复旦大学：《微电子学与固体电子学》教学资源（硕士士学位论文）数字电视调谐器中的信道选择滤波器设计.pdf_大学文库

I 目录图目录 ····························································································III 表目录 ·····························································································V 摘要 ···························································································VII Abstract ·························································································IX 第一章绪论 ····················································································1 1.1 研究背景 ··············································································1 1.2 研究内容 ··············································································2 1.3 论文组织 ··············································································3 第二章系统指标分析与设计·······························································5 2.1 接收机输入信号特性 ·······························································5 2.1.1 接收机载噪比 ·································································5 2.1.2 最小输入功率和最大输入功率 ············································6 2.2 模拟基带性能指标 ··································································8 2.2.1 增益控制 ·······································································8 2.2.2 噪声系数和线性度 ··························································11 2.2.3 滤波器设计指标 ·····························································15 2.3 滤波器的系统考虑 ·································································16 2.3.1 滤波器的类型和结构选择 ·················································16 2.3.2 滤波器级联 ···································································17 2.3.3 增益控制和带宽调节 ·······················································18 第三章线性度和噪声的优化设计························································21 3.1 系统非线性的表征 ·································································21 3.1.1 非线性时不变系统的级联 ·················································21 3.1.2 三阶交调的伏尔特拉级数表示 ···········································22 3.2 双二次结构的线性度和噪声 ·····················································22 3.2.1 双二次结构的线性度 ·······················································22 3.2.2 双二次结构的噪声 ··························································28 3.2.3 线性度和噪声的优化及参数确定········································29 3.3 不同带宽下的线性度和噪声 ·····················································31 第四章非理想特性··········································································35 4.1 前言 ···················································································35

II 4.2 非理想效应 ··········································································35 4.2.1 非理想积分器模型 ··························································35 4.2.2 非理想双二次结构的传递函数 ···········································37 4.3 电路设计与改进 ····································································39 4.3.1 增益误差的减小及有效品质因数的矫正·······························39 4.3.2 跨导放大器线性度模型及设计 ···········································42 第五章频率校准与直流消除······························································47 5.1 滤波器参数稳定性 ·································································47 5.2 片上频率校准电路回顾 ···························································48 5.2.1 基于锁相环的校准 ··························································49 5.2.2 基于开关电容技术的校准 ·················································50 5.2.3 基于计数器和数字逻辑的校准 ···········································50 5.3 改进的频率校准方法 ······························································51 5.3.1 电路结构 ······································································51 5.3.2 误差分析 ······································································53 5.4 直流失调消除 ·······································································55 5.4.1 接收机中的直流失调 ·······················································55 5.4.2 直流失调消除电路 ··························································56 第六章芯片实现及测试结果······························································59 6.1 电路实现 ·············································································59 6.2 测试结果与对比 ····································································60 6.2.1 线性度测试结果 ·····························································60 6.2.2 频带选择测试结果 ··························································61 6.2.3 增益及带内纹波测试结果 ·················································62 6.2.4 性能总结与对比 ·····························································63 第七章总结与展望··········································································65 7.1 总结 ···················································································65 7.2 研究展望 ·············································································65 参考文献···························································································67 致谢 ·······························································································71

