一种5 Gb/s双信道并行时钟数据恢复电路

基于0.18 μm CMOS工艺,设计了一种双信道并行时钟数据恢复(CDR)电路,它由1个锁相环(PLL)型CDR和1个相位选择/相位插值(PS/PI)型CDR结合实现。与传统的并行CDR相比,该CDR电路不需要本地参考时钟。PLL型CDR中环形压控振荡器的延迟单元采用电感峰化技术,拓展了带宽,实现了较高的振荡频率;电荷泵采用自举基准和运放,改善了充放电电流匹配。PS/PI型CDR中Bang-Bang型鉴相器结构简单,具有较好的鉴相功能;PS/PI电路比传统结构少2个相位选择器。仿真结果表明,当输入并行数据速率为5 Gb/s时,恢复出的2组时钟与数据的峰峰抖动值分别为6.1 ps,8.1 ps和8.7 ps,11.2 ps。电路核心模块的功耗为172.4 mW,整体电路版图面积为(1.7×1.585) mm²。     

3.125Gb/s基于PS/PI型的时钟与数据恢复电路设计

基于标准0.18μmCMOS工艺,设计了一种全速率PS/PI型时钟与数据恢复(CDR)电路。该电路主要由bang-bang型鉴相器、数字控制模块、分接器、相位选择器以及相位插值器等模块构成。根据本CDR的特点,提出了一种在分接器后对超前、滞后信息进行统计比较得到一组低速信号来解决高速模拟电路和低速数字电路之间的接口问题。     

5Gb/s0.18μm CMOS半速率时钟与数据恢复电路设计

基于具体的系统需求,采用标准0.18μm CMOS工艺,设计了一种半速率bang-bang型时钟与数据恢复(CDR)电路。该CDR电路主要由改进型半速率鉴相器、带粗控端的环形压控振荡器(VCO)以及信道选择器等模块构成。其中,改进型半速率鉴相器通过增加四个锁存器,不但能获得较好的鉴相性能,还能使分接输出的两路数据自动实现相位对齐。带粗控端的环形VCO能够解决高振荡频率范围需求与低调谐增益需求之间的矛盾。信道选择器则能解决信道交叉出错问题。仿真结果表明,电路工作正常,在1.8V电压下,电路功耗为140mW,恢复出的时钟和数据抖动峰峰值分别为3.7ps和5ps。     

适用于连续数据速率CDR的相位插值器研制

通过对相位插值器电路进行建模分析,得到了相位插值器的线性度与输入信号之间相位差、输入信号上升时间和输出节点时间常数的关系.根据分析得到的结论,提出了一种新型的应用于连续数据速率时钟数据恢复电路的相位插值器,通过在相位插值器之前插入延时可控的缓冲器,使其输入信号的上升时间可以跟踪数据速率的改变,在保证线性度的同时,降低电路的噪声敏感度和功耗.芯片采用Charlerd 0.13 μm低功耗1.5/3.3 V工艺流片验证,面积为0.02 mm2,数据速率3.125 Gb/s时,功耗为8.5 mW.     

一种环路带宽自适应调整的时钟数据恢复电路

针对SONTE OC-192、PCIE3.0、USB3.2等协议在串行时钟数据恢复时对抖动容限、环路稳定时间的要求,提出了一种环路带宽自适应调整、半速率相位插值的时钟数据恢复电路(CDR)。设计了自适应控制电路,能适时动态调整环路带宽,实现串行信号时钟恢复过程中环路的快速稳定,提高了时钟数据恢复电路抖动容限。增加了补偿型相位插值控制器,进一步降低了数据接收误码率。该CDR电路基于55 nm CMOS工艺设计,数据输入范围为8～11.5 Gbit/s。采用随机码PRBS31对CDR电路的仿真测试结果表明,稳定时间小于400 ns,输入抖动容限大于0.55UI@10 MHz,功耗小于23 mW。     

