A 12-bit 100 MS/s pipelined SAR ADC with addition-only digital error correction

A new redundant successive approximation register (SAR) ADC architecture with digital error correction is presented to avoid the comparator offset issue and subtraction operations. A 2-channel 12-bit 100 MS/s SAR ADCs based on the proposed architecture with voltage-controlled delay lines based time-domain comparator is designed in a 65 nm CMOS technology. Simulation results show that at a supply voltage of 1.2 V, the SAR ADC achieves a signal-to-noise-and-distortion ratio (SNDR) of 70.81 dB (11.47 ENOB), a spurious free dynamic range (SFDR) of 80.33 dB for a near Nyquist input at 100 MS/s, while dissipating 11 mW from a 1.2-V supply, giving a FOM of 38.8 fJ/Conversion-step.     

A 7-bit 40 MS/s single-ended asynchronous SAR ADC in 65 nm CMOS

This paper presents a 7-bit 40 MS/s single-ended asynchronous SAR ADC intended for in-probe use in medical applications, which requires small area and good power efficiency. A single-ended architecture is proposed for a moderate resolution for its simplicity. Together with a double reference technique, the architecture reduces the area of the technology-limited large capacitors. The speed is optimized by an asymmetric delay line embedded in the asynchronous digital logic, enabling a sampling frequency of 40 MS/s. The prototype is fabricated in a 65 nm CMOS technology. Measurement shows that at 1 V supply and 40 MS/s, the ADC achieves an SNDR of 39.73 dB and an ENOB of 6.3 bit, while consuming 298.6 µW, resulting in an energy efficiency of 94.74 fJ/conversion-step. The core circuit layout only occupies 0.017 mm².     

An 8-bit 208 MS/s SAR ADC in 65 nm CMOS

An 8-bit low-power 208MS/s SAR analog-to-digital converter is presented. To achieve a high-speed and low-power operation, a reused terminating capacitor switching procedure is proposed. The proposed switching procedure halves the capacitors leading to a significant power saving over the conventional one. Moreover, the proposed architecture relaxes the settling time of DAC and subsequently improves the conversion rate. The ADC has been simulated in SMIC 65 nm 1.2 V CMOS technology. At a 1.2-V supply and 208 MS/s, the ADC consumes 2.7 mW and achieves an SNDR of 49.6 dB, an SFDR of 61.0 dB with 100 MHz inputs.     

A zero-crossing based 10-bit 100 MS/s pipeline ADC with controlled current in 90 nm CMOS

In this paper, a new charging technique for low power zero-crossing based circuit pipeline analog-to-digital converters (ADCs) is presented. The charging current sources are implemented as voltage-controlled current sources in order to charge the sampling capacitors based on the error signal. Using this method, the ADC power consumption is reduced while improving the accuracy. The necessary current control block is shared between consecutive stages further reducing the power consumption and die area. The proposed technique is applied to a 10-bit 100 MS/s pipeline ADC designed in a 90 nm CMOS technology with 1 V power supply. Circuit level simulation results using Cadence Spectre show a signal-to-noise and distortion ratio of 55.6 dB with 3.56 mW power consumption resulting in a figure of merit of 72.3 fJ/conv.step without employing any calibration technique.     

An 80-MS/s 14-bit pipelined ADC featuring 83 dB SFDR

An 80-MS/s 14-bit pipelined analog-to-digital converter (ADC) is presented in this paper. After gain error and offset extraction from prototype measurement, the improved circuit achieves spurious free dynamic range (SFDR) of 82.9 dB and signal-to-noise-and-distortion ratio (SINAD) of 64.1 dB for a 30.5 MHz input, maintained within 6 dB performance deterioration up to 170 MHz input. Differential nonlinearity (DNL) is 0.66 LSB and integral nonlinearity (INL) is 2.5 LSB. Low-jitter clock amplifier and buffers with balanced loads are used to reduce the jitter and skew between different stages. An on-chip voltage reference generator is schemed with low impedance to reduce noise and spurs of reference signals. The ADC is fabricated in a 0.18-μm CMOS process with core area of 3.86 mm², and consumes 518 mW at 1.8 V supply.     

An area-and-power-efficient 8.4-bit ENOB 30 MS/s SAR ADC in 65 nm CMOS

Area and power consumption are two main concerns for the electronics towards the digitalization of in-probe 3D ultrasound imaging systems. This work presents a 10-bit 30 MS/s successive approximation register analog-to-digital converter, which achieves good area efficiency as well as power efficiency, by using a symmetrical MSB-capacitor-split capacitor array with customized small-value finger capacitors. Moreover, simplified dynamic digital logic and a dynamic comparator have been designed. Fabricated in a 65 nm CMOS technology, the core circuit only occupies 0.016 mm². The ADC achieves a signal-to-noise ratio of 52.2 dB, and consumes 61.3 μW at 30 MS/s from a 1 V supply voltage, resulting in a figure of merit (FoM) of 6.2 fJ/conversion-step. The FoM defined by including the area is 0.1 mm² fJ/conversion-step.     

