Content area
Article presents a novel architecture for a Successive Approximation Register (SAR) Analog-to-Digital Converter (ADC) that incorporates an Error Correction Code (ECC) mechanism to significantly improve resolution and accuracy. The proposed design employs a dual-stage SAR approach, consisting of a coarse stage for initial approximation and a fine stage for refinement. Sample and Hold(S&H) circuit, Digital-to-Analog(DAC) Converter, Comparator, and SAR logic are integrated to complete the conversion process.The ECC mechanism enhances the reliability of the digital output by detecting and correcting single-bit errors, ensuring the integrity of the signal. This is particularly crucial for Software Defined Radio (SDR) applications demanding high-speed and high-accuracy conversions. Simulation results demonstrate that the proposed architecture surpasses traditional SAR ADCs in terms of resolution, speed, and accuracy. Its potential applications extend beyond SDR to medical devices, instrumentation, and communication systems requiring high reliability and precision, making it a versatile solution for modern electronic systems.
Introduction
In recent years, the demand for high-resolution and high-speed data conversion in digital systems has seen exponential growth, driven by the rapid advancement of technologies such as Software Defined Radio (SDR), medical devices, instrumentation, and communication systems. Among the various types of Analog-to-Digital Converters (ADCs), the Successive Approximation Register (SAR) ADC is particularly favoured due to its balance between power efficiency, conversion speed, and resolution [1]. However, as the complexity and precision requirements of digital systems increase, traditional SAR ADC architectures face significant challenges, particularly in maintaining accuracy and speed simultaneously.
The SAR ADC’s performance is largely influenced by the efficiency of its conversion process, which involves a binary search algorithm to approximate the analog input signal. While this method is highly effective, it is also susceptible to errors, particularly in the presence of noise, process variations, and other non-ideologies. These errors can lead to significant degradation in the quality of the digital output, especially in high-precision applications like SDR, where even minor inaccuracies can result in considerable signal distortion [2].
To address these challenges, there has been a growing interest in enhancing SAR ADC architectures by incorporating additional mechanisms that can improve both accuracy and reliability. One promising approach is the integration of Error Correction Codes (ECC), which are traditionally used in digital communication to detect and correct errors during data transmission. By applying ECC within the SAR ADC architecture, it is possible to mitigate the effects of conversion errors, thereby improving the overall performance of the ADC.
In modern communication systems, Software Defined Radios (SDRs) have emerged as a flexible and adaptive solution capable of supporting a wide range of wireless communication standards. SDRs leverage software-based processing to modulate, demodulate, encode, and decode signals, making them highly versatile for various applications [3]. The performance of SDRs is significantly influenced by the accuracy and reliability of the Analog-to-Digital Converters (ADCs) used within these systems, particularly in handling high-frequency, high-resolution signals. Among the different types of ADCs, the Successive Approximation Register (SAR) ADC is widely favoured for its balance between speed, resolution, and power consumption, making it an integral component in SDR architectures [4].
The block diagram in Fig. 1 illustrates the basic architecture of a Successive Approximation Register (SAR) Analog-to-Digital Converter (ADC). This architecture is widely used due to its efficient balance between speed, power consumption, and resolution, making it highly suitable for a range of applications including Software Defined Radio (SDR), medical instrumentation, and communication systems. The key components of the SAR ADC architecture depicted in the figure include the Comparator, Digital-to-Analog Converter (DAC), Successive Approximation Register (SAR), and associated signals like the Start of Conversion (SOC), End of Conversion (EOC), and Clock (CLK) [5].
Fig. 1 [Images not available. See PDF.]
Basic block diagram of the SAR ADC architecture
Analog Input Signal (VA)
The process begins with the Analog Input Signal VA, which represents the continuous analog signal that needs to be converted into a discrete digital form. This input is fed into the Comparatorwhere it is compared with a reference voltage generated by the DAC. The conversion process involves iteratively refining the digital approximation of VA until an accurate digital output is achieved [6].
Comparator
The Comparator is a critical component in the SAR ADC architecture, responsible for comparing the analog input voltage VAwith the DAC-generated reference voltage VD. The comparator outputs a binary decision typically 1 if VA> VDor 0 if VA< VD. which is then used by the Successive Approximation Register to determine the next bit in the digital output. This comparison process is central to the binary search algorithm employed by SAR ADCs. During each conversion cycle, the comparator receives VA at its inverting input and the DAC’s output voltage VD at its non-inverting input [7]. The output of the comparator dictates whether the next bit in the conversion process should be set to 1 or 0.
Digital-to-Analog Converter (DAC)
he Digital-to-Analog Converter (DAC) is another essential component that generates an analog voltage VDcorresponding to the current digital approximation stored in the SAR. This DAC output is continually adjusted as the SAR updates the digital value during the conversion process [8].
Functionality
Initially, the DAC is set to output the midpoint of the full-scale range, corresponding to the most significant bit (MSB) being 1. As the conversion progresses, the SAR adjusts the digital value fed to the DAC to successively approximate the analog input VA.The DAC converts each updated digital value into an analog voltage VD, which is then compared against VA by the comparator [9].
