Successive approximation register (SAR) analog to digital converter (ADC) with partial loop-unrolling

Abstract

A receiver system that includes an ADC for converting analog values to digital representations. A digital representation is a sum of discrete values some of which are non-binary scaled and the other are binary scaled. The ADC includes dedicated comparators to determine whether to add or to subtract the non-binary scaled values. A comparator is used to determine whether to add or to subtract the binary scaled values. The ADC further calibrates offset voltages of the comparators to substantially remove dead zone and conversion errors, without compromising the conversion speed. The calibration can be performed both in foreground and background.

Description

BACKGROUND

1. Field of the Disclosure

This disclosure pertains in general to digital transceivers, and more specifically to analog to digital converters.

2. Description of the Related Art

Serial/Deserializers (SerDes's) are widely used in applications that involve gigabit rate links. Compared to parallel links, serial communication has many advantages such as no crosstalk noise, occupies less area, and consumes less power. Analog to digital converters (ADC) are used in SerDes's to digitize analog signals to achieve channel equalization. However, ADCs can be power hungry and present challenges in integrating ADCs in SerDes′, especially in high speed operation.

SUMMARY

Integration of high speed, low power and area efficient ADCs is a key differentiator in the design of the receiver (RX) path of an ADC-based Serializer-Deserializer (SerDes) transceiver. An ADC converts analog values to digital representations. A digital representation is a sum of discrete values some of which may be non-binary scaled and/or binary scaled. The ADC includes dedicated comparators to determine whether to add or to subtract the non-binary scaled values. A comparator is used to determine whether to add or to subtract the binary scaled values. The ADC further calibrates offset voltages of the comparators to substantially remove dead zone and conversion errors, without compromising the conversion speed. The calibration can be performed both in foreground and background.

Other aspects include components, devices, systems, improvements, methods, processes, applications and other technologies related to the foregoing.

BRIEF DESCRIPTION OF THE DRAWINGS

The teachings of the embodiments disclosed herein can be readily understood by considering the following detailed description in conjunction with the accompanying drawings.

FIG. 1A is a block diagram illustrating an example partial loop-unrolling successive approximation register (SAR) analog-to-digital converter (ADC), according to one embodiment.

FIG. 1B is a timing diagram of an example ADC, according to one embodiment.

FIG. 2A is a block diagram illustrating an example ADC that outputs a 7-bit digital signal using 8 conversion cycles, according to one embodiment.

FIG. 2B is a block diagram illustrating an example ADC that outputs a 7-bit digital signal using 8 conversion cycles, according to one embodiment.

FIG. 3 is a block diagram illustrating an example partial loop-unrolling SAR ADC with comparator offset calibration, according to one embodiment.

DETAILED DESCRIPTION

The Figures and the following description relate to various embodiments by way of illustration only. It should be noted that from the following discussion, alternative embodiments of the structures and methods disclosed herein will be readily recognized as viable alternatives that may be employed without departing from the principles discussed herein. Reference will now be made in detail to several embodiments, examples of which are illustrated in the accompanying figures. It is noted that wherever practicable similar or like reference numbers may be used in the figures and may indicate similar or like functionality.

Operation

FIG. 1A is a block diagram illustrating an example partial loop-unrolling successive approximation register (SAR) analog-to-digital converter (ADC), according to one embodiment. The example SAR ADC 100 outputs an N-bit digital signal. The example partial loop-unrolling SAR ADC 100 generates the N-bit digital signal in Nc quantization cycles. The number of cycles Nc (Nc=m+n) is greater than the number of digits to introduce redundancy in the conversion process of the ADC 100 by means of non-binary scaling of the weights. Out of the Nc cycles, m cycles are non-binary cycles where the discrete values are non-binary scaled and n cycles are binary cycles where the discrete values are binary scaled. The most significant bits (MSBs) are determined in m cycles and the least significant bits (LSBs) are determined in n cycles. A loop-unrolled approach is applied to make the determination in the m cycles. That is, for each of the m most significant bits (MSBs), an individual comparator determines whether the discrete value should be added or subtracted. For the n cycles, one comparator is used to determine whether to add or to subtract the discrete values for all n least significant bits (LSBs). The m cycles are also referred herein as “loop-unrolled cycles” and the n cycles are also referred herein as “SAR cycles.” By doing this, the proposed ADC achieves an optimum power consumption, operation speed, and area occupation.