一种500 Mbps至4 Gbps连续速率的多模式CDR电路

提出了一种连续速率的时钟数据恢复(CDR)电路,可覆盖500 Mbps到4 Gbps数据率。该CDR电路在130 nm互补金属氧化物半导体(CMOS)工艺下实现,基于相位插值(PI)原理,采用数字投票电路和相位控制逻辑替代电荷泵和模拟滤波器以方便工艺移植。为缩小片上锁相环(PLL)输出时钟频率范围,同时避免PI电路处于非线性区,该CDR电路采用多种速率模式切换的方式将采样时钟频率限定在500 MHz~1 GHz之间。PI电路为7 bit精确度,线性度良好,4 Gbps数据率时,恢复时钟的峰峰值抖动约为25.6 ps。该CDR误码率在10-10以下,可跟踪最大±976.6 ppm的数据频偏,功耗约为13.28 mW/Gbps,测试芯片大小为5 mm2,其中CDR芯核部分为0.359 mm2。     

1.25～3.125 Gb/s连续数据速率CDR设计

设计了一款工作速率为1.25～3.125 Gb/s的连续可调时钟数据恢复(CDR)电路,可以满足多种通信标准的设计需求.CDR采用相位插值型双环路结构,使系统可以根据应用需求对抖动抑制和相位跟踪能力独立进行优化.针对低功耗和低噪声的需求,提出一种新型半速率采样判决电路,利用电流共享和节点电容充放电技术,数据速率为3.125 Gb/s时,仅需要消耗50 μA电流.芯片采用0.13 μm工艺流片验证,面积0.42 m㎡,功耗98 mw,测试结果表明,时钟数据恢复电路接收PRBS7序列时,误码率小于10-12.     

2.488 Gbit/s时钟数据恢复电路的设计

利用Cadence集成电路设计软件,基于SMIC 0.18 μm 1P6M CMOS工艺,设计了一款2.488 Gbit/s三阶电荷泵锁相环型时钟数据恢复(CDR)电路.该CDR电路采用双环路结构实现,为了增加整个环路的捕获范围及减少锁定时间,在锁相环(PLL)的基础上增加了一个带参考时钟的辅助锁频环,由锁定检测环路实时监控频率误差实现双环路的切换.整个电路由鉴相器、鉴频鉴相器、电荷泵、环路滤波器和压控振荡器组成.后仿真结果表明,系统电源电压为1.8V,在2.488 Gbit/s速率的非归零(NRZ)码输入数据下,恢复数据的抖动峰值为14.6 ps,锁定时间为1.5μs,功耗为60 mW,核心版图面积为566 μm×448μm.     

基于DLL的3.5GHz时钟校准电路设计

设计了一种基于某65 nm CMOS工艺的3.5 GHz时钟校准电路,应用于高速高精度DAC中。该电路采用延迟锁相环结构,优化DAC内部的数字和模拟通路时钟信号,使数据在3.5 GHz速率下完成正确转换,有效提高了系统时钟的稳定性。电源电压为1.2 V/3.3 V,时钟相位调节精度为2 ps/LSB,目标锁定相位可调,带有时钟占空比调制功能,最大功耗小于60 mW。     

一种数据时钟恢复电路的研究与设计

提出了一种支持双数据率的数据时钟恢复电路,对电路中的鉴相器、环路滤波器、压控振荡器等进行了详细的分析研究和设计.基于0.18μm CMOS工艺,在电源电压1.8V下对电路进行仿真.仿真结果显示,电路在2.7 Gb/s和1.62 Gb/s随机流下的抖动峰峰值分别为14 ps和12ps,功耗为80 mW.测试结果显示,时钟恢复电路在2.7 Gb/s和1.62 Gb/s随机流下的抖动峰峰值分别为38 ps和27 ps.     

