时钟抖动对射频脉宽调制性能的影响分析

在全数字发信机系统中,射频脉宽调制(RF-PWM)将基带调制信号的幅度与相位信息编码为输出脉冲的宽度和位置。由于数字信号处理器件的非理想特性,其时钟信号的上升沿和下降沿存在抖动误差,影响RF-PWM的输出信号质量。基于3种RF-PWM实现方案,本文通过公式推导确定了时钟抖动引入的非线性失真项,并给出了时钟抖动影响下不同方案输出脉冲信号底噪的数学解析式。最后利用Matlab软件,对不同方案在时钟抖动条件下的基波、奇次谐波和底噪进行仿真验证,结果证明理论推导正确;同时对信号的矢量幅度误差(EVM)和邻信道功率比(ACPR)进行仿真,分析出时钟抖动对信号带内外性能的影响。结果表明,时钟抖动引入的非线性失真主要体现为底噪的抬高;不同RF-PWM实现方案时钟抖动的影响特性各有不同,其中五电平方案对时钟抖动影响具有抑制效果,且随时间分辨力的增大而增大。     

一种用于光学条纹相机的种子源设计

介绍了一种基于锁相环及频率合成方法产生高重复频率正弦同步扫描种子源产生技术。利用锁相环实现了正弦信号与触发光脉冲的同步跟踪, 并通过频率合成实现对正弦小信号的频率、相位、幅度的调制。调制相位可实现扫描时间的延迟, 调节振幅可实现不同扫描速度。电路系统进行了实验测试, 获得频率可达250 MHz、时间抖动小于10 ps的稳定正弦同步扫描种子源, 证明设计达到了预期目标, 满足光学条纹相机对种子源频率、幅度、抖动的高精度需求。     

PulseCore用扩谱法减少电磁干扰

1998年成立于加州San Jose的PulseCore半导体公司用一个扫频正弦信号对时钟信号进行调制,把能量集中的频谱变成分散在以时钟频率为中心的一个较宽的频带上,幅射出去的总能量不变,但频谱幅度降低了,从而减少时钟信号产生的电磁干扰。     

基于不同Q值F-P滤波器的40GHz全光帧时钟提取

对用不同Q值的F-P滤波器提取40 GHz光帧时钟的方案进行了实验比较.实验表明,用高Q值F-P滤波器提取的帧时钟信号好,但是建立、消失时间长;用低Q值F-P滤波器提取的帧时钟,信号质量较差,但可以保证帧时钟的快速建立和快速消失.实验还表明,半导体放大器(SOA)的自增益调制效应(SGM)可以有效地提高时钟的建立速度,对于时钟信号的抖动也有一定程度的改善.抖动的改善与滤波器的Q值有直接的关系.用上述方案对40 GHz的帧信号进行了全光帧时钟提取,得到了建立时间为15个信号周期,消失时间为115个信号周期,抖动为2.35 ps的帧时钟信号.     

低功率抽运光纤参量振荡器的时钟提取抖动性能

采用两次光调制技术将非归零(NRZ)码数据转换成归零(RZ)码光信号后,利用光纤参量振荡器(FOPO)结构实现了较低光功率抽运下的参量波长转换和时钟提取功能。实验表明,对于波长转换间隔为1.6nm的10Gbit/s时钟提取,优化的输入抽运光功率范围是8～14dBm;当输入信号的幅度抖动和相位抖动分别大于2.28mV和3.5ps时,该时钟提取系统可实现抖动抑制功能,其输入/输出抖动转移曲线斜率约为0.29和0.16。     

基于DDS的SDH时钟抖动信号源的设计

提出了基于DDS的带抖动时钟信号产生算法,介绍了其实现原理,分析了产生的带抖动时钟信号在频域和时域的性能,并通过仿真验证。使用该设计的抖动信号源产生带有O.172所规定的抖动频率及幅度范围的低频带抖动信号可对数字设备进行抖动性能的测试。     

PDH/SDH通信网中抖动发生及检测方法

采用DDS作为低频可控调制信号源对高频时钟信号进行调制,产生可控抖动信号源,利于锁相环技术对抖动信号进行检测和测量.实践证明,该设计方案被成功应用于PDH/SDH数字传输分析仪的抖动测试电路中,电路性能优异.     