III 图目录图 2-1 MBRAI 测试标准中接收机系统框图.................................................... 5 图 2-2 接收机输入信号功率谱 ....................................................................... 7 图 2-3 接收机的增益分配和各级载噪比变化.................................................. 8 图 2-4 模拟基带输入信号功率谱及信道选择.................................................. 9 图 2-5 阻带衰减对功率百分比的影响 .......................................................... 10 图 2-6 邻道抑制对功率百分比的影响 .......................................................... 10 图 2-7 信噪失真比随输入功率的变化 .......................................................... 11 图 2-8 基带噪声和交调失真对系统信噪失真比的影响 ................................. 13 图 2-9 基带噪声系数对前端噪声系数指标的影响......................................... 14 图 2-10 基带三阶交调对前端三阶交调指标的影响 ........................................ 14 图 2-11 Tow-Thomas II 双二次结构电路图 ................................................... 16 图 2-12 零极点配对方式................................................................................ 17 图 2-13 六阶滤波器系统方案......................................................................... 18 图 2-14 双二次结构可变增益和带宽的实现 ................................................... 19 图 3-1 伏尔特拉算子对输入信号的作用....................................................... 21 图 3-2 非线性系统和线性系统的级联 .......................................................... 21 图 3-3 双二次结构非线性分析模型.............................................................. 23 图 3-4 双二次结构等效非线性级联系统....................................................... 23 图 3-5 双二次结构小信号半边等效电路图 ................................................... 24 图 3-6 线性度理论计算值和仿真值的比较 ................................................... 27 图 3-7 双二次结构噪声分析模型.................................................................. 28 图 3-8 噪声理论计算值和仿真值的比较....................................................... 30 图 3-9 双二次结构的线性度和噪声.............................................................. 31 图 3-10 滤波器带宽的连续调节..................................................................... 33 图 4-1 理想积分器及其信号流图.................................................................. 35 图 4-2 非理想积分器信号流图 ..................................................................... 36 图 4-3 非理想双二次结构信号流图.............................................................. 37 图 4-4 Ra 和 Rc 电阻阵列的传统实现电路 .................................................... 40 图 4-5 Rc 电阻阵列改进方案一 .................................................................... 41 图 4-6 Rc 电阻阵列改进方案二 .................................................................... 41

IV 图 4-7 两种跨导放大器的频率响应.............................................................. 43 图 4-8 两种跨导放大器的功耗比较.............................................................. 43 图 4-9 前馈零点补偿型跨导放大器电路原理图 ............................................ 44 图 4-10 差分跨导单元 ................................................................................... 44 图 4-11 前馈零点补偿型跨导放大器的非线性模型 ........................................ 45 图 4-12 前馈零点补偿型跨导放大器开环频率响应仿真结果 .......................... 46 图 5-1 滤波器截止频率偏移造成的影响....................................................... 48 图 5-2 使用 VCF 和 VCO 的锁相环路频率校准电路 .................................... 49 图 5-3 基于计数器的频率校准电路.............................................................. 50 图 5-4 改进的频率校准电路......................................................................... 51 图 5-5 频率校准电路的充电计数过程 .......................................................... 51 图 5-6 自动频率校准的工作过程.................................................................. 52 图 5-7 计数误差分析 ................................................................................... 53 图 5-8 直接变频接收机中的三种自混现象 ................................................... 55 图 5-9 跨导放大器等效输入失调电压 .......................................................... 56 图 5-10 直流失调消除电路 ............................................................................ 57 图 5-11 直流失调消除环路的高通特性 .......................................................... 57 图 5-12 直流失调下滤波器的频率响应 .......................................................... 58 图 6-1 信道选择滤波器电路系统框图 .......................................................... 59 图 6-2 调谐器模拟基带芯片照片.................................................................. 59 图 6-3 双音测试频谱分析仪截图.................................................................. 60 图 6-4 0 dB 增益下的双音测试结果............................................................. 60 图 6-5 频带选择特性测试结果 ..................................................................... 61 图 6-6 增益范围和带内纹波测试结果 .......................................................... 62 图 6-7 不同增益模式下的增益误差.............................................................. 62

