Timing calibration techniques for discrete time circuits
By comparing the first and second amplitudes of the signal in the discrete time circuit and adjusting the sampling time, the problems of gain error and non-ideal conditions at high sampling rates are solved, and the signal-to-noise ratio and calibration consistency are improved, and the work is reduced Consumption.
Patent Information
- Application Number
- CN202410634909.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2023-05-23
- Filing Date
- 2024-05-22
- Publication Date
- 2024-11-26
AI Technical Summary
The prior art is difficult to effectively correct gain errors and other non-ideal conditions in discrete time circuits at high sampling rates, especially the effects caused by incomplete stability and aging, resulting in a decrease in signal-to-noise ratio and a decrease in calibration consistency.
By sampling the first and second amplitudes of the signal in the discrete time circuit, comparing and adjusting the sampling time, using a variable delay clock and a gain estimator, dynamically adjusting the sampling based on conditions such as signal-to-noise ratio, distortion, gain or stray response, etc., the sampling is dynamically adjusted based on conditions such as signal-to-noise ratio, distortion, gain or stray response, etc. Time, add jitter terms to maintain convergence, ensuring that the sampling time is in the best position.
It realizes maintaining signal-to-noise ratio and calibration consistency at high sampling rates, reducing power consumption, enhancing circuit stability and performance, and being able to dynamically respond to process, voltage and temperature changes.
Smart Images

Figure CN119030538A_ABST
Abstract
Description
Technical Field
[0001] The present disclosure relates generally to calibration techniques and, more particularly, to timing calibration techniques in discrete-time circuits. Background Art
[0002] As the sampling rate of the converter increases, circuits with less than perfect stability parameters can be used to reduce power consumption and still achieve the desired high sampling rate. The less than perfect stability of these circuits often leads to gain errors and other non-ideal conditions in the converter and discrete-time circuits.
[0003] In some converters, gain errors can be corrected in the digital domain using dither-based calibration techniques. These techniques detect the actual gain of the circuit in the analog domain and apply the appropriate digital correction corresponding to that gain. Gain error correction is performed on the interstage gain amplifiers of pipelined analog-to-digital (ADC) converters to correct for interstage gain errors. This correction can also be performed at the output of the track-and-hold (T / H) or buffer circuits to multiple sub-ADCs to correct for gain mismatches.
[0004] Although these techniques can correct for gain errors and nonlinearities, they cannot correct for other effects that may be caused by incomplete settling. In addition, the sampling point may not generally be at the optimum point on the output waveform of the amplifier or the input waveform of the sampler. Summary of the invention
[0005] A simplified overview of one or more aspects is given below to provide a basic understanding of these aspects. This overview is not an extensive overview of all contemplated aspects, and is neither intended to identify the key or critical elements of all aspects, nor to delineate the scope of any or all aspects. Its sole purpose is to present some concepts of one or more aspects in a simplified form as a prelude to a more detailed description presented later.
[0006] A method for operating a discrete-time circuit according to one aspect of the present disclosure may include: sampling a first amplitude of a signal at a first time, sampling a second amplitude of the signal at a second time, comparing the first amplitude to the second amplitude; and adjusting the sampling time of the signal based at least in part on a difference between the first amplitude and the second amplitude.
[0007] This method also optionally includes delaying / advancing a sampling time or delaying / advancing a signal, at least one of the delayed sampling time and the delayed signal including a random delay amount, adjusting the sampling time based at least in part on satisfying a condition of a discrete time circuit, the condition including a signal-to-noise ratio (SNR), distortion, gain, or spurious response, adjusting the sampling time of the signal including delaying a sampling instant of the signal when a difference between a first amplitude and a second amplitude is negative or positive, and adjusting the sampling time of the signal including maintaining the sampling time of the signal when the difference between the first amplitude and the second amplitude is less than a threshold.
[0008] The method may also optionally include adding a jitter term to the sampling time, the value of the jitter term maintaining convergence of the adjustment of the sampling time, sampling the signal a first plurality of times in succession at a first time to obtain a plurality of first amplitudes, sampling the signal a second plurality of times in succession at a second time to obtain a plurality of second amplitudes, comparing a first estimate of the gain obtained from the plurality of first amplitudes with a second estimate of the gain obtained from the plurality of second amplitudes, and adjusting the sampling time of the signal further includes making an adjustment based at least in part on a difference between the first estimate obtained from the plurality of first amplitudes and the second estimate obtained from the plurality of second amplitudes.
