荧光检测器信号检测系统设计及杂散光研究
|
摘要
液相色谱系统是影响最大、发展最快、应用最广泛的现代分析系统之一,其广泛应用于石油化工、有机合成、生理生化、医药卫生,乃至空间探索等几乎所有的应用科学领域,而荧光检测器作为液相色谱系统的“眼睛”,是液相色谱系统的核心部件,其性能的好坏直接关系着定性定量分析结果的可靠性和准确性。针对液相色谱系统中荧光检测器主要依靠进口,国内研发落后于实际需求的现状,本文理论设计、性能调试、实验验证等方面深入的研究了荧光检测器信号检测系统的设计及其杂散光的抑制。
本文主要完成的工作包括以下几个方面：
1.荧光是一种极其微弱的光信号,而微弱信号的探测一直是检测系统设计中非常困难的问题,设计、调试中的微小疏忽,都会导致荧光的电信号被湮没在各种电学噪声中。因此,本文建立了从光电倍增管到ADC采样电路的完整电路模型,详细的对电路各处的信号、噪声的带宽以及幅值进行分析,并最终通过试验验证：本文设计调试的采样系统电路噪声仅31.7μV,光电流检测精度可达6.4pA,折合最弱光强分辨能力为25.6mlm,性能达到国外同类优秀产品水平。
2.现有的荧光检测器信号采集系统模数转换均使用△—∑调制技术的24位集成ADC,其拥有高精度、高集成度的优点,本文使用CS5532实现了常规的模数转换系统,经实验测试,系统有效数字精度达到18位。但是,高精度带来的不利因素是极低的字输出速度,并进一步制约了荧光检测器的波长扫描速度。针对上述问题,本文又提出并实现了使用高速ADC过采样,并利用硬件数字滤波模块提高信噪比的模数转换方案。新的方案可以根据系统的需要,实现60～250k任意值的字输出速度,并到达16～22位的数字有效精度。新方案的意义在于：使设计者和使用者可以灵活的选择最佳的系统速度-精度方案,克服了常规方案波长扫描速度的限制。
3.为解决常见的基于MCU的控制系统实时性差、运算速度慢、集成度低的问题,本文引入并实现了SOPC控制系统。这是一种在芯片内部集成用户自定义外设模块,实现高实时性、高运算能力、高集成度的控制系统设计方案。本文设计ADC自定义驱动,克服了传统ADC驱动信号不连续造成采样速度损失的问题,系统采样速度增快10%；设计用户自定义UART控制器,总线带宽利用率增大3倍,大大减轻CPU与总线负担；设计嵌入式硬件IIR数字滤波协处理模块,进一步提高系统信噪比,经试验验证,其能有效抑制采样系统53%的噪声。新系统众多功能集成在一个单片芯片中,大大简化外部的连接设计,减小设计、制作成本。
4.针对荧光检测器中单色器的杂散光对系统性能的影响,本文首次在ASAP中对荧光检测器中单色器进行光线追迹与杂散光分析,首次引入单色器光谱传递函数并通过实验验证理论模型的正确性。在此基础上,通过对系统光路中“关键表面”的分析,提出了利用视场光阑和孔径光阑、多级叶片结构、光栅入射光抑制结构等方案进行杂散光抑制,通过ASAP模拟,上述结构的杂散光抑制比分别达到2.7×10-57.6×10-5以及0.33×10-5,为实际改进工作提供了有益的参考。
综上所述,本文详细分析并设计了荧光信号检测系统电路；实现了高精度信号采样系统；提出一种新的精度-速度可调的模数转换方案；引入SOPC控制系统克服传统控制系统的速度与性能瓶颈；并首次利用软件模拟分析了荧光检测器单色器杂散光的来源,提出了有效的杂散光抑制方案。最终,实现了一台简单的荧光检测器样机,样机在350nm位置水的拉曼峰信噪比为63。
Liquid chromatography system is the most influential, fastest-growing and widely used modern analytical systems, which enjoys its popularity in petroleum chemical in-dustry, organic synthesis, physiology, biochemistry, medicine, health, and even space exploration and other applications in almost all fields of science. The fluorescence detector, as the "eye" of the system, is a key component of liquid chromatography system, which is crucial for reliability and accuracy of qualitative and quantitative analysis results. However, domestic fluorescence detector demand rely mainly on im-ports, for the absence of effective domestic R & D, this paper make a in-depth study of the fluorescence detector signal detection system as well as its stray light inhibition.