摘要移动数字电视调谐器需要具有出色的信道选择特性，以保证其具有从众多干扰信号中将有用信号筛选出来的功能。本文围绕线性度和噪声优化、多带宽、可变增益、截止频率校准等几方面，介绍了一款用于数字电视调谐器中的高线性度，增益可调，带宽可调的信道选择滤波器。本文完成的主要工作如下：首先，本文详细讨论了如何对滤波器参数进行优化的设计方法。运用参数归一化的方法，将多变量的最优化问题转换成单变量的最优化问题。这不仅简化了优化过程，同时也使得结果更富有参考意义。基于这一方法，对有源RC双二次结构级联滤波器的噪声和线性度进行优化设计。其次，设计实现了带宽和增益可调的功能并解决了由非理想效应所引起的增益误差和带内纹波问题。多标准多频段的应用要求信道选择滤波器具有带宽可调的特性：而输入信号较大的动态范围需要模拟基带提供足够的增益变化范围以优化输出信号的信噪比。设计中通过切换电阻阵列中的电阻和电容阵列中的电容来实现多频带的选择和增益调节。改进的电阻匹配阵列使增益误差在整个增益范围内得到有效控制。根据不同增益对分段电阻的调节来补偿品质因素的恶化减小了高增益下的带内纹波。双二次结构中通过运用高增益、高线性度的前馈零点补偿型跨导放大器来增加滤波器的线性度，同时避免了额外补偿电容的使用，减小了芯片面积。最后，对原有频率校准电路进行了改进。改进后的频率校准电路使得对由于工艺偏差所造成的频率偏移的校准更加高效。所设计的直流偏移消除电路克服了直接变频接收机中存在的直流失调问题。测试结果显示，滤波器的带内P3在0dB增益时超过31dBm,增益范围 0~54dB,增益步长6dB,频率范围从0.25MHz~4MHz连续可调。高增益时，带内纹波小于1.4dB,增益误差和频率校准误差分别小于3.4%和5%。该设计在 0.18-m CMOS工艺下制造，1.8伏供电电压下功耗为12.6mW,面积1.28mm2。关键词：信道选择滤波器、高线性度、增益可调、多频带、连续可调、数字电视调谐器中图分类号：TN4 VII

VII 摘要移动数字电视调谐器需要具有出色的信道选择特性，以保证其具有从众多干扰信号中将有用信号筛选出来的功能。本文围绕线性度和噪声优化、多带宽、可变增益、截止频率校准等几方面，介绍了一款用于数字电视调谐器中的高线性度，增益可调，带宽可调的信道选择滤波器。本文完成的主要工作如下：首先，本文详细讨论了如何对滤波器参数进行优化的设计方法。运用参数归一化的方法，将多变量的最优化问题转换成单变量的最优化问题。这不仅简化了优化过程，同时也使得结果更富有参考意义。基于这一方法，对有源 RC 双二次结构级联滤波器的噪声和线性度进行优化设计。其次，设计实现了带宽和增益可调的功能并解决了由非理想效应所引起的增益误差和带内纹波问题。多标准多频段的应用要求信道选择滤波器具有带宽可调的特性；而输入信号较大的动态范围需要模拟基带提供足够的增益变化范围以优化输出信号的信噪比。设计中通过切换电阻阵列中的电阻和电容阵列中的电容来实现多频带的选择和增益调节。改进的电阻匹配阵列使增益误差在整个增益范围内得到有效控制。根据不同增益对分段电阻的调节来补偿品质因素的恶化减小了高增益下的带内纹波。双二次结构中通过运用高增益、高线性度的前馈零点补偿型跨导放大器来增加滤波器的线性度，同时避免了额外补偿电容的使用，减小了芯片面积。最后，对原有频率校准电路进行了改进。改进后的频率校准电路使得对由于工艺偏差所造成的频率偏移的校准更加高效。所设计的直流偏移消除电路克服了直接变频接收机中存在的直流失调问题。测试结果显示，滤波器的带内 IIP3 在 0 dB 增益时超过 31 dBm，增益范围 0~54 dB，增益步长 6 dB，频率范围从 0.25 MHz~4 MHz 连续可调。高增益时，带内纹波小于 1.4 dB，增益误差和频率校准误差分别小于 3.4%和 5%。该设计在 0.18-μm CMOS 工艺下制造，1.8 伏供电电压下功耗为 12.6 mW，面积 1.28 mm2。关键词：信道选择滤波器、高线性度、增益可调、多频带、连续可调、数字电视调谐器中图分类号：TN4