The ADC 100 includes a sampling switch 101, an adder 102, a series of comparators 103_1-m, a comparator 104, a shift register 106, a digital-to-analog converter (DAC) 108, and a digital register 110. The comparators 103_1-m are also collectively referred to as the comparators 103. As illustrated, one terminal of the sampling switch 101 is the input terminal of the ADC 100 and the other terminal of the sampling switch 101 is connected to the adder 102. FIG. 1B is a timing diagram of the ADC 100. FIGS. 1A and 1B are described in connection with each other to illustrate the operation of the ADC 100.

The ADC 100 is driven by a clock signal to take samples of an incoming analog signal V_in(t). The ADC 100 can sample the incoming analog signal at rising edges of the clock signal. That is, at each rising edge of the clock signal, the ADC 100 takes a sample of the incoming analog signal and converts the sampled value to a N-bit digital symbol representing the sampled value as shown in Equation (1):

$\begin{array}{cc}{V}_{\mathrm{ADC}}=\sum _1^{N_c}{b}_i\times w_i,& \left(1\right)\end{array}$ where i is an integer in the range of [1, Nc] and represents one of the Nc cycles, b_i is a logical signal determined in the i cycle, and w_i is the weight (i.e., discrete value) in the i cycle. The logical signal b_i has values indicating whether the corresponding weight w_i should be added or subtracted.

The sampling switch 101 is controlled according to the clock signal. As illustrated in FIG. 1B, in a clock cycle T, the sampling switch 101 is switched off according to the rising edge of the clock signal and stays off for a half clock cycle (0−T/2) when the clock signal is in its high state. When the sampling switch 101 is turned off, the ADC 100 samples and converts the incoming analog signal to a N-bit digital symbol. The sampling switch 101 is switched on according to the falling edge of the clock signal and stays on for the half-cycle (T/2−T) when the clock signal is in its low state. When the sampling switch 101 is on, the ADC 100 is in tracking mode where it pauses the conversion until the next clock cycle.

The adder takes as input the sampled value V_in(zT) and the output of the DAC 108 and outputs a difference signal (V_in(zT)−V_DAC), where z is an integer. The output of the DAC 108 represents an existing digital conversion of the sampled value. The difference signal represents the remaining analog value to be converted. In a loop-unrolled cycle i, the difference signal is provided as an input to a comparators 103_i and serves as the basis for the comparator 103_i to determine the logical signal b_i. In a SAR cycle i, the difference signal is provided as an input to the comparator 104 and serves as the basis for the comparator 104 to determine the logical signal b_i.

To determine the logical signal b_i (i.e., whether to add or subtract a discrete value w_i), the comparator 103_i (or the comparator 104) compares the sampled value V_in(zT) to the DAC's output V_DAC. If the sampled value V_in(zT) is greater than (less than) the DAC's output V_DAC, the comparator outputs the logical signal b_i having a positive (negative) value indicating that the discrete value w_i is to be added (subtracted). The outputs of the comparator 103_i (or the comparator 104) can be determined according to Equations (2)-(3):b_i=+1 if V_in(zT)>V_DAC  (2),b_i=−1 if V_in(zT)<V_DAC  (3).

In addition to generating the logical signal b_i, the comparator 103_i outputs a ready signal (rdy_i) that enables the next comparator 103_(i+1) (or 104) to make the determination. For example, as illustrated, the ready signal rdy₁ output by the comparator 103-1 that determines the logical signal b₁ enables the comparator 103-2 to make the determination on the logical signal b₂. The ready signal rdy_m output by the comparator 103_m enables the comparator 104 to make the determination on the logical signal b_(m+1).

In each SAR cycle, the comparator 104 outputs the logical signal b_i to the shift register 106. At a given time, the shift register 106 stores one or more logical signals that have been outputted by the comparator 104 and provides the logical signals to the DAC 108. Additionally, the comparator 104 generates a ready signal that triggers the comparator 104 to reset and to make the determination for the next cycle after a delay.

As described above, the DAC's 108 output V_DAC is the existing digital conversion of the sampled analog value. In a cycle i, the DAC's output can be determined according to Equation (4):

$\begin{array}{cc}{V}_{\mathrm{DAC}}=\sum _1^{i-1}{b}_j\times w_j,& \left(4\right)\end{array}$ where j is an integer in the range of [1, i−1], b_j is a logical signal determined in the j cycle, and w_j is the weight in the j cycle. As shown in Equation (4), the DAC's output is updated when a comparator determines to add or to subtract a discrete value.

The digital register 110 outputs the N-bit digital symbol. The digital register 110 combines the logical signals generated by the comparators 103-104 with their corresponding weights to determine the N-bit converted digital signal according to Equation (1).

