滕云田， 胡星星， 王喜珍， 王曉美， 盧紅婭， 王喆， 張旸

中國(guó)地震局地球物理研究所， 北京　100081

用多通道AD分級(jí)采集擴(kuò)展地震數(shù)據(jù)采集器的動(dòng)態(tài)范圍

滕云田， 胡星星*， 王喜珍， 王曉美， 盧紅婭， 王喆， 張旸

中國(guó)地震局地球物理研究所， 北京100081

摘要地震數(shù)據(jù)采集是地震信號(hào)數(shù)字化必不可少的環(huán)節(jié)，動(dòng)態(tài)范圍是其一個(gè)重要的性能指標(biāo).實(shí)際地震信號(hào)的動(dòng)態(tài)范圍在160 dB以上，而目前普遍使用的24位地震數(shù)據(jù)采集器動(dòng)態(tài)范圍相對(duì)較小且在50 Hz采樣率時(shí)最大只達(dá)到135 dB，致使24位地震數(shù)據(jù)采集器在實(shí)際使用中對(duì)小信號(hào)分辨率不夠，不能有效提取地震信息；在大地震時(shí)又容易使數(shù)據(jù)采集器出現(xiàn)飽和限幅失真的現(xiàn)象而失去地震監(jiān)測(cè)記錄功能.本文針對(duì)在地震監(jiān)測(cè)和地震研究中需要具有高分辨率和高動(dòng)態(tài)范圍的地震數(shù)據(jù)采集器這個(gè)亟待解決的問(wèn)題，提出一種采用多通道AD轉(zhuǎn)換器并行分級(jí)采集的方法，討論了通道間失配及其標(biāo)定.對(duì)研制實(shí)驗(yàn)樣機(jī)的測(cè)試表明，其動(dòng)態(tài)范圍在50 Hz采樣時(shí)可以達(dá)到157 dB以上，線性度優(yōu)于0.005%.

關(guān)鍵詞動(dòng)態(tài)范圍； 地震數(shù)據(jù)采集器； 高分辨率； 通道失配； 標(biāo)定

In the multi-channel AD converter sample grading method, several ADC channels are put in parallel, and simulated input signals are sent to each channel. Synchronous sampling and digital conversion are made on the full range input signal in each channel with different input voltage ranges in each channel for conversion. Meanwhile, the ranges of measurement and the input range of AD converters are matched by modulating waveforms and electrical level displacements in simulated preprocessing circuits in the front ends of each channel. The results are put out in the format of 32-bits converted digital signals after processing and fitted by the CPU-control process unit and multi switch MUX.

The method of multi-channel AD converter sample grading not only improves the resolution of seismic data acquisition for minor signals, but also enables the acquisition system to record great number of seismic information without saturation in advance to prevent from amplitude limiting. This method can help meet requirements for the large dynamic ranges for data acquisition in seismic monitoring, with low costs and easily to achieve technologically.

