Counter based on loadable true single-phase clock flip-flop

Through the counter architecture of loadable True Single-Phase Clock (L-TSPC) flip-flops in CMOS image sensors, integrating counting, subtraction and readout functions, the performance problems of traditional ADCs under high-speed clock and low power consumption limitations are solved. , realizes efficient multi-stage pixel conversion, and improves the performance and efficiency of image sensors.

CN115037892BActive Publication Date: 2024-11-12SHENZHEN GOODIX TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202210157168.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Priority Date
2021-03-04
Filing Date
2022-02-21
Publication Date
2024-11-12
Estimated Expiration
2042-02-21

AI Technical Summary

Technical Problem

The ADC in existing CMOS image sensors is difficult to achieve high-performance pixel conversion under high-speed clock and low power consumption limitations. Traditional architectures consume a lot of power and area, and the two-stage conversion process takes time.

Method used

Using a loadable True Single-Phase Clock (L-TSPC) trigger counter architecture, it supports multi-stage pixel conversion by integrating counting, subtraction and readout functions in the counter, and generates using the combination of L-TSPC triggers and data triggers. High resolution count values ​​and optimize counter operation through trigger loader blocks and move-out blocks.

Benefits of technology

It realizes high-speed and high-resolution pixel conversion, reduces power and area consumption, improves conversion speed and efficiency, supports multi-stage conversion cycles, and enhances the performance of image sensors.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115037892B_ABST
    Figure CN115037892B_ABST
Patent Text Reader

Abstract

Techniques for implementing a counter architecture to support high-speed, high-resolution pixel conversion, such as for CMOS image sensor applications, are disclosed. An embodiment implements a counter block that uses a loadable true single-phase clock (L‑TSPC) flip-flop as at least a portion of the counter flip-flop. Some embodiments support efficient two-phase pixel conversion by integrating counting, subtraction, and shifting out in a counter. For example, an embodiment may perform a first high-speed pixel conversion phase to obtain a first conversion count. Prior to a second phase, an initial counter may be pre-decremented by the amount of the first conversion count. An embodiment may then perform a second high-speed pixel conversion phase to obtain a second conversion count. Since the second conversion count has been pre-decremented by the first conversion count, the second conversion count represents a final two-phase conversion result. An embodiment may read out this final two-phase conversion result as a digital output of the counter.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates generally to image sensors and more particularly to a counter circuit with a loadable true-single-phase-clocking (TSPC) flip-flop for pixel conversion in a complementary metal-oxide semiconductor (CMOS) image sensor. Background Art

[0002] Many electronic devices have cameras and other features that rely on digital image sensors. For example, most modern smartphones have one or more digital cameras that rely on digital image sensing hardware and software to capture and process images. Such applications typically use complementary metal oxide semiconductor (CMOS) image sensors (CIS) to perform image sensing. Over time, consumers have come to expect higher performance from these image sensors, including higher resolution and lower noise. Furthermore, particularly in portable electronic devices (e.g., with fixed battery capacity), it is desirable to provide such features without adversely affecting power consumption and dynamic range. For example, analog power drives a significant portion (if not the majority) of the power consumption of modern CISs. Therefore, implementing a high-performance CIS in a portable electronic device may involve designing analog-to-digital converters (ADCs) and other components within strict power efficiency and noise constraints.

[0003] The ADC in a CMOS image sensor converts analog pixel information into a digital code ("pixel conversion"). Many applications place stringent design requirements on such ADCs. For example, many applications need to operate at very high-speed clocks (e.g., in the gigahertz (GHz) range) and in an environment with strict constraints on power consumption and area consumption. Traditional architectures typically include a ramp-based ADC that compares the analog pixel output voltage (corresponding to the signal strength detected by the pixel) with a reference ramp voltage. For example, a comparator converts the pixel voltage information into a pulse whose width is proportional to the pixel voltage level. A counter measures the pulse width by counting the number of clock cycles that pass between the starting edge and the ending edge of the pulse. In this way, the size of the count generated by the counter can be a digital value related to the analog pixel output value. The resolution of the output depends on the counter clock. Therefore, implementing pixel conversion in a high-performance image sensor may require supporting a very high-speed counter clock.

[0004] Certain applications require additional pixel conversion functionality. In some cases, the implementation of pixel conversion in a high-performance image sensor may involve multiple pixel conversion stages. For example, a first conversion stage is performed without any analog pixel voltage information to obtain a baseline for the system's low-frequency noise and / or other artifacts; a second stage is then performed using the analog pixel information (and, incidentally, any artifacts common to the first conversion stage). The final conversion value can then be obtained by subtracting the first stage value from the second stage value to effectively eliminate any effects of these artifacts. In some cases, implementing pixel conversion in a high-performance image sensor may further involve efficiently reading out the output from a large number of counters. For example, an image sensor may have thousands of rows, each with thousands of pixels, so the conversion speed is affected by the architecture's ability to support efficient readout of counter values ​​at high speed. Summary of the invention

[0005] Embodiments provide a counter architecture to support high-speed, high-resolution pixel conversion, such as for CMOS image sensor applications. Embodiments implement a counter block that uses a loadable true-single-phase-clocking ("L-TSPC") trigger for at least a portion of the counter trigger. Some embodiments support efficient multi-stage pixel conversion by integrating counting, subtraction, and readout in the counter. For example, an embodiment may perform a first high-speed pixel conversion stage to obtain a first conversion count. Before the second stage, the initial counter may be pre-subtracted by the number of first conversion counts. The embodiment may then perform a second high-speed pixel conversion stage to obtain a second conversion count. Since the second conversion count has been pre-subtracted by the first conversion count, the second conversion count represents the final two-stage conversion result. The multi-stage conversion may include two or more conversion stages, and an embodiment may read out the final multi-stage conversion result as a digital output of the counter.

[0006] According to a group of embodiments, a counter based on a loadable true single-phase clock (L-TSPC) trigger is provided. The counter based on the L-TSPC trigger includes: an N-bit counter block, a trigger loader block, and a shift block. The N-bit counter block includes: M L-TSPC triggers and NM data flip-flops (data flip-flops, referred to as "DFFs"), the N-bit counter block is used to generate a count value, the count value indicates the N-bit number of clock cycles passed by the input clock signal received at the clock input node, so that the state of the L-TSPC trigger represents the M least significant bits (LSBs) of the count value, and the state of the NM data flip-flops (DFFs) represents the NM most significant bits (MSBs) of the count value, where M and N are positive integers, and each L-TSPC trigger can be loaded to a certain value based on a respective set / reset signal; the TSPC holding sub-block is used to couple with the L-TSPC trigger to hold the M LSBs of the count value. The trigger loader block is used to operate in multiple modes, so that: the trigger loader block is used to obtain M LSB reset values ​​and NM MSB reset values ​​in reset mode to generate corresponding set / reset signals to configure each L-TSPC trigger into a loading mode corresponding to a corresponding bit in the LSB reset value, and load each bit in the NM MSB reset values ​​into a corresponding one of the NM DFFs; the trigger loader module is used to generate corresponding set / reset signals to configure the L-TSPC trigger into a clock divider mode in run mode; the trigger loader module is used to obtain M LSBs of the count value from the TSPC hold submodule in hold mode, and generate corresponding set / reset signals to configure each L-TSPC trigger into a loading mode corresponding to the LSB corresponding bit of the reloaded count value. The shift-out block is coupled to the TSPC hold subblock to output the M LSBs of the count value, and is coupled to the DFF to output the NM MSBs of the count value.

[0007] According to another set of embodiments, a method for pixel conversion counting using a counter based on a loadable true single-phase clock (L-TSPC) flip-flop is provided. The method includes: operating an N-bit counter block having M L-TSPC flip-flops and NM data flip-flops (DFFs) within a conversion time frame defined by a ramp start signal and a ramp stop signal to update a count value by configuring the L-TSPC flip-flops and the DFFs to operate in a clock divider mode to count the number of clock cycles of an input clock signal that pass within the conversion time frame, so that the L-TSPC flip-flops represent the M least significant bits (LSBs) of the count value and the DFFs represent the NM most significant bits (MSBs) of the count value; after operating the N-bit counter block, holding the LSB of the count value by latching it in a TSPC latch coupled to the L-TSPC flip-flops; outputting the count value by outputting the LSB of the count value from the TSPC latch and outputting the MSB of the count value from the DFFs. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and, together with the description, serve to explain the principles of the invention.