The partial loop-unrolling SAR ADC 100 achieves a power efficient digital conversion of analog signals while maintaining a high conversion speed, a low area occupation, as well as optimized noise performance and error tolerance. Because dedicated comparators are used for making determinations for the MSBs, these comparators do not need to be reset during the conversion. This is beneficial for the conversion speed. In addition, the noise and power consumption of the comparators used in the loop-unrolling cycles can be optimized for every cycle of comparison, for example based on the amount of redundancy available at that cycle. Furthermore, because one comparator is reused in all SAR cycles to make determinations for LSBs that are more subject to comparator noises, the comparator noise can be optimized such that it is negligible compared to quantization noise.

In one embodiment, the ADC 100 outputs a 7-bit digital signal and the full-scale (FS) is 64. The ADC 100 employs 8 cycles to digitize an analog sample according to Equation (5):V_ADC=(b₁·34+b₂·18+b₃·10+b₄·5)+(b₅·4+b₆·2+b₇·1+b₈·0.5)  (5).In this particular embodiment, 4 cycles are loop-unrolled cycles where dedicated comparators determine the logical signals b₁-b₄, and 4 cycles are SAR cycles where one comparator determines the logical signals b₅-b₈. Other combinations of weights are also possible.

In the illustrated example, the ADC 100 samples the incoming analog signal at rising edges of the clock signal. The ADC 100 can also sample the incoming analog signal at falling edges of the clock signal or at both rising and falling edges of the clock signal.

FIG. 2A is a block diagram illustrating an example ADC 200 that outputs a 7-bit digital signal using 8 conversion cycles. Out of the 8 conversion cycles, 4 are loop-unrolled cycles and 4 are SAR cycles. The ADC 200 includes switches 208-209, a capacitive array 202, comparators 103_1-4, 104, a clock generator 204, and a shift register 106. The capacitive array 202a-g implements the adder 102 and the DAC 108. The capacitive array 202 is a charge-redistribution capacitive array and includes capacitive units 202a-g. The capacitive units 202a-g have the same structure but different capacitances. Take the capacitive unit 202a for example, the capacitive unit 202a includes capacitors 212a, 213a and switches 214a-219a. The capacitors 212a and 213a have the same capacitance. The bottom-plate of the capacitor 212a is coupled to one terminal of the switch 208 and one input terminal of the comparators 103_1-4 and 104, and the bottom-plate of the capacitor 213a is coupled to one terminal of the switch 209 and the other input terminal of the comparators 103_1-4 and 104. The other terminals of the switches 208 and 209 are coupled to the ground. In addition, the top-plate of the capacitor 212a (213a) is coupled to the input signal V_in via the switch 216a (217a), a positive reference signal V_R via the switch 214a (215a), and a negative reference signal −V_R via the switch 218a (219a).

The capacitances of the capacitors C₁-C₇ of the capacitive array 202a-g are scaled in a non-binary way and the sum of the capacitances C₁-C₇ is the total capacitance of the DAC. The capacitances can be based on Metal-Oxide-Metal (MOM) structures.

The capacitive array 202 uses a bottom-plate sampling scheme. The bottom-plates of the capacitors 212a through 212g are coupled together and are coupled to the ground via the switch 208. The bottom-plates of the capacitors 213a through 213g are coupled together and are coupled to the ground via the switch 209. The switches 208-209 are controlled according to the clock signal T. In the illustrated example, the signal start_conv that is the inverse of the clock signal T is provided to the comparator 103a to trigger the comparator 103a to make a comparison. In the tracking phase, the ADC 200 tracks the analog signal, and in the conversion phase, the ADC 200 samples and converts the analog signal.

In the tracking phase (i.e. the signal start_conv is low and the clock signal T is high), the switches 208-209 are on. The switches 216a-g are on to provide the analog signal V_in to the top-plates of the capacitors 212a-g. The switches 217a-g are on to provide the analog signal −V_in to the top-plates of the capacitors 213a-g. The switches 214a-g, 215a-g coupling the top plates of the capacitors 212a-g, 213a-g to the positive reference voltage V_R and the switches 218a-g, and 219a-g coupling the top plates of the capacitors 212a-g, 213a-g to the negative reference voltage −V_R are off.