Abstract The function that choosing wanted signal from numerous interferences of mobile TV Tuner is guaranteed by its excellent channel selection property. Focusing on issues such as linearity and noise optimization,multiple bandwidth, variable gain,cutoff frequency tuning and so on,this thesis presents a channel select filter used in TV Tuner with high linearity,variable gain and adjustable bandwidth characteristic. The main works of this thesis are as following: First,the design methodology for optimizing filter parameters is discussed in detail.A multivariable optimization problem is converted to a univariate optimization problem by using parameter normalization approach.This method not only simplifies the optimization process but also gets a more intuitionistic result.Linearity and noise performance of Active-RC cascade biquad is optimized basing on this method. Second,the design realizes adjustable bandwidth and gain and solves the problem of gain error and in band ripple caused by non-ideal effect. Multi-standard and multi-band application requires adjustable band width of the channel select filter.While the big dynamic range of input power requires the base band to have enough gain variation range for optimization the SNR (Signal Noise Ratio)of output signal.Band selection and gain adjustment are realized by switching resistors in resistor arrays and capacitors in capacitor arrays in this design.The revised match resistor array makes gain error effectively controlled in the whole gain range.Tuning segmented resistor according different gain to compensate Q (Quality)factor degradation decreases in band ripple under high gain.FFZC (Feed Forward Zero Compensation)OTA with high gain and high linearity is used in biquad to increase the linearity performance of filter while avoiding extra compensation capacitors use. Third,improve the frequency tuning circuit.The correction on frequency shift due to process variation is more efficient by using the improved frequency tuning circuit.DC offset problem in DCR (Direct Conversion Receiver)is prohibited by DC offset cancellation circuit. X

IX Abstract The function that choosing wanted signal from numerous interferences of mobile TV Tuner is guaranteed by its excellent channel selection property. Focusing on issues such as linearity and noise optimization, multiple bandwidth, variable gain, cutoff frequency tuning and so on, this thesis presents a channel select filter used in TV Tuner with high linearity, variable gain and adjustable bandwidth characteristic. The main works of this thesis are as following: First, the design methodology for optimizing filter parameters is discussed in detail. A multivariable optimization problem is converted to a univariate optimization problem by using parameter normalization approach. This method not only simplifies the optimization process but also gets a more intuitionistic result. Linearity and noise performance of Active-RC cascade biquad is optimized basing on this method. Second, the design realizes adjustable bandwidth and gain and solves the problem of gain error and in band ripple caused by non-ideal effect. Multi-standard and multi-band application requires adjustable band width of the channel select filter. While the big dynamic range of input power requires the base band to have enough gain variation range for optimization the SNR (Signal Noise Ratio) of output signal. Band selection and gain adjustment are realized by switching resistors in resistor arrays and capacitors in capacitor arrays in this design. The revised match resistor array makes gain error effectively controlled in the whole gain range. Tuning segmented resistor according different gain to compensate Q (Quality) factor degradation decreases in band ripple under high gain. FFZC (Feed Forward Zero Compensation) OTA with high gain and high linearity is used in biquad to increase the linearity performance of filter while avoiding extra compensation capacitors use. Third, improve the frequency tuning circuit. The correction on frequency shift due to process variation is more efficient by using the improved frequency tuning circuit. DC offset problem in DCR (Direct Conversion Receiver) is prohibited by DC offset cancellation circuit

第一章绪论第一章绪论 1.1研究背景随着无线通信技术的发展，越来越多的频段被划分给不同的应用。数字电视信号被划分在超高频Very High Frequency,VHF)、甚高频(Ultra-High Frequency,UHF)和L波段三个频段[1]。其中每一个频段又包括多个频道。无线信道这种多频段多频道信号共存的现象使得接收机在接收数字电视信号的过程中，需要滤除其它频道或者频段的干扰信号，提取出所需频道的有用信号，以方便后面数字解调模块对信号的进一步处理。接收机的这种信道选择能力称为选择性，一般通过带通滤波器的带通特性来实现。而完成接收机信道选择功能的滤波器就被称为信道选择滤波器。信道选择滤波器通带和阻带之间的滚降就决定了接收机的选择性。在所需频道信号很小、邻道存在强干扰信号的情况下，为了防止阻塞及交调的产生，接收机除了应具有较大的动态范围以外，还必须具有很好的选择性，将强干扰信号滤除。因多种原因，各国各地区的数字电视标准不尽相同。协议的多样性决定了各地数字电视信号带宽的多样性。甚至同一协议中也存在多种信号带宽的应用 [1]3]。无线终端的发展，使多频段、多协议、可配置的接收机成为趋势。因其架构简单、灵活性高、功耗低等优点，直接变频架构成为移动、便携式接收机的主流架构4][5]。在此架构中，低通滤波器取代带通滤波器成为完成信道选择功能的模块。为了能兼容各种标准，信道选择滤波器的带宽需要随信号带宽的改变而改变。带宽的可调节性成为多频段、多带宽接收机的一个重要特征。实现信道选择滤波器的方法有多种。从集成的角度来看，可以分三类。一种是片外滤波，另一种是片上滤波，第三种则是两者的结合6][8]。片外滤波可以在射频前端使用声表面波(Surface Acoustic Wave,SAW)滤波器。声表面波滤波器具有很好的线性度、很好的选择性和低噪声，且不消耗功耗。但是这种实现方式的成本很高，而且集成度小，增加了印刷电路板(Printed Circuit Board,PCB) 的面积和元件数目。更重要的是，这种滤波器的带宽是固定的，并不适合于信号带宽可变的情况。由于处于射频前端，声表面波滤波器对邻道强干扰信号的滤除作用并不明显。片上实现信道选择滤波器的方式有多种，从1978年第一个全集成滤波器的诞生[9]到1985年第一篇实现截止频率自动校准的论文[10]，集成滤波器的实现技术在三十多年的发展中已经日趋成熟。射频前端使用预选择带通滤波器、低中频架构中模拟基带使用带通滤波器以及直接变频接收机中使用低通滤波器等方案已经在商用产品和论文中频繁出现。带宽连续可调、增益可变、低噪