[0009] An apparatus according to one aspect of the present disclosure may include a sampling block, a variable delay clock coupled to the sampling block, a gain estimator coupled to the sampling block; and a delay block coupled to the gain estimator, wherein an amount of delay of at least one of the delay block and the variable delay clock is based at least in part on a difference in amplitudes of signals sampled in the sampling block at at least two different times.
[0010] Such an apparatus may also optionally include an amount of delay applied to the input signal, the amount of delay being a random amount of delay based at least in part on satisfying a condition of the discrete time circuit, the condition including at least one of a signal-to-noise ratio (SNR), distortion, gain, or spurious response, the amount of delay being varied by adding a jitter term, the value of the jitter term maintaining convergence of the sampling time, and
[0011] The difference in amplitude of the signal sampled in the sampling block comprises a difference in estimates of the amplitude of the signal sampled at at least two different times.
[0012] To accomplish the above-mentioned and related purposes, one or more aspects include the features hereinafter fully described and particularly pointed out in the claims. The following description and the accompanying drawings set forth in detail certain illustrative features of one or more aspects. However, these features are indicative of only some of the various ways in which the principles of the various aspects may be employed, and the present description is intended to include all such aspects and their equivalents. BRIEF DESCRIPTION OF THE DRAWINGS
[0013] Figure 1A and1B is a graph of sedimentation patterns at sampling moments according to an exemplary aspect of the present disclosure.
[0014] Figure 2 is a block diagram according to example aspects of the present disclosure.
[0015] Figures 3A to 3D is a diagram of a method of sampling a signal according to an exemplary aspect of the present disclosure.
[0016] Figure 4 is a block diagram of a timing calibration circuit according to an exemplary aspect of the present disclosure.
[0017] Figure 5 is a graph of a stable distribution of an input signal according to an exemplary aspect of the present disclosure.
[0018] Figure 6 is a graph of sampling instants shifted toward a peak of an input signal according to an exemplary aspect of the present disclosure.
[0019] Figure 7 is a graph of a clock signal delayed or advanced relative to an input signal according to an exemplary aspect of the present disclosure.
[0020] Figure 8 is a graph of convergence of timing errors of a clock signal delayed or advanced relative to an input signal according to an exemplary aspect of the present disclosure.
[0021] Fig. 9 is a flow chart according to an exemplary aspect of the present disclosure. DETAILED DESCRIPTION
[0022] The detailed description set forth below in conjunction with the accompanying drawings is intended as a description of various configurations and is not intended to represent the only configuration in which the concepts described herein may be practiced. The detailed description includes specific details used to provide a comprehensive understanding of the various concepts. However, it is apparent to those skilled in the art that these concepts may be practiced without these specific details. In some cases, well-known structures and components are shown in block diagram form to avoid confusing these concepts.
[0023] Aspects of the present disclosure allow the gain of an amplifier in a discrete time circuit to be sampled at a desired point throughout the life of the circuit. Aspects of the present disclosure also allow the desired point to be selected based on amplitude, signal-to-noise ratio, or other parameters, and allow the sampling point to be dynamically selected and / or moved based on process, voltage, and temperature parameters.
[0024] Overview
[0025] Figure 1A and 1BSettling patterns and sampling moments according to exemplary aspects of the present disclosure are shown.
[0026] Figure 1A An example of an output signal 100 of an amplifier in a discrete time circuit is shown, wherein the output signal 100 is sampled at a given point in the output signal 100 curve. Figure 1A , the amplitude of the output signal 100 is on the y-axis and time is on the x-axis. The sampling time 102 is shown at or near the peak of the output signal 100, which is the desired or optimal point to sample the output signal 100. The slope of the output signal 100 is approximately zero at the desired or optimal sampling point of the output signal 100.
[0027] However, as process, voltage, and temperature (PVT) conditions vary, the sampling instant 102 in a discrete time circuit may occur before or after the optimum point for the output signal 100. Aging of amplifier and clock circuits may also be a problem in fine geometry processes; such aging of circuits may also affect the relative timing of the sampling instants 102 within a given system. In the presence of such timing variations, the desired performance may not be maintained at high speeds.