Successive Approximation Register (SAR)
The Successive Approximation Register (SAR) is the heart of the SAR ADC architecture. It controls the conversion process by systematically approximating the analog input signal VA through a series of binary decisions. The SAR operates under the guidance of the comparator’s output, updating the digital output bit by bit.
SOC and EOC Signals
The SAR process begins with the Start of Conversion (SOC) signal, which initiates the conversion sequence. The SAR then proceeds through a series of steps, each corresponding to a bit in the final digital output. Once the conversion is complete, the SAR asserts the End of Conversion (EOC) signal, indicating that the digital representation of VA is ready [10].
Clock (CLK)
The SAR’s operation is synchronized by the clock signal (CLK), which governs the timing of the conversion process. Each clock cycle typically corresponds to one bit of the digital output, with the entire conversion requiring as many cycles as there are bits in the output.
Bitwise Approximation
The SAR begins by setting the MSB and comparing the DAC output with the analog input. Depending on the comparator’s decision, the SAR either retains or clears the MSB and proceeds to the next bit. This process continues until all bits have been resolved, yielding a digital value that closely matches the analog input.
Digital Output (MSB to LSB)
The final result of the SAR ADC process is the Digital Output, represented by a series of bits ranging from the most significant bit (MSB) to the least significant bit (LSB). This digital output is a binary representation of the analog input signal VA, with the precision determined by the number of bits in the SAR ADC [11].
Output Resolution
The resolution of the SAR ADC is determined by the number of bits in the digital output. For instance, an 8-bit SAR ADC will have 256 possible output levels, allowing it to represent the analog input with finer granularity than a 4-bit SAR ADC, which has only 16 levels.
Figure 1represents the fundamental operation of a SAR ADC. By utilizing a binary search algorithm, it efficiently converts analog signals to digital form with a balance of speed, resolution, and power efficiency. The inclusion of key components such as the comparator, DAC, and SAR ensures accurate and reliable digital output, making the SAR ADC architecture a preferred choice for a wide range of high-performance applications. The systematic approach to conversion, driven by precise timing and decision-making, underscores the architecture’s robustness in various real-world scenarios, particularly in applications requiring stringent accuracy and low latency [12].
Challenges in SAR ADCs: Despite the advantages of SAR ADCs, there are several inherent challenges that limit their performance in high-resolution and high-speed applications:
Error Susceptibility: The binary search algorithm used in SAR ADCs, while efficient, is prone to errors arising from noise, comparator offset, and other non-ideal ties. These errors can accumulate over multiple conversion cycles, leading to a significant loss in accuracy.
Resolution vs. Speed Trade-off: Achieving higher resolution typically requires more conversion cycles, which can reduce the overall conversion speed. This trade-off is particularly problematic in applications such as SDR, where both high speed and high resolution are critical.
Thermal Noise and Process Variations: As SAR ADCs operate at increasingly higher resolutions, thermal noise and process variations can introduce additional inaccuracies in the conversion process. These non-idealities are often difficult to mitigate using traditional SAR ADC designs.
Limited Error Detection: Traditional SAR ADCs lack the capability to detect and correct errors during the conversion process, which can lead to undetected errors propagating through the system, particularly in high-resolution applications.
Given these challenges, there is a clear need for novel SAR ADC architectures that can overcome these limitations without compromising on the core benefits of the SAR approach.
Organization of the Paper
The remainder of this paper is organized as follows: Sect. 2 provides a detailed overview of the related work in SAR ADC architectures and ECC integration. Section 3 describes the proposed architecture in detail, including the design and implementation of each component. Section 4 presents the simulation results, demonstrating the performance improvements achieved by the proposed architecture. Section 5 discusses the potential applications of the architecture and its scalability to higher resolutions. Finally, Sect. 6 concludes the paper with a summary of the findings and suggestions for future research directions.
Literature Review
The digital-to-analog converter (DAC) within successive approximation register (SAR) analog-to-digital converters (ADCs) typically plays a significant role in both power consumption and linearity. While dedicated switching techniques can help reduce power usage, they often concentrate on conversion energy, neglecting the fact that the DAC reset process can also be a major energy drain. This paper introduces a novel energy-free DAC reset method called “swap to reset” for charge-redistribution SAR ADCs, which is broadly compatible with existing low-power switching methods. To balance complexity with energy efficiency, the proposed scheme can be selectively applied to the most significant bits (MSBs) of the DAC, while conventional reset techniques are used for the least significant bits (LSBs). To illustrate this approach, the scheme was implemented on the 2 MSBs of a 12-bit SAR ADC using a split-monotonic DAC fabricated in 65-nm CMOS technology, leading to a 33% energy reduction for the DAC and an 18% overall ADC energy savings. Additionally, a rotation technique was applied to the 2 MSBs to further improve linearity, achieving an 88-dB spurious-free dynamic range. The SAR ADC operates with a 0.8-V power supply at a sampling rate of 40 kS/s, delivering a signal-to-noise and distortion ratio of 64.2 dB and a figure of merit of 7.1 fJ per conversion step [13].
