A 10-bit 200-MHz CMOS video DAC for HDTV applications

This paper describes a 10-bit 200-MHz CMOS current steering digital-to-analog converter (DAC) for HDTV applications. The proposed 10-bit DAC is composed of a unit decoded matrix for 6 MSBs and a binary weighted array for 4 LSB’s, considering linearity, power consumption, routing area, and glitch energy. A new switching scheme for the unit decoded matrix is developed to improve linearity further. Cascade current sources and differential switches with deglitch latch improve dynamic performance. The measured differential nonlinearity (DNL) and integral nonlinearity (INL) are 0.3 LSB and 0.2 LSB, respectively. The converter achieves a spurious-free dynamic range (SFDR) of above 55 dB over a100-MHz bandwidth and low glitch energy of 1.5 pVs. The circuit is fabricated in a 0.25 μm CMOS process and occupies 0.91 mm². When operating at 200 M Sample/s, it dissipates 82 mW from a 3.3 V power supply.     

A 56-mW 23-mm/sup 2/ single-chip 180-nm CMOS GPS receiver with 27.2-mW 4.1-mm/sup 2/ radio

A 56-mW 23-mm/sup 2/ GPS receiver with CPU-DSP-64 kRAM-256 kROM and a 27.2-mW 4.1-mm/sup 2/ radio has been integrated in a 180-nm CMOS process. The SoC GPS receiver, connected to an active antenna, provides latitude, longitude, height with 3-m rms precision with no need of external host processor in a [-40, 105]/spl deg/C temperature range. The radio draws 17 mA from a 1.6-1.8-V voltage supply, takes 11 pins of a VFQFPN68 package, and needs just a few passives for input match and a crystal for the reference oscillator. Measured radio performances are NF=4.8 dB, Gp=92 dB, image rejection > 30 dB, -112 dBc/Hz phase noise @ 1 MHz offset from carrier. Though GPS radio linearity and ruggedness have been made compatible with the co-existence of a microprocessor, radio silicon area and power consumption is comparable to state-of-the-art stand-alone GPS radio. The one reported here is the first ever single-chip GPS receiver requiring no external host to achieve satellite tracking and position fix with a total die area of 23 mm/sup 2/ and 56-mW power consumption.     

A 12-bit intrinsic accuracy high-speed CMOS DAC

A 12-bit intrinsic accuracy digital-to-analog (D/A) converter integrated in a standard digital 0.5 μm CMOS technology is presented. It is based on a current steering doubly segmented 6+2+4 architecture and requires no calibration, no trimming, or dynamic averaging. The differential nonlinearity (DNL) and integral nonlinearity (INL) are 0.3 and 0.6 least significant bits (LSB's), respectively. The measured glitch energy is 1.9 pV.s. For a 12-bit resolution, the converter reaches an update rate of 300 MS/s. By reducing the voltage supply of the latches to 2.0 V, the glitch energy is reduced to sub-pV.s, and the update rate reaches 500 MS/s, for a resolution of 8 bits. The worst case power consumption is 320 mW, and it operates from a single 3.3 V voltage supply. The die area is 3.2 mm²     

Nyquist电流型7位20MHz采样率CMOS数模转换器

介绍了一种高速7位DAC的设计及芯片测试结果,该DAC选取高5位单位电流源,低2位二进制电流源的分段结构。考虑了电流源匹配、毛刺降低以及版图中误差补偿等方面的问题来优化电路。流片采用0.35μmChartered双层多晶四层金属工艺,测试结果表明在20 MH z的采样频率下,微分非线性度和积分非线性度分别小于±0.2 LSB和±0.35 LSB。该DAC的满幅建立时间是20 ns,芯片面积为0.17 mm×0.23 mm。电源电压为3.3 V,功耗为3 mW。     

A 14-bit intrinsic accuracy Q2 random walk CMOS DAC

In this paper, a 14-bit, 150-MSamples/s current steering digital-to-analog converter (DAC) is presented. It uses the novel Q² random walk switching scheme to obtain full 14-bit accuracy without trimming or tuning. The measured integral and differential nonlinearity performances are 0.3 and 0.2 LSB, respectively; the spurious-free dynamic range is 84 dB at 500 kHz and 61 dB at 5 MHz. Running from a single 2.7-V power supply, it has a power consumption of 70 mW for an input signal of 500 kHz and 300 mW for an input signal of 15 MHz. The DAC has been integrated in a standard digital single-poly, triple-metal 0.5-μm CMOS process. The die area is 13.1 mm²     