1引言

地震數(shù)據(jù)采集系統(tǒng)是地震信號(hào)數(shù)字化必不可少的環(huán)節(jié)，其性能優(yōu)劣直接關(guān)系到最終獲得地震信號(hào)的質(zhì)量，并最終影響數(shù)據(jù)處理結(jié)果.對(duì)地震數(shù)據(jù)采集器而言，動(dòng)態(tài)范圍是一個(gè)極其重要的性能指標(biāo)(袁子龍和曹廣華，2000)，它表示采集器在把模擬地震信號(hào)轉(zhuǎn)換為數(shù)字輸出信號(hào)的過(guò)程中，所能夠不失真轉(zhuǎn)換的最大輸入信號(hào)和最小輸入信號(hào)的幅度跨度.地震觀測(cè)信號(hào)具有很大的動(dòng)態(tài)范圍：它包含了來(lái)自幾千公里之外的微弱地震、核震信號(hào)、地脈動(dòng)及地球背景噪聲等微弱信號(hào)(葛洪魁等，2013)，直到要近場(chǎng)監(jiān)測(cè)8級(jí)以上的大地震；地震勘探中需要檢測(cè)來(lái)自地層深處的微弱信號(hào)，又要采集來(lái)自地表的強(qiáng)大信號(hào)，總的信號(hào)幅度跨度超過(guò)160 dB(袁子龍和曹廣華，2000；孫嫻和羅桂娥，2008).現(xiàn)代寬頻帶地震計(jì)的噪聲水平在1 μV以下( 李威和陳祖斌，2006；Evans et al., 2010; Ringler et al., 2011)，最大輸出幅度達(dá)到了±20 V以上，動(dòng)態(tài)范圍大于140 dB，如Kinemetrics公司的寬頻帶地震計(jì)KS-2000的動(dòng)態(tài)范圍為155 dB，加拿大Nanometrics公司的Titan加速度計(jì)，動(dòng)態(tài)范圍達(dá)到165 dB.而現(xiàn)代廣為使用的基于ΔΣAD(Analog-to-Digital Convert)轉(zhuǎn)換技術(shù)的高性能24位地震數(shù)據(jù)采集器的動(dòng)態(tài)范圍只有135 dB@50SPS(Samples Per-second)，因而導(dǎo)致了在實(shí)際地震監(jiān)測(cè)中，數(shù)據(jù)采集器在記錄地震時(shí)對(duì)微小信號(hào)分辨率不夠，信噪比不高，信號(hào)幅度和數(shù)據(jù)采集器的分辨率處于同一量級(jí)，不能有效的提取地震信息；而在大地震時(shí)又容易使數(shù)據(jù)采集器出現(xiàn)飽和限幅失真的現(xiàn)象(王翠芳等，2010)，致使在抗震救災(zāi)最需要第一手觀測(cè)資料的時(shí)候這些最重要的近源地震臺(tái)站卻幾乎失去了地震監(jiān)測(cè)的功能，同時(shí)在地震研究上也失去了記錄大震信號(hào)的珍貴地震資料的機(jī)會(huì).因而具有高分辨率和高動(dòng)態(tài)范圍的地震數(shù)據(jù)采集器是當(dāng)今地震學(xué)發(fā)展領(lǐng)域中亟待解決的問(wèn)題.

自從有了現(xiàn)代計(jì)算機(jī)技術(shù)的發(fā)展，信號(hào)處理就由模擬領(lǐng)域逐漸向數(shù)字領(lǐng)域轉(zhuǎn)變，而模數(shù)轉(zhuǎn)換技術(shù)因?yàn)樘幱陉P(guān)鍵環(huán)節(jié)而得以迅速發(fā)展，比如出現(xiàn)了超導(dǎo)相位調(diào)制-解調(diào)高分辨率ADC(the superconductor phase modulation-demodulation ADC)(Razavi, 1995; Norsworthy et al., 1996; Marques et al., 1998; Medeiro et al., 1999；Geerts et al., 2000)和采用ΔΣ拓?fù)浣Y(jié)構(gòu)的高精度ADC(Rylov and Robertazzi, 1995; Rylov et al., 1999; Mukhanov et al., 1999；Mukhanov et al., 2001).但由于受電子技術(shù)發(fā)展水平和器件材料的制約，目前分辨率(或動(dòng)態(tài)范圍)和轉(zhuǎn)換速率仍然是數(shù)據(jù)轉(zhuǎn)換器難以突破的技術(shù)瓶頸(Mukhanov et al., 2001).因而，在現(xiàn)有轉(zhuǎn)換器的基礎(chǔ)上，出現(xiàn)了不同的應(yīng)用技術(shù)來(lái)擴(kuò)展轉(zhuǎn)換器的部分性能指標(biāo).如在現(xiàn)代數(shù)字通信領(lǐng)域，為克服轉(zhuǎn)換器的速度瓶頸，采用了time-interleaved ADC即多通道分時(shí)并行交替采樣的流水結(jié)構(gòu)AD轉(zhuǎn)換技術(shù)(Black, 1980; Lee et al., 2007; Saleem and Vogel, 2011; Li, 2013b; Xu and Duan, 2014)，即利用幾個(gè)速度低但精度高的ADC分時(shí)并行交替采樣，以在保證轉(zhuǎn)換精度的同時(shí)獲得采樣速率的提高(Saleem and Vogel, 2010).在24位AD以前，還采用過(guò)采樣(Candy and Temes, 1987；王萍和李小京，2002)、并聯(lián)ADC、組合瞬時(shí)浮點(diǎn)放大器IFP(Instantaneous Floating Point Gain Amplifier)(羅運(yùn)先等，2006)等應(yīng)用技術(shù)措施.在過(guò)采樣中，采樣頻率每提高一倍，動(dòng)態(tài)范圍可增加3dB(李國(guó)，2005；李江等，2013).N個(gè)并聯(lián)ADC通道可獲得10×lgN(dB)信噪比的增加(李江等，2013).隨著基于過(guò)采樣的ΔΣ24位高精度AD轉(zhuǎn)換器的出現(xiàn)，這些方法帶來(lái)的改善效果有限，且隨著AD轉(zhuǎn)換器分辨率的提高，通道間的器件失配(mismatch)、偏移失配、增益失配、噪聲等變得突出，反過(guò)來(lái)降低了系統(tǒng)動(dòng)態(tài)范圍、線性等性能.因而目前在地震監(jiān)測(cè)與勘探領(lǐng)域，普遍采用的仍是由單個(gè)高精度AD轉(zhuǎn)換器設(shè)計(jì)的24位數(shù)據(jù)采集器(陳祖斌等，2006).