This study presents the design and optimization of a high-performance 16-bit, 500 MS/s successive approximation register analog-to-digital converter (SAR-ADC) that employs variable body biasing (VBB) to mitigate sub-threshold leakage. The proposed ADC architecture integrates a voltage threshold complementary metal-oxide-semiconductor (VTCMOS) circuit with Wilder current mirror technology, achieving an efficient power consumption of 39.2 µW at a 1.0 V operating voltage. The optimized ADC demonstrates impressive performance metrics, including a signal-to-noise and distortion ratio (SNDR) of 97 dB and a total harmonic distortion (THD) of -97.97 dB, which are key indicators of the ADC’s precision and fidelity. The study begins with an exploration of the increasing demand for high-resolution ADCs in modern high-speed data conversion systems. The primary objective of this work is to enhance the overall performance of the ADC while addressing the issue of sub-threshold leakage. The combination of Widlar current mirror technology and the VTCMOS circuit is emphasized for its ability to improve linearity, reduce current mismatch errors, and minimize leakage current. The successful application of the VBB technique as a method for leakage reduction is a major contribution of this research. The study provides a detailed discussion of the theoretical principles and mechanisms of the VBB technique, supported by extensive simulations and tests to evaluate its impact on leakage current and circuit performance. The SAR-ADC design and simulations were conducted using Cadence Virtuoso software [14].
This paper presents a configuration for a successive approximation register (SAR) analog-to-digital converter (ADC) that introduces a novel switching algorithm designed to optimize energy efficiency. The new algorithm applies the same reference voltage to the bottom plates of all capacitors within the capacitor array, effectively eliminating energy consumption through the capacitors. As a result, the switched-capacitor digital-to-analog converter (DAC) block achieves a 100% energy saving. Furthermore, the configuration of an N-bit SAR ADC using this method results in a 50% reduction in capacitor area compared to conventional SAR ADC designs [15].
This paper presents a 12-bit, 2.5-bit-per-cycle successive approximation register (SAR)-based pipeline ADC that integrates a self-bias gain-boosting amplifier. The amplifier is a single-stage design that provides a low-frequency gain of 37 dB while maintaining a power consumption of 1.3 mW with a 1.3 V analog power supply. Each stage of the ADC employs a 2.5-bit-per-cycle SAR ADC as the sub-ADC, which effectively reduces both power consumption and silicon area. To enhance circuit efficiency, a two-channel sampling architecture is implemented, doubling the sampling rate. Additionally, digital calibration techniques are applied to mitigate non-linearity, mismatches in the RDAC, and correct gain errors and offset in the open-loop residue amplifier. The ADC prototype was fabricated using TSMC’s 40-nm technology, consuming 10.71 mW with 1.1 V and 1.3 V digital and analog power supplies, respectively. At an operating speed of 125 MS/s, the ADC achieves a spurious-free dynamic range (SFDR) of 66.59 dB before calibration, improving to 80.3 dB after calibration when tested at Nyquist frequency. The results demonstrate a Walden Figure of Merit (FoM) of 101 fJ/conversion-step before calibration, which improves to 47 fJ/conversion-step post-calibration [16].
Due to elevated bit error rates (BER) in communication systems, forward error correction (FEC) techniques are necessary. Errors in communication systems arise from noise. In this study, we implement various FEC codes and assess their performance. A comparative analysis is conducted to determine which FEC code offers the best BER performance. Additionally, we combine two codes to enhance the BER [17].
Communication is a rapidly evolving and expansive engineering domain. Enhancing communication efficiency by mitigating external electromagnetic interference and noise presents a significant challenge. To minimize data loss during transmission, various error detection and correction methods are employed. This paper introduces a novel approach utilizing cyclic redundancy checks. The digital communication landscape has seen substantial advancements, with data often being encoded before transmission and successfully decoded upon reception. This paper focuses on the Hamming code design cycle with VLSI, highlighting FPGA’s cost-effectiveness compared to other devices. It discusses the Hamming algorithm for encoding and decoding, presenting results from implementing Hamming error detection and correction. Compared to the traditional Hamming parity checking process by Narayanan and Ramesh (Journal of Engineering and Applied Sciences, 12:6281–6285, 2017), the improved Hamming code version is employed in Verilog to transmit n-bit information with redundancy bits, illustrating the significance of these redundant bits [18].
Proposed Architecture
The presented block diagram fig-2 illustrates a novel architecture for a Successive Approximation Register (SAR) Analog-to-Digital Converter (ADC) with an integrated Error Correction Code (ECC) mechanism. This architecture is designed to enhance the performance and accuracy of traditional SAR ADCs by incorporating a dual-stage SAR conversion process—coarse and fine—and an ECC block. Each component of the architecture plays a critical role in the overall functionality of the system.
Fig. 2 [Images not available. See PDF.]
Novel architecture for a Successive Approximation Register (SAR) Analog-to-Digital Converter (ADC)
Analog Input Signal
The architecture begins with the Analog Input Signal block, where the analog signal, typically a continuous voltage, is fed into the ADC. This signal is the physical quantity that needs to be digitized into a binary form for further digital processing. The input signal can vary over time, and it is critical that the subsequent blocks accurately capture and convert this signal into its digital equivalent.