Performance improvement in tensile-strained In/sub 0.5/Al/sub 0.5/As/In/sub x/Ga/sub 1-x/As/In/sub 0.5/Al/sub 0.5/As metamorphic HEMT

This paper proposes a In/sub 0.5/Al/sub 0.5/As/In/sub x/Ga/sub 1-x/As/In/sub 0.5/Al/sub 0.5/As (x=0.3-0.5-0.3) metamorphic high-electron mobility transistor with tensile-strained channel. The tensile-strained channel structure exhibits significant improvements in dc and RF characteristics, including extrinsic transconductance, current driving capability, thermal stability, unity-gain cutoff frequency, maximum oscillation frequency, output power, power gain, and power added efficiency.     

Very low-noise Al/sub 0.3/Ga/sub 0.7/As/Ga/sub 0.65/In/sub 0.35/As/GaAs single quantum-well pseudomorphic HEMTs

AlGaAs/GaInAs/GaAs pseudomorphic HEMTs with an InAs mole fraction as high as 35% in the channel has been successfully fabricated. The device exhibits a maximum extrinsic transconductance of 700 mS/mm. At 18 GHz, a minimum noise figure of 0.55 dB with 15.0 dB associated gain was measured. At 60 GHz, a minimum noise figure as low as 1.6 dB with 7.6 dB associated gain was also obtained. This is the best noise performance yet reported for GaAs-based HEMTs.<>     

A 12 /spl times/ 12 In/sub 0.53/Ga/sub 0.47/As-In/sub 0.52/Al/sub 0.48/As avalanche photodiode array

We report a 12 /spl times/ 12 In/sub 0.53/Ga/sub 0.47/As-In/sub 0.52/Al/sub 0.48/As avalanche photodiode (APD) array. The mean breakdown voltage of the APD was 57.9 V and the standard deviation was less than 0.1 V. The mean dark current was /spl sim/2 and /spl sim/300 nA, and the standard deviation was /spl sim/0.19 and /spl sim/60 nA at unity gain (V/sub bias/ = 13.5 V) and at 90% of the breakdown voltage, respectively. External quantum efficiency was above 40% in the wavelength range from 1.0 to 1.6 /spl mu/m. It was /spl sim/57% and /spl sim/45% at 1.3 and 1.55 /spl mu/m, respectively. A bandwidth of 13 GHz was achieved at low gain.     

Al/sub 0.3/Ga/sub 0.7/As/In/sub x/Ga/sub 1-x/As (0/spl les/x/spl les/0.25) doped-channel field-effect transistors (DCFET's)

The properties of both lattice-matched and strained doped-channel field-effect transistors (DCFET's) have been investigated in AlGaAs/In/sub x/Ga/sub 1-x/As (0/spl les/x/spl les/0.25) heterostructures with various indium mole fractions. Through electrical characterization of grown layers in conjunction with the dc and microwave device characteristics, we observed that the introduction of a 150-/spl Aring/ thick strained In/sub 0.15/Ga/sub 0.85/As channel can enhance device performance, compared to the lattice-matched one. However, a degradation of device performance was observed for larger indium mole fractions, up to x=0.25, which is associated with strain relaxation in this highly strained channel. DCFET's also preserved a more reliable performance after biased-stress testings.<>     

Morphological and electrical characterization of SixCryCzBv thin films

Si_xCr_yC_zB_v thin films with several compositions have been studied for integration of high precision resistors in 0.8 μm BICMOS technology. These resistors, integrated in the back-end of line, have the advantage to provide high level of integration and attractive electrical behavior in temperature, for analog devices. The film morphology and the structure have been investigated through transmission electron microscopy analysis and have been then related to the electrical properties on the base of the percolation theory. According to this theory, and in agreement with experimental results, negative thermal coefficient of resistance (TCR) has been obtained for samples with low Cr content, corresponding to a crystalline volume fraction below the percolation threshold.Samples with higher Cr content exhibit, instead, a variation of the TCR as a function of film thickness: negative TCR values are obtained for thickness lower than 5 nm, corresponding to a crystalline volume fraction below the percolation threshold; positive TCR are obtained for larger thickness, indicating the establishment of a continuous conductive path between the Cr rich grains. This property seems to be determinant in order to assure the possibility to obtain thin film resistors almost independent on the temperature.     

