residual analytical delay unit is shown. Detailed Implementation

[0024] In the following description, many technical details are presented to help the reader better understand this application. However, those skilled in the art will understand that the technical solutions claimed in this application can be implemented even without these technical details and various variations and modifications based on the following embodiments.

[0025] Through extensive and in-depth research, the inventors have proposed a time-digital conversion-assisted successive approximation analog-to-digital conversion system. This system combines SAR successive approximation quantization with TDC time-domain quantization based on comparator decision time. It utilizes a pre-constructed code weight relationship between the SAR quantization code and the time quantization code as an internal calibration benchmark, enabling the SAR quantization code to directly serve as a "weight scale" for TDC delay calibration. During calibration, by sequentially adjusting the delay units within the TDC, an equivalent quantization weight correspondence is established between the TDC time quantization code in different conversion cycles and the corresponding SAR quantization code or the time quantization code in adjacent conversion cycles. This achieves self-calibration of the TDC time quantization scale without requiring additional high-precision time references, duplicate delay chains, or dedicated PVT tracking circuits, and improves the conversion accuracy and stability of the SAR / TDC hybrid analog-to-digital conversion system under variations in process technology, voltage, and temperature.

[0026] Compared with existing technologies, this application has at least the following beneficial effects and advantages: By combining SAR successive approximation quantization with TDC time-domain quantization based on comparator decision time, high-order quantization is performed rapidly using SAR in the front-end stage, and residual information is further extracted using the comparator decision time in the low-order conversion stage, thereby improving low-order quantization accuracy while maintaining conversion speed; furthermore, by setting redundant or overlapping equivalent quantization weight relationships between SAR quantization codes and TDC time quantization codes, and using these equivalent quantization weight relationships as the internal calibration benchmark of TDC, the time shift delay unit and residual analysis delay unit in TDC can be directly calibrated based on the system's own quantization results, thereby... Reducing reliance on additional reference delay chains, matching circuits, or dedicated PVT tracking circuits helps lower overhead, power consumption, and area. Simultaneously, since the calibration process is based on the code weight correspondence in the actual comparator, multiplexer, and TDC signal path, calibration errors caused by mismatch between the reference path and the actual operating path can be reduced, improving the accuracy and stability of the TDC time quantization scale under varying process, power supply voltage, and temperature conditions. Furthermore, by establishing the time code weight correspondence between different conversion cycles, multiple delay units can be calibrated step-by-step, forming a consistent time quantization scale, which helps improve the overall conversion accuracy, robustness, and environmental adaptability of the hybrid SAR-TDC analog-to-digital conversion system.

[0027] Terminology Explanation SAR: Successive Approximation Register. Like a binary search, it gradually approximates the input voltage by changing one capacitor weight at a time and observing the comparator's sign.

[0028] CDAC: A capacitor array that combines sampling and digital-to-analog conversion. After sampling, it generates a test voltage by switching the reference terminal, gradually reducing the input residual.

[0029] VTC: Voltage to Time Conversion. Dynamic comparators flip quickly when faced with large input differences and slowly when faced with very small input differences. Therefore, the comparator's decision time itself carries information about the residual magnitude.

[0030] TDC: Time-to-Digital Converter. It compares the comparator decision time with the internal delay unit and outputs discrete-time quantized code to supplement the lowest few bits of quantization.

[0031] Redundancy: Different SAR codes and TDC codes can represent the same or adjacent simulation weights. This "overlapping area" can both absorb errors from previous periods and serve as an internal scale for calibrating TDC delays.

[0032] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.