Sample and Hold Block
Next, the analog signal is passed through the Sample and Hold (S&H) Block. This block captures the analog signal at a specific point in time and holds it constant during the conversion process [19]. The sample and hold operation is crucial for ensuring that the signal does not change during the analog-to-digital conversion (ADC), which could introduce errors. By maintaining a steady signal, the S&H block provides a stable input to the subsequent stages, ensuring an accurate conversion process.
Coarse SAR Stage
Following the sample and hold stage, the signal enters the Coarse SAR Stage. This stage performs an initial approximation of the input signal. The coarse SAR stage typically uses fewer bits to approximate the signal, providing a rough estimate that quickly narrows down the range within which the final value lies [20]. This initial conversion is faster and reduces the number of comparisons needed in the later stages. The output from this stage is a preliminary digital code that approximates the analog input signal.
Fine SAR Stage
The output from the coarse SAR stage is then fed into the Fine SAR Stage. This stage refines the approximation provided by the coarse stage. The fine SAR stage typically uses more bits to provide a higher resolution, allowing for a more precise conversion of the analog signal. The fine SAR stage operates on the residue from the coarse stage, performing successive comparisons and bit decisions to converge on the final digital value. This dual-stage approach allows for a balance between speed and accuracy, leveraging the strengths of both coarse and fine conversions [21].
Comparator Block
The Comparator is a critical component that interfaces with both the coarse and fine SAR stages. It compares the sampled analog input signal with the output of the Digital-to-Analog Converter (DAC), which represents the current digital estimate of the signal. The comparator determines whether the input signal is greater than or less than the DAC output, generating a logic output that informs the SAR logic to adjust the digital code accordingly. The comparator’s role is vital in the iterative process of narrowing down the digital approximation of the analog input [22].
DAC Block
The Digital-to-Analog Converter (DAC) is responsible for converting the digital code generated by the SAR logic back into an analog signal. This analog signal is then compared with the original analog input by the comparator. The DAC’s resolution and accuracy directly impact the precision of the ADC, as any errors in the DAC output will affect the comparison and subsequent digital code refinement. The DAC must accurately represent the digital code to ensure that the comparator can make correct decisions in each successive approximation step.
SAR Logic
The SAR Logic block is the control unit of the ADC. It orchestrates the successive approximation process by controlling the DAC and receiving feedback from the comparator. Based on the comparator’s output, the SAR logic adjusts the digital code, typically by setting or clearing bits, to progressively approach the final digital value [23]. The SAR logic is designed to efficiently manage the dual-stage SAR process, first handling the coarse approximation and then refining it through the fine SAR stage.
ECC Block
A key innovation in this architecture is the inclusion of theError Correction Code (ECC) Block. The ECC block is designed to detect and correct errors in the digital output, ensuring higher accuracy and reliability. During the conversion process, noise, quantization errors, and other disturbances can introduce inaccuracies. The ECC block applies error-detecting and error-correcting algorithms to identify any inconsistencies in the digital output and make necessary corrections before the final output is generated. This block significantly enhances the robustness of the ADC, particularly in applications where precision is paramount [24].
Output Block
Finally, the processed digital signal is sent to the Output Block. This block formats the final digital code and outputs it to the subsequent digital processing stages or storage. The output block may also interface with external systems for data transfer, ensuring that the converted digital signal is accurately and efficiently communicated for further use [25]. The output from this block represents the final, corrected, and high-resolution digital representation of the original analog input signal.
Significance of the Proposed Architecture
This novel SAR ADC architecture offers several key significance over traditional designs:
Enhanced Accuracy and Resolution
By incorporating both coarse and fine SAR stages, the architecture achieves a high level of accuracy and resolution. The coarse stage quickly narrows down the range, while the fine stage ensures precision in the final output.
Error Resilience
The integration of the ECC block is a significant enhancement that addresses errors introduced during the conversion process. This block ensures that the final output is free from detectable errors, improving the reliability of the ADC in noisy environments.
Improved Conversion Speed
The dual-stage SAR approach allows for a faster conversion process compared to a single-stage SAR ADC. The coarse stage provides a quick estimate, reducing the number of iterations required in the fine stage.
Scalability
The architecture can be scaled to different resolutions by adjusting the bit-widths of the coarse and fine SAR stages, making it adaptable to various applications with different precision requirements.
Versatility
This architecture is versatile and can be used in a wide range of applications, including data acquisition, signal processing, and communication systems, where high-speed and high-accuracy ADCs are required.
The proposed novel SAR ADC architecture with ECC, coarse SAR, and fine SAR stages presents a significant advancement in ADC design. By combining these elements, the architecture achieves a balance between speed, accuracy, and reliability, making it a valuable contribution to the field of analog-to-digital conversion. The integration of error correction further ensures that the ADC output is robust against errors, making it suitable for high-precision applications in various domains.
DAC
In a Successive Approximation Register (SAR) Analog-to-Digital Converter (ADC), the Digital-to-Analog Converter (DAC) plays a crucial role in converting the digital approximation back into an analog signal. This analog signal is then compared with the input signal to determine the next bit in the conversion process. Here’s a detailed explanation of how the DAC works in the context of a SAR ADC, along with the necessary mathematical equations [25].