The main work of this paper including:
1. Fluorescence is an extremely weak optical signal, while the design of weak sig-nal detection system involves considerable carefulness of any minor electrical noise that may be introduced in from anywhere. This paper set up the model of the whole detection circuit, from the photomultiplier tubes to the ADC sampling module, and a detailed analysis of bandwidth and amplitude of the signal and noise is made and then verified by experiment. The final overall circuit noise is limited in 31.7μV,and the detection accuracy of light-current is up to 6.4pA, equivalent to 25.6mlm light detection precision. The system's performance is as equal as the similar high-level foreign products.
2. AΔ-Σtype ADC is the mainstream design in fluorescence detection signal ac-quisition system for its high precision and integration. This paper uses CS5532 ADC to successfully set up a 24 bit ADC system, whose digital accuracy reaches 18 bit. However,the low word output rate is the by-product of the high precision, which is the main restriction of the scanning speed of the system. To address this problem, this paper has proposed and implements a new ADC system based on over-sampling and matching hardware digital filter module. The new sheme could offer a 60 (～) 250k word output rate and flexible 16 (～) 22 bit effective precision according to different demands. The significant advantage of this new ADC design was to enable the flexibility of make the best compromise between system precision and wavelength scanning speed.
3. The most popular control system is based on MCU which suffers from poor real-time performance, slow operation speed and low system integration. This paper introduces and implements the SOPC control system instead of MCU, whose on-chip user-defined peripheral modules enjoys excellent realtime performance, strong computing power and high-integration. In This paper, a custom-defined ADC driver, which avoid ADC-speed-loss in traditional driving design, is de-signed to speeds up the systematic sampling rate for 10%. A user-defined UART controller which enable at-one-time 24bit data-bus transferring is designed to increase the bus bandwidth utilization by 3 times. And also, an embedded hard-ware IIR digital co-processor is designed which can effectively inhibit the noise by 53%. The new system integrates these different features in a single chip, which greatly simplifies external circuit design and reduces costs.
4. To study the influence of the stray light from monochromator on system perfor-mance, this paper, for the first time, builds up a model of the monochromator in ASAP for ray-tracing and stray light analysis, and for the first time, introduces monochromator- spectral-transfer-function to define its influence on optical sig-nal. Some experiments are made to verify the correctness of the simulation mod-els. And on the analysis of the "critical surface" in the optical path, this paper puts forward several stray light suppression structure as using field diaphragm and aperture diaphragm, multi-level leaf structure, and unique structures for grating incident light,and the simulation shows stray light rejection ratio of above structure are 2.7×10(-5),7.6×10(-5), and 0.33×10(-5)respectively, which could provide some meaningful guide for practical improvements. end (enumerate) In summary, this paper makes an in-depth study and design of the signal detec-tion system of the fluorescence detector, presents a precision-speed adjustable new ADC scheme, introduces SOPC control system which break though the speed and performance bottlenecks of traditional control system, and for the first time, analyze the stray light from fluorescence detector monochromator in software, as well as puts forward several effective suppression structurer. In the end, this paper developed a fluorescence detector prototype, whose SNR of Raman peak of water under 350nm excitation light reaches 63.
本文主要完成的工作包括以下几个方面：
1.荧光是一种极其微弱的光信号,而微弱信号的探测一直是检测系统设计中非常困难的问题,设计、调试中的微小疏忽,都会导致荧光的电信号被湮没在各种电学噪声中。因此,本文建立了从光电倍增管到ADC采样电路的完整电路模型,详细的对电路各处的信号、噪声的带宽以及幅值进行分析,并最终通过试验验证：本文设计调试的采样系统电路噪声仅31.7μV,光电流检测精度可达6.4pA,折合最弱光强分辨能力为25.6mlm,性能达到国外同类优秀产品水平。