因此為了適應(yīng)地震觀測(cè)高動(dòng)態(tài)范圍的需要，本文研究了一種采用多片模數(shù)轉(zhuǎn)換器對(duì)輸入信號(hào)分級(jí)采集來(lái)擴(kuò)展地震數(shù)據(jù)采集器動(dòng)態(tài)范圍的方法.并研制了實(shí)驗(yàn)樣機(jī)，進(jìn)行了噪聲、滿幅測(cè)量范圍、線性等主要性能指標(biāo)的測(cè)試.

2多AD轉(zhuǎn)換器的分級(jí)采集

在常規(guī)的采集方法中，采集器所能達(dá)到的最高動(dòng)態(tài)范圍取決于其所采用的AD轉(zhuǎn)換芯片.目前常用于地震數(shù)據(jù)采集的高精度24位ΣΔAD轉(zhuǎn)換芯片主要有CS5371/CS5372/CS5376芯片組、ADS1255/ADS1256、以及ADS1281/ADS1282等.其中ADS1281/ADS1282是32位數(shù)據(jù)格式輸出，但其精度仍是24位的.這些內(nèi)置高階(四階)ΣΔ調(diào)制器的24位低速高精度的AD轉(zhuǎn)換器所能達(dá)到的最大動(dòng)態(tài)范圍在50 Hz采樣率時(shí)約為135 dB(RMS值).一些進(jìn)口的高性能地震數(shù)據(jù)采集器如美國(guó)REFTEK公司的130B、加拿大Nanometrics公司的Taurus、英國(guó)Guralp公司的CMG-DM24S31AM，以及國(guó)內(nèi)一些公司生產(chǎn)的高精度24位地震數(shù)據(jù)采集器，采用的基本上是CS5371/CS5372/CS5376芯片組(陳祖斌等，2006)或其上代產(chǎn)品CS5321芯片組.在一般情況下，24位地震數(shù)據(jù)采集器設(shè)置最大輸入幅度為±20 V(差分)，對(duì)小信號(hào)的分辨率約為3 μV左右.為提高對(duì)小信號(hào)的分辨率，一些數(shù)據(jù)采集系統(tǒng)往往在通道前端設(shè)置程控增益放大器PGA(Programmable Gain Amplifier)(陳祖斌等，2006)，但在提高分辨率的同時(shí)也降低了采集系統(tǒng)的滿幅輸入范圍.比如轉(zhuǎn)換器ADS1255在內(nèi)置的前置可編程增益放大器PGA增益為1時(shí)，短路輸入噪聲是0.629 μV(RMS值，50SPS)，在增益為64時(shí)，分辨率提高至0.122 μV，但滿幅輸入電壓范圍也從±5V(差分)下降至只有±78.125 mV, 容易導(dǎo)致大信號(hào)的溢出而飽和失真(廖聲剛，2005).而且在地震觀測(cè)中，采集器增益大小的設(shè)置需要專業(yè)觀測(cè)人員根據(jù)監(jiān)測(cè)現(xiàn)場(chǎng)實(shí)際情況對(duì)信號(hào)可能的幅度范圍進(jìn)行預(yù)先估計(jì)，給實(shí)際使用帶來(lái)不便，也不適用于如地震臺(tái)站這樣需要全范圍信號(hào)監(jiān)測(cè)的場(chǎng)合.因而一些地震臺(tái)網(wǎng)只能通過(guò)架設(shè)兩套地震儀器(一套設(shè)置高倍放大倍數(shù)用于高分辨率的小信號(hào)觀測(cè)，另一套設(shè)置大輸入范圍用于強(qiáng)震信號(hào)觀測(cè))來(lái)做到全動(dòng)態(tài)范圍的地震監(jiān)測(cè)(Romeo and Spinelli, 2013)，不僅增加了觀測(cè)成本的投入，在數(shù)據(jù)分析時(shí)也要大、小信號(hào)進(jìn)行分別處理，造成客觀上使用的不便.