A 19-mW 2.6-mm/sup 2/ L1/L2 dual-band CMOS GPS receiver

This paper presents the design and implementation of an L1/L2 dual-band global positioning system (GPS) receiver. Dual-conversion with a low-IF architecture was used for dual-band operation. The receiver is composed of an RF preamplifier, down-conversion mixers, a variable-gain channel filter, a 2-bit analog-to-digital converter, and the full phase-locked-loop synthesizer including an on-chip voltage controlled oscillator. Fabricated in a 0.18-/spl mu/m CMOS technology, the receiver exhibits maximum gain of 95 dB and noise figures of 8.5 and 7.5 dB for L1 and L2, respectively. An on-chip variable-gain channel filter provides IF image rejection of 20 dB and gain control range over 60 dB. The receiver consumes 19 mW from a 1.8-V supply while occupying a 2.6-mm/sup 2/ die area including the ESD I/O pads.     

A 14-bit 1.8-V 20-mW 1-mm2 CMOS DAC

A low-voltage low-power small-area and high-resolution digital-to-analog converter (DAC) for mixed-signal applications is Introduced. A binary weighted current steering DAC is a power-efficient architecture, because almost all the current taken from the supply is used for the output signal. The current steering architecture is also highly suitable for high-speed operation. Typically, the architecture suffers from poor linearity characteristics, but the problem can be prevented with a novel calibration method, where the currents generated for the most significant bits are fine tuned. As a result, a very compact and low-power solution can be implemented by using a low-voltage digital technology     

A 10-bit 250-MS/s binary-weighted current-steering DAC

This paper studies the impact of segmentation on current-steering digital-to-analog converters (DACs). Segmentation may be used to improve the dynamic behavior of the converter but comes at a cost. A method for reducing the segmentation degree is given. The presented chip, a 10-bit binary-weighted current-steering DAC, has >60 dB SFDR at 250 MS/s from DC to Nyquist. At 62.5 MHz signal frequency and 250 MS/s, we operated the device in 9-bit unary, 1-bit binary-weighted mode. The obtained 60 dB SFDR in this measurement demonstrates that the binary nature of the converter did not limit the SFDR. The chip draws 4 mW from a dual 1.5 V/1.8 V supply plus load currents. The active area is less than 0.35 mm/sup 2/ in a standard 1P-5M 0.18-/spl mu/m 1.8-V CMOS process. Both INL and DNL are below 0.1 LSB.     

A 12-bit 320-MSample/s current-steering CMOS D/A converter in 0.44 mm/sup 2/

A 12-bit 320-MSample/s current-steering D/A converter in 0.18-/spl mu/m CMOS is presented. In order to achieve high linearity and spurious free dynamic range (SFDR), a large degree of segmentation has been used, with the seven most significant bits (MSBs) being implemented as equally weighted current sources. A "design-for-layout" approach has allowed this to be done in an area of just 0.44 mm/sup 2/. The increased switching noise associated with a high degree of segmentation has been reduced by a new latch architecture. Differential nonlinearity of /spl plusmn/0.3 LSB and integral nonlinearity of /spl plusmn/0.4 LSB have been measured. Low-frequency SFDR of 95 dB has been achieved, while SFDR at 320 MS/s remains above 70 and 60 dB for input frequencies up to 10 and 60 MHz, respectively. The converter consumes a total of 82 mW from 1.8-V and 3.3-V supplies. The validity of the techniques used has been demonstrated by fabricating the converter in two separate 0.18-/spl mu/m processes with similar results measured for both.     

A 130 nm CMOS 6-bit Full Nyquist 3 GS/s DAC

This paper presents a 6-bit very high-speed, low-power digital-to-analog converter (DAC). It is based on a current steering binary weighted architecture and achieves 10-bit static linearity without calibration. Due to the use of a pseudo-segmented structure instead of a thermometer decoder, the operating speed of the converter can be up to 4.5 GS/s. The DAC occupies 0.4 mm $times$ 0.5 mm in a standard 130 nm CMOS technology. A spurious-free dynamic range (SFDR) of more than 36 dB has been measured over the complete Nyquist interval at sampling frequencies up to 3 GS/s. The power consumption at a 3 GHz clock frequency for a near-Nyquist sinusoidal output signal equals 29 mW .      