[0028] Figure 1B A movement 104 from sampling instant 102 to sampling instant 106 is shown. Figure 1B , the amplitude of the output signal 100 is on the y-axis and the time is on the x-axis. The shift 104 of the sampling instant 102, although shown as shifting the sampling instant 102 to the left, may also shift the sampling instant 102 to the right without departing from the scope of the present disclosure.
[0029] The shift 104 from the sampling instant 102 to the sampling instant 106 may be caused by one or more conditions within the circuit of the discrete time circuit as described above, such as PVT, aging, etc. Figure 1B As shown, sampling instant 106 now occurs at a point or region of the output signal 100 where the output signal 100 of the discrete time circuit has a greater slope than sampling instant 102. While some effects of shift 104 can be corrected in the digital domain, such as gain error, etc., other effects of shift 104 do not necessarily need to be corrected, such as clock jitter, etc. These effects may reduce the signal-to-noise ratio (SNR) of the circuit. Sampling at a high slope region of the output signal 100, combined with a shorter settling time of the circuit, may also increase the overall sensitivity of the converter to PVT variations, which will reduce the consistency and effectiveness of the calibration of the entire system.
[0030] Good layout practices and component matching can be used to reduce the amount of movement 104. However, these may have only limited success due to the different locations and different data / clock processing employed in different paths, such as path length, number of inverters used to achieve a particular drive strength, etc. In addition, the location of the sampling instant 106 relative to the output signal 100 also depends on the propagation delay of the input signal in the amplifier, matching amplifier delays, amplifier network clocks, and receive sampling network clocks on multiple ADCs, and the variability of temperature, power supply, process, and duty cycle to within 1-2 picocon (ps), which is difficult to use consistently over the entire life cycle of a given amplifier. In addition, there are statistical variations in different circuit paths, which may also increase the variation of about 2ps.
[0031] To account for all of these variables, a margin on the order of 5-10ps can be used to ensure reasonable performance. At high sample rates, such as 20 Gsample / s, the sample rate may drop by 30% and / or the noise / spurious performance may drop by about 10dB when the available settling time is taken into account. These degradations are unacceptable and will increase at higher sample rates.
[0032] In one aspect of the present disclosure, techniques are provided to calibrate the sampling instant 102 to be at a desired or "optimal" position and still account for PVT variations and aging. In one aspect of the present disclosure, such techniques can also help correct for gain errors and other non-idealities. These techniques can enable timing calibration of discrete time sampling circuits where insufficiently stable amplifiers are used. Such techniques can also allow for increased sampling rates and can also reduce the power consumption of such circuits.
[0033] For discrete time signals, the error caused by timing offset is not orthogonal to the gain error. For systems with random inputs or with jitter injected on the input, the desired signal is well behaved with respect to time. Since timing error causes gain error, timing variation causes gain variation. Although the exact gain estimate may have inconsistencies due to poor settling, the approximate relationship between gain and settling time can be captured in the gain estimate with sufficient accuracy. Therefore, the reduction of timing error can be treated as a gain maximization problem. The timing can be adjusted by formulating the problem as a gain maximization or gain error minimization problem.
[0034] Gain maximization / sample timing delay
[0035] Figure 2 A block diagram is shown according to exemplary aspects of the present disclosure.
[0036] Diagram 200 shows a device that receives a signal 202 (Vin) as an input to a sampling block 204. The sampling block 204 is coupled to a variable delay clock 206 that receives a clock signal 208 input and a delay input 210.
[0037] Data 212 is the output of sampling block 204 and is used as an input to gain estimator 214. The gain estimation output of gain estimator 214 is coupled to delay block 216. Delay block 216 generates delay input 210 which is used as an input to variable delay clock 206.
[0038] Delay block 216 allows the amplifier to be sampled with a variable delay, where delay block 216 shifts the sampling time (instant) based on the amplitude difference at different sampling times.