[0009] Figure 1 A portion of an illustrative pixel analog-to-digital converter (ADC) is shown.

[0010] Figure 2 Shows Figure 1 A simplified diagram of an illustrative signal.

[0011] Figure 3 A novel loadable true single-phase clock (L-TSPC) flip-flop based counter architecture according to embodiments described herein is shown.

[0012] Figures 4A to 4E Operation of an illustrative L-TSPC flip-flop based counter in accordance with various embodiments is shown while performing a two-phase pixel conversion cycle.

[0013] Figure 5 A flow chart of an illustrative method for pixel transition counting using a loadable true single-phase clock flip-flop based counter is shown in accordance with various embodiments.

[0014] In the drawings, similar components and / or features may have the same reference label. In addition, various components of the same type may be distinguished by following the reference label with a second label that distinguishes the similar components. If only the first reference label is used in the specification, the description applies to any similar components having the same first reference label, regardless of the second reference label. DETAILED DESCRIPTION

[0015] In the following description, many specific details are provided to thoroughly understand the present invention. However, it should be understood by those skilled in the art that the present invention can be implemented without one or more of these details. In other examples, features and technologies known in the art will not be described for the sake of brevity.

[0016] Figure 1 A portion of an illustrative pixel analog-to-digital converter (ADC) 100 is shown. As shown, a portion of the ADC 100 includes a pixel ramp voltage generator 140 coupled to a plurality of comparators 110, which are coupled to a plurality of counters 120. The pixel ramp voltage generator 140 generates a ramp voltage (Vramp) 105, which can be a voltage that starts from a low starting level and charges to a high ending level in a substantially linear manner. For example, a current generator starts a charging current flowing through a load (e.g., a capacitive load), thereby charging the load. As the load charges, the voltage across the load can increase substantially linearly over time.

[0017] As shown, a ramp voltage 105 may be received at one of the inputs (e.g., the negative input) of each comparator 110. In many typical ADC architectures for image sensors, the same ramp voltage 105 may be shared by multiple comparators 110, such as all comparators 110 for each row of pixels. The other input (e.g., the positive input) of each comparator 110 may be coupled to a corresponding pixel response voltage (Vp) 107 of the corresponding pixel. For example, the pixel response voltage 107 is an analog output of the pixel corresponding to the intensity (e.g., number of photons) of light detected by the photodetector element of the pixel. Each comparator 110 compares its respective pixel response voltage 107 with the ramp voltage 105. The ramp voltage 105 may be configured so that the ramp starts at a level that is assumed to be lower than any pixel response voltage 107, and so that the ramp ends at a level that is assumed to be higher than any pixel response voltage 107. At some point in the pixel conversion cycle, the linearly changing level of the ramp voltage 105 crosses the level of the pixel response voltage 107, thereby triggering a state change of the output of the corresponding comparator 110.

[0018] In such a configuration, the output of each comparator 110 is actually a pulse that starts at a first time point corresponding to the start of the ramp voltage increase and ends at a second time point corresponding to the crossing of the ramp voltage 105 level and the pixel response voltage 107 level. Therefore, different levels of the pixel response voltage 107 will result in different timings of the state change of the comparator 110 output, and the pulse width between the first time point and the second time point is temporally related to the measurement of the pixel response voltage 107 level. As shown, the comparator 110 is coupled to the counter 120, which is used to generate a digital output 130, which indicates the measured value of the pulse width, thereby indicating the measured value of the pixel response voltage 107 level.

[0019] In this way, each counter 120 can be coupled to the output of a corresponding one of the comparators 110 to receive an output signal as a stop signal 123. Each counter 120 also receives a start signal 125 and an input clock signal 127. The start signal 125 indicates the start of the ramp (i.e., when the ramp voltage begins to increase from its starting level). A change in state (e.g., assertion or deassertion) of the start signal 125 triggers the counter 120 to start counting pulses of the input clock signal 127. Counting continues until the counter 120 detects a change in state of the stop signal 123, corresponding to a change in state of the output of the comparator 110. At this point, the counter 120 can stop counting and generate a digital output 130 indicating the obtained count value.

[0020] To make it clearer, Figure 2 Shows Figure 1 1 is a simplified diagram of an illustrative signal. As shown, a change in state (e.g., a rising edge) of start signal 125 triggers ramp voltage 105 to begin rising from a certain low start level. Ramp voltage 105 may continue to rise (e.g., rise substantially linearly) for a period of time. At some point, as ramp voltage 105 increases, it crosses the level of illustrative pixel response voltage 107. For example, pixel response voltage 107 corresponds to the intensity of the photodetector detected at the pixel corresponding to pixel response voltage 107. As shown in FIG. Figure 1 As shown in FIG. 1 , when the rising level of the ramp voltage 105 crosses the pixel response voltage 107 level, the output state of the corresponding comparator 110 changes, and the output can be used as a stop signal 123. Figure 2 It can be seen that the stop signal 123 changes state in response to the level of the ramp voltage 105 crossing the level of the pixel response voltage 107 .

[0021] Also shown is an input clock signal 127. The input clock signal 127 may be any suitable periodic signal, but is typically a square wave having a duty cycle of approximately 50%. Figure 1 As described, the counter 120 may be triggered to start counting cycles of the input clock signal 127 (e.g., a rising edge, or other detectable state change) in response to the start signal 125, and to stop counting cycles of the input clock signal 127 in response to the stop signal 123. The number of cycles of the input clock signal 127 that pass between the start signal 125 and the stop signal 123 may be used as a measure of the pixel response voltage 107 and may be output by the counter 120 as a digital output 130.

[0022] The process of starting the ramp, detecting the corresponding timing of the ramp level crossing of all pixels in a row, and ending the ramp can be considered as a pixel row conversion cycle. Many image sensor designs use a two-stage conversion cycle, that is, two conversion stages are used for each pixel in each row in each cycle to generate an ADC result. For example, in the first stage, the reset voltage on each pixel line is converted. The reset voltage does not include image information, but it tends to include various noise artifacts, such as low-frequency noise from the pixel ramp voltage generator 140 and the comparator 110. In the second stage, the reset voltage plus the image information is converted. By subtracting the two outputs, the reset voltage is canceled and the image information is retained. Because some noise is common in the two stages, the subtraction is beneficial to reduce the noise impact of the pixel ramp voltage generator 140 and the comparator 110 (e.g., most of the low-frequency noise) as long as the noise does not change significantly between the two conversion stages.

[0023] Conventional pixel conversion processes often have various limitations. For example, conventional architectures for high-performance image sensors may include thousands of comparators 110 and counters 120, which consume considerable power and area. In addition, each pixel conversion cycle consumes considerable time. For example, in a conventional two-stage pixel conversion process, the conversion of each pixel may involve obtaining a first conversion count (i.e., a count of the cycles of the input clock signal 127 between the start signal 125 and the stop signal 123), reading the first conversion count, obtaining a second conversion count, reading the second conversion count, and subtracting the two conversion counts to obtain the final digital conversion result. In order to support high resolution in high-performance applications, it may be necessary to use a higher-bit counter (e.g., at least 10 bits) and a high-speed clock. However, the maximum clock speed, the maximum number of counter bits, and / or other design features are usually limited by other aspects of the architecture, such as the maximum clock speed of the flip-flops used in the counter, the maximum shift-out speed, etc. Therefore, even with a relatively high-speed conventional architecture, the two-stage pixel conversion process consumes considerable time and / or other resources.