[0033] One embodiment of this application relates to a time-to-digital conversion-assisted successive approximation analog-to-digital converter system, wherein the successive approximation analog-to-digital converter system has a multi-channel architecture. Figure 1 The structure of a single-channel sub-ADC is shown, each sub-channel including a successive approximation ADC circuit 10, a time-to-digital converter circuit 20, and a calibration circuit (not shown). The successive approximation ADC circuit 10 includes a capacitor-type digital-to-analog converter (CDAC), a comparator 11, and successive approximation (SAR) logic. The input of the comparator 11 is coupled to the input voltage VIN via a sampling switch. The CDAC includes a capacitor array, with one end of each capacitor coupled to the input of the comparator 11, and the other end of each capacitor coupled to a common-mode voltage via a sampling switch and simultaneously coupled to different reference power supplies via a switching switch. During the sampling phase, the sampling switch is closed, the switching switch is open, and the other end of each capacitor is coupled to the common-mode voltage. The capacitor array completes the sampling of the input voltage. After sampling, the sampling switch is opened, and the sampled signal is held in the capacitor array. Then, the conversion phase begins. The CDAC is configured to perform a successive approximation conversion on the sampled signal according to the control of the SAR logic, so that the SAR logic produces a successive approximation quantization result and the input of the comparator 11 forms a residual signal corresponding to the sampled signal. The SAR logic switches the capacitor combination of the CDAC based on the positive or negative result of each output of comparator 11, causing the residual signal at the comparator input to gradually approach zero. The time-to-digital converter circuit 20 is configured to quantize the decision time of the comparator 11 in judging the residual signal, to generate a time-quantized result. At least a portion of the quantization codes in the successive approximation quantization result have the same equivalent quantization weight as at least a portion of the time quantization codes in the time quantization result. The calibration circuit is configured to adjust at least one delay parameter in the time-to-digital converter circuit 20 based on the equivalent quantization weight relationship between the successive approximation quantization result and the time quantization result, so that the corresponding time quantization code and the successive approximation quantization code satisfy a preset weight correspondence.

[0034] The time-to-digital conversion circuit 20 includes multiple time shift units and multiple residual analytical delay units. The calibration circuit is configured to adjust at least one of the multiple time shift units and multiple residual analytical delay units so that the corresponding time quantization code and the successive approximation quantization code satisfy equivalent quantization weights.

[0035] In one embodiment, the time-to-digital converter 20 includes n time shift units, n residual resolution delay units, n+1 flip-flops, and n+1 clock delay units. The inputs of the n time shift units and one of the clock delay units are coupled to SAR logic via switches. The outputs of the n time shift units are simultaneously coupled to the input of one of the residual resolution delay units and the data input of one of the flip-flops. The remaining residual resolution delay units are sequentially coupled to the output of the aforementioned residual resolution delay unit. The data inputs of the remaining flip-flops are coupled to the outputs of the n residual resolution delay units. The remaining clock delay units are sequentially coupled to the output of the aforementioned clock delay unit. The output of the aforementioned clock delay unit is coupled to the clock input of the aforementioned flip-flop. The outputs of the remaining clock delay units are sequentially coupled to the clock inputs of the remaining flip-flops. The n+1 flip-flops output n+1 time quantization codes for each transition in each of the multiple low-noise transition cycles in the lower bits.

[0036] Figure 2A schematic diagram of a time-to-digital converter circuit 20 in one embodiment is shown. The time-to-digital converter circuit 20 includes a first time shift unit TM1, a second time shift unit TM2, a third time shift unit TM3, a first residual analysis delay unit TL1, a second residual analysis delay unit TL2, a third residual analysis delay unit TL3, a first flip-flop DFF1, a second flip-flop DFF2, a third flip-flop DFF3, a fourth flip-flop DFF3, a first clock delay unit TC1, a second clock delay unit TC2, a third clock delay unit TC2, and a fourth clock delay unit TC4. The input of the first time shift unit TM1 is coupled to one end of a first switch S1, the input of the second time shift unit TM2 is coupled to one end of a second switch S2, the input of the third time shift unit TM3 is coupled to one end of a third switch S3, and the input of the fourth clock delay unit TC4 is coupled to one end of a fourth switch S4. The other ends of the first to fourth switches S1-S4 are all coupled to SAR logic. The first to fourth switches S1-S4 are selected to be turned on, forming a multiplexer (MUX). The outputs of the first time shift unit TM1, the second time shift unit TM2, and the third time shift unit TM3 are simultaneously coupled to the input of the third residual analytical delay unit TL3 and the data input of the fourth flip-flop DFF3. The second residual analytical delay unit TL2 and the first residual analytical delay unit TL1 are sequentially coupled to the output of the third residual analytical delay unit TL3. The data input of the third flip-flop DFF3 is coupled to the output of the third residual analytical delay unit TL3, and the data input of the second flip-flop DFF2 is coupled to the output of the second residual analytical delay unit TL2. The data input of the first flip-flop DFF1 is coupled to the output of the third residual analytical delay unit TL3. The input is coupled to the output of the first residual analytical delay unit TL1. The third clock delay unit TC3, the second clock delay unit TC2, and the first clock delay unit TC1 are sequentially coupled to the output of the fourth clock delay unit TC4. The output of the fourth clock delay unit TC4 is coupled to the clock input of the fourth flip-flop DFF3. The output of the third clock delay unit TC3 is coupled to the clock input of the third flip-flop DFF3. The output of the second clock delay unit TC2 is coupled to the clock input of the second flip-flop DFF2. The output of the first clock delay unit TC1 is coupled to the clock input of the first flip-flop DFF1.