At the end of the tracking phase (i.e. the signal start_conv transitions from low to high and the clock signal T transitions from high to low), the switches 208-209, the switches 216a-g, and the switches 217a-g are switched off to sample the analog signal V_in. The switches 214a-g, 215a-g 218a-g, and 219a-g are operated in such a way that the capacitors 212a-g, 213a-g are connected in a balanced way to the reference voltages V_R and −V_R. The difference between the signals provided to the terminals of the comparators 103a through 104 is the opposite of the analog value to be converted.

During the conversion phase (i.e. signal start_conv high), the capacitors 212a-g and 213a-g are progressively connected to the reference voltage V_R or −V_R according to the comparators decisions that have been determined. The remaining analog value to be converted (i.e., the difference between the input signal and the DAC output signal) is provided to each comparator. The DAC 200 applies the SAR algorithm to approximate the digital conversion to the analog signal by reducing this difference voltage to substantially zero. The error between the analog value and the converted digital value is the quantization error.

FIG. 2B is a block diagram illustrating an example ADC 250 that outputs a 7-bit digital signal using 8 conversion cycles. The example ADC 250 includes switches 208-209, a capacitive array 252, comparators 103a-d, 104, a clock generator 204, and a shift register 106. The capacitive array 252a-g implements the adder 102 and the DAC 108.

Compared to the ADC 200 illustrated in FIG. 2A, the example ADC 250 uses a top-plate sampling scheme. The switch 208 (209) couples the bottom-plates of the capacitors 212a-g (213a-g) to the input analog signal Vin (negative analog signal −Vin) rather than the ground. The top-plates of the capacitors 212a-g, 213a-g are no longer coupled to the input analog signals Vin and −Vin. In the tracking phase, the input analog signals are provided to the bottom-plates of the capacitors 212a-g (213a-g), rather than the top-plates of the capacitors 212a-g (213a-g) as illustrated in FIG. 2A. At the same time the capacitors 212a-g and 213a-g are connected in a balanced way to V_R and −V_R. At the end of the tracking phase, the switches 208-209 are switched off to sample the input signal. During the SAR conversion phase, the capacitors 212a-g, 213a-g are progressively connected to V_R or −V_R according to the decisions of the comparators that have been determined.

Compared to the ADC 200, the input analog signals are substantially not attenuated by parasitic capacitance of the comparators that is between the input of the comparators and the ground when being converted by the ADC 250. As a result, the ADC 250 has an improved power consumption and area occupation compared to the ADC 200. The parasitic capacitance is the sum of the parasitic capacitance between the bottom-plates of the capacitors 212a-g, 213a-g and the input parasitic capacitance of the comparators. The attenuation of the reference voltages can be resolved by increasing the reference voltage value.

Calibration of Offset Voltages

Mismatches in offset voltages of the comparators (hereinafter also referred to as “offset mismatches”) may cause dead zones in ADCs and subject ADCs to conversion errors which compromise the overall ADC performance. For example, offset mismatches may cause an increase in distortion components and noise floor. As further described below, the offset voltages of the comparators are calibrated to substantially minimize the quantization error.

FIG. 3A is a block diagram illustrating an example partial loop-unrolling SAR ADC 300 with comparator offset calibration, according one embodiment. The example ADC 300 includes a sampling switch 101, an adder 102, a series of comparators 103_1-m, a comparator 104, a shift register 106, a digital-to-analog converter (DAC) 108, a digital register 110, a calibration module 302, and compensators 3041-m. A compensator 304 includes a voltage source, a current source, or a programmable resistance (not shown) to a compensation offset to a comparator 103. The calibration module 302 detects a voltage offset for each comparator 103_1-m. The compensators 304_1-m provide voltage offset compensation to the comparators 103a-m according to the determined voltage offset. The description of a sampling switch 101, an adder 102, a series of comparators 103_1-m, a comparator 104, a shift register 106, a digital-to-analog converter (DAC) 108, and a digital register 110 is provided in connection with FIG. 1.

The calibration module 302 determines voltage offsets that result in dead zones and redundant comparisons. To determine the voltage offset for a comparator 103i, the calibration module 302 adjusts the offset voltage provided to the comparator 103i until the outputs of the comparators 103i-103m and 104 do not include predetermined patterns that correspond to dead zones. The calibration module 302 records the amount of voltage offset provided to the comparator. This amount is the amount of the voltage offset to be compensated. In various embodiments, the calibration module 302 determines the amount of voltage offset that results in dead zones before determining the residual amount of voltage offset that can be measured by analyzing SAR redundant comparisons. The total amount of offset voltage to be compensated is the sum of both amounts.