[0024] Embodiments described herein include a novel counter architecture to support high-speed, high-resolution pixel conversion, such as in CMOS image sensor applications. Embodiments implement a counter block that uses a loadable true single-phase clock (L-TSPC) trigger for at least a portion of the counter trigger. Some embodiments support efficient two-stage pixel conversion by integrating counting, subtraction, and shifting out in the counter. For example, an embodiment may perform a first high-speed pixel conversion stage to obtain a first conversion count. Before the second stage, the initial counter may be pre-decremented by the number of first conversion counts. The embodiment may then perform a second high-speed pixel conversion stage to obtain a second conversion count. Since the second conversion count has been pre-decremented by the first conversion count, the second conversion count represents the final two-stage conversion result. The embodiment may read out this final two-stage conversion result as a digital output of the counter. Although some embodiments are described herein with reference to a two-stage pixel conversion cycle, the new techniques described herein may be used to support pixel conversion cycles with any suitable number of stages. For example, some embodiments may be configured to support single-stage conversions, while other embodiments may be configured to support more than two conversion stages. For example, each pixel conversion cycle may include a first group of one or more conversion phases to obtain and / or pre-subtract the conversion count in the absence of a pixel signal (e.g., to analyze and reduce noise), and a second group of one or more conversion phases to count and / or pre-subtract the conversion count with the pixel signal.

[0025] Figure 3A novel counter architecture 300 based on a loadable true single-phase clock (L-TSPC) flip-flop according to an embodiment described herein is shown. An embodiment of the counter architecture 300 includes an N-bit counter block 310, a flip-flop loader block 340, and a shift-out block 360. As shown, the N-bit counter block 310 includes N flip-flops arranged as a clock divider chain. For example, in a chain formed by N flip-flops, each flip-flop can be configured to have its output state switched with each rising edge (or, each falling edge) of its input, and its input can be the same as the previous flip-flop in the chain (or the first flip-flop in the chain using the input clock signal 127). In such a configuration, the output of each flip-flop switches at half the frequency of its previous flip-flop; if the frequency of the input clock signal 127 is F, the output of each j-th flip-flop switches at a frequency of F / 2^j. For example, if the input clock signal 127 is a 2.5 GHz clock signal, the output of the first flip-flop in the N-bit counter block 310 is 1.25 GHz, the output of the second flip-flop in the N-bit counter block 310 is 612.5 megahertz (MHz), and so on. Thus, the N outputs from the N flip-flops in the N-bit counter block 310 correspond to an N-bit count of the number of cycles of the input clock signal 127 since the count began (e.g., since the states of all flip-flops in the N-bit counter block 310 were reset to '0'). For this N-bit count, the first flip-flop in the N-bit counter block 310 (i.e., the flip-flop directly coupled to the input clock signal 127) will be the least significant bit, and the last flip-flop in the N-bit counter block 310 will be the most significant bit.

[0026] As shown, the N-bit counter block 310 includes a loadable true single-phase clock (L-TSPC) flip-flop 312. An embodiment of the L-TSPC flip-flop 312 is described in U.S. patent application Ser. No. 17 / 164,722, entitled “Loadable True Single-Phase Clock Flip-Flop,” filed on Feb. 1, 2021, the entire contents of which are incorporated herein by reference. Typically, each L-TSPC flip-flop 312 is configured to operate in one of at least a normal mode, a set load mode, and a reset load mode in response to a set of control signals. In the set load mode, control signals (e.g., set and reset signals) may be used to maintain the output (e.g., Q-bar) of the L-TSPC flip-flop 312 in a first state (e.g., HIGH); and in the reset load mode, control signals may be used to maintain the output of the L-TSPC flip-flop 312 in a second state (e.g., LOW) that is the inverse of the first state. In the set and reset load mode, the output of the L-TSPC flip-flop 312 remains in a determined state and is not affected by the changing state of the input clock signal (eg, or the absence of such a signal).

[0027] In normal mode, the L-TSPC flip-flop 312 can substantially emulate the operation of a conventional TSPC flip-flop, such that the output state of the L-TSPC flip-flop 312 can be a function of changes in the state of the input clock signal. For example, by coupling the output of the L-TSPC flip-flop 312 with its input, the L-TSPC flip-flop 312 can be effectively used as a clock divider to support very high speed clocks. However, it is well known that conventional TSPC flip-flops (and similarly, the L-TSPC flip-flop 312 when operating in normal mode) tend to malfunction when operating below a relatively high minimum clock frequency. If the clock to the flip-flop stops or becomes too slow, the output of the TSPC flip-flop can quickly transition to an undefined state due to parasitic capacitance and other electrical factors, and leakage between the power supply level and the ground level may occur through the TSPC flip-flop. Therefore, the TSPC flip-flop tends to be used in high-speed, always-on clock circuits, such as a divider of a phase-locked loop (PLL) in a high-speed clock.

[0028] Traditionally, TSPC flip-flops are often not used in digital counter circuits, such as N-bit counter block 310, where the clock frequency is divided again by each flip-flop stage of the counter. For example, even if the frequency of input clock signal 127 is fast enough to support TSPC flip-flop operation, the input clock is less than five percent before the fifth flip-flop in the counter and may be too slow for reliable TSPC operation. Furthermore, in a CMOS image sensor environment, there may be thousands of such flip-flops used to support the conversion of each row of pixels, which may increase any problems arising from TSPC flip-flop leakage (e.g., increased power consumption and heat, reduced performance, etc.). Therefore, TSPC flip-flops are not typically used in counter circuits.

[0029] As shown, the N-bit counter block 310 includes an L-TSPC flip-flop 312 and a data flip-flop (DFF) 314. The remaining flip-flops in the N-bit counter block 310 can be implemented as DFFs 314, or any other suitable flip-flop architecture. Then, in some implementations, the M least significant bits (LSBs) of the N-bit counter block 310 are implemented as L-TSPC flip-flops 312, and the remaining NM most significant bits (MSBs) of the N-bit counter block 310 are implemented as DFFs 314. As described above, if the input clock signal 127 frequency is "F", each k-th L-TSPC flip-flop 312k will be clocked at a divided frequency of F / 2^(k-1). Therefore, the number of L-TSPC flip-flops 312 (here, an integer M is used as an example) can be selected to ensure that F / 2^(M-1) is higher than the predetermined minimum frequency for reliable L-TSPC flip-flop 312 operation. For example, if F is high enough and / or N is low enough, the N-bit counter block 310 may be implemented with only L-TSPC flip-flops 312 (ie, without DFFs 314 ) such that F / 2^(N−1) supports reliable operation of the Nth L-TSPC flip-flop 312 .

[0030] During some stages of the pixel conversion process, the N-bit counter block 310 may generate a count value that indicates an N-bit number of clock cycles of the input clock signal 127 received at the clock input node. The output states of the flip-flops in the N-bit counter block 310 may indicate the count value. For example, the state of the L-TSPC flip-flop 312 indicates the M LSBs (q <m:1>, or q <m-1:0>), the states of NM DFFs represent the NM MSBs of the count value (q<N:m+1> , or q <n-1:m>). As described more fully herein, the N-bit counter block 310 also includes a TSPC hold sub-block 320 for coupling with the L-TSPC flip-flop 320 to hold the M LSBs of the count value. For example, when the input clock signal 127 stops at the input of the first L-TSPC flip-flop 312 a, the TSPC hold sub-block 320 can latch the M LSBs, e.g., corresponding to the stop signal 123 being asserted at the output of the connected comparator 110 (not shown).

[0031] Embodiments of the trigger loader block 340 are configured to operate in multiple modes, including at least a reset mode, a run mode, and a hold mode. In the reset mode, the trigger loader block 340 is used to obtain M LSB reset values ​​and NM MSB reset values. These values ​​can be all "0" or any other suitable N-bit initial group for initializing the N-bit counter block 310. The trigger loader block 340 can generate corresponding set / reset signals to configure each L-TSPC flip-flop 312 into a loading mode corresponding to a corresponding bit in the LSB reset value, and can load each of the NM MSB reset values ​​to the corresponding NM DFFs 314. For example, generating the corresponding set / reset signals can involve generating a corresponding set signal and a corresponding reset signal for each L-TSPC flip-flop 312. For a given L-TSPC flip-flop 312, asserting the set signal and deasserting the reset signal can set the output of the L-TSPC flip-flop 312 to "1" at the "Q" output (and to "0" at the "Q_bar" output); and asserting the reset signal and deasserting the set signal can set the output of the L-TSPC flip-flop 312 to "0" at the Q output (and to "1" at the "Q_bar" output).