Highly linear Al/sub 0.3/Ga/sub 0.7/N-Al/sub 0.05/Ga/sub 0.95/N-GaN composite-channel HEMTs

We report an Al/sub 0.3/Ga/sub 0.7/N-Al/sub 0.05/Ga/sub 0.95/N-GaN composite-channel HEMT with enhanced linearity. By engineering the channel region, i.e., inserting a 6-nm-thick AlGaN layer with 5% Al composition in the channel region, a composite-channel HEMT was demonstrated. Transconductance and cutoff frequencies of a 1 /spl times/100 /spl mu/m HEMT are kept near their peak values throughout the low- and high-current operating levels, a desirable feature for linear power amplifiers. The composite-channel HEMT exhibits a peak transconductance of 150 mS/mm, a peak current gain cutoff frequency (f/sub T/) of 12 GHz and a peak power gain cutoff frequency (f/sub max/) of 30 GHz. For devices grown on sapphire substrate, maximum power density of 3.38 W/mm, power-added efficiency of 45% are obtained at 2 GHz. The output third-order intercept point (OIP3) is 33.2 dBm from two-tone measurement at 2 GHz.     

Comparison between Pt/TiO2/Pt and Pt/TaOX/TaOY/Pt based bipolar resistive switching devices

Nonvolatile memories have emerged in recent years and have become a leading candidate towards replacing dynamic and static random-access memory devices. In this article, the performances of TiO₂ and TaO₂ nonvolatile memristive devices were compared and the factors that make TaO₂ memristive devices better than TiO₂ memristive devices were studied. TaO₂ memristive devices have shown better endurance performances (10⁸ times more switching cycles) and faster switching speed (5 times) than TiO₂ memristive devices. Electroforming of TaO₂ memristive devices requires~4.5 times less energy than TiO₂ memristive devices of a similar size. The retention period of TaO₂ memristive devices is expected to exceed 10 years with sufficient experimental evidence. In addition to comparing device performances, this article also explains the differences in physical device structure, switching mechanism, and resistance switching performances of TiO₂ and TaO₂ memristive devices. This article summarizes the reasons that give TaO₂ memristive devices the advantage over TiO₂ memristive devices, in terms of electroformation, switching speed, and endurance.     

Electrooptic ferroelectric Na/sub 0.5/K/sub 0.5/NbO/sub 3/ films

We report on waveguiding and electrooptic properties of epitaxial Na/sub 0.5/K/sub 0.5/NbO/sub 3/ films grown by radio-frequency magnetron sputtering on Al/sub 2/O/sub 3/(11_02) single crystal substrates. High optical waveguiding performance has been demonstrated in infrared and visible light. The in-plane electrooptic effect has been recorded in transmission using a transverse geometry. At dc fields, the effective linear electrooptic coefficient was determined to 28 pm/V, which is promising for modulator applications.     

A 1 cm/spl times/1 cm In/sub 0.53/Ga/sub 0.47/As-In/sub 0.52/Al/sub 0.48/As avalanche photodiode array

We report a 1 cm/spl times/1 cm array of 100 In/sub 0.53/Ga/sub 0.47/As-In/sub 0.52/Al/sub 0.48/As avalanche photodiodes (APD). The average breakdown voltage was 28.7 V with a standard deviation of less than 0.5 V. The distribution of breakdown voltage across the area followed a radial pattern consistent with a slight epitaxial growth nonuniformity. The mean dark current at a gain of 10, or 6.1 A/W, was 10.3 nA, and none of the 100 APDs had a dark current of more than 25 nA. The bandwidth at a gain of 10 was 6.2 GHz, and the maximum gain-bandwidth product was 140 GHz. This technology is ideally suited for next-generation three-dimensional imaging applications.