[0037] The successive approximation conversion of CDAC includes multiple high-speed conversion cycles of the high bits and multiple low-noise conversion cycles of the low bits. The time-to-digital converter 20 is configured to generate multiple time quantization codes in each of the multiple low-noise conversion cycles. The successive approximation quantization codes generated sequentially by the SAR logic in the multiple low-noise conversion cycles have the same equivalent quantization weight as the highest-order time quantization codes generated sequentially by the time-to-digital converter 20 in the multiple low-noise conversion cycles. In one embodiment, the high-speed conversion cycle includes 6 conversion cycles, and the low-noise conversion cycle includes 3 conversion cycles. The equivalent quantization weights of the successive approximation quantization codes generated by the SAR logic in the 3 low-noise conversion cycles are 16, 8, and 4, respectively. The time-to-digital converter 20 generates 4 time quantization codes in each low-noise conversion cycle of the 3 conversion cycles. The equivalent quantization weights among the 4 time quantization codes are 4, 3, 2, and 1, respectively. Furthermore, the equivalent quantization weights of the highest-order bit time quantization codes generated by the time-to-digital converter 20 in the 3 low-noise conversion cycles are 16, 8, and 4, respectively, which are the same as the equivalent quantization weights of the successive approximation quantization codes.

[0038] It should be understood that the above values ​​are only illustrative examples. In this application, the successive approximation quantization codes generated sequentially by the SAR logic in multiple low-noise conversion cycles are set to have the same equivalent quantization weight as the highest bit time quantization codes generated sequentially by the time-to-digital converter in multiple low-noise conversion cycles. The self-calibration of the TDC time quantization scale is achieved according to the corresponding weight relationship. In other embodiments of this application, other numbers of high-speed conversion cycles, other numbers of low-noise conversion cycles, and different numbers of time quantization codes generated in each low-noise conversion cycle are all within the scope of the concept protected by this invention, and this invention does not limit them.

[0039] Figure 3 The internal timing diagram of a sub-ADC for one channel is shown, with a conversion cycle spanning 9 clock cycles: the first 6 cycles are in high-speed (high-noise) mode, and the following 3 cycles are in low-speed (low-noise) mode. The SAR logic generates successive approximation quantization codes B7, B8, and B9 sequentially in the 3 low-noise conversion cycles. The fourth flip-flop DFF3, the third flip-flop DFF3, the second flip-flop DFF2, and the first flip-flop DFF1 output the fourth time quantization code F4(i), the third time quantization code F3(i), the second time quantization code F2(i), and the first time quantization code F1(i) respectively in each of the 3 conversion cycles, where i represents the cycle number.

[0040] The SAR logic outputs the comparator clock CKC to comparator 11 and the time-to-digital converter circuit 20. The falling edge of the conversion phase of the comparator clock CKC determines the triggering time of the fourth flip-flop DFF4 for time quantization, i.e., it determines the rising edge of the CKFF signal. The rising edge of the conversion phase of the comparator clock CKC determines the timing when the first time shift unit TM1, the second time shift unit TM2, and the third time shift unit TM3 select the output signal to the fourth flip-flop DFF3, i.e., it determines the rising edge of the DFF signal.

[0041] The calibration circuit adjusts at least one delay parameter in the time-to-digital conversion circuit 20 based on the equivalent quantization weight relationship between the successive approximation quantization result and the time quantization result, and further includes steps 1 to 4.