[0032] In the run mode, the trigger loader block 340 can be used to generate corresponding set / reset signals to configure the L-TSPC flip-flops into the clock divider mode. For example, deasserting the set and reset signals of a particular L-TSPC flip-flop 312 can configure the L-TSPC flip-flop 312 to operate in the clock divider (e.g., normal) mode. As described above, in this mode, the N-bit counter block 310 can effectively be used as a counter. In the hold mode, the trigger loader block 340 can be used to obtain the M LSBs of the count value from the TSPC hold sub-block 320 and generate corresponding set / reset signals to configure each L-TSPC flip-flop 312 into the load mode corresponding to the corresponding bit in the LSB of the reloaded count value. For example, at the end of a count cycle (e.g., when the flip-flop loader block 340 switches from the run mode to the hold mode), the M LSBs of the count value are latched by the TSPC hold sub-block 320 and fed to the flip-flop loader block 340, which generates corresponding set / reset signals to reload the same M LSBs back into the L-TSPC flip-flop 312. In this way, the internal nodes and the output of the L-TSPC flip-flop 312 can be maintained in a determined state even if the clock to the flip-flop is stopped. In an alternative implementation, in the hold mode, the flip-flop loader block 340 can load the L-TSPC flip-flop 312 with any other suitable value (e.g., reload with a reset value, etc.). However, reloading the L-TSPC flip-flop 312 with the M LSBs can support certain functions, such as allowing counting to continue from where it stopped.

[0033] An embodiment of the shift-out block 360 is configured to generate the digital output 130. For example, the shift-out block 360 is coupled to the TSPC hold sub-block 320 to output the M LSBs of the count value, and the shift-out block 360 is coupled to the DFF 314 to output the NM MSBs of the count value. In some implementations, the shift-out block 360 is controlled by the output enable signal 363. For example, asserting the output enable signal 363 results in the readout of N count value bits.

[0034] Some embodiments of the L-TSPC flip-flop based counter architecture 300 also include a pre-subtractor block 330 coupled to the N-bit counter block 310. The pre-subtractor block 330 can receive a count value and output an inverted count value. For example, the pre-subtractor block 330 can be coupled to the TSPC holding sub-block 320 to receive the M LSBs of the count value, and can be coupled to the DFF 314 to receive the NM MSBs of the count value. The pre-subtractor block 330 can invert the bits of the count value, such as by latching the bits and outputting the inverse of each bit (e.g., the inverted count value can effectively be the binary inverse of the count value). The inverted count value can then be provided as an input to the flip-flop loader block 340. In such an embodiment, the flip-flop loader block 340 can also be used to selectively operate in a pre-subtraction mode to obtain an inverted count value from the pre-subtractor block 330. In this mode, the trigger loader block 340 can generate corresponding set / reset signals to configure each L-TSPC trigger 312 into a loading mode corresponding to a corresponding bit in the M LSBs of the inverted count value, and load each of the NM MSBs of the inverted count value into a corresponding one of the NM DFFs 314.

[0035] The embodiment of the counter architecture 300 based on the L-TSPC trigger may further include a counter controller 370, which may generate any suitable control signal 375 and provide it to the various components of the counter based on the L-TSPC trigger. As shown, the counter controller 370 may be coupled to at least the N-bit counter block 310 and the trigger loader block 340. In some embodiments, the counter controller 370 uses the control signal 375 to indicate the operation of the N-bit counter block 310 and the trigger loader block 340 to implement a single-stage conversion. For example, in a first time frame, the counter controller 370 uses the control signal 375 to indicate the trigger loader block 340 to operate in a reset mode so that the count value in the N-bit counter block 310 is reset according to the LSB reset value and the MSB reset value (e.g., all reset to "0"). In a second time frame, the counter controller 370 uses the control signal 375 to indicate the trigger loader block 340 to operate in a run mode so that the N-bit counter block 310 updates the count value to indicate the clock operation duration counted in the second time frame. In the third time frame, the counter controller 370 uses the control signal 375 to instruct the TSPC hold sub-block 320 to hold the M LSBs of the count value and instruct the flip-flop loader block 340 to operate in the hold mode to reload the LSBs of the count value to the L-TSPC flip-flop 312 .

[0036] In some embodiments, the counter controller 370 uses the control signal 375 to instruct the operation of the N-bit counter block 310, the trigger loader block 340, and the pre-subtractor block 330 to implement the two-stage conversion. For example, in a first time frame, the counter controller 370 uses the control signal 375 to instruct the trigger loader block 340 to operate in the reset mode so that the count value in the N-bit counter block 310 is reset according to the LSB reset value and the MSB reset value. In a second time frame, the counter controller 370 uses the control signal 375 to instruct the trigger loader block 340 to operate in the operation mode so that the N-bit counter block 310 updates the count value to a first updated count value, which indicates the first stage clock operation duration (i.e., the first conversion count) counted in the second time frame. In the third time frame, the counter controller 370 uses the control signal 375 to instruct the TSPC hold sub-block 320 to hold the M LSBs of the first updated count value, and instructs the trigger loader block 340 to operate in the hold mode to reload the LSBs of the first updated count value to the L-TSPC flip-flop 312. In the fourth time frame, the counter controller 370 uses the control signal 375 to instruct the trigger loader block 340 to operate in the pre-decrement mode to update the count value to a second updated count value indicating the inverted count value obtained from the pre-subtractor block 330. In the fifth time frame, the counter controller 370 uses the control signal 375 to instruct the trigger loader block 340 to operate in the run mode to cause the N-bit counter block 310 to update the count value to a third updated count value indicating the second stage clock operation duration counted in the fifth time frame (i.e., the second conversion count, which has been pre-decremented by the first conversion count). In the sixth time frame, the counter controller 370 uses the control signal 375 to instruct the TSPC hold sub-block 320 to hold the M LSBs of the third updated count value and instruct the flip-flop loader block 340 to operate in the hold mode to reload the LSBs of the third updated count value to the L-TSPC flip-flop 312 .

[0037] In some embodiments, the counter controller 370 is also used to at least partially indicate the operation of the shift-out block 360. For example, the output enable signal 363 can be one of the control signals 375 provided by the counter controller 370. In such an embodiment, after obtaining the final conversion count in the single-stage or two-stage pixel conversion cycle, the counter controller 370 can use the control signal 375 to instruct the shift-out block 360 to output the count value from the N-bit output count value. For example, in the above-mentioned single-stage conversion cycle, the shift-out can be performed in the fourth time frame, at which time, the counter controller 370 can use the control signal 375 to instruct the shift-out block 360 to output the count value from the N-bit counter block 310. In the above-mentioned two-stage conversion cycle, the shift-out can be performed in the seventh time frame, at which time, the counter controller 370 uses the control signal 375 to instruct the shift-out block 360 to output the third updated count value from the N-bit counter block 310. Some embodiments of the two-stage conversion cycle can also include an optional shift-out after obtaining the first conversion count (e.g., after the third time frame). Other implementations can support additional functions. For example, the implementation may be designed to support pixel conversion cycles with more than two phases. As another example, the implementation may be designed to continue counting from the previous conversion count result (e.g., by using the hold mode of the trigger loader block 340 and the load mode of the L-TSPC trigger 312 to maintain the previous conversion count result until the count resumes). As another example, the implementation may be designed to start counting from any desired initial value (e.g., by using the reset mode of the trigger loader block 340 and the load mode of the L-TSPC trigger 312 to set the flip-flop in the N-bit counter block 310 to the desired initial value).

[0038] Embodiments of the counter architecture 300 and / or any of its components may be implemented in any suitable manner. In some embodiments, the counter controller 370 and / or the trigger loader block 340 include: a central processing unit (CPU), an application specific integrated circuit (ASIC), an application specific instruction set processor (ASIP), a graphics processing unit (GPU), a physical processing unit (PPU), a digital signal processor (DSP), a field programmable gate array (FPGA), a programmable logic device (PLD), a controller, a microcontroller unit, a reduced instruction set (RISC) processor, a complex instruction set processor (CISC), a microprocessor, etc., or any combination thereof, or implemented thereon. In addition, components may be combined or separated without departing from the scope of the present invention. In addition, each instance of the counter architecture 300 may include its own instances of certain components, or certain components may be shared between instances. For example, a single counter controller 370 may be used to generate control signals 375 for a large number of counter architecture 300 instances (i.e., the counter controller 370 is shared).