基于DVB-C的符号同步电路优化设计

根据DVB-C标准,采用反馈环路实现符号同步,其中时钟误差检测器选择M&M算法.通过复用将符号同步电路中时钟误差检测器和环路滤波器的乘法器数目由六个减少到两个,减少了硬件规模.通过对所设计的符号同步环路仿真,环路能够在采样时钟抖动为200×10-6时完成对64QAM(正交幅度调制)信号的符号同步,并能在较短时间内达到收敛.     

雷达发射机射频性能分析

现代雷达发射机多半采用O型管放大链来放大固定载频信号,射频信号通过放大链会产生频率域或时间域失真。频率域失真是由放大链本身的非线性所引起的,而时间域失真主要是由幅度调制和相位调制引起的。     

雷达发射机射频性能分析

现代雷达发射机多半采用O型管放大链来放大固定载频信号，射频信号通过放大链会产生频率域或时间域失真。频率域失真是由放大链本身的非线性所引起的，而时间域失真主要是由幅度调制和相位调制引起的。     

All-Optical Clock Recovery for 40-Gb/s MZM-Generated NRZ-DPSK Signals Using a Self-Pulsating DBR Laser

All-optical clock recovery for the nonreturn-to-zero differential phase-shift-keying (NRZ-DPSK) modulation format is demonstrated experimentally at 40 Gb/s using a self-pulsating distributed Bragg reflector laser. The use of a Mach–Zehnder modulator to generate the NRZ-DPSK signal yields a modulated signal spectrum with a weak clock tone. The self-pulsating laser is able to directly recover a clock signal with a root-mean-square timing jitter of 760 fs and an extinction ratio of 13 dB; a preprocessing stage to enhance the clock tone is not required. The timing jitter of the recovered clock signal is characterized for different values of the input signal optical signal-to-noise ratio and for varying amounts of waveform distortion due to polarization-mode dispersion.      

Timing Jitter Measurement of Intrinsic Random Jitter and Sinusoidal Jitter Using Frequency Division

This paper presents a new method for measuring random timing jitter or sinusoidal timing jitter in signals of telecommunication devices. The method uses a divide-by-M circuit to reduce the frequency and the number of clock samples, and applies the Hilbert transform to measure the timing jitter. This new frequency division method is validated with experimental data from a serializer-deserializer device and a modulated signal source generating a 2.5 GHz FM signal.     

Distortion in the Frequency Demodulator Using Feedback

The response of a frequency demodulator using feedback (FMFB) to a frequency modulated signal is analyzed. Canonical equations of operation are obtained. Harmonic distortion is calculated for the case of a sinusoidal modulating signal. Intermodulation distortion is calculated assuming a noise-like modulation. Design curves are presented. The special case of harmonic and intermodulation distortion in a discriminator is also presented. It is shown that the results obtained in this paper by treating the discriminator as a degenerate FMFB compare favorably with those obtained by other authors. However, the results presented here do not require digital computation.     

Approximate dynamic model for evaluating distortion in a semiconductor laser under overmodulation

Overmodulation of a directly modulated semiconductor laser causes distortion in the optical signal. This distortion can be analyzed, in a first approach, using the model of a static limiter for the laser. In this letter, the dynamic effects in overmodulation of a semiconductor laser are addressed for a sinusoidal modulating current and an improved model is proposed by introducing an hysteresis. A time delay is taken into account for evaluating additional dynamic distortion arising when the optical intensity rises above threshold.     

Race-free clocking of CMOS pipelines using a single global clock

A novel, single-phase, race-free CMOS circuit clocking technique, which provides and uses complementary logic, is presented. The clock can be driven by a sinusoidal waveform. This avoids any requirement for transmitting the very-high-frequency components associated with fast clock edges. Such circuits are therefore less sensitive to clock distortion caused by transmission-line effects     

A 10-bit wide-band CMOS direct digital RF amplitude modulator

This paper describes a low-distortion wide-band CMOS direct digital RF amplitude modulator, which uses a 10-bit linear interpolation current-steering digital-to-analog converter (DAC) and a Gilbert-cell-based mixer to generate an amplitude modulated RF signal directly. The linear interpolation increases the attenuation of the DAC's image components. The reconstruction filter is, therefore, eliminated. The DAC's differential current signals are directly sent to the mixer, which improves the linearity of the modulated RF signal. Thus, the RF transmitter structure is simplified, and the low distortion is achieved. This modulator is suitable for system-on-chip (SOC) design and is easily scalable. The chip was fabricated in a 0.35-/spl mu/m 3.3-V double-poly triple-metal CMOS process. The core size of the chip is 0.52 mm/spl times/0.68 mm. With a 3.3-MHz modulation signal, a 50-MHz clock, and a 1-GHz carrier, the distortion components are below -53.81 dBc, and the attenuation of the image signal is 47.45 dB. The output power is -6.5 dBm, and the total power consumption is 159.8 mW.     