[0042] Step 1: Adjust the third time shift unit TM3 so that the fourth time quantization code F4(7) generated in the first low-noise transition cycle corresponds to and is equal to the successive approximation quantization code B8 generated in the second low-noise transition cycle. Then, adjust the second residual analysis delay unit TL2 and the third residual analysis delay unit TL3 so that the second time quantization code F2(7) generated in the first low-noise transition cycle corresponds to and is equal to the successive approximation quantization code B9 generated in the third low-noise transition cycle.

[0043] Step 2: Adjust the third time shift unit TM3 so that the second time quantization code F2(7) generated in the first low-noise conversion cycle is equal to the successive approximation quantization code B8 generated in the second low-noise conversion cycle. Then, adjust the third residual analysis delay unit TL3 to the first residual analysis delay unit TL1 in sequence. Adjust the third residual analysis delay unit TL3 so that the sum of the third time quantization code F3(7) generated in the first low-noise conversion cycle, the successive approximation quantization code B8 generated in the second low-noise conversion cycle, and the successive approximation quantization code B9 generated in the third low-noise conversion cycle is equal to the sum of the two codes. Adjust the second residual analysis delay unit TL2 so that the second time quantization code F2(7) generated in the first low-noise conversion cycle is equal to the successive approximation quantization code B8 generated in the second low-noise conversion cycle. Adjust the first residual analysis delay unit TL1 so that the first time quantization code F1(7) generated in the first low-noise conversion cycle is equal to the successive approximation quantization code B9 generated in the third low-noise conversion cycle. Step 3: Adjust the second time shift unit TM2 so that the second time quantization code F2(8) generated in the second low-noise conversion cycle is equal to the successive approximation quantization code B9 generated in the third low-noise conversion cycle.

[0044] Step 4: Adjust the first residual analysis delay unit TL1 so that the second time quantization code F2(9) generated in the third low-noise conversion cycle is equal to the first time quantization code F1(8) generated in the second low-noise conversion cycle.

[0045] After the calibration circuit adjusts the first to third residual analysis delay units TL3, the delay of the third residual analysis delay unit TL3 is: The delay of the second residual analytical delay unit TL2 is The first residual analytical delay unit TL1 is ,in is the time constant.

[0046] To better understand the technical solution of this application, a specific example is provided below. The details listed in this example are mainly for ease of understanding and are not intended to limit the scope of protection of this application.

[0047] refer to Figure 1 As shown, the input voltage VIN first enters the CDAC. During the sampling phase, the CDAC stores the input voltage. During the conversion phase, the SAR logic switches the capacitor combination of the CDAC according to the positive or negative result of each output of comparator 11, so that the residual signal at the input of comparator 11 gradually approaches zero.

[0048] The output of comparator 11 is fed to the SAR logic to determine which direction the CDAC should switch in next. Simultaneously, the clock or toggle time of comparator 11 is sent to the 2.5-bit Vernier TDC via the MUX. The preceding conversion cycles primarily use the comparison result, while the subsequent lower-order conversion cycles use both the comparison result and the decision time. The time quantization code generated by the TDC and the SAR quantization code are added digitally with weights to form the final digital result of the sub-ADC.

[0049] The dynamic latch comparator incorporates internal positive feedback regeneration. The larger the residual signal at the input, the faster the positive feedback forms a clear high or low level; conversely, the smaller the residual signal, the closer the two internal nodes are to equilibrium, and the longer it takes to establish the final result. Therefore, the comparator output (0 or 1) indicates the sign of the residual, while the time it takes to output the absolute value of the residual signal determines the outcome. This allows the same comparator to perform two tasks: quickly determining the sign in the first few conversion cycles, and precisely observing the decision time in low-noise mode in the later conversion cycles. This eliminates the need for a separate high-linearity VTC, saving power and area, and avoiding mismatch caused by an additional analog path.

[0050] refer to Figure 2As shown, TM1, TM2, and TM3 are time shift units used to move the lowest effective time weight from the previous cycle to the current cycle; TL1, TL2, and TL3 are residual analytical delay units that divide the decision time within a low-bit conversion cycle into several intervals. The F4(i), F3(i), F2(i), and F1(i) output by the triggers are the time quantization codes for the i-th cycle.