[0039] Figures 4A-4E FIG. 4 shows the operation of an illustrative L-TSPC trigger-based counter 400 when performing a two-phase pixel conversion cycle according to various embodiments. The L-TSPC trigger-based counter 400 may be Figure 3 Implementation of an L-TSPC flip-flop based counter architecture 300 . Figures 4A-4E Each corresponds to one or more specific time frames of the pixel conversion process. Figures 4A-4E In the diagram, solid lines generally represent components and signal paths that are being used in a particular time frame of the diagram, while dashed lines generally represent components and signal paths that are not being used in a particular time frame of the diagram. The use of solid and dashed lines is not intended to limit the scope of the embodiments. Instead, the use of solid and dashed lines is intended only to provide additional clarity and represents only a simplified subset of possible implementations.

[0040] In the illustrated implementation, an 11-bit counter block is implemented using five L-TSPC flip-flops 312 and six DFFs 314 (ie, N=11 and M=5, such that NM=6). Figure 3 The TSPC holding sub-block 320 is implemented as five TSPC latches 322, each coupled to a corresponding one of the five L-TSPC flip-flops 312. The latches of the TSPC holding sub-block are referred to as "TSPC latches" only to clearly distinguish these latches from other latches in the counter 400 based on the L-TSPC flip-flops; the TSPC latches 322 can be implemented using any suitable type of latch without any specific relationship to the TSPC architecture. In one embodiment, each TSPC latch 322 is essentially a pair of inverters in series, and is also coupled to the corresponding L-TSPC flip-flop 312 through a corresponding switch. The switch can be controlled by a first control signal 375a. When the 11-bit counter operates as a counter, the first control signal 375a can set the switch to be turned on (ON). In the on state, the input of each pair of inverters in series is coupled to the output of its corresponding L-TSPC flip-flop 312. When counting stops, the first control signal 375a can switch the state to set the switch to be turned off (OFF). In the off state, the input of each pair of series inverters is decoupled from the output of its corresponding L-TSPC flip-flop 312; instead, the output of each pair of series inverters is coupled to its own input, so that each TSPC latch 322 essentially latches the last read output value of its corresponding L-TSPC flip-flop 312. Other latch architectures may be used for the TSPC latch 322. For example, the first control signal 375a may be coupled to the "enable" input of the D latch, and so on.

[0041] In the illustrated embodiment, Figure 3 The pre-subtractor block 330 can be implemented as a set of 11 subtractor latches 332. Again, the latches of the pre-subtractor block are referred to as "subtractor latches" only to clearly distinguish these latches from other latches in the counter 400 based on the L-TSPC flip-flop; the subtractor latches 332 can be implemented using any suitable type of latch. The black circle at the output of the subtractor latch 332 indicates that the output of the subtractor latch 332 is an inverted version of the input signal. In some embodiments, this inversion is achieved by inverting the output of the subtractor latch 332 using an inverter or other component. In other embodiments, the architecture of each subtractor latch 332 includes an inverted output. As shown, the subtractor latch 332 can be controlled in response to a second control signal 375b.

[0042] In the embodiment shown, Figure 3 The trigger loader block 340 can be implemented to include a reset sub-block 345, a multiplexer (MUX) 342, and a load logic sub-block 350. An embodiment of the reset sub-block 345 generates an LSB reset value and an MSB reset value. In some embodiments, the reset sub-block 345 is a set of registers, or similar devices, on which a set of initial values ​​are stored. In some embodiments, the reset sub-block 345 is a set of hard-wired nodes, such as nodes hard-wired to a ground reference. In some embodiments, the reset sub-block 345 includes a programmable interface so that the reset value can be changed (e.g., by hardware, software, etc.).

[0043] Embodiments of MUX 342 may include 11 MUXs 342, each corresponding to a respective flip-flop of the 11-bit counter. For example, a first group of MUXs including the first to M-th MUXs (i.e., the first five MUXs 342, or MUXs 342a-MUX 342e in the illustrated implementation) corresponds to M L-TSPC flip-flops 312, and a second group of MUXs including the M+1st to N-th MUXs (i.e., the remaining six MUXs 342, or MUXs 342f-MUX 342k in the illustrated implementation) corresponds to NM DFFs 314. Each of the embodiments of MUX 342 may include a plurality of inputs and an output, wherein the output effectively corresponds to one of the inputs selected in response to one or more control signals (e.g., the third, fourth, and / or fifth control signals 375c, 375d, and / or 375e). Embodiments can be configured to ensure that one of the inputs is always selected (i.e., to ensure that there is no situation where all control signals 375 coupled to any given MUX 342 are low at the same time, and / or more than one control signal 375 is high at the same time). In some embodiments, this is achieved by configuring the counter controller 370 to at least generate control signals 375c, 375d, and 375e to meet these standards. Additionally or optionally, the signal can be logically interconnected in a manner that ensures that these standards are met. For example, as shown in the figure, the counter controller 370 is configured to ensure that only one of these control signals is always valid, so that the first group of MUX 342a-342e is coupled to control signals 375c, 375d, and 375e respectively. Similarly, as shown in the figure, the second group of MUX 342f-342k is each coupled to the control signal 375c and the inverse of the control signal 375c (illustrated as the control signal 375c coupled to the second selection input by an inverter). This can ensure that one of the selectable inputs is always high. For example, if the counter controller 370 ensures that at a point in time, only one of the control signals 375c, 375d, and 375e is at a high level, it can ensure that when the control signal 375c is HIGH, the first input of the MUX 342f–342k is selected; when either the control signal 375d or 375e is HIGH (and the control signal 375c must be LOW), the second input of the MUX 342f–342k is selected.

[0044] For each MUX 342, a first input of its input may be coupled to the reset sub-block 345. For example, a first input of each of the 11 MUXs 342 receives a corresponding bit of an 11-bit reset value, wherein a first group of MUXs 342a..e receives 5 LSBs, and a second group of MUXs 342f..k receives 6 MSBs. For each MUX 342, a second input of its input may be coupled to the subtractor latch 332 of the pre-subtractor block. For example, a second input of each of the 11 MUXs 342 receives a corresponding bit of an 11-bit inverted count value, wherein a first group of MUXs 342a..e receives 5 LSBs, and a second group of MUXs 342f..k receives 6 MSBs. For each of the first group of MUXs 342a..e, a third input of its input may be coupled to the 11-bit counter block. For example, the third input of each of the five MUXs 342a..e receives a corresponding bit of the five LSBs of the count value output by the TSPC latch 322. As described above, in some embodiments, the flip-flop loader block can operate in different modes, which can correspond to specific configurations of the MUXs 342. For example, the flip-flop loader block can be configured in a reset mode by at least setting each MUX 342 to generate its output from its first input (i.e., each MUX 342 outputs a corresponding bit corresponding to the reset value); the flip-flop loader block can be configured in a pre-subtraction mode by at least setting each MUX 342 to generate its output from its second input (i.e., each MUX 342 outputs a corresponding bit corresponding to the inverted count value); and the flip-flop loader block can be configured in a hold mode by at least setting each of at least a first group of MUXs 342a..e to generate its output from its third input (i.e., each MUX 342a..e outputs a corresponding bit corresponding to the held count value).

[0045] Embodiments of the load logic sub-block 350 may be used to generate respective set / reset signals based on respective MUX outputs. For example, each L-TSPC flip-flop 312 may be configured to one of run (e.g., clock divider), set-load, or reset-load modes based on respective states of the set and reset signals received by the L-TSPC flip-flop 312. The load logic sub-block 350 may receive a bit value from the MUX 342 and may convert at least the LSB into respective set / reset signals of the L-TSPC flip-flop 312, thereby deterministically forcing the state of each L-TSPC flip-flop 312 output. In some embodiments, the load logic sub-block 350 has similar components to support the loading of the DFF 314 to deterministically force the state of each DFF 314 output according to respective set / reset signals. The operation of the load logic sub-block 350 may be controlled by the sixth and seventh control signals 375f and 375g. For example, the sixth control signal 375 f may control the operation of the load logic sub-block 350 with respect to loading the L-TSPC flip-flop 312 , and the seventh control signal 375 g may control the operation of the load logic sub-block 350 with respect to loading the DFF 314 .