To determine the amount of offset voltage that causes a dead zone in the ADC 300, the calibration module 302 injects an offset voltage to a comparator 103i. This amount of offset voltage V_osi exceeds the redundancy range (−ε_toli, ε_toli) that is available at a conversion step i. The redundancy range can be determined according to Equation (10). As a result, a dead zone appears in the input-output characteristic of the ADC 300. The calibration module 302 sets the offset voltage for all comparators 103i-104 to zero. The calibration module 302 adjusts the offset voltage until the comparators' outputs do not include dead zone patterns. For example, the calibration module 302 injects an offset voltage to a comparator 103i, and observes the output of the comparators 103k-104. A dead zone pattern appears when the subsequent comparators 103k-104 have outputs with opposite sign with respect to the output of the comparator 103i as shown in Equations (6)-(7).b_i=+1 and b_k=−1 for ∀k>i  (6)b_i=−1 and b_k=+1 for ∀k>i  (7).

If a dead zone pattern as indicated in Equations (6)-(7) appears, the calibration module 302 determines that the amount of offset voltage to be compensated for the comparator 103i has the same sign as the output of the comparator 103i. A small increase/decrease in the compensation offset is applied to the comparator 103i. At the steady state, this amount of offset voltage compensation prevents the dead zone of the comparator 103i. The calibration module 302 determines the amounts of offset voltage compensation to remove dead zones for all comparators 103_1-m. If all the comparators' 103_1-m dead zone is compensated, the offsets of the comparators 103_1-m do not contribute to conversion errors of the ADC 300.

For each comparator 103i, the calibration module 302 determines whether there is any residual voltage offset mismatch between the comparator 103i and 103j. The residual offset can be measured by identifying redundant comparisons in subsequent comparators. The calibration module 302 may apply a full-scale signal to the input of the ADC 300. For each comparator 103i, the calibration module 302 identifies whether there is any subsequent comparator 103j that has the opposite output with respect to the comparator 103i and whether the weighted sum of all the intermediate comparators' outputs is opposite to the weighted output of the comparator 103i as shown in Equations (8)-(9):

$\begin{array}{cc}\mathrm{if}{b}_i=+1\mathrm{and}{b}_j=-1\mathrm{and}\sum _{k=i+1}^{j-1}{b}_k{w}_k=-b_i{w}_i,& \left(8\right)\\ \mathrm{if}{b}_i=-1\mathrm{and}{b}_j=+1\mathrm{and}\sum _{k=i+1}^{j-1}{b}_k{w}_k=-b_i{w}_i.& \left(9\right)\end{array}$

If the comparator 103j exists, the calibration module 302 adjusts the offset voltage provided to the comparator 103i until the above two sequences according to equations (8)-(9) have the same occurrence rate. For example, for every comparator 103i, the calibration module 302 decreases (or increases) the offset voltage if the occurrence rate of the outputs of the comparators as shown in equation (8) is larger than (or less than) the occurrence rate the outputs of the comparators as shown in equation (9). The calibration module 302 determines this amount of offset voltage is to be compensated for the comparator. This amount of offset voltage is the offset mismatch between the comparators 103i and 103j. This is because if the difference (V_in−V_dac,i) is small during a comparison i, the ADC 300 converts the same difference (V_in−V_dac,i) in one of the subsequent conversion step j. For every redundant cycle i, the sum of DAC weights w_n (n>i) added or subtracted in the following cycles is greater than w_i and the difference between these two terms defines the available redundancy range ε_toli as shown in Equation (10):

$\begin{array}{cc}\sum _{n=i+1}^{N_C}{w}_n>w_i\mathrm{and}{\varepsilon }_{\mathrm{toli}}=\sum _{n=i+1}^{N_C}{w}_n-w_i.& \left(10\right)\end{array}$ In the redundant comparison case, if the comparator 103i is affected by an offset error V_osi that is different from the one which affects the comparator 103j V_osj, the two decisions b_i and b_j are different. The calibration module 302 makes the determination for all comparators 103_1-m. If the offset voltage mismatches between the comparators' 103_1-m is compensated, the offsets of the comparators 103_1-m do not subject the comparators to other errors. As such, the ADC's redundancy range can absorb errors (e.g. noise, reference voltage ripple, etc.) from other sources.

The calibration can be performed in the foreground or in the background. During foreground calibrations, the calibration module 302 applies a dedicated calibration signal to the input of the ADC. For example, an arbitrary analog signal that may improve convergence speed of the calibration is applied to the input of the ADC. The foreground calibration may be performed during the startup of the ADC. During background calibrations, the calibration module 302 uses signals applied to the ADC. The calibration module 302 may apply foreground calibration to track the comparators' offsets drifting during time slots when no signal is applied to the ADC in the application. The calibration can be performed without using any dedicated clock cycles or resetting any comparator at the end of the conversion and thus does not compromise the conversion speed and power consumption of the ADC.