[0051] The term 2.5-bit does not simply mean two and a half physical bits, but rather that the TDC provides approximately four subdivision levels and retains additional out-of-bounds detection capability. The outermost time quantization code F4(i) participates in quantization and also acts as an alarm code for "too early" or "too late".

[0052] refer to Figure 3 As shown, the nine conversion cycles of the successive approximation analog-to-digital converter (ADC) are divided into six high-speed conversion cycles and three low-noise conversion cycles. The first six cycles correspond to larger high-order weights, and the residual signal is still relatively large. Therefore, the comparator adopts a high-speed mode: increasing the tail current or regeneration capability to quickly determine the positive or negative result. This mode has higher input reference noise, but its impact on the high-order result is relatively small. The last three cycles correspond to very small low-order residuals. At this time, the comparator switches to a low-speed, low-noise mode, allowing for a longer decision time to reduce input reference noise. The extended decision process is not an ineffective wait, because the TDC quantizes this time into additional low-order information.

[0053] CKC is the comparator clock. The interval marked C performs voltage domain SAR conversion, and the interval marked R resets the comparator and related nodes. After the last three cycles, the comparator's toggle time simultaneously becomes the input for time domain conversion. DFF is the data signal fed into the outermost flip-flop DFF4 of TDC, carrying the actual time when the comparator completes its decision. CKFF is the corresponding trigger clock; the interval marked T performs time-to-digital conversion, and the interval marked R resets the related nodes of TDC. TDC actually compares the order of DFF and CKFF after different delay paths.

[0054] If all nine cycles use high-speed mode, the last few bits will be significantly affected by comparator noise; if all cycles use low-noise mode, the conversion time and power consumption will increase significantly. The dual-mode design allocates performance according to the residual scale of each cycle: high-bit priority for speed, low-bit priority for noise, while using TDC to convert the long decision time of the low-bit bits into effective information.

[0055] refer to Figure 4 As shown in (a) above, the horizontal axis Vin represents the differential input residual of the comparator, and a logarithmic scale is used in the figure; the vertical axis tcmp represents the time from when the comparator is triggered to when it produces a reliable digital result. The decision time of the dynamic comparator in the small-signal region can be approximated as: T0 is a fixed delay that still exists even when the input is large. V0 is a characteristic voltage used for normalization, related to the regeneration time constant. Due to the existence of ln(·), the decision time increases by an approximately fixed time difference as the input residual decreases proportionally; therefore, it appears as a near-linear line on a graph with the horizontal axis being logarithmic and the vertical axis being linear.

[0056] TDC does not need to directly recover a continuous voltage; it only needs to determine which interval the decision time falls within. To ensure that adjacent time quantization codes correspond to specified amplitude code weights, the three residual resolution delays are set according to the adjacent code weight ratio: the delay of TL3 is... The delay of TL2 is The delay of TL1 is For example, when two magnitude levels differ by a factor of two, the decision time difference is approximately... When the grade ratio is 4:3 or 3:2, the required time difference is respectively and This transforms the nonlinear time characteristics of the comparator into discrete digital codes that can be added according to weights.

[0057] Figure 4 Table (b) in the table is a normalized example of the SAR and TDC output weights. The SAR bit B7 weight in the 7th cycle is 16, corresponding to the weights of the four time quantization codes of TDC: 16, 12, 8, and 4. The weights in the 8th cycle are halved, becoming 8, 6, 4, and 2 respectively. The weights in the 9th cycle are halved again, becoming 4, 3, 2, and 1 respectively. These values ​​are not arbitrarily arranged, but simultaneously satisfy two types of relationships. The first type is binary scaling between cycles, for example, B7:B8:B9 = 16:8:4. The second type is the overlap between SAR and TDC, for example, F4(7) and B7 are both 16, F2(7) and B8 are both 8, and F1(7) and B9 are both 4. The same weight can be generated by different paths, so the statistical results of the two paths can be compared. If the two are inconsistent, the difference is not an unknown input signal, but mainly reflects the TDC delay deviation.