[0046] As described above, embodiments may use control signal 375 to direct operation of L-TSPC flip-flop based counter 400 to implement a pixel conversion cycle, such as a two-phase conversion cycle. Figure 4A An illustrative representation of a counter 400a based on an L-TSPC flip-flop in a first time frame is provided, wherein the counter is effectively reset or initialized. The third control signal 375c is asserted, thereby generating a corresponding output according to the first input selection of each MUX 342, wherein the first input of each MUX 342 corresponds to a corresponding bit of the reset value from the reset sub-block 345. In response to the sixth and seventh control signals 375f and 375g, the load logic sub-block 350 can convert the bits of the reset value received from the MUX 342 into appropriate load signals (e.g., set / reset signals) to output to the L-TSPC flip-flops 312 and DFF 314 of the counter block. Thus, the flip-flops of the counter block are initialized to the reset value.

[0047] Figure 4B An illustrative representation of a counter 400b based on an L-TSPC flip-flop in a second time frame is provided, wherein a first conversion count has been acquired. The sixth and seventh control signals 375f and 375g, or any other suitable logic, may be used to instruct the load logic sub-block 350 to generate a load signal (e.g., a set / reset signal) to configure the flip-flops of the counter to a normal (e.g., clock divider) mode. For example, the load logic sub-block 350 may configure the set and reset inputs of all L-TSPC flip-flops 312 and DFF 314 to all "0s". This may cause each flip-flop to respond to its clock input to achieve the desired counter function. In the context of pixel conversion, at the end of the second time frame, the output state of the flip-flop represents an 11-bit binary count value of the number of clock cycles that have elapsed between the start and stop of the clock (e.g., the start and stop of the ramp).

[0048] Figure 4C An illustrative representation of the counter 400c based on the L-TSPC flip-flop in the third time frame is provided, wherein the count value obtained from the second time frame is held (e.g., for subsequent shifting out and / or pre-decrementing). The first control signal 375a is asserted to latch the LSB of the count value in the TSPC latch 322. The fifth control signal 375e is asserted to select and generate a corresponding output according to the third input of each of the first group of MUXs 342a..e, wherein the third input of each of the first group of MUXs 342a..e corresponds to the corresponding bit of the LSB of the count value received from the corresponding one of the TSPC latches 322. In response to the sixth control signal 375f, the load logic sub-block 350 can convert the LSB of the count value received from the first group of MUXs 342a..e into an appropriate load signal (e.g., a set / reset signal) to output to the L-TSPC flip-flop 312. As described above, in the absence of a high-speed clock input, the state of the L-TSPC flip-flop 312 may quickly become uncertain, which may bring undesirable current leakage and / or other problems. By using the load logic sub-block 350 to reload the count value back to the L-TSPC flip-flop 312, the L-TSPC flip-flop 312 can be maintained in a determined state and such leakage and / or other problems can be avoided. In some embodiments, signaling is no longer provided to the DFF 314 within this time frame. Since the output from the most effective L-TSPC flip-flop 312e (coupled with the least effective DFF 314a) remains unchanged, the DFF 314 will retain their respective final count value bits. For example, as shown, the seventh control signal 375g is not used to load the DFF 314. It is worth noting that in the configuration shown, the inverse of the control signal 375c must be high (HIGH) to select the second input of each MUX 342f..k. However, since the load logic sub-block 350 does not load anything to the DFF 314, the selected input of the MUX 342f..k is substantially unaffected.

[0049] Figure 4D An illustrative representation of a counter 400d based on an L-TSPC flip-flop in a fourth time frame is provided, wherein a pre-subtraction is performed. The second control signal 375b may be asserted to latch the count value in the subtractor latch 332 so that the output of the pre-subtractor block is the inverted count value. The fourth control signal 375d is asserted to generate a corresponding output based on the second input selection of each MUX 342, which corresponds to the corresponding bit of the inverted count value received from the corresponding one of the subtractor latches 332. As described above, the second input selection of the MUX 342f..k is controlled by the inverse of the control signal 375c (rather than by the fourth control signal 375d). However, the effect at this stage is the same, as shown by "(Ctrl4 375d)" at the second selection input of the MUX 342f..k in the figure. In response to the sixth and seventh control signals 375f and 375g, the load logic sub-block 350 can convert the bits of the inverted count value received from the MUX 342 into appropriate load signals (e.g., set / reset signals) for output to the L-TSPC flip-flop 312 and the DFF 314. In effect, this preloads the counter with the binary complement of the count value from the first conversion phase (before the second phase of the conversion cycle). Adding the second conversion count to the binary complement of the first conversion count corresponds to subtracting the first conversion count from the second conversion count. For example, if the first conversion count indicates 29 clock cycles, loading the inverted count value into the counter block effectively starts the next count at -30 (i.e., the binary complement of 29 is negative 30).

[0050] The example can return Figure 4B To illustrate, we can now Figure 4B An illustrative representation of the L-TSPC flip-flop based counter 400b is provided in the fifth time frame, where a second conversion count has been obtained. The second conversion count is obtained in substantially the same manner as the first conversion count in the second time frame, except that the count is started from the inverted count value in the fifth time frame, rather than from the reset value as in the second time frame. Such an embodiment can be implemented by returning Figure 4C To continue, we can now Figure 4C 400c in the sixth time frame, wherein the count value obtained from the fifth time frame is maintained. The second conversion count may be maintained in the sixth time frame in substantially the same manner as the first conversion count is maintained in the third time frame. In effect, the return value in the fifth and sixth time frames is Figure 4B and 4C This results in obtaining and maintaining a final conversion count that represents the second stage conversion count from which the first stage conversion count has been pre-subtracted.

[0051] Figure 4E An illustrative representation of a counter 400e based on an L-TSPC flip-flop in the seventh time frame is provided, wherein the final conversion count value obtained from the sixth time frame is read out. The shift-out block 360 is shown as including a buffer and a transmission gate (T-gates). The transmission gate can be implemented in any suitable manner, for example, a tri-state buffer. For example, a first group of buffers and transmission gates are used to generate a first group of outputs 130b of the LSB of the final conversion count from the TSPC latch 322, and a second group of buffers and transmission gates are used to generate a second group of outputs 130a of the MSB of the final conversion count from the DFF 314. Some embodiments do not include buffers and / or transmission gates. For example, in some embodiments, the output of the flip-flop can directly drive other components of other systems, such as a higher-level shift-out system coupled to many such counters. In other embodiments, the flip-flop directly drives the transmission gate without intermediate buffering. In some embodiments, the components are configured in the seventh time frame to be substantially the same as in the third and sixth time frames. For example, the control signal 375 is configured to cause the LSB of the count value to be latched in the TSPC latch 322, the output of the first set of MUXs 342a..e represents the LSB of the latched count value, and the load logic sub-block 350 is configured to reload the L-TSPC flip-flop 312 with the LSB of the latched count value. In an alternative implementation, the load logic sub-block 350 may continue to provide appropriate set / reset signals to the L-TSPC flip-flop 312 even when the output of the MUX 342 is no longer available. In such an implementation, the control signal 375 may be effectively configured to disable the MUX 342 and the corresponding signal path. An embodiment may assert the output enable signal 363 to provide a shift out of the final conversion count value from the TSPC latch 322 and the DFF 314 through the buffer and transmission gate of the shift-out block 360.

[0052] Figure 5 A flow chart of an illustrative method 500 for counting pixel transitions using a loadable true single-phase clock (L-TSPC) flip-flop based counter is shown in accordance with various embodiments. Embodiments of the method 500 may use Figure 3 and Figures 4A-4E 5. The method 500 may be performed by the counter architecture 300 and related components shown in the figure. In general, embodiments of the method 500 may proceed from the beginning of a pixel conversion cycle to the end of a pixel conversion cycle. The pixel conversion cycle includes one or more iterations of each of a pre-set phase 510, a counting phase 520, a holding phase 530, and an output phase 540. Some embodiments provide a single-phase pixel conversion, which may follow the following sequence: a pre-set phase 510, a counting phase 520, a holding phase 530, and an output phase 540. Other embodiments provide a two-phase pixel conversion, which may follow the sequence: a pre-set phase 510, a counting phase 520, a holding phase 530, a pre-set phase 510, a counting phase 520, a holding phase 530, and an output phase 540. Some two-phase pixel conversion embodiments may include the additional performance of an output phase 540 after the first holding phase 530.