[0039] The following algorithm may be used to control the timing delay from delay block 216 such that the gain is maximized:
[0040]
[0041] Where t is the sampling time (delay adjustment for the sampling time), G is the gain estimate, and D d is optional digital jitter added in the digital domain to prevent the algorithm from "getting stuck" on a wrong maximum. Alternatively, the update steps can be taken uniformly in the ascending direction:
[0042] ΔG=G(i)-G(i-1)
[0043] ΔT=t(i)-t(i-1)+D d (i)
[0044] update_dir = sign(ΔG×ΔT)
[0045] t(i+1)=t(i)+μ×update_dir
[0046] Additionally, the update step can be scaled by changes in gain (ΔG) or time (ΔT):
[0047] t(i+1)=t(i)+μ×abs(ΔG)×update_dir,
[0048] t(i+1)=t(i)+μ×abs(ΔT)×update_dir.
[0049] To prevent the algorithm from getting stuck after convergence, if necessary, the digital dither D d is added to the algorithm. This additional jitter term can be applied, but not necessarily, to any timing adjustments applied in the analog domain.
[0050] If a jitter term is applied to the time adjustment from delayed clock 206, the jitter term can be made small enough based on the slope of the sampled signal and the desired performance so as to keep the sampling instant of signal 202 in a desired region and / or have a negligible impact on performance when the sampling instant is at a desired location.
[0051] In one aspect of the present disclosure, variations of these equations can be used to maximize a desired amount or minimize error. The adjustment of the delay can be in the direction of higher gain, or can meet another desired condition, such as maximizing signal-to-noise ratio (SNR), minimizing distortion, minimizing spurious responses, and / or other conditions. Without departing from the scope of the present disclosure, each desired condition can adjust the direction of the delay in a different direction and can correspond to a different algorithm than shown above.
[0052] The gain estimate from the gain estimator 214 may be obtained by using a jitter-based (correlation-based) technique or by using a power estimate (or absolute value estimate) of the input signal or any other method. Other variations of the algorithm may also be used, where G does not necessarily represent gain, but some other parameter that needs to be maximized, such as SNR, or some parameter that needs to be minimized, such as spurious levels or distortion.
[0053] Timing Calibration Under Random Timing Perturbations
[0054] FIG. 3A to FIG. 3D A method for sampling a signal according to an exemplary aspect of the present disclosure is shown. Figures 3A-3D In the graph, the amplitude of the signal is on the y-axis and time is on the x-axis.
[0055] In one aspect of the present disclosure, the change of sampling instant 106 may be performed over multiple samples of signal 100. Furthermore, the amplitude or any other desired parameter of signal 100 may be compared at two different sampling instants over a large number of samples of signal 100.
[0056] Based on the comparison results over a large number of samples, the sampling instant of signal 100 may be moved to a desired point, such as an approximately maximum amplitude of signal 100, an approximately maximum signal-to-noise ratio, etc. For example, if the sampling instant is desired at or near the peak of a settling curve (step response) of signal 100, then a delay (e.g., a delay in delay clock 206 or signal 100 itself) may be adjusted until the signal sampled at two different times over a large number of samples has approximately the same amplitude estimate.
[0057] Figure 3AA first sampling instant 300 at time t and a second sampling instant at time t+Δt are shown. In one aspect of the present disclosure, the signal 100 is sampled at either the sampling instant 300 or the sampling instant 302. The sampling instant of a given signal 100 alternates between the sampling instant 300 and the sampling instant 302. If desired, the alternation of the sampling instants 300 and 302 can be randomly alternated. The gains at the sampling instants 300 and 302 are estimated after collecting a desired number of samples at the sampling instants 300 and 302. Using these gain estimates at the sampling instants 300 and 302, the slope of the gain is determined, which allows the position at which the signal 100 is to be sampled to be calibrated by moving the delayed clock 206 and / or the signal 100 relative to the sampling instant 300 or the sampling instant 302.
[0058] Figure 3B A first sampling instant 300 and a second sampling instant 302 are shown, wherein the amplitude of the signal 100 at the second sampling instant 302 is higher than the amplitude of the signal 100 at the first sampling instant 300. In such an example, the slope of the amplitude of the signal 100 is positive, e.g., has a value greater than zero. When the slope of the amplitude of the signal 100 is positive, in one aspect of the present disclosure, the delay of the sampling instant can be moved in direction 304, e.g., a longer delay of the sampling instant or a shorter delay of the signal 100 (e.g., in this case, the signal 100 needs to be advanced relative to the delay clock 206). In such an aspect, the sampling times are referred to as being "delayed" in time.