2.1多通道AD并行分級(jí)采集的實(shí)現(xiàn)

擴(kuò)展采集器的動(dòng)態(tài)范圍包含兩方面的含義：一是降低數(shù)據(jù)采集器等效輸入噪聲，以提高對(duì)微弱信號(hào)的分辨力；二是提高采集器的最大輸入電壓范圍，以使采集器在大信號(hào)輸入時(shí)不至出現(xiàn)飽和限幅失真.多通道AD轉(zhuǎn)換分級(jí)采集實(shí)現(xiàn)動(dòng)態(tài)范圍擴(kuò)展的電路結(jié)構(gòu)原理框圖如圖1所示.N通道高精度24位模數(shù)轉(zhuǎn)換器對(duì)全范圍模擬輸入信號(hào)進(jìn)行并行采集，居于每通道前端的模擬預(yù)處理電路分別為增益固定的前置放大器或起電平位移功能的精密電阻網(wǎng)絡(luò)，用于調(diào)整該通道的電壓輸入范圍.前置放大電路可以提高該通道對(duì)小信號(hào)的分辨率(Yin and Ghovanloo, 2007)，降低等效輸入噪聲或提高小信號(hào)的信噪比SNR.電阻網(wǎng)絡(luò)衰減大信號(hào)和實(shí)現(xiàn)電平位移以匹配AD滿幅輸入范圍.采集控制處理單元和數(shù)字多路開關(guān)MUX則對(duì)采集結(jié)果進(jìn)行數(shù)字處理，擬合成32位高動(dòng)態(tài)的轉(zhuǎn)換數(shù)字信號(hào)輸出.MUX的輸出可表示如(1)式所示：

圖1　多通道AD分級(jí)采集的原理框圖Fig.1　Multi-Channels ADC

(1)

其中，

X0

2.2動(dòng)態(tài)范圍分析

圖2　前置輸入緩沖/放大電路Fig.2　Analog signal preparation amplifier

放大器的噪聲en主要由輸入電壓噪聲en_V、輸入電流噪聲en_i以及電阻熱噪聲en_R組成，計(jì)算如下：

(2)

這里由圖3可得：

=220.97nVRMS

=133.60nVRMS.