[0053] The embodiment begins at stage 512 (in the pre-set stage 510) by resetting the count value to an initial reset value in an N-bit counter block having M L-TSPC flip-flops and NM data flip-flops (DFFs). The embodiment may initialize the N-bit counter block by forcing each output of the L-TSPC flip-flops and the DFFs to correspond to a corresponding bit of a predetermined N-bit reset value, such that the count value after operation corresponds to the number of clock cycles of the input clock signal that have passed within the conversion time frame plus the reset value. At stage 522 (in the count stage 520), the embodiment may operate the N-bit counter block in the conversion time frame defined by the ramp start signal and the ramp stop signal to update the count value. Updating the count value may involve configuring the L-TSPC flip-flops and the DFFs to operate in a clock divider mode to count the number of clock cycles of the input clock signal that have passed over the conversion time frame, such that the L-TSPC flip-flops represent the M least significant bits (LSBs) of the count value, and the DFFs represent the NM most significant bits (MSBs) of the count value. Configuring the L-TSPC flip-flop and the DFF to operate in a clock divider mode may involve generating (e.g., via a flip-flop loader) set / reset signals to configure respective set and reset inputs of the L-TSPC flip-flop and the DFF so that respective outputs of the L-TSPC flip-flop and the DFF are functions of respective clock inputs.

[0054] In stage 532 (in the hold stage 530), an embodiment may hold the LSB of the count value after operation by latching the LSB of the count value in a TSPC latch coupled to the L-TSPC flip-flop. As described herein, the hold step may also include reloading each L-TSPC flip-flop with the corresponding bit of the LSB of the count value. For example, an embodiment may generate (e.g., via a flip-flop loader) a set / reset signal to configure the corresponding set and reset inputs of the L-TSPC flip-flops so that the corresponding output of each L-TSPC flip-flop is each deterministically forced to a state based on the corresponding bit of the LSB of the count value. In some embodiments, the count value obtained in stage 522 and held in stage 532 may be output in stage 542 (in the output stage 540) by outputting the LSB of the count value from the TSPC latch and by outputting the MSB of the count value from the DFF.

[0055] Some embodiments return to the pre-setup stage 510 for the second phase of the conversion cycle. Some such embodiments return to the pre-setup stage 510 directly from stage 532 (i.e., without outputting a count value at stage 542), while other such embodiments return to the pre-setup stage 510 after outputting a count value at stage 542. The second iteration of the pre-setup stage 510 may be used to perform a pre-subtraction. At stage 514 (in the pre-setup stage), an embodiment may pre-subtract the count value held in stage 532 by inverting the count value to produce an inverted count value and resetting the N-bit counter block to force the outputs of the L-TSPC flip-flop and the DFF to each correspond to a corresponding bit of the inverted count value.

[0056] Then, an embodiment of method 500 may continue to obtain an updated count value at stage 524 and maintain the updated count value at stage 534 in a manner similar to stages 522 and 532, respectively. For example, at stage 524, an embodiment may operate the N-bit counter block in a second conversion time frame after the pre-subtraction to update the count value by configuring the L-TSPC flip-flop and the DFF to operate in a clock divider mode to count the number of clock cycles of the input clock signal that have passed in the second conversion time frame. After the counting in stage 524, the count value corresponds to the number of clock cycles of the input clock signal that have passed in the second conversion time frame minus the number of clock cycles of the input clock signal that have passed in the first conversion time frame (i.e., the pre-subtraction count at stage 522). At stage 534, an embodiment may maintain the LSB of the count value in the TSPC latch by updating the TSPC latch according to the second operation. In some embodiments, the count value obtained in stage 524 and held in stage 534 may be output in stage 544 (in output stage 540 ) by outputting the LSB of the count value from the TSPC latch and by outputting the MSB of the count value from the DFF.

[0057] It should be understood that when an element or component is referred to as "connected to" or "coupled to" another element or component in this article, it can be connected or coupled to another element or component, or connected or coupled to another element or component through an intermediate element or component. On the contrary, when an element or component is referred to as "directly connected to" or "directly coupled to" another element or component, there is no intermediate element or component between them. It should be understood that although the terms "first", "second", "third", etc. can be used to describe various elements and components in this article, these elements, components, and regions should not be limited by these terms. These terms are only used to distinguish an element or component from another element or component. Therefore, without departing from the teachings of the present invention, the first element and component discussed in the text can be referred to as the second element or component. As used herein, the terms "logical low", "low state", "low level", "logical low level", "low" or "0" are used interchangeably. The terms "logical high", "high state", "high level", "logical high level", "high" or "1" are used interchangeably.

[0058] As used herein, the terms "a", "an" and "the" may include singular and plural references. It should also be understood that when used in this specification, the terms "comprising", "including", "having" and their variations specify the presence of the features, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, elements, components and / or groups thereof. Conversely, when used in this specification, the term "consisting of" specifies the features, steps, operations, elements and / or components, and excludes additional features, steps, operations, elements and / or components. In addition, as used herein, the term "and / or" may refer to and cover any possible combination of one or more of the relevant listed items.

[0059] Although the present invention is described herein with reference to illustrative embodiments, the description is not intended to be construed as limiting. On the contrary, the purpose of the illustrative embodiments is to enable those skilled in the art to better understand the spirit of the present invention. In order not to confuse the scope of the present invention, many details of well-known processes and manufacturing techniques are omitted. With reference to the description, various modifications of the illustrative embodiments and other embodiments will be apparent to those skilled in the art. Therefore, the appended claims are intended to cover any such modifications.

[0060] In addition, some features of the preferred embodiments of the present invention can be advantageously used without the corresponding use of other features. Therefore, the above description should be considered as only an illustration of the principle of the present invention, rather than a limitation thereof. Those skilled in the art should understand the variations of the above-described embodiments that fall within the scope of the present invention. Therefore, the present invention is not limited to the specific embodiments and descriptions discussed above, but is limited by the following claims and their equivalents.

Claims

1. A counter based on a loadable true single-phase clock L-TSPC trigger, comprising: N-bit counter block, including: M L-TSPC flip-flops and NM data flip-flops DFF, the N-bit counter block is used to generate a count value, the count value indicating an N-bit number of clock cycles of an input clock signal received at a clock input node, wherein the state of the L-TSPC flip-flops represents the M least significant bits (LSBs) of the count value, and the state of the NM data flip-flops DFF represents the NM most significant bits (MSBs) of the count value, wherein M and N are positive integers, and each L-TSPC flip-flop is loadable to a determined value based on a respective set / reset signal; and a TSPC holding sub-block, coupled to the L-TSPC flip-flop to hold the M LSBs of the count value; The trigger loader block is used to run in multiple modes such that: The flip-flop loader block is used to obtain M LSB reset values ​​and NM MSB reset values ​​in a reset mode to generate corresponding set / reset signals to configure each L-TSPC flip-flop into a loading mode corresponding to a corresponding bit in the LSB reset value, and load each bit in the NM MSB reset values ​​into a corresponding one of the NM DFFs; The flip-flop loader block is used to generate the corresponding set / reset signal in the run mode to configure the L-TSPC flip-flop in the clock divider mode; and The flip-flop loader block is used to obtain the M LSBs of the count value from the TSPC hold sub-block in a hold mode and generate corresponding set / reset signals to configure each L-TSPC flip-flop to a load mode corresponding to a LSB corresponding bit of the reloaded count value; and A shift-out block is coupled to the TSPC holding sub-block to output the M LSBs of the count value and is coupled to the DFF to output the NM MSBs of the count value.

2. The L-TSPC trigger-based counter according to claim 1, further comprising: A counter controller is coupled to the N-bit counter block and the flip-flop loader block and is configured to: Instructing the flip-flop loader block to operate in the reset mode to reset the count value in the N-bit counter block according to the LSB reset value and the MSB reset value in a first time frame; In a second time frame, instructing the trigger loader block to operate in the operation mode so that the N-bit counter block updates the count value to indicate a clock operation duration counted in the second time frame; and In a third time frame, the TSPC hold sub-block is instructed to hold the M LSBs of the count value, and the flip-flop loader block is instructed to operate in the hold mode to reload the LSBs of the count value to the L-TSPC flip-flops.