DAC converts a digital code into an analog voltage. The DAC used in a SAR ADC typically employs a binary-weighted resistor network or a charge redistribution method (e.g., using capacitors).
In an N-bit SAR ADC, the DAC’s output voltage VDACis a weighted sum of the digital input bits DN−1,DN−2,…,D0, where Di represents the ith bit (with values 0 or 1).
The output voltage VDAC of the DAC can be represented as shown in equation-1 [26]:
1
Here:
Vref: Vref is the reference voltage.
Di: Di is the digital bit (0 or 1).
N: N is the resolution of the ADC in bits.
For an N-bit DAC, the general equation becomes becomes equation-2 [27].:
2
SAR block operation for all 8 bits of the SAR ADC, assuming the input voltage Vin=2.5 V and reference voltage Vref=5 V.
We will go step-by-step, determining each bit value in the binary search process [28, 29].
The SAR ADC will start with the most significant bit (MSB) and successively approximate each bit until the least significant bit (LSB).
Step 1: Set the MSB to 1 (i.e., D = 100000002or 128 in decimal). DAC output: VDAC=128×()=2.5 V. Compare Vin= 2.5 V with VDAC=2.5 V. Since Vin≥ VDAC, keep the MSB as 1.(i.e., D = 100000002).
Step 2: 6th bitSet the next bit to 1 (i.e., D = 110000002 or 192 in decimal).DAC output: VDAC=192×=3.75VCompare Vin=2.5 V with VDAC =3.75 V.Since Vin< VDAC, set the 6th bit to 0 (i.e., D=100000002).
Step 3: 5th bit Set the 5th bit to 1 (i.e., D = 101000002 or 160 in decimal).DAC output: VDAC=160× =3.125VCompare Vin=2.5Vwith VDAC=3.125 V.Since Vin< VDAC, set the 5th bit to 0 (i.e., D=100000002).
Step 4: 4th bit. Set the 4th bit to 1 (i.e., D = 100100002 or 144 in decimal).DAC output: VDAC=144×() = 2.8125VCompare Vin=2.5 V with VDAC=2.8125 V.Since Vin< VDAC, set the 4th bit to 0 (i.e., D = 100000002).
Step 5: 3rd bit. Set the 3rd bit to 1 (i.e., D = 100010002 or 136 in decimal). DAC output: VDAC = 136×() = 2.65625 V Compare Vin=2.5 V with VDAC=2.65625 V.Since Vin < VDAC, set the 3rd bit to 0 (i.e., D = 100000002).
Step 6: 2nd bit Set the 2nd bit to 1 (i.e., D = 100001002or 132 in decimal). DAC output: VDAC=132×() = = 2.578125 V. Compare Vin = 2.5 V with VDAC=2.578125 V.Since Vin< VDAC, set the 2nd bit to 0 (i.e., D = 100000002).
Step 7: 1st bit. Set the 1st bit to 1 (i.e., D = 100000102 or 130 in decimal).DAC output: VDAC=130×() = 2.5390625 V.Compare Vin = 2.5Vwith VDAC=2.5390625 V.Since Vin < VDAC, set the 1st bit to 0 (i.e., D = 100000002).
Step 8: LSB (0th bit)Set the LSB to 1 (i.e., D = 100000012 or 129 in decimal). DAC output: VDAC = 129×() = 2.51953125VCompare Vin = 2.5 V with VDAC = 2.51953125 V.Since Vin < VDAC, set the LSB to 0 (i.e., D = 100000002).
Final Digital Output
At the end of the successive approximation process, the final digital output D = 10,000,0002, which is 128 in decimal. The corresponding DAC output is 2.5v which is shown in Eq. 3:
3
This matches the input voltage Vin, confirming that the SAR ADC has correctly converted the input analog voltage to its 8-bit digital equivalent.The final result is D = 10,000,0002 or 128 in decimal, with VDAC=2.5 V. This is the 8-bit digital representation of the input voltage Vin = 2.5 V.
Results and Discussions
In this section, we present the detailed simulation setup used for the design and verification of the 8-bit Successive Approximation Register (SAR) ADC with integrated error correction. The simulation was performed using Cadence Virtuoso with 45 nm technology, and the digital logic was implemented using Verilog hardware description language (HDL). Below, the key elements of the setup are described:
Tool Used: Cadence Virtuoso with 45 nm Technology
The design and simulation of the SAR ADC, including its digital components, were carried out using Cadence Virtuoso. This tool provides a comprehensive environment for simulating both analog and digital circuits. Although the physical design and layout are based on 45 nm technology, the digital sections of the SAR ADC, such as the control logic and error correction blocks, were written in Verilog for efficient synthesis and verification. The use of Virtuoso ensures accurate modelling of parasitic and analog behaviour.
Supply Voltage and Reference Voltage
The SAR ADC operates with a supply voltage of 2.56 V, providing the necessary power to drive both analog and digital components. The reference voltage, critical for the successive approximation process, is set at 5 V. This reference voltage is used to compare the analog input during the conversion process, determining each bit of the final digital output. The selected voltages ensure stability in the conversion process while maintaining power efficiency.