[0059] Figure 3C A first sampling instant 300 and a second sampling instant 302 are shown, wherein the amplitude of the signal 100 at the second sampling instant 302 is lower than the amplitude of the signal 100 at the first sampling instant 300. In such an example, the slope of the amplitude of the signal 100 is negative, e.g., has a value less than zero. When the slope of the amplitude of the signal 100 is negative, in one aspect of the present disclosure, the delay of the sampling instant can be moved in a direction 306, e.g., a shorter delay of the sampling instant or a longer delay of the signal 100. In such an aspect, the sampling time is said to be "advanced" in time.
[0060] Figure 3DA first sampling instant 300 and a second sampling instant 302 are shown, wherein the amplitude of the signal 100 at the second sampling instant 302 is approximately equal to the amplitude of the signal 100 at the first sampling instant 300. In such an example, the slope of the amplitude of the signal 100 is approximately equal to zero. When the slope of the amplitude of the signal 100 is approximately zero, or within a desired tolerance or threshold depending on the slope of the sampled signal and the desired performance. In one aspect of the present disclosure, the delay of the sampling instant can be left alone because the desired sampling instant has been found. In this way, the movement of the sampling instant can be in the direction 308, for example, if any change is desired, the minimum change of the sampling instant can be used.
[0061] In one aspect of the present disclosure, the signal 100 is sampled in various situations, for example, Figure 3A The first sampling time 300 and the second sampling time 302 are shown, and according to the slope of the amplitude (or other desired parameters), the sampling time is advanced or delayed, such as Figure 3B and 3C As shown, until the slope reaches a stationary point or a plateau of the desired parameter, or reaches Figure 3D The sampling point of the signal 100 determined by the iterative process is then used. Due to aging of the circuit generating the signal 100, temperature changes, etc., in one aspect of the present disclosure, the sampling point of the signal 100 can be moved.
[0062] In one aspect of the present disclosure, a large number of samples, for example, a large number of first sampling instants 300 and second sampling instants 302 of the signal 100 over time, may be averaged together before a sampling point of the mobile signal 100. Thus, for example, and not by way of limitation, 100 to 100 thousand first sampling instants 300 and second sampling instants 302 may be taken and averaged, and these averages are used to determine the slope of the signal 100 before a sampling point of the mobile signal 100, wherein the number of samples collected at the sampling instants 300 and 302 may be approximately the same, and if the gain estimates at the first sampling instants 300 and the second sampling instants 302 are within a desired tolerance or indicate a zero slope of the signal 100, then a slope of 100 has been reached. Figure 3D of the still point.
[0063] In one aspect of the present disclosure, any desired sampling point of the signal 100 (or signal 202) can be reached by determining the desired slope between the first sampling instant 300 and the second sampling instant 302, determining the normalized offset between the peaks of the two sampling instants, or determining the time distance between the plateau of the signal 100 and the desired sampling point. Many possibilities for the sampling instants are possible within the scope of the present disclosure. In one aspect of the present disclosure, an offset can be added to the gain estimation block 214 so that the sampling instants converge to the slope corresponding to the added offset value.
[0064] In one aspect of the present disclosure, the signal 402 may be sampled at two different points by varying the sampling time from one sample to another, or by varying the timing of the signal 402 relative to the clock signal 408 from one sample to another. Similarly, a correction in timing may be applied to the delay block 406 or the input signal 402 as a delay correction. The signal 402 may be sampled at two different points by varying the sampling time from one sample to another, or by varying the timing of the signal 402 relative to the clock signal 408 slowly (not necessarily sample by sample). For example, the sampling instant may be randomly varied from 300 to 302. This random selection may be updated at a fraction of the sampling frequency.
[0065] Figure 4 A block diagram of a timing calibration circuit according to an exemplary aspect of the present disclosure is shown.
[0066] Figure 4 System 400 is shown showing signal 402 (Vin) as an input to sampling block 404. Sampling block 404 is coupled to variable delay clock 406 which receives clock signal 408 input and delay signal 410.