3. The L-TSPC trigger-based counter according to claim 2, wherein: The counter controller is further configured to instruct the shift-out block to output the count value from the TSPC holding sub-block and the DFF in a fourth time frame.

4. The L-TSPC trigger-based counter according to claim 1 , further comprising: a pre-subtractor block, coupled to the N-bit counter block, configured to receive the count value and output an inverted count value, The trigger loader block is further configured to selectively operate in a pre-subtraction mode to obtain the inverted count value from the pre-subtractor block to generate the corresponding set / reset signal to configure each L-TSPC trigger to a loading mode corresponding to a corresponding bit in the M LSBs of the inverted count value, and to load each bit in the NM MSBs of the inverted count value into a corresponding one of the NM DFFs.

5. The L-TSPC trigger-based counter according to claim 4, further comprising: a counter controller coupled to the N-bit counter block, the flip-flop loader block, and the pre-subtractor block, and configured to: Instructing the flip-flop loader block to operate in the reset mode to reset the count value in the N-bit counter block according to the LSB reset value and the MSB reset value in a first time frame; In a second time frame, instructing the trigger loader block to operate in the operation mode so that the N-bit counter block updates the count value to a first updated count value, wherein the first updated count value indicates a first stage clock operation duration counted in the second time frame; in a third time frame, instructing the TSPC hold sub-block to hold the M LSBs of the first updated count value, and instructing the flip-flop loader block to operate in the hold mode to reload the LSBs of the first updated count value into the L-TSPC flip-flop; instructing the flip-flop loader block to operate in the pre-subtraction mode to update the count value to a second updated count value indicative of the inverted count value obtained from the pre-subtractor block, the inverted count value being the inverse of the first updated count value, in a fourth time frame; In a fifth time frame, instructing the trigger loader block to operate in the operation mode so that the N-bit counter block updates the count value to a third updated count value, wherein the third updated count value indicates a second stage clock operation duration counted in the fifth time frame; and In a sixth time frame, the TSPC hold sub-block is instructed to hold the M LSBs of the third updated count value, and the flip-flop loader block is instructed to operate in the hold mode to reload the LSBs of the third updated count value to the L-TSPC flip-flop.

6. The L-TSPC trigger-based counter according to claim 5, wherein: The counter controller is further configured to instruct the shift-out block to output the third updated count value from the TSPC holding sub-block and the DFF in a seventh time frame.

7. The L-TSPC trigger-based counter according to claim 1, wherein: The trigger loader block includes: N multiplexers MUX, each MUX having a respective MUX output generated from one selected from at least a respective first MUX input and a respective second MUX input, wherein: For each of the first group of MUXs including the first to M-th MUXs, the corresponding first MUX input is coupled to a corresponding bit in the LSB reset value, and the corresponding second MUX input is coupled to a corresponding bit in the LSB of the count value output by the TSPC hold sub-block; For each of the second group of MUXs including the M+1th to Nth MUXs, the corresponding first MUX input is coupled to a corresponding bit in the MSB reset value, and the corresponding second MUX input is coupled to a corresponding bit in the MSB of the count value output by the DFF; the flip-flop loader block being for setting each of the N MUXs in the reset mode so as to generate each of the respective MUX outputs from the respective first MUX inputs; and The flip-flop loader block is for setting each of the N MUXs in the hold mode so as to generate each of the corresponding MUX outputs from the corresponding second MUX input.

8. The L-TSPC trigger-based counter according to claim 7, wherein: Each of the M L-TSPC flip-flops corresponds to one of the first group of MUXs including the first to M-th MUXs, and the flip-flop loader block further includes: A loading logic block is used to generate the corresponding set / reset signal based on the corresponding MUX output of the first group of MUXs including the first to M-th MUXs, thereby loading each of the L-TSPC flip-flops to output a state, wherein the state indicates a corresponding one of the corresponding MUX outputs in the first group of MUXs including the first to M-th MUXs.

9. The L-TSPC trigger-based counter according to claim 7, wherein: The trigger loader block also includes: a reset block for generating the LSB reset value and the MSB reset value, wherein, for each of the first group of MUXs including the first to M-th MUXs, the corresponding first MUX input is coupled to a corresponding bit in the LSB reset value generated by the reset block, and Wherein, for each of the second group of MUXs including the M+1th to Nth MUXs, the corresponding first MUX input is coupled to a corresponding bit in the MSB reset value generated by the reset block.

10. The L-TSPC flip-flop based counter according to claim 1, wherein: The removal block includes: first one or more transmission gates coupled to an output of the TSPC hold sub-block to selectively output the LSB of the count value; A second one or more transmission gates are coupled to the output of the DFF to selectively output the NM MSBs of the count value.

11. The L-TSPC trigger-based counter according to claim 1, wherein: N is 11 and M is 5.

12. A method for counting pixel transitions using a counter based on a loadable true single-phase clock L-TSPC flip-flop, comprising: Within a conversion time frame defined by a ramp start signal and a ramp stop signal, operating an N-bit counter block having M L-TSPC flip-flops and NM data flip-flops DFF to update a count value by configuring the L-TSPC flip-flops and the DFFs to operate in a clock divider mode to count the number of clock cycles of an input clock signal that pass within the conversion time frame, such that the L-TSPC flip-flops represent the M least significant bits (LSBs) of the count value and the DFFs represent the NM most significant bits (MSBs) of the count value; maintaining the LSB of the count value by latching the LSB of the count value in a TSPC latch coupled to the L-TSPC flip-flop after said running N-bit counter block; and The count value is output by outputting the LSB of the count value from the TSPC latch and the MSB of the count value from the DFF.

13. The method according to claim 12, further comprising: Prior to the running N-bit counter block, the N-bit counter block is initialized by forcing the outputs of the L-TSPC flip-flop and the DFF to each correspond to a corresponding bit of a predetermined N-bit reset value, so that the count value after the running N-bit counter block corresponds to the number of clock cycles of the input clock signal that have passed within the conversion time frame plus the reset value.

14. The method according to claim 12, wherein: The maintaining step also includes generating a plurality of set / reset signals by a trigger loader to configure corresponding set and reset inputs of the L-TSPC flip-flops, reloading each of the L-TSPC flip-flops with a corresponding bit of the LSB of the count value, so that a corresponding output of each of the L-TSPC flip-flops is deterministically forced to a state based on a corresponding bit of the LSB of the count value.

15. The method according to claim 12, wherein: The running step is a first running step, the holding step is a first holding step, the conversion time frame is a first conversion time frame, and the method further includes: After the first hold step, pre-decrementing the count value by negating the count value to generate an inverted count value and resetting the N-bit counter block to force the outputs of the L-TSPC flip-flop and the DFF to each correspond to a corresponding bit of the inverted count value; in a second conversion time frame after the pre-subtraction of the count value, running the N-bit counter module for a second time to update the count value by configuring the L-TSPC flip-flop and the DFF to operate in the clock divider mode to count the number of clock cycles of the input clock signal that have passed in the second conversion time frame, so that the count value after the second update corresponds to the number of clock cycles of the input clock signal that have elapsed in the second conversion time frame minus the number of clock cycles of the input clock signal that have elapsed in the first conversion time frame; and maintaining the LSB of the count value in the TSPC latch by updating the TSPC latch a second time in accordance with the second run of the N-bit counter block, The step of outputting the count value includes: outputting the count value after holding the TSPC latch for the second time.

16. The method according to claim 15, wherein: The step of outputting the count value comprises: After the first holding, outputting the count value for the first time; and The count value after the second holding is output for the second time.

17. The method according to claim 12, wherein: Configuring the L-TSPC flip-flop and the DFF to operate in the clock divider mode includes generating a plurality of set / reset signals by a flip-flop loader to configure respective set and reset inputs of the L-TSPC flip-flop and the DFF so that respective outputs of the L-TSPC flip-flop and the DFF are functions of respective clock inputs.