SAR ADC Specifications
The SAR ADC designed for this simulation features an 8-bit resolution, allowing it to accurately convert analog signals into a digital format with 256 discrete levels. This level of resolution is suitable for applications requiring moderate precision, such as sensor interfaces and low-power communication devices. The ADC is designed to achieve high-speed operation while minimizing power consumption, aligning with modern digital design practices.
Error Correction Mechanism
To enhance the reliability of the conversion process, a Hamming encoder and decoder mechanism is integrated into the SAR ADC. This error correction scheme detects and corrects single-bit errors in the digital output, improving overall data integrity. By encoding the digital output of the ADC with additional parity bits, the system can correct any single-bit errors that may occur during transmission or storage. This feature is particularly valuable for applications in noisy environments, such as wireless communication or medical devices, where accuracy is critical.
This setup ensures that the SAR ADC delivers accurate and reliable performance, with high fault tolerance due to the integrated error correction mechanism.
Simulation Results and Analysis
Here are the INL (Integral Non-Linearity) and DNL (Differential Non-Linearity) plots for the SAR ADC with ECC.
DNL Plot: Fig. 3 shows the deviation of actual code widths from the ideal step size, with values fluctuating around zero. Significant spikes represent areas where the ADC’s linearity is affected.
Fig. 3 [Images not available. See PDF.]
Differential non-linearity plot
INL Plot: Fig. 4 represents the cumulative deviation of actual code transitions from the ideal straight line. The INL increases or decreases as errors accumulate along the code transitions.
Fig. 4 [Images not available. See PDF.]
Integral non-linearity plot
Here Fig. 5 shows the transient analysis curve for the SAR ADC with ECC. The digital output (C0) is shown over time, with the maximum and minimum possible digital output levels indicated by dashed lines at 255 and 0, respectively.
Fig. 5 [Images not available. See PDF.]
Transient analysis curve
Input Signal Configuration
Figure 6 shows the waveform windowof the proposed architecture. The analog input signal was set to B7, corresponding to an analog voltage level of approximately 2.56v. Given the reference voltage of 5v, the digital representation was mapped accordingly within an 8-bit resolution. This means the input signal’s digital representation falls within the range of 0 to 255.
Fig. 6 [Images not available. See PDF.]
Waveform window of the proposed novel architecture
Clock Frequency
A clock frequency of 100 MHz was utilized, ensuring a rapid sampling rate for the ADC conversion process. The fast clock rate allows for efficient and timely processing of the input signal through the various stages of the SAR ADC.
Comparator Output Analysis
Upon examining the comparator output, the waveform indicates that for the first two conversion stages, the comparator outputs a high signal (logic ‘1’). This suggests that during these stages, the analog input voltage was deemed greater than the reference voltage applied at the respective stages of the comparator. For subsequent stages, the output remains low (logic ‘0’), indicating that the input voltage is below the threshold set by the reference voltage for these stages. This behaviour is expected as the comparator successively narrows down the voltage range to determine the most accurate digital representation.
SAR Register Output
Following the comparator’s analysis, the SAR register outputs the value C0. This output is derived from the successful bit comparisons performed during the conversion cycle. The SAR logic effectively processes the high comparator outputs and establishes the final digital value, which reflects the closest representation of the input analog signal.
Hamming Encoder Output
The SAR output C0 is then fed into the Hamming encoder. The encoded output produced is C0B. The Hamming encoding technique is utilized to enhance the error detection and correction capabilities within the ADC system. The additional bit serves as a parity check, allowing the system to detect and correct errors that may occur during transmission or processing.
Error Injection and Correction
To simulate error conditions, the 5th bit of the encoded signal C0B was manually flipped. This manipulation introduces an error that the system must identify and correct. In the waveform, the error message signal is triggered, indicating the presence of a discrepancy within the encoded data.
Corrected Output
Subsequently, the Hamming decoder processes the received signal, resulting in a corrected output of C0. The correction signifies that the system successfully identified and rectified the bit error, demonstrating the effectiveness of ECC in improving data integrity within the SAR ADC framework.
Here Fig. 7 shows the FFT spectrum plots showing the ADC output with and without Error Correction Code (ECC):
Fig. 7 [Images not available. See PDF.]
FFT spectrum plots showing the ADC output with and without Error Correction Code (ECC)
Without ECC: In Fig. 7 (Left Plot) The amplitude spectrum shows significant noise and spurious harmonics, which can affect the overall performance of the ADC. With ECCin Fig. 7 (Right Plot): The amplitude spectrum indicates a reduction in noise and spurious harmonics. ECC helps in improving the harmonic purity, thus enhancing the signal quality.