2.现有的荧光检测器信号采集系统模数转换均使用△—∑调制技术的24位集成ADC,其拥有高精度、高集成度的优点,本文使用CS5532实现了常规的模数转换系统,经实验测试,系统有效数字精度达到18位。但是,高精度带来的不利因素是极低的字输出速度,并进一步制约了荧光检测器的波长扫描速度。针对上述问题,本文又提出并实现了使用高速ADC过采样,并利用硬件数字滤波模块提高信噪比的模数转换方案。新的方案可以根据系统的需要,实现60～250k任意值的字输出速度,并到达16～22位的数字有效精度。新方案的意义在于：使设计者和使用者可以灵活的选择最佳的系统速度-精度方案,克服了常规方案波长扫描速度的限制。
3.为解决常见的基于MCU的控制系统实时性差、运算速度慢、集成度低的问题,本文引入并实现了SOPC控制系统。这是一种在芯片内部集成用户自定义外设模块,实现高实时性、高运算能力、高集成度的控制系统设计方案。本文设计ADC自定义驱动,克服了传统ADC驱动信号不连续造成采样速度损失的问题,系统采样速度增快10%；设计用户自定义UART控制器,总线带宽利用率增大3倍,大大减轻CPU与总线负担；设计嵌入式硬件IIR数字滤波协处理模块,进一步提高系统信噪比,经试验验证,其能有效抑制采样系统53%的噪声。新系统众多功能集成在一个单片芯片中,大大简化外部的连接设计,减小设计、制作成本。
4.针对荧光检测器中单色器的杂散光对系统性能的影响,本文首次在ASAP中对荧光检测器中单色器进行光线追迹与杂散光分析,首次引入单色器光谱传递函数并通过实验验证理论模型的正确性。在此基础上,通过对系统光路中“关键表面”的分析,提出了利用视场光阑和孔径光阑、多级叶片结构、光栅入射光抑制结构等方案进行杂散光抑制,通过ASAP模拟,上述结构的杂散光抑制比分别达到2.7×10-57.6×10-5以及0.33×10-5,为实际改进工作提供了有益的参考。
综上所述,本文详细分析并设计了荧光信号检测系统电路；实现了高精度信号采样系统；提出一种新的精度-速度可调的模数转换方案；引入SOPC控制系统克服传统控制系统的速度与性能瓶颈；并首次利用软件模拟分析了荧光检测器单色器杂散光的来源,提出了有效的杂散光抑制方案。最终,实现了一台简单的荧光检测器样机,样机在350nm位置水的拉曼峰信噪比为63。
Liquid chromatography system is the most influential, fastest-growing and widely used modern analytical systems, which enjoys its popularity in petroleum chemical in-dustry, organic synthesis, physiology, biochemistry, medicine, health, and even space exploration and other applications in almost all fields of science. The fluorescence detector, as the "eye" of the system, is a key component of liquid chromatography system, which is crucial for reliability and accuracy of qualitative and quantitative analysis results. However, domestic fluorescence detector demand rely mainly on im-ports, for the absence of effective domestic R & D, this paper make a in-depth study of the fluorescence detector signal detection system as well as its stray light inhibition.
The main work of this paper including:
1. Fluorescence is an extremely weak optical signal, while the design of weak sig-nal detection system involves considerable carefulness of any minor electrical noise that may be introduced in from anywhere. This paper set up the model of the whole detection circuit, from the photomultiplier tubes to the ADC sampling module, and a detailed analysis of bandwidth and amplitude of the signal and noise is made and then verified by experiment. The final overall circuit noise is limited in 31.7μV,and the detection accuracy of light-current is up to 6.4pA, equivalent to 25.6mlm light detection precision. The system's performance is as equal as the similar high-level foreign products.
2. AΔ-Σtype ADC is the mainstream design in fluorescence detection signal ac-quisition system for its high precision and integration. This paper uses CS5532 ADC to successfully set up a 24 bit ADC system, whose digital accuracy reaches 18 bit. However,the low word output rate is the by-product of the high precision, which is the main restriction of the scanning speed of the system. To address this problem, this paper has proposed and implements a new ADC system based on over-sampling and matching hardware digital filter module. The new sheme could offer a 60 (～) 250k word output rate and flexible 16 (～) 22 bit effective precision according to different demands. The significant advantage of this new ADC design was to enable the flexibility of make the best compromise between system precision and wavelength scanning speed.
3. The most popular control system is based on MCU which suffers from poor real-time performance, slow operation speed and low system integration. This paper introduces and implements the SOPC control system instead of MCU, whose on-chip user-defined peripheral modules enjoys excellent realtime performance, strong computing power and high-integration. In This paper, a custom-defined ADC driver, which avoid ADC-speed-loss in traditional driving design, is de-signed to speeds up the systematic sampling rate for 10%. A user-defined UART controller which enable at-one-time 24bit data-bus transferring is designed to increase the bus bandwidth utilization by 3 times. And also, an embedded hard-ware IIR digital co-processor is designed which can effectively inhibit the noise by 53%. The new system integrates these different features in a single chip, which greatly simplifies external circuit design and reduces costs.