[0058] By utilizing the VTC operating characteristics of the comparator, a hybrid architecture of SAR and TDC is realized. The time-domain transformation is highly sensitive to PVT (process, voltage, temperature) variations. Unlike previous TDCs that relied on PVT tracking or matching circuits, the calibration algorithm proposed in this invention utilizes the inherent redundancy between the SAR bit and the TDC bit. First, TM3 is adjusted until F4(7) is aligned with B8 (B8 corresponds to half of the normal weight). Next, TL3 is adjusted until F2(7) matches B9; then TM3 is readjusted so that F2(7) is equal to B8 (thus restoring TM3 to its normal delay state). Through this alignment operation, F4(7) can correspond to B7; combined with the comparator's output signal, overflow or underflow detection can be achieved. After establishing this benchmark, the delay parameters TL1 to TL3 of the TDC can be determined by comparing B8 and B9 with the TDC output results. For example, in the 7th SAR cycle, the TDC output value F3(7) should be equal to the sum of B8 and B9, and the value of TL3 can be determined accordingly. Similarly, TL2 is determined by comparing F2(7) with B8, and TL1 is determined by comparing F1(7) with B9. Finally, the parameter values ​​of TM2 and TM1 are updated by comparing B9 with F2(8) and F1(8) with F2(9), respectively.

[0059] Figure 5 A schematic diagram of the calibration algorithm is shown. SAR outputs B7-B9 and TDC outputs F1(i)-F4(i) are first combined according to their respective weights, and then the difference e is calculated. Since the two codes being compared ideally represent the same analog quantity, the average difference of multiple samples should be close to zero. If the difference is consistently positive or negative, it indicates that the corresponding delay unit is too long or too short. μ in the diagram represents the update step size. The calibration circuit can use sign gradient, successive approximation, or binary search: the delay code is increased or decreased according to the sign of e, repeated sampling and averaging are performed until the error falls within the allowable window. The averaging function suppresses random noise from the comparator and variations in the input samples.

[0060] Figure 5 The green box indicates the delay unit currently being adjusted. The red bidirectional arrows indicate two code weight positions that need to be actively aligned by the calibration loop; the blue bidirectional arrows indicate self-alignment positions naturally established by the proportional relationship after the previous step. The dashed lines transfer the reference established in the previous step to the next. During calibration, code weights are calibrated sequentially for each low-noise conversion cycle, progressing from the outermost to the innermost. The specific calibration process includes the following steps.

[0061] Step 1: Establish a coarse reference First, adjust TM3 to align the outermost time quantization code F4(7) of the 7th cycle with the weight of B8, so that e(TM3) = B8 - F4(7) is close to zero. At this time, a semi-weighted reference is used, which makes it easier to move the outermost boundary into the observable range. Then adjust the total delay of TL2 and TL3 to align F2(7) with B9.

[0062] Step 2: Calibrate the internal scale of the 7th period Adjust TM3 again to align F2(7) with B8, restoring the normal time shift position. At this time, F4(7) corresponds to B7, and the outermost code can handle overflow / underflow detection.

[0063] Then, the three analytical delays are calculated using the known code weights: F3(7) = B8 + B9 to mark TL3; F2(7) = B8 to mark TL2; and F1(7) = B9 to mark TL1. In this way, the four time positions of code weights 16, 12, 8, and 4 in the 7th period are all determined.

[0064] Step 3: Calibrate the 8th cycle time shift TM2 The weight of the 8th cycle is half that of the 7th cycle. Adjust TM2 to align F2(8) with B9, both with a weight of 4. Since TL1-TL3 have been established in the previous step, the other time positions of the 8th cycle will be constrained proportionally.

[0065] Step 4: Calibrate TM1 using adjacent period self-alignment Finally, adjust TM1 to align F1(8) of the 8th cycle with F2(9) of the 9th cycle, with both having a weight of 2. This cross-cycle equality locks the last time shift unit and simultaneously establishes the 4, 3, 2, 1 scale of the 9th cycle.