第一章绪论 1 第一章绪论 1.1 研究背景随着无线通信技术的发展，越来越多的频段被划分给不同的应用。数字电视信号被划分在超高频 (Very High Frequency ， VHF) 、甚高频 (Ultra-High Frequency，UHF)和 L 波段三个频段[1]。其中每一个频段又包括多个频道。无线信道这种多频段多频道信号共存的现象使得接收机在接收数字电视信号的过程中，需要滤除其它频道或者频段的干扰信号，提取出所需频道的有用信号，以方便后面数字解调模块对信号的进一步处理。接收机的这种信道选择能力称为选择性，一般通过带通滤波器的带通特性来实现。而完成接收机信道选择功能的滤波器就被称为信道选择滤波器。信道选择滤波器通带和阻带之间的滚降就决定了接收机的选择性。在所需频道信号很小、邻道存在强干扰信号的情况下，为了防止阻塞及交调的产生，接收机除了应具有较大的动态范围以外，还必须具有很好的选择性，将强干扰信号滤除。因多种原因，各国各地区的数字电视标准不尽相同。协议的多样性决定了各地数字电视信号带宽的多样性。甚至同一协议中也存在多种信号带宽的应用 [1]-[3]。无线终端的发展，使多频段、多协议、可配置的接收机成为趋势。因其架构简单、灵活性高、功耗低等优点，直接变频架构成为移动、便携式接收机的主流架构[4][5]。在此架构中，低通滤波器取代带通滤波器成为完成信道选择功能的模块。为了能兼容各种标准，信道选择滤波器的带宽需要随信号带宽的改变而改变。带宽的可调节性成为多频段、多带宽接收机的一个重要特征。实现信道选择滤波器的方法有多种。从集成的角度来看，可以分三类。一种是片外滤波，另一种是片上滤波，第三种则是两者的结合[6]-[8]。片外滤波可以在射频前端使用声表面波(Surface Acoustic Wave，SAW)滤波器。声表面波滤波器具有很好的线性度、很好的选择性和低噪声，且不消耗功耗。但是这种实现方式的成本很高，而且集成度小，增加了印刷电路板(Printed Circuit Board，PCB) 的面积和元件数目。更重要的是，这种滤波器的带宽是固定的，并不适合于信号带宽可变的情况。由于处于射频前端，声表面波滤波器对邻道强干扰信号的滤除作用并不明显。片上实现信道选择滤波器的方式有多种，从 1978 年第一个全集成滤波器的诞生[9]到 1985 年第一篇实现截止频率自动校准的论文[10]，集成滤波器的实现技术在三十多年的发展中已经日趋成熟。射频前端使用预选择带通滤波器、低中频架构中模拟基带使用带通滤波器以及直接变频接收机中使用低通滤波器等方案已经在商用产品和论文中频繁出现。带宽连续可调、增益可变、低噪