[0067] Data 412 is the output of sampling block 404 and is used as an input to pre-gain calibrator 414. Pre-gain calibrator 414 also receives pre-gain coefficient 416 for sampling time t1 and pre-gain coefficient value 418 for sampling time t2 from circuit 420. Circuit 420 may also receive a dither input 422.
[0068] Pre-gain calibrator 414 outputs signal 424 to gain (multiplication) block 426, which also receives gain 428 at sampling time t1 and gain 430 at sampling time t2 from circuit 432. Circuit 432 may also receive dither input 422.
[0069] Output 434 of summing block 426 is provided as input to post gain calibrator 436, which also receives post gain coefficient 438 for sampling time t1 and post gain coefficient value 440 for sampling time t2 from circuit 442. Gain post calibrator 436 provides output 444 from system 400 for use as a gain estimate for a given system.
[0070] Output 434 is also provided to circuit 446, which provides gains 428 and 430 of signal 402 to gain estimator 448. Within gain estimator 448, a gain estimate G1 for time t1 is completed in circuit 450, and a gain estimate G2 for time t2 is completed in circuit 452. In one aspect of the present disclosure, these gain estimates G1 and G2 can be made in consecutive samples. This can be used for slowly varying signals. For signals whose amplitude varies with sampling, a large number of samples can be taken at sampling time 300 and sampling time 302 so that the gain at these two sampling times can be accurately estimated, or an average can be taken over multiple samples as needed. The gain estimate at sampling time t1 is G1, while the gain estimate at t2 is G2. G1 is compared to G2 in comparator 454. If the embodiment is an averaging embodiment, in which a large number of samples are taken before comparison, counter 456 can be used to count the number of samples and / or reset the comparison performed in gain estimator 448.
[0071] The output of comparator 454 is provided to delay block 458, which may further average / filter the output and adjust the delay of signal 402 and / or the delay of the sampling instant of signal 402, such as with respect to Figures 3A-3D Delay block 458 generates a delayed input 460 which is used as an input to a summation block 462 .
[0072] The summing block 462 receives the delay input 460 and the jitter input 422 and adds these signals together to provide a delayed signal 410 to the variable delay clock 406. This feedback of the delayed signal 410 then changes the sampling instant of the signal 402, as with respect to Figures 3A-3D as described.
[0073] In one aspect of the disclosure, the clock signal 408 may be altered or “perturbed,” such as delayed or advanced, to make the system 400 less sensitive to input conditions by dithering the input 422 .
[0074] Since the signal 402 cannot be sampled at t1 and t2 (i.e., sampling instant 300 and sampling instant 302) in the same signal, a statistical method of a large number of samples can be used. In one aspect of the present disclosure, the sampling instant can be randomly changed between two points from one sample to another sample, or at a speed that can be a small fraction of the sampling speed. The maximum peak-to-peak value of a large number of samples at the sampling points can be compared, or based solely on the comparison sign of the points, an error signal (delay input 460) can be generated, which is then used to adjust the delay of the variable delay clock 406. Although only one of the sampling instants can be moved, in one aspect of the present disclosure, the two sampling instants (i.e., sampling instant 300 and sampling instant 302) can be moved simultaneously through the delay input 460. The timing difference between the sampling instant 300 and the sampling instant 302 can be maintained by the summing block 462 while adjusting the delay. The summation block 462 shifts the clock through the dither input 422 which randomly shifts the sampling instant from the sampling instant 300 to the sampling instant 302 and the delay input 460 which shifts the sampling instant 300 and the sampling instant 302 together to a desired point.
[0075] In one aspect of the disclosure, coefficients 416, 418, 428, 430, 438, and 440 may be used anywhere in system 400, and the method described for system 400 may be used at the output of a sample-and-hold or track-and-hold system or at any stage in such a system. The method described in the disclosure may be applied to any dynamic amplifier behavior where incomplete settling makes it desirable to control the sampling instant. Furthermore, the method described in the disclosure may be applied regardless of whether the amplifier settling behavior is underdamped, overdamped, or integrator-like. In the case of severe underdamping, this method may be adjusted to handle multiple peaks and designed to converge to an appropriate peak of signal 402 as desired, e.g., the maximum peak, the peak with the longest settling time, the maximum SNDR, etc. In one aspect of the disclosure, the sampling instant shifting method may be used to achieve a specific desired gain value, rather than maximizing the gain as described above.