An 87 fJ/conversion-step 12 b 10 MS/s SAR ADC using a minimum number of unit capacitors

This work proposes a 12 b 10 MS/s 0.11 μm CMOS successive-approximation register ADC based on a C-R hybrid DAC for low-power sensor applications. The proposed C-R DAC employs a 2-step split-capacitor array of upper seven bits and lower five bits to optimize power consumption and chip area at the target speed and resolution. A V_CM-based switching method for the most significant bit and reference voltage segments from an insensitive R-string for the last two least significant bits minimize the number of unit capacitors required in the C-R hybrid DAC. The comparator accuracy is improved by an open-loop offset cancellation technique in the first-stage pre-amp. The prototype ADC in a 0.11 μm CMOS process demonstrates the measured differential nonlinearity and integral nonlinearity within 1.18 LSB and 1.42 LSB, respectively. The ADC shows a maximum signal-to-noise-and-distortion ratio of 63.9 dB and a maximum spurious-free dynamic range of 77.6 dB at 10 MS/s. The ADC with an active die area of 0.34 mm² consumes 1.1 mW at 1.0 V and 10 MS/s, corresponding to a figure-of-merit of 87 fJ/conversion-step.     

A 1 V 186-μW 50-MS/s 10-bit subrange SAR ADC in 130-nm CMOS process

This paper presents a 10-bit 50-MS/s subrange successive-approximation register (SAR) analog-to-digital converter (ADC) composed of a 4-bit SAR coarse ADC and a 6-bit SAR fine ADC. In the coarse ADC, multi-comparator SAR architecture is used to reduce the digital logic propagation delay, and a traditional asynchronous SAR ADC with monotonic switching method is used as the fine ADC. With that combination, power dissipation also can be much reduced. Meanwhile, a modified SAR control logic is adopted in the fine ADC to speed up the conversion and other techniques, such as splitting capacitors array, are borrowed to reduce the power consumption. Fabricated with 1P8M 130-nm CMOS technology, the proposed SAR ADC achieves 51.6-dB signal to noise and distortion ratio (SNDR) and consumes 186 μ W at 50 MS/s with a 1-V supply, resulting in a figure of merit (FOM) of 12 fJ/conversion-step. The core area is only 0.045 mm².     

A 1.2-V 19.2-mW 10-bit 30-MS/s pipelined ADC in 0.13-μm CMOS

A 10-bit 30-MS/s pipelined analog-to-digital converter(ADC) is presented.For the sake of lower power and area,the pipelined stages are scaled in current and area,and op amps are shared between the successive stages. The ADC is realized in the 0.13-μm 1-poly 8-copper mixed signal CMOS process operating at 1.2-V supply voltage. Design approaches are discussed to overcome the challenges associated with this choice of process and supply voltage, such as limited dynamic range,poor analog characteristic device...     

A mismatch-error minimized four-channel time-interleaved 11 b 150 MS/s pipelined SAR ADC

This work proposes a four-channel time-interleaved 11 b 150 MS/s pipelined SAR ADC based on various analog techniques to minimize mismatches between channels without any calibration scheme. The proposed ADC eliminates an input SHA to reduce offset mismatches, while the pipelined SAR architecture solves the problem of limited input bandwidth as observed in conventional SHA-free ADCs. In addition, a shared residue amplifier between four channels minimizes various mismatches caused by amplifiers in the first-stage MDACs. Two types of references for the residue amplifier and the SAR ADCs isolate the reference instability problem due to different functional requirements, while the shared residue amplifier uses only a single reference during the amplifying mode of each channel to reduce a gain mismatch. For high performance of the SAR ADC, high-frequency clocks with a controllable duty cycle are generated on chip without external, complicated, high-speed multi-phase clocks. The prototype 11 b ADC in a 0.13 μm CMOS shows a measured DNL and INL of 0.31 LSB and 1.18 LSB, respectively, with an SNDR of 59.3 dB and an SFDR of 67.7 dB at 100 MS/s, and an SNDR of 54.5 dB and an SFDR of 65.5 dB at 150 MS/s. The ADC with an active die area of 2.42 mm² consumes 46.8 mW at 1.2 V and 150 MS/s.     

An 8.38?fJ/conversion-step 0.6?V 8-b 4.35?MS/s asynchronous SAR ADC in 65?nm CMOS

An asynchronous reference-free successive-approximation register analog-to-digital converter (ADC) in 65?nm CMOS is presented. In order to fit for low supply voltage design in advanced digital sub-nanometer process, a differential time-domain comparator is implemented. It tracks process variation with unit transistor of input device as well as charging capacitor, and enables differential configuration by using RS trigger instead of D-flip-flop. The proposed architecture is digital circuits heavily involved; therefore it will benefit from technology scaling. The power dissipation is further improved by means of asynchronous architecture, which is adjusted for low supply voltage operation, as well as reference-free configuration. The prototype ADC is fabricated in SMIC 65?nm digital CMOS process. It has signal to noise and distortion ratio of 46.7?dB at the sampling rate of 4.35?MS/s while consuming 6.6???W from 0.6-V power supply. It maintains linearity over 7-bit when the input frequency is lower than 5?MHz. The figure of merit of 8.38?fJ/conversion-step and die area of 0.011?mm² is achieved.     