(2) 輸入電流噪聲en_i由寬帶噪聲eiBB和1/f噪聲eif組成：

≈304.26nVRMS

(3)

其中由圖3可得

圖3　LT1028放大器的輸入電壓噪聲頻譜密度曲線Fig.3　Noise density of LT1028

≈217.82nVRMS

≈212.44nVRMS

(3)電阻熱噪聲en_R為

≈1626.28nVRMS

(4)

(4)放大器總的輸出噪聲電壓en為

≈1647.53nVRMS

(5)

其等效輸入噪聲電壓為

(6)

當(dāng)最大輸入為±40 V(差分)時(shí)，在帶寬20 Hz時(shí)系統(tǒng)所能達(dá)到的最大動(dòng)態(tài)范圍SNR(Signal-to-Noise Ratio)為

≈174.8dB

(7)

實(shí)際電路還要考慮放大電路制作工藝、電源噪聲、熱電偶電勢(shì)引起的噪聲、PCB板漏電流、電磁輻射噪聲、器件本身的非理想?yún)?shù)(如有限輸入阻抗、有限共模抑制比、輸入偏置電流及偏置電壓等)等因素的影響，使得等效輸入噪聲要比計(jì)算值高一些，但現(xiàn)代高精度低噪聲放大器的分辨率還是可輕易做到小于0.5 μV的(Harrison and Charles, 2003; Mosheni and Najafi, 2004; Yin and Ghovanloo, 2007).

3實(shí)驗(yàn)樣機(jī)研制

3.1電路設(shè)計(jì)

現(xiàn)代高精度24位AD轉(zhuǎn)換器可達(dá)130 dB以上的動(dòng)態(tài)范圍，采用兩個(gè)AD并行連接，對(duì)輸入信號(hào)進(jìn)行兩級(jí)分割采集.模數(shù)轉(zhuǎn)換器芯片采用內(nèi)帶四階ΣΔ調(diào)制器的高精度24位AD芯片ADS1255，該芯片具有可達(dá)23位無(wú)噪聲分辨率、最大只有±0.001%非線性的優(yōu)良特性，采樣率在2.5SPS～30kSPS(samples per-second)之間可以軟件設(shè)置，片上集成有可編程低噪聲前置增益放大器PGA和數(shù)字濾波器，可按設(shè)置采樣率直接輸出24位的二進(jìn)制數(shù)據(jù)，因而使用起來(lái)非常方便，外圍電路簡(jiǎn)潔，有利于提高系統(tǒng)的可靠性和信噪比.小信號(hào)采集通道采用ADS1255芯片內(nèi)置的可編程放大器進(jìn)行放大預(yù)處理以提高分辨率，在設(shè)置放大倍數(shù)為64倍、采樣率為50SPS時(shí)其等效輸入短路噪聲僅有0.122 μV.為擴(kuò)大動(dòng)態(tài)范圍，地震計(jì)內(nèi)部模擬電路一般采用±15 V的較高工作電壓，現(xiàn)代軌到軌(rail-to-rail)輸入輸出運(yùn)算放大器輸出信號(hào)擺幅接近電源電壓，即地震計(jì)輸出信號(hào)范圍可達(dá)±30 V(差分)(Melton, 1976; Muramatu, 1995).而通常24位地震數(shù)據(jù)采集器為兼顧大信號(hào)輸入和小信號(hào)分辨率，把最大滿幅(full-scale)輸入信號(hào)范圍設(shè)置在差分±20 V，因而在遇有大地震時(shí)，數(shù)據(jù)采集器比地震計(jì)要先進(jìn)入飽和限幅狀態(tài).為能最大避免輸入信號(hào)過(guò)大而引起數(shù)據(jù)采集器限幅失真的情況，在大信號(hào)采集通道輸入端加入電阻網(wǎng)絡(luò)進(jìn)行電平位移和衰減，使AD輸入信號(hào)范圍由±5 V擴(kuò)展到±40 V(差分).數(shù)字部分的采集控制、數(shù)據(jù)處理、數(shù)據(jù)存儲(chǔ)及通信傳輸?shù)瓤刂茊卧獎(jiǎng)t采用32位高性能低功耗的嵌入式ARM9微處理器作為控制處理單元，并采用嵌入式Linux操作系統(tǒng)為采集系統(tǒng)多任務(wù)處理的軟件平臺(tái).