[0066] Traditional TDC (Time-of-Conversion) methods often use replication delay chains or dedicated matching circuits to track process, power supply voltage, and temperature variations, but local mismatches still exist between the replicated path and the actual comparator path. This invention compares the SAR / TDC code weights generated within the same actual comparator, the same MUX, and the same TDC path, fixing the common delay as mostly canceling out common-mode terms. Therefore, the calibration results directly target the actual operating path. The chip can perform foreground calibration upon power-up and recalibrate when temperature or power supply changes exceed thresholds. If the system allows for the insertion of training samples, slow background updates can also be performed within the normal data stream.

[0067] The term “coupled to” and its derivatives may be used in this document. “Coupled” can mean two or more elements in direct physical or electrical contact. However, “coupled” can also mean two or more elements in indirect contact with each other, but still cooperating or interacting with each other, and can mean one or more other elements coupled or connected between elements referred to as being coupled to each other.

[0068] The various technical features disclosed in the above-described invention, the various technical features disclosed in the following embodiments and examples, and the various technical features disclosed in the accompanying drawings can be freely combined to form various new technical solutions (all of which should be considered as having been recorded in this specification), unless such a combination of technical features is technically infeasible. For example, in one example, feature A+B+C is disclosed, and in another example, feature A+B+D+E is disclosed. Features C and D are equivalent technical means that serve the same function, and technically only one needs to be used; it is impossible to use both simultaneously. Feature E can be technically combined with feature C. Therefore, the solution A+B+C+D should not be considered as having been recorded because it is technically infeasible, while the solution A+B+C+E should be considered as having been recorded.

[0069] All references to this specification are considered to be incorporated integrally into the disclosure of this application so that they can serve as the basis for modifications if necessary. Furthermore, it should be understood that the above descriptions are merely preferred embodiments of this specification and are not intended to limit the scope of protection of this specification. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of one or more embodiments of this specification should be included within the scope of protection of one or more embodiments of this specification.

[0070] In some cases, the actions or steps described in the claims and specification can be performed in a different order than that shown in the embodiments and still achieve the desired results. Furthermore, the processes depicted in the drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

Claims

1. A time-digital conversion-assisted successive approximation analog-to-digital conversion system, characterized in that, include: A successive approximation analog-to-digital converter circuit includes a capacitive digital-to-analog converter, a comparator, and successive approximation logic. The capacitive digital-to-analog converter is configured to perform successive approximation conversion on a sampled signal according to the control of the successive approximation logic, so that the successive approximation logic generates a successive approximation quantization result, and the input terminal of the comparator forms a residual signal corresponding to the sampled signal. A time-to-digital converter circuit is configured to quantize the decision time of the comparator in judging the residual signal to generate a time quantization result, wherein at least a portion of the quantization code in the successive approximation quantization result has the same equivalent quantization weight as at least a portion of the time quantization code in the time quantization result. The calibration circuit is configured to adjust at least one delay parameter in the time-to-digital conversion circuit based on the equivalent quantization weight relationship between the successive approximation quantization result and the time quantization result, so that the corresponding time quantization code and the successive approximation quantization code satisfy a preset weight correspondence.

2. The system as described in claim 1, characterized in that, The time-to-digital conversion circuit includes multiple time shift units and multiple residual analytical delay units. The calibration circuit is configured to adjust at least one of the multiple time shift units and multiple residual analytical delay units so that the corresponding time quantization code and the successive approximation quantization code satisfy the equivalent quantization weight.

3. The system as described in claim 2, characterized in that, The time-to-digital conversion circuit includes n time shift units, n residual resolution delay units, n+1 flip-flops, and n+1 clock delay units. The inputs of the n time shift units and one of the clock delay units are coupled to the successive approximation logic via switches. The outputs of the n time shift units are simultaneously coupled to the input of one of the residual resolution delay units and the data input of one of the flip-flops. The remaining residual resolution delay units are sequentially coupled to the output of the remaining residual resolution delay unit. The data inputs of the remaining flip-flops are respectively coupled to the outputs of the n residual resolution delay units. The remaining clock delay units are sequentially coupled to the output of the remaining clock delay unit. The output of the remaining clock delay unit is coupled to the clock input of the remaining flip-flops. The outputs of the remaining clock delay units are sequentially coupled to the clock inputs of the remaining flip-flops. The n+1 flip-flops output n+1 time quantization codes for each transition in each of the multiple low-noise transition cycles in the lower bits.