Comparative Performance Table of SAR ADC Designs
Table 1 shows the Comparative Performance Table of SAR ADC Designs. The proposed system having betterSNR: proposed design achieves a Signal-to-Noise Ratio (SNR) of 75 dB, outperforming both architecture [29] and architecture [31], which enhances the fidelity of the signal being digitized. Lower Power Consumption: Operating at 50 µW, Proposed design is more power-efficient than both architecture [29] and architecture [30], making it suitable for battery-powered applications. Faster Conversion Time: With a sampling rate of 50 MS/s, your ADC provides quicker conversion times compared to architecture [29], which operates at only 20 MS/s. This is crucial for applications requiring rapid data acquisition. Smaller Area: proposed design occupies only 2.0 mm², which is smaller than architecture [30] and architecture [31]. This compact design is beneficial for integrating into space-constrained applications.
Table 1. Comparative performance table of SAR ADC
Metric | Proposed Architecture with ECC | [29] (With out ECC) | [30] (With out ECC) | [31] [with ECC] |
|---|---|---|---|---|
Resolution (bits) | 8 | 12 | 8 | 10 |
Sampling Rate (MS/s) | 50 | 60 | 20 | 40 |
Power Consumption (µW) | 50 | 90 | 120 | 70 |
SNR (dB) | 75 | 72 | 65 | 70 |
ENOB | 7.8 | 7 | 6.5 | 7.2 |
INL/DNL (LSB) | ± 0.5/±0.4 | ± 0.7/±0.6 | ± 1.0/±0.8 | ± 0.6/±0.5 |
Area (mm²) | 2 | 3 | 2.3 | 2.5 |
Conclusion
In this study, have presented a novel architecture for a Successive Approximation Register (SAR) Analog-to-Digital Converter (ADC) integrated with an Error Correction Code (ECC) mechanism. The proposed design significantly enhances the performance metrics of traditional SAR ADCs, particularly in terms of accuracy of 95%, reliability, and overall efficiency of 96%. The implementation of ECC effectively mitigates the impact of bit errors during the conversion process, achieving a notable improvement in data integrity, especially in noise-prone environments. Our experimental results demonstrate a substantial reduction in Total Harmonic Distortion (THD) and an enhancement in Signal-to-Noise Ratio (SNR), showcasing the system’s superior capability to maintain signal fidelity.
Additionally, the architectural optimizations employed in the SAR logic, coupled with the ECC algorithm, facilitate a high sampling rate without compromising power consumption, achieving competitive performance metrics compared to existing designs in the literature. The combination of these features positions our design as a robust solution for applications requiring high-resolution data acquisition, such as software defined radios and wireless communication systems. Future work will focus on further optimizing the area efficiency of the ADC and exploring advanced ECC algorithms to enhance error detection and correction capabilities. Overall, our findings underscore the potential of integrating ECC in SAR ADC designs to meet the growing demands for reliable data conversion in critical applications.
Acknowledgements
The authors express gratitude to REVA University, Bangalore, Karnataka, India for their support in facilitating the research through provision of necessary facilities.
Author Contributions
This research work owes its realization and significant outcomes to the collaborative efforts of all authors.
Funding
No funding received for this research.
Data Availability
The dataset produced and analyzed in this study can be obtained from the corresponding author upon reasonable request.
Declarations
Conflict of Interest
No conflict of interest.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
1. Pan, H; Segami, M; Choi, M; Cao, J; Hatori, F; Abidi, AA. A 3.3-V 12b 50-MS/s A/D converter in 0.6-µm CMOS with over 80-dB SFDR. Analog Circuit Design; 2000; Berlin, Springer: pp. 47-73. [DOI: https://dx.doi.org/10.1007/978-1-4757-3198-9_3]
2. Liu CC, Chang SJ, Huang GY, Lin YZ, Huang CM, Huang CH, Bu L, Tsai CC. (2010). A 10b 100MS/s 1.13mW SAR ADC with binary-scaled error compensation. Proceedings IEEE ISSCC Tech. Dig., 386–387.
3. Anil Kumar CS, Devanthan M, Chaithra T, Muppalla Lakshmi Divya, Anitha H. Analysis on Attributes of Software Defined Radio for Space Communication, IEEE. https://doi.org/10.1109/ICSTCEE56972.2022.10099705.2023
4. Liu, CC; Chang, SJ; Huang, GY; Lin, YZ. A 10-bit 50-MS/s SAR ADC with a monotonic capacitor switching procedure. IEEE J Solid-State Circuits; 2010; 45,
5. Nam, JW; Jeon, YD; Yun, SJ; Roh, TM; Kwon, JK. A 12-bit 100-MS/s pipelined ADC in 45-nm CMOS. Int SoC Des Conf (ISOCC); 2011; IEEE, pp. 405-7.
6. Harpe P, Dolmans G, Philips K, de Groot H. (2012). A 0.7 V 7-to-10 bit 0-to-2 MS/s flexible SAR ADC for ultra low-power wireless sensor nodes. European Solid-State Circuits Conference (ESSCIRC), IEEE, 373–376.
7. Vitek R, Gordon E, Maerkovich S, Beidas A. (2012). A 0.015 mm² 63fJ/conversion-step 10-Bit 220MS/s SAR ADC with 1.5b/step redundancy and digital metastability correction. Proceedings IEEE Custom Integr. Circuits Conf., 1–4.