4. To study the influence of the stray light from monochromator on system perfor-mance, this paper, for the first time, builds up a model of the monochromator in ASAP for ray-tracing and stray light analysis, and for the first time, introduces monochromator- spectral-transfer-function to define its influence on optical sig-nal. Some experiments are made to verify the correctness of the simulation mod-els. And on the analysis of the "critical surface" in the optical path, this paper puts forward several stray light suppression structure as using field diaphragm and aperture diaphragm, multi-level leaf structure, and unique structures for grating incident light,and the simulation shows stray light rejection ratio of above structure are 2.7×10(-5),7.6×10(-5), and 0.33×10(-5)respectively, which could provide some meaningful guide for practical improvements. end (enumerate) In summary, this paper makes an in-depth study and design of the signal detec-tion system of the fluorescence detector, presents a precision-speed adjustable new ADC scheme, introduces SOPC control system which break though the speed and performance bottlenecks of traditional control system, and for the first time, analyze the stray light from fluorescence detector monochromator in software, as well as puts forward several effective suppression structurer. In the end, this paper developed a fluorescence detector prototype, whose SNR of Raman peak of water under 350nm excitation light reaches 63.
引文
[1]TEWETT M. Ber Dtsch Bot Ges[J]. Chemical Physics Letters,1906,24(10):316.
[2]李彤,张庆合,张维冰.高效液相色谱仪器系统[M].北京：化学工业出版社,2005.
[3]云自厚,欧阳津,张晓彤.液相色谱检测方法[M].北京：化学与应用化学出版中心,2005.
[4]卢佩章,张玉奎,梁鑫淼.高效液相色谱及其专家系统[M].沈阳：辽宁科学技术出版社,1992.
[5]HAKAMATA T. Photomultiplier tubes basics and applicaitons[M]. 静岡県浜松 巿:Hamamatsu Photonics K.K. elctron tube division,2006.
[6]FLYCKT S O, MARMONIER C. Photomultiplier tubes principles and applica-tions [M]. Brive, France:photonis,2002.
[7]曾庆勇.微弱信号检测[M].杭州：浙江大学出版社,1994.
[8]童诗白.模拟电子技术基础[M].北京：人民教育出版社,1980.
[9]OPPENHEIM A V,刘树棠.信号与系统[M].西安：西安交通大学出版社,1985.
[10]BAKER B C. Comparision of niose performance between a FET transimpedance amplifier and a swithched integrator[R].[S.1.]: Texas Instruments,1994. http: //www.ti.com/.
[11]GRAEME J. Op amp performance analysis[R].[S.1.]:Texas Instruments,1993. http://www.ti.com/.
[12]YEAGER J, ANNE M, HEUSCH-TUPTA. Low Level Measurements Hand-book,Precision DC Current,Voltage and Resistance Measurements [M]. Cleve-land, Ohio:Keithley,1998.
[13]JUNG W G. Op Amp applications handbook[M]. Boston:Amsterdam,2005.
[14]PROAKIS J G. Digital signal processing[M]. Upper Saddle River, N.J.:Pearson Prentice Hall,2007.
[15]王建校,危建国.SOPC设计基础与实践[M].西安：西安电子科技大学出版社,2006.
[16]DRIVE A I. Avalon Interface Specifications[R].[S.1.]:altera,2009. http: //www.altera.com/.
[17]侯建军,郭勇.SOPC技术基础教程[M].北京：清华大学出版社,2008.
[18]GAN W S. Embedded Signal Processing with the Micro Signal Architecture[M]. Canada:John Wiley and Sons.Inc,2007.
[19]DRIVE A I. DSP Builder Reference Manual[R].[S.1.]:altera,2008. http: //www.altera.com/.
[20]DRIVE A I. DSP Builder Flow User Guide[R].[S.1.]:altera,2009. http: //www.altera.com/.
[21]廖胜.光学系统杂光抑制研究[D].成都：电子科技大学,2003.
[22]原育凯.光学系统杂散光的消除方法[J].大气与环境光学学报,2007,2(1)：6-10.
[23]BASS M. Handbook of Optics[M]. New York: McGRAW-HILL.inc,1995.
[24]李诗铁.分光光度计中杂散光的产生、危害及防护[J].计量与测试技术,2000,27(4)：31.
[25]PETERSON G L. Stray light calculation methods with optical ray trace soft-ware[J]. Proc. SPIE,1999,3780:132-137.
[26]BREAULT R P, ETC. ASAP 8.0 Primer[M]. Arizona: Breault Research Organi-zation,2003.