4. The system as described in claim 1, characterized in that, The successive approximation conversion sequentially includes multiple high-speed conversion cycles of the high-order bits and multiple low-noise conversion cycles of the low-order bits. The time-to-digital converter is configured to generate multiple time quantization codes in each of the multiple low-noise conversion cycles. The successive approximation quantization codes generated sequentially by the successive approximation logic in the multiple low-noise conversion cycles have the same equivalent quantization weight as the highest-order time quantization codes generated sequentially by the time-to-digital converter in the multiple low-noise conversion cycles.

5. The system as described in claim 4, characterized in that, The low-noise conversion cycle includes 3 conversion cycles. The equivalent quantization weights of the successive approximation quantization codes generated by the successive approximation logic in the 3 conversion cycles are 16, 8, and 4 respectively. The time-to-digital conversion circuit generates 4 time quantization codes in each of the 3 conversion cycles. The equivalent quantization weights of the 4 time quantization codes are 4, 3, 2, and 1 respectively.

6. The system as described in claim 5, characterized in that, The time-to-digital conversion circuit includes first to third time shift units, first to third residual analysis delay units, first to fourth flip-flops, and first to fourth clock delay units. The inputs of the first to third time shift units and the input of the fourth clock delay unit are respectively coupled to the successive approximation logic via switches. The outputs of the first to third time shift units are simultaneously coupled to the input of the third residual analysis delay unit and the data input terminal of the fourth flip-flop. The second and first residual analysis delay units are sequentially coupled to the output terminal of the third residual analysis delay unit. The data input terminals of the third to first flip-flops are respectively coupled to the output terminals of the third to first residual analysis delay units. The third to first clock delay units are sequentially coupled to the output terminal of the fourth clock delay unit. The output of the fourth to first clock delay unit is coupled to the clock input terminal of the fourth to first flip-flops. The fourth to the first triggers output the fourth to the first time quantization codes in each of the three conversion cycles.

7. The system as described in claim 6, characterized in that, The successive approximation logic outputs the comparator clock to the comparator and the time-to-digital converter circuit. The falling edge of the conversion phase of the comparator clock determines the triggering time of the fourth flip-flop for time quantization. The rising edge of the conversion phase of the comparator clock determines the timing when the output signals of the first to third time shift units are sent to the fourth flip-flop.

8. The system as described in claim 6, characterized in that, The calibration circuit adjusts at least one delay parameter in the time-to-digital conversion circuit based on the equivalent quantization weight relationship between the successive approximation quantization result and the time quantization result, further comprising: The third time shift unit is adjusted so that the fourth time quantization code generated in the first low-noise conversion cycle corresponds to and is equal to the successive approximation quantization code generated in the second low-noise conversion cycle. Adjust the second and third residual analysis delay units so that the second time quantization code generated in the first low-noise conversion cycle corresponds to the successive approximation quantization code generated in the third low-noise conversion cycle; The third time shift unit is adjusted so that the second time quantization code generated in the first low-noise conversion cycle corresponds to and is equal to the successive approximation quantization code generated in the second low-noise conversion cycle. Adjust the third residual analysis delay unit so that the third time quantization code generated in the first low-noise conversion cycle is equal to the sum of the successive approximation quantization codes generated in the second and third low-noise conversion cycles; Adjust the second residual analysis delay unit so that the second time quantization code generated in the first low-noise conversion cycle corresponds to the successive approximation quantization code generated in the second low-noise conversion cycle; Adjust the first residual analysis delay unit so that the first time quantization code generated in the first low-noise conversion cycle corresponds to the successive approximation quantization code generated in the third low-noise conversion cycle. Adjust the second time shift unit so that the second time quantization code generated in the second low-noise transition cycle corresponds to and is equal to the successive approximation quantization code generated in the third low-noise transition cycle; Adjust the first residual resolution delay unit so that the second time quantization code generated in the third low-noise conversion cycle is equal to the first time quantization code generated in the second low-noise conversion cycle.

9. The system as described in claim 8, characterized in that, After the calibration circuit adjusts the first to third residual analysis delay units, the delay of the third residual analysis delay unit is... The delay of the second residual analysis delay unit is The delay of the first residual analysis delay unit is ,in is the time constant.