A 1.2 V 10-bit 5MS/s low power CMOS cyclic ADC based on double-sampling technique

This paper presents a 1.2 V 10-bit 5MS/s low power cyclic analog-to-digital converter (ADC). The strategy to minimize the power adopts the double-sampling technique. At the front-end, a timing-skew-insensitive double-sampled Miller-capacitance-based sample-and-hold circuit (S/H) is employed to enhance the dynamic performance of the cyclic ADC. Double sampling technique is also applied to multiplying digital-to-analog converter (MDAC). This scheme provides a better power efficiency for the proposed cyclic ADC. Furthermore, bootstrapped switch is used to achieve rail-to-rail signal swing at low-voltage power supply. The prototype ADC, fabricated in TSMC 0.18 μm CMOS 1P6 M process, achieves DNL and INL of 0.32LSB and 0.45LSB respectively, while SFDR is 69.1 dB and SNDR is 58.6 dB at an input frequency of 600 kHz. Operating at 5MS/s sampling rate under a single 1.2 V power supply, the power consumption is 1.68 mW.     

14位50MS/s100mW0.18μm CMOS流水线ADC

实现了一种14位40MS/s CMOS流水线A/D转换器(ADC)。在1.8V电源电压下,该ADC功耗仅为100mW。基于无采样/保持放大器前端电路和双转换MDAC技术,实现了低功耗设计,其中,无采样/保持放大器前端电路能降低约50%的功耗,双转换MDAC能降低约10%的功耗。该ADC采用0.18μm CMOS工艺制作,芯片尺寸为2.5mm×1.1mm。在40MS/s采样速率、10MHz模拟输入信号下进行测试,电源电压为1.8V,DNL在±0.8LSB以内,INL在±3.5LSB以内,SNR为73.5dB,SINAD为73.3dB,SFDR为89.5dBc,ENOB为11.9位,THD为-90.9dBc。该ADC能够有效降低SOC系统、无线通信系统及数字化雷达的功耗。     

A 12-bit 250 MS/s pipeline ADC with 78 dB SFDR in 0.13-µm CMOS

A 12-bit 250 MS/s pipeline ADC is presented and implemented in 0.13 µm CMOS process. To reduce the load capacitance of each pipeline stage and save area, the inter-metal capacitors are adopted as input sampling capacitors of the comparators. A fully integrated reference buffer associated with a simulation scheme is proposed to improve the settling speed and PSRR of the differential reference voltage. To reduce the overall power a low cost foreground calibration for capacitor mismatches is employed. The single-stage telescopic with gain-boosting amplifiers and an improved bias is applied in each stage due to its high power efficiency. Additionally, the timing in the sampling phase is optimised to achieve high sampling linearity. Even harmonics induced by parasitic capacitance are analysed profoundly and mitigated at the level of layout. The measured SNDR and SFDR are 63 and 78 dB with 38.1 MHz input, respectively, and remain 63 and 77 dB with Nyquist input. The ADC core area is 1.6 mm² and consumes 165 mW (reference buffer included, LVDS excluded) at 250 MS/s under 1.3 V.     

A CMOS switch-capacitor 14-bit 100 Msps pipeline ADC with over 90 dB SFDR

This article presents a design of 14-bit 100?Msamples/s pipelined analog-to-digital converter (ADC) implemented in 0.18?µm CMOS. A charge-sharing correction (CSC) is proposed to remove the input-dependent charge-injection, along with a floating-well bulk-driven technique, a fast-settling reference generator and a low-jitter clock circuit, guaranteeing the high dynamic performance of the ADC. A scheme of background calibration minimises the error due to the capacitor mismatch and opamp non-ideality, ensuring the overall linearity. The measured results show that the prototype ADC achieves spurious-free dynamic range (SFDR) of 91?dB, signal-to-noise-and-distortion ratio (SNDR) of 73.1?dB, differential nonlinearity (DNL) of +0.61/?0.57?LSB and integrated nonlinearity (INL) of +1.1/?1.0?LSB at 30?MHz input and maintains over 78?dB SFDR and 65?dB SNDR up to 425?MHz, consuming 223?mW totally.     

A 1.0-V 6-b 40 MS/s time-domain flash ADC in 0.18 μm CMOS

This paper presents a 6-bit low power low supply voltage time-domain comparator. The conventional voltage comparison is moved to time-domain so as to remove pre-amplifier and latch, which enables its feasibility to low supply voltage. The voltage-to-time converter is realized by the proposed linear pulse-width-modulation. The set-up time of the D flip-flop determines the sampling rate of the converter. The resistive averaging relaxes the matching requirement of the parallel comparison cells. The total input capacitance is decreased to less than 40fF in this architecture. The above digital-intensive setting makes the analog-to-digital converter (ADC) benefit from technology scaling in both power consumption and sampling rate. The prototype ADC is fabricated in SMIC 0.18 μm CMOS process. At 40 MS/s and 1.0-V supply, it consumes 540 μW and achieves an effective-number-of-bit of 5.43, resulting in a figure-of-merit of 0.31 pJ/conversion-step and active area of 0.1 mm².