一种低功耗10位流水线结构的CMOS A/D转换器的设计

本文设计了一种可满足视频速度应用的低电压低功耗10位流水线结构的CMOS A/D转换器.该转换器由9个低功耗运算放大器和19个比较器组成,采用1.5位/级共9级流水线结构,级间增益为2并带有数字校正逻辑.为了提高其抗噪声能力及降低二阶谐波失真,该A/D转换器采用了全差分结构.全芯片模拟结果表明,在3V工作电压下,以20MHz的速度对2MHz的输入信号进行采样时,其信噪失调比达到53dB,功率消耗为28.7mW.最后,基于0.6μm CMOS工艺得到该A/D转换器核的芯片面积为1.55mm2.     

A 35-mW 3.6-mm/sup 2/ fully integrated 0.18-/spl mu/m CMOS GPS radio

A single-chip CMOS Global Positioning System (GPS) radio has been integrated using only a couple of external passive components for the input matching network and one external reference for the synthesizer. The receiver downconverts the GPS L1 signal at 1575.42 MHz to an IF of 9.45 MHz. The complete front-end and frequency synthesizer section have been integrated: low noise amplifier, image rejection mixer, IF active filter, and the full phase-locked loop synthesizer, including voltage-controlled oscillator and loop filter. The front-end measured performances are 81-dB maximum gain, 5.3-dB noise figure, and >30-dB image rejection. The synthesizer features a phase noise of -95 dBc/Hz at 1-MHz offset and a total integrated phase noise of less than 7/spl deg/ rms in the 500-Hz-1.5-MHz band. The front-end and the synthesizer draw, respectively, 11 and 9 mA from a 1.8-V supply. The architecture of the front-end and synthesizer has been geared to high level of integration and reduction of silicon area at the lowest possible power consumption. Consequently, the one reported here is the smallest and most integrated CMOS GPS receiver reported so far.     

一种带分段随机温度计码的10位电流舵数模转换器

介绍了一种基于分段随机温度计码的动态匹配算法。该算法可以有效抑制电流源失配造成的谐波失真,因此可以降低对电流源匹配的需求。在此算法基础上,针对芯片面积,优化了电流源尺寸选取与分段位数的选择。在SMIC 0.13μm CMOS工艺中实现了一款10位电流舵数模转换器(Digial-to-analog converter,DAC),单通道的面积为0.05mm2。测试结果显示,微分非线性(Differential non-linearity,DNL)与积分非线性(Integral nonlinearity,INL)分别为0.58LSB和0.56LSB,无杂散动态范围(Spurious free dynamic range,SFDR)最高可达80dBc。单通道DAC在1.2V数字/模拟电源电压下整体功耗小于3mW。     

A 125-mm/sup 2/ 1-Gb NAND flash memory with 10-MByte/s program speed

A single 3-V only, 1-Gb NAND flash memory has been successfully developed. The chip has been fabricated using 0.13-/spl mu/m CMOS STI technology. The effective cell size including the select transistors is 0.077 /spl mu/m/sup 2/. To decrease the chip size, a new architecture is introduced. The in-series connected memory cells are increased from 16 to 32. Furthermore, as many as 16 k memory cells are connected to the same wordline. As a result, the chip size is decreased by 15%. A very small die size of 125 mm/sup 2/ and an excellent cell area efficiency of 70% are achieved. As for the performance, a very fast programming and serial read are realized. The highest program throughput ever of 10.6-MByte/s is realized: 1) by quadrupling the page size and 2) by newly introducing a write cache. In addition, the garbage collection is accelerated to 9.4-MByte/s. In addition, the write cache accelerates the serial read operation and a very fast 20-MByte/s read throughput is realized.     