[0076] In one aspect of the present disclosure, an addition may be added to signal 402 to dither signal 402. In the case where system 400 has no external inputs, the additional input, which may be a dithering input, may serve as signal 402 (i.e., input) to system 400. In such an aspect, calibration of sampling instant 300 and sampling instant 302 may be performed by such additional inputs to system 400. Thus, the calibration method presented as part of the present disclosure may operate even in the absence of external inputs to system 400. Furthermore, instead of controlling the timing of variable delay clock 406, the algorithm may be used to control the timing of a clock, track and hold, or a stage of a track and hold system, and / or control some parameters of an amplifier to control its delay, reset time, etc.
[0077] When random timing perturbations are injected into the clock signal 408, the frequency of the random timing perturbations may be slow to avoid the effects of rapid changes in timing and hardware variations.
[0078] Figure 5 A stable distribution of an input signal according to an exemplary aspect of the present disclosure is shown.
[0079] Graph 500 shows signal 502 which is a plot of the settling behavior of an amplifier over time with amplitude on the y-axis and time on the x-axis.
[0080] Figure 6 A graph showing sampling instants moving towards a peak of an input signal according to an exemplary aspect of the present disclosure.
[0081] Graph 600 shows line 602 and line 604 starting at diverging values and reaching similar values at the instant of sampling. Figure 6 In FIG. 6 , the delay of the signal is shown on the y-axis and time is shown on the x-axis. Line 602 shows a negative slope of convergence, similar to the Figure 3D As discussed, line 604 shows a converging positive slope, similar to Figure 3C discussed.
[0082] Figure 7 A clock signal delayed or advanced relative to an input signal according to an exemplary aspect of the present disclosure is shown.
[0083] Graph 700 shows signal 702, which is a graph of the stable behavior of the amplifier over time, with amplitude on the y-axis and time on the x-axis. Signal 704 shows an "early" clock, i.e., a clock signal that samples signal 702 before the peak of signal 702. Signal 706 shows a "late" clock, i.e., a clock signal that samples signal 702 after the peak of signal 702. Figure 7 , the delay of signals 704 and 706 is shown on the y-axis and time is shown on the x-axis.
[0084] Signal 708 illustrates the convergence of signal 704 and signal 706 after applying features of the present disclosure to signal 704, signal 706, and / or signal 702. Signal 708 illustrates that the clock signal has been shifted relative to signal 702 and now samples signal 702 at a peak amplitude.
[0085] Figure 8 The convergence of the timing error of a clock signal that is delayed or advanced relative to an input signal according to an exemplary aspect of the present disclosure is shown. Figure 8 In the example, the delay of the signal is shown on the y-axis and time is shown on the x-axis.
[0086] Graph 800 shows line 802, which is a convergence graph of an early clock signal, line 804, which is a convergence graph of a late clock signal, and line 806, which is a convergence graph of a late clock signal that is further delayed relative to the clock signal used in line 804. Lines 802, 804, and 806 start at divergent values and reach similar values at the sampling instant. Line 802 shows a similar pattern to that about Figure 3C The positive convergence slope discussed, while line 804 and line 806 show similar Figure 3D The negative convergence slope in question.
[0087] Fig. 9 A flow chart according to an exemplary aspect of the present disclosure is shown.
[0088] Diagram 900 shows blocks 902 , 904 , 906 , and 908 .
[0089] Block 902 represents sampling a first amplitude of a signal at a first time. Block 902 may be performed as described with respect to Figure 2 The described sampling block 204 is performed.
[0090] Block 904 represents sampling the second amplitude of the signal at a second time. Block 904 may be performed as described with respect to Figure 2 The described sampling block 204 is performed.
[0091] Block 906 represents comparing the first amplitude with the second amplitude. Block 906 may be performed as described with respect to Figure 2 The gain estimator 214 described above is performed.
[0092] Block 908 represents adjusting a sampling time of the signal based at least in part on a difference between the first amplitude and the second amplitude. Block 908 may be performed as described with respect to Figure 2 The variable delay clock 206 described above is implemented.