[27]HENDRIK ROTHE A K, DOROTHEE HUESER, RINDER T. Investigations of smooth surfaces by measuring the BRDF with a stray light sensor in comparison with PSD curves evaluated from topography of large AFM scans[J]. Proc. SPIE, 1999,3619:112-120.
[28]张以谟.应用光学[M].沈阳：机械工业出版社,1982.
[29]BREAULT R P. Problems and Techniques In Stray Radiation Suppression[J]. Proc. SPIE,1977,107:2-23.
[30]FRENIERE E R. First-order Design Of Optical Baffles[J]. Proc. SPIE,1987,257: 19-28.
[2]李彤,张庆合,张维冰.高效液相色谱仪器系统[M].北京：化学工业出版社,2005.
[3]云自厚,欧阳津,张晓彤.液相色谱检测方法[M].北京：化学与应用化学出版中心,2005.
[4]卢佩章,张玉奎,梁鑫淼.高效液相色谱及其专家系统[M].沈阳：辽宁科学技术出版社,1992.
[5]HAKAMATA T. Photomultiplier tubes basics and applicaitons[M]. 静岡県浜松 巿:Hamamatsu Photonics K.K. elctron tube division,2006.
[6]FLYCKT S O, MARMONIER C. Photomultiplier tubes principles and applica-tions [M]. Brive, France:photonis,2002.
[7]曾庆勇.微弱信号检测[M].杭州：浙江大学出版社,1994.
[8]童诗白.模拟电子技术基础[M].北京：人民教育出版社,1980.
[9]OPPENHEIM A V,刘树棠.信号与系统[M].西安：西安交通大学出版社,1985.
[10]BAKER B C. Comparision of niose performance between a FET transimpedance amplifier and a swithched integrator[R].[S.1.]: Texas Instruments,1994. http: //www.ti.com/.
[11]GRAEME J. Op amp performance analysis[R].[S.1.]:Texas Instruments,1993. http://www.ti.com/.
[12]YEAGER J, ANNE M, HEUSCH-TUPTA. Low Level Measurements Hand-book,Precision DC Current,Voltage and Resistance Measurements [M]. Cleve-land, Ohio:Keithley,1998.
[13]JUNG W G. Op Amp applications handbook[M]. Boston:Amsterdam,2005.
[14]PROAKIS J G. Digital signal processing[M]. Upper Saddle River, N.J.:Pearson Prentice Hall,2007.
[15]王建校,危建国.SOPC设计基础与实践[M].西安：西安电子科技大学出版社,2006.
[16]DRIVE A I. Avalon Interface Specifications[R].[S.1.]:altera,2009. http: //www.altera.com/.
[17]侯建军,郭勇.SOPC技术基础教程[M].北京：清华大学出版社,2008.
[18]GAN W S. Embedded Signal Processing with the Micro Signal Architecture[M]. Canada:John Wiley and Sons.Inc,2007.
[19]DRIVE A I. DSP Builder Reference Manual[R].[S.1.]:altera,2008. http: //www.altera.com/.
[20]DRIVE A I. DSP Builder Flow User Guide[R].[S.1.]:altera,2009. http: //www.altera.com/.
[21]廖胜.光学系统杂光抑制研究[D].成都：电子科技大学,2003.
[22]原育凯.光学系统杂散光的消除方法[J].大气与环境光学学报,2007,2(1)：6-10.
[23]BASS M. Handbook of Optics[M]. New York: McGRAW-HILL.inc,1995.
[24]李诗铁.分光光度计中杂散光的产生、危害及防护[J].计量与测试技术,2000,27(4)：31.
[25]PETERSON G L. Stray light calculation methods with optical ray trace soft-ware[J]. Proc. SPIE,1999,3780:132-137.
[26]BREAULT R P, ETC. ASAP 8.0 Primer[M]. Arizona: Breault Research Organi-zation,2003.
[27]HENDRIK ROTHE A K, DOROTHEE HUESER, RINDER T. Investigations of smooth surfaces by measuring the BRDF with a stray light sensor in comparison with PSD curves evaluated from topography of large AFM scans[J]. Proc. SPIE, 1999,3619:112-120.
[28]张以谟.应用光学[M].沈阳：机械工业出版社,1982.
[29]BREAULT R P. Problems and Techniques In Stray Radiation Suppression[J]. Proc. SPIE,1977,107:2-23.
[30]FRENIERE E R. First-order Design Of Optical Baffles[J]. Proc. SPIE,1987,257: 19-28.