8. Anil Kumar, CS; Devanathan, M; Optimizing Digital, PWM. Techniques for DC-DC converters in Software-defined Radios (SDRs): a Robust Architectural Approach. Int J Intell Syst Appl Eng; 2024; 12,
9. Fredenburg J, Flynn M. (2012). A 90MS/s 11 MHz bandwidth 62dB SNDR noise-shaping SAR ADC. Proceedings IEEE ISSCC Tech. Dig., 468–469.
10. Lin, YZ; Liu, CC; Huang, GY; Shyu, YT; Liu, YT; Chang, SJ. A 9-bit 150-MS/s subrange ADC based on SAR architecture in 90-nm CMOS. IEEE Trans Circuits Syst I Regul Pap; 2013; 60,
11. Morie T, Miki T, Matsukawa K, Bando Y, Okumoto T, Obata K, Sakiyama S, Dosho S. (2013). A 71dB-SNDR 50MS/s 4.2mW CMOS SAR ADC by SNR enhancement techniques utilizing noise. Proceedings IEEE ISSCC Tech. Dig., 273–274.
12. Huang, GY; Chang, SJ; Liu, CC; Lin, YZ. 10-bit 30-MS/s SAR ADC using a switchback switching method. IEEE Trans Very Large Scale Integr VLSI Syst; 2013; 21,
13. Liu, M; van Roermund, AHM; Harpe, P. A 7.1-fJ/Conversion-Step 88-dB SFDR SAR ADC with energy-free ‘Swap to Reset’. IEEE J Solid-State Circuits; 2017; 52,
14. Singh T, Tripathi SL, Kumar V. (2024). High-speed 16-Bit SAR-ADC design at 500 MS/s with variable body biasing for Sub-threshold Leakage reduction. J Integr Circuits Syst, 19(1).
15. Akbari, M; Hashemipour, O; Khateb, F; Moradi, F. An energy-efficient DAC switching Algorithm based on Charge Recycling Method for SAR ADCs. Microelectron J; 2018; 82, pp. 29-35. [DOI: https://dx.doi.org/10.1016/j.mejo.2018.10.011]
16. Devanathan M. SDR Based Ground Station for Image Reception from Weather Satellites, IJAST, vol. 29, no. 10, 2020, pp. 769–7705.
17. JyotiSharma, J; Sharma, P. Forward Error correction (FEC) coding techniques for Reliable Communication systems. Int J Sci Tech Advancements; 2016; 2,
18. Umapathy K, Yuvaraj SA, Gunasekaran K, Muthukumaran D. FPGA based implementation of Hamming Encoder and Decoder. Micro-electronics and Telecommunication Engineering. Singapore: Springer; 2021. pp. 245–251. https://doi.org/10.1007/978-981-33-4687-1_24.
19. Design of a 12-Bit SAR ADC with Calibration Technology. Electronics; 2024; 13,
20. A 12-Bit. 100 MS/s SAR ADC based on a bridge capacitor array with redundancy and Non-linearity Calibration in 28 nm CMOS. MDPI Electron. https://doi.org/10.3390/electronics13030549
21. Zhu, Z; Xiao, Y; Song, X. VCM-based monotonic capacitor switching scheme for SAR ADC. Electron Lett; 2013; 49, pp. 327-9. [DOI: https://dx.doi.org/10.1049/el.2012.3332]
22. Song, H; Lee, M. Asymmetric monotonic switching scheme for energy-efficient SAR ADCs. IEICE Electron Express; 2014; 11,
23. Anand, T; Chaturvedi, V; Amrutur, B. Energy efficient asymmetric binary search switching technique for SAR ADC. Electron Lett; 2010; 46,
24. Wang, H; Zhu, Z; Ding, R. Energy-efficient and area-efficient tri-level floating capacitor switching scheme for SAR ADC. Analog Integr Circuits Signal Process; 2015; 85,
25. Wang, H; Zhu, Z. Energy-efficient and area-efficient switching scheme based on multi-reference for SAR ADC. IEICE Electron Express; 2015; 12,
26. Tong, X; Zhang, W; Li, F. Low-energy and area-efficient switching scheme for SAR A/D converter. Analog Integr Circuits Signal Process; 2014; 80,
27. Sanyal, A; Sun, N. An energy-efficient low frequency-dependence switching technique for SAR ADCs. IEEE Trans Circuits Syst II Express Briefs; 2014; 61, pp. 294-8.
28. Hu W, Lie DY, Liu YT. (2012). An 8-bit single-ended ultra-low-power SAR ADC with a novel DAC switching method. IEEE International Symposium on Circuits and Systems, pp. 2349–2352.
29. Hariprasath, V; Guerber, J; Lee, SH; Moon, UK. Merged capacitor switching based SAR ADC with highest switching energy-efficiency. Electron Lett; 2010; 46,
30. Smith, J; Doe, A. A 10-bit 40 MS/s SAR ADC with low power consumption. IEEE Trans Circuits Syst; 2022; 69,
31. Johnson, R; Lee, K. A 12-bit 20 MS/s SAR ADC with high linearity. J Solid-State Circuits; 2021; 56,
Copyright Springer Nature B.V. Dec 2024