[0093] The previous description is provided to enable any person skilled in the art to practice the various aspects described herein. It will be apparent to those skilled in the art that various modifications to the exemplary aspects and aspects throughout this disclosure will be apparent, and the concepts disclosed herein can be applied in other contexts and for different purposes. Therefore, the claims are not intended to be limited to the exemplary aspects presented throughout this disclosure, but should be given the full scope consistent with the language claims. All structural and functional equivalents of the elements of the exemplary aspects described throughout this disclosure that are known or will be known later by those of ordinary skill in the art are intended to be covered by the claims. In addition, nothing disclosed herein is intended to be dedicated to the public, regardless of whether such disclosure is explicitly listed in the claims. No claim element shall be interpreted under the provisions of Section 112(f) of Title 35 of the United States Code or similar laws of applicable jurisdictions, unless the phrase "method of..." is explicitly used to state the element, or in the case of a method claim, the phrase "step of..." is used to narrate the element.
Claims
1. A method for operating a discrete-time circuit, comprising: sampling a first amplitude of a signal at a first time; sampling a second amplitude of the signal at a second time; comparing the first amplitude to the second amplitude; and A sampling time of the signal is adjusted based at least in part on a difference between the first amplitude and the second amplitude.
2. The method according to claim 1, wherein adjusting the sampling time of the signal further comprises at least one of the following: Delay the sampling time or delay the signal. 3 . The method of claim 2 , wherein delaying at least one of the sampling time and delaying the signal comprises a random amount of delay.
4. The method of claim 1, wherein adjusting the sampling time of the signal further comprises at least one of the following: Advance the sampling time or advance the signal. The method of claim 1 , wherein adjusting the sampling time is based at least in part on satisfying a condition of the discrete-time circuit.
6. The method of claim 5, wherein the condition comprises at least one of signal-to-noise ratio, distortion, gain, or spurious response.
7. The method according to claim 1, wherein adjusting the sampling time of the signal comprises: When the difference between the first amplitude and the second amplitude is negative, a sampling time of the signal is delayed.
8. The method of claim 1, wherein adjusting the sampling time of the signal comprises: When the difference between the first amplitude and the second amplitude is positive, the sampling time of the signal is advanced.
9. The method of claim 1 , wherein adjusting the sampling time of the signal comprises: When the difference between the first amplitude and the second amplitude is less than a threshold, the sampling time of the signal is maintained.
10. The method according to claim 1, further comprising: A jitter term is added to the sampling time. The method of claim 10 , wherein the value of the jitter term maintains convergence of the adjustment of the sampling time.
12. The method according to claim 1, further comprising: Performing a first plurality of continuous samplings on the signal at a first time to obtain a plurality of first amplitudes; Performing a second plurality of consecutive samplings on the signal at the second time to obtain a plurality of second amplitudes; comparing a first estimate of a gain of a signal obtained from the plurality of first amplitudes with a second estimate of a gain of a signal obtained from the plurality of second amplitudes; and Wherein adjusting the sampling time of the signal further comprises adjusting based at least in part on a difference between a first estimate obtained from the plurality of first amplitudes and a second estimate obtained from the plurality of second amplitudes.
13. An apparatus for operating a discrete-time circuit, comprising: Sampling blocks; a variable delay clock coupled to the sampling block; a gain estimator coupled to the sampling block; and A delay block is coupled to the gain estimator, wherein an amount of delay of at least one of the delay block and the variable delay clock is based at least in part on a difference in amplitude of signals sampled in the sampling block at at least two different times. The apparatus of claim 13 , wherein the delay amount is applied to an input signal. The apparatus of claim 14 , wherein the delay amount is a random delay amount.
16. The apparatus of claim 13, wherein the amount of delay is based at least in part on satisfying a condition of the discrete-time circuit.
17. The apparatus of claim 16, wherein the condition comprises at least one of signal-to-noise ratio, distortion, gain, or spurious response.
18. The apparatus of claim 13, wherein the delay amount is varied by adding a jitter term. The apparatus of claim 18 , wherein the value of the jitter term is used to implement a sampling instant.
20. The apparatus of claim 13, wherein the difference in amplitude of the signal sampled in the sampling block comprises a difference in estimates of amplitude of the signal sampled at at least two different times.