A 10-b, 500-MSample/s CMOS DAC in 0.6 mm2

A 10-b current steering CMOS digital-to-analog converter (DAC) is described, with optimized performance for frequency domain applications. For sampling frequencies up to 200 MSample/s, the spurious free dynamic range (SFDR) is better than 60 dB for signals from DC to Nyquist. For sampling frequencies up to 400 MSample/s, the SFDR is better than 55 dB for signals from DC to Nyquist. The measured differential nonlinearity and integral nonlinearity are 0.1 least significant bit (LSB) and 0.2 LSB, respectively. The circuit is fabricated in a 0.35-μm, single-poly, four-metal, 3.3 V, standard digital CMOS process and occupies 0.6 mm². When operating at 500 MSample/s, it dissipates 125 mW from a 3.3 V power supply. This DAC is optimized for embedded applications with large amounts of digital circuitry     

A digitally self-calibrating 14-bit 10-MHz CMOS pipelined A/Dconverter

A digitally self-calibrating pipelined analog-to-digital converter (ADC) featuring 1.5-bit/stage structure is presented. The integral (INL) and differential nonlinearity (DNL) errors are removed using a novel digital calibration algorithm, which also eliminates missing codes that can occur with other calibration algorithms near the extremes of the input range. After calibration, the measured DNL is ±0.6 LSB and the INL is ±2.5 LSB at the 14-bit level. Sampling at a 10-MHz rate, the chip dissipates 220 mW and (post-calibration) yields a signal-to-noise ratio of 77 dB and a spurious-free dynamic range of 95 dB with 4.8-MHz sine wave input signal. The chip is fabricated in 0.5-μm CMOS double-poly double-metal process, measures 3.8 mm × 3.3 mm (150 mil × 130 mil), and operates from a single 5-V supply     

A 21-mW 8-b 125-MSample/s ADC in 0.09-mm/sup 2/ 0.13-/spl mu/m CMOS

This paper presents an 8-b two-step subranging analog-to-digital (ADC) using interpolation, averaging, offset compensation, and pipelining techniques to accomplish an effective number of bits of 7.6 b at 125 MSample/s. The 0.13-/spl mu/m CMOS ADC occupies 0.09 mm/sup 2/ and consumes 21 mW.     

A 10-bit 50-MS/s redundant SAR ADC with split capacitive-array DAC

A new architecture for successive-approximation register analog-to-digital converters (SAR ADC) using generalized non-binary search algorithm is proposed to reduce the complexity and power consumption of the digital circuitry. The proposed architecture is based on the split capacitive-array DAC with a simple switching logic as compared to the conventional non-binary SAR ADC architecture. A 10-bit 50-MS/s SAR ADC is designed based on the proposed architecture in a 0.18 μm CMOS technology. Simulation results show that at a supply voltage of 1.2 V, the SAR ADC achieves a peak signal-to-noise-and-distortion ratio of 59.5 dB, and a power consumption of 1.3 mW, resulting in a figure of merit of 33 fJ/conversion-step.     

A 10-bit 1-GSample/s Nyquist current-steering CMOS D/A converter

In this paper, a 10-bit 1-GSample/s current-steering CMOS digital-to-analog (D/A) converter is presented. The measured integral nonlinearity is better than ±0.2 LSB and the measured differential nonlinearity lies between -0.08 and 0.14 LSB proving the 10-bit accuracy. The 1-GSample/s conversion rate has been obtained by an, at transistor level, fully custom-designed thermometer decoder and synchronization circuit. The layout has been carefully optimized. The parasitic interconnect loads have been estimated and have been iterated in the circuit design. A spurious-free dynamic range (SFDR) of more than 61 dB has been measured in the interval from dc to Nyquist. The power consumption equals 110 mW for a near-Nyquist sinusoidal output signal at a 1-GHz clock. The chip has been processed in a standard 0.35-μm CMOS technology and has an active area of only 0.35 mm²     

A 10-bit 200-MS/s CMOS parallel pipeline A/D converter

This paper describes a 10-bit 200-MS/s CMOS parallel pipeline analog-to-digital (A/D) converter that can sample input frequencies above 200 MHz. The converter utilizes a front-end sample-and-hold (S/H) circuit and four parallel interleaved pipeline component A/D converters followed by a digital offset compensation. By optimizing for power in the architectural level, incorporating extensively parallelism and double-sampling both in the S/H circuit and the component ADCs, a power dissipation of only 280 mW from a 3.0-V supply is achieved. Implemented in a 0.5-μm CMOS process, the circuit occupies an area of 7.4 mm². The converter achieves a differential nonlinearity and integral nonlinearity of ±0.8 LSB and ±0.9 LSB, respectively, while the peak spurious-free-dynamic-range is 55 dB and the total harmonic distortion better than 46 dB at a sampling rate of 200 MS/s