Upon reading this disclosure, those of skill in the art will appreciate still additional alternative designs. Thus, while particular embodiments and applications of the present disclosure have been illustrated and described, it is to be understood that the embodiments are not limited to the precise construction and components disclosed herein and that various modifications, changes and variations which will be apparent to those skilled in the art may be made in the arrangement, operation and details of the method and apparatus of the present disclosure disclosed herein without departing from the spirit and scope of the disclosure as defined in the appended claims.

Claims

1. An analog to digital converter to convert analog signals into digital symbols comprising: an input terminal to receive an input analog signal;a plurality of comparators coupled in series and together configured to determine a plurality of discrete values, a comparator's decision based in part upon a previous comparator's decision;a digital register configured to output a digital symbol by combining the discrete values determined by the comparators; anda calibration module configured to determine an offset amount for each of the comparators, the offset voltage including a first amount to prevent a deadzone in the analog to digital converter and a second amount to prevent redundant comparisons in the comparators.

2. The analog to digital converter of claim 1, wherein the analog to digital converter further comprises a plurality of compensators, each of the plurality of compensators coupled to a corresponding comparator and configured to inject the offset amount to an input of the corresponding comparator.

3. The analog to digital converter of claim 2, wherein the compensators comprise at least one of a voltage source, a current source, and a programmable resistance.

4. The analog to digital converter of claim 1, wherein the calibration module is configured to determine a plurality of the first amounts for the plurality of comparators before determining the plurality of the second amounts for the plurality of comparators.

5. The analog to digital converter of claim 1, wherein the calibration module is configured to determine the first amount for a first comparator before determining the first amount for a second comparator, wherein the first comparator's output is coupled to the second comparator.

6. The analog to digital converter of claim 1, wherein the calibration module is configured to determine the first amount for a comparator by: adjusting an amount of the offset provided to an input of the comparator while monitoring the comparator's output and all subsequent comparators' output;responsive to determining that all subsequent comparator's output is opposite to the comparator's output, recording the amount of the offset.

7. The analog to digital converter of claim 1, wherein the calibration module is configured to determine that an amplitude of the first amount is the amount of the offset recorded and a sign of the first amount is the comparator's output.

8. The analog to digital converter of claim 1, wherein each comparator corresponds to a weight and wherein the calibration module is configured to determine the second amount for a first comparator by: applying an arbitrary signal to the input terminal; andidentifying a second comparator from subsequent comparators, the second comparator and the first comparator having opposite outputs and a weighted sum of one or more comparators between the first comparator and the second comparator opposite to a weighted output of the first comparator.

9. The analog to digital converter of claim 8, wherein the calibration module is further configured to determine the second amount by: adjusting an amount of the offset provided to an input of the comparator while monitoring a first occurrence rate of a first occurrence rate of the second comparator's output being negative and the first comparator's output being positive and the weighted outputs of the all comparators between the first comparator and the second comparator opposite to the weighted output of the first comparator and a second occurrence rate of the second comparator's output being positive and the first comparator's output being negative and the weighted outputs of the all comparators between the first comparator and the second comparator opposite to the weighted output of the first comparator;responsive to determining that the first occurrence rate equaling to the second occurrence rate, recording the amount of the offset.

10. The analog to digital converter of claim 9, wherein the calibration module is configured to determine the second amount is the amount of the offset recorded.

11. The analog to digital converter of claim 1, wherein the plurality of comparators include multiple first comparators and a second comparator, each first comparator corresponds to a non-binary scaled discrete value and is configured to determine the non-binary scaled discrete value, the second comparator corresponds to binary scaled discrete values and is configured to determine the binary scaled discrete values.

12. The analog to digital converter of claim 1, wherein the plurality of comparators are configured to determine the discrete values in sequence and each of the comparators determines the corresponding discrete value by comparing a sampled value of the analog signal to an existing digital representation of the sampled value.

13. The analog to digital converter of claim 1, wherein the calibration module is configured to apply an arbitrary signal to the input terminal and wherein the calibration module is configured to determine the offset amount using the arbitrary signal.

14. The analog to digital converter of claim 1, wherein the calibration module is configured to determine the offset amount concurrently while the analog to digital converter converts the input analog signal.