Multicast-enabled high-speed VCSEL technology for flexible data center networks

We experimentally demonstrate the multiple signal modulation on a single class 10 G vertical cavity surface emitting laser (VCSEL) carrier at 1 310 nm for next generation multicast-enabled data center networks. A 10 Gbit/s data signal is directly modulated onto a single mode VCSEL carrier. To maximize carrier spectral efficiency, a 2 GHz reference frequency (RF) clock tone is simultaneously modulated on the VCSEL phase attribute. The inherent VCSEL orthogonal polarization bistability with changing bias current is further exploited in transmission of a polarization based pulse per second (PPS) timing clock signal. Therefore, we simultaneously transmit a 10 Gbit/s directly modulated data, 2 GHz phase modulated RF and a polarization-based PPS clock signals using a single mode 10 GHz bandwidth VCSEL carrier. It is the first time that a single class 10 G VCSEL carrier is reported to transmit a directly modulated data, phase modulated RF clock and polarization based PPS timing signal simultaneously in a single wavelength. A of G.652 single mode fibre (SMF) transmission over 3.21 km is experimentally attained. A receiver sensitivity of ?15.60 dBm is experimentally obtained for the directly modulated 10 Gbit/s data signal. A 3.21-km-long SMF transmission introduces a penalty of 0.23 dB to the data signal. The contribution of a 2 GHz phase modulated RF and a polarization-based PPS clock signal to this penalty is found to be 0.03 dB. An RF single-side band (SSB) phase noise values of ?82.36 dBc/Hz and ?77.97 dBc/Hz are attained without and with simultaneous directly modulated data and polarization-based PPS clock signals respectively for a 3.21-km-long SMF transmission. This work provides an alternative efficient and cost effective technique for simultaneous high-speed multiple information transmission to different network nodes within a data center network through shared network infrastructure.     

应用频率积分相位解调测量径向畸变

为了测量光学成像系统的径向畸变,采用载频条纹模板,应用瞬时频率积分法提取因径向畸变而产生的径向调制相位;推导了条纹径向调制相位与瞬时频率的关系式,并导出径向调制相位和径向畸变位移关系;采用小波频率估计提取畸变条纹径向瞬时频率,并对其进行积分获得畸变条纹的径向调制相位;应用径向调制相位和立方卷积插值算法对畸变图像进行了校正,得出了详细的理论分析和实验结果。结果表明,上述方法是可行的。     

Bandpass filtering of high-speed forwarded clocks

Bandpass channel filtering is shown to attenuate both random and determinstic jitter components in high frequency clock signals. A fully differential, tunable LC bandpass filter is developed and employed to reduce Gaussian and sinusoidal distributed jitter and duty cycle distortion.     

基于等光频细分重采样的调频干涉测距方法

为了解决调频连续波(FMCW)激光器调制非线性导致的测量信号频谱展宽降低激光干涉测距精度的问题, 采用一种基于等光频细分重采样的调频干涉测距方法, 进行了理论分析和实验验证, 获得了双光路测距系统对不同位置目标信号等光频细分重采样后的波形数据, 并进行了频谱分析。结果表明, 通过等光频细分重采样的方法, 使用细分后的时钟信号点对距离大于辅助干涉光路光程差的目标测量信号进行重采样, 消除了激光器的调制非线性的影响, 并且避免了采样点数不足引起信号失真的问题; 在4.3m测量范围内, 等光频细分重采样测距系统与激光干涉仪相比最大残余误差不超过±18.46μm, 最大测量标准差为23.39μm; 该方法使用的辅助干涉光路光程差很短, 受环境的影响较小, 可以获得稳定的时钟信号, 并且可以减少双光路FMCW测距系统的体积与成本。该研究为长距离、高精度调频连续波测量提供了实用参考。