圖4　兩通道TI結(jié)構(gòu)ADC模型Fig.4　Model of a two-channel TI-ADC

3.2數(shù)據(jù)處理

兩通道AD在同一時(shí)鐘驅(qū)動(dòng)下同步并行地采集輸入模擬信號(hào)，數(shù)字選擇開關(guān)根據(jù)輸入信號(hào)的幅度選擇通道數(shù)據(jù)，經(jīng)控制處理器數(shù)據(jù)處理后組合成32位輸出.在多通道并行采集結(jié)構(gòu)中，如直接并行轉(zhuǎn)換再平均的數(shù)據(jù)采集系統(tǒng)、TI(Time-Interleaved)結(jié)構(gòu)分時(shí)并行采集系統(tǒng)(如圖4所示)(Saleem and Vogel, 2011)等，不同通道元器件參數(shù)及狀態(tài)的不完全匹配，會(huì)產(chǎn)生增益、偏移以及采集相位等不匹配并導(dǎo)致采集精度降低(Petraglia and Mitra, 1991; Kurosawa et al., 2001; Elbornsson et al., 2003; Vogel, 2005； Saleem and Vogel, 2011).如由于通道間的失配(mismatch)，TI-ADC結(jié)構(gòu)是一種時(shí)變系統(tǒng)(a time-varying system)，這會(huì)引起實(shí)際輸入信號(hào)的假頻信號(hào)，從而導(dǎo)致信納比SINAD(signal-to-noise and distortion ratio，信號(hào)對(duì)噪聲及失真比)及無(wú)失真動(dòng)態(tài)范圍SFDR(spurious free dynamic range)等性能指標(biāo)的大幅度降低( Leger et al.,2004; Vogel 2005; EI-Chammas and Murmann,2009； Saleem and Vogel,2011).因而要采用標(biāo)定技術(shù)(calibration techniques)來(lái)糾正由這些不匹配因素所帶來(lái)的AD轉(zhuǎn)換錯(cuò)誤(Mendel and Vogel, 2006; Lee, 2007; Marelli, 2009; Saleem and Vogel, 2011; Rao et al., 2012; Xu and Duan, 2014).在超過(guò)14位分辨率的流水結(jié)構(gòu)ADC中，往往采用數(shù)字標(biāo)定(前景或背景)來(lái)減小對(duì)器件的匹配和對(duì)放大器增益的要求(Siragusa and Galton, 2004; Liu,2005; Bogner et al.,2006; Lee et al.,2007)，包括增益標(biāo)定(Seo et al., 2005; Huang and Levy, 2006; Vogel et al., 2008; Saleem and Volgel, 2011)、時(shí)序失配(timing dismatches)(Johansson and Lowenborg, 2002; EI-Chammas and Murmann, 2009; Saleem and Vogel, 2011; Li, 2013b)、頻響失配標(biāo)定(Tsai et al., 2006; Satarzadeh, et al., 2007; Saleem and Vogel, 2011)等.此外還提出了split delta-sigma ADC結(jié)構(gòu)的背景標(biāo)定法(McNeill et al., 2005; Lee and Temes, 2006).

在用兩通道并行分級(jí)采集中，最容易出現(xiàn)也是最關(guān)鍵的誤差是“交越失配”，主要表現(xiàn)在兩通道的直流偏移不匹配(零點(diǎn)誤差)和增益或電壓靈敏度不匹配兩個(gè)方面.但由于采用了多通道AD的并行采集，使得各通道的偏移和增益標(biāo)定是靜態(tài)的，即標(biāo)定值在整個(gè)信號(hào)范圍內(nèi)都是固定的，因而可以獲得很高的標(biāo)定精度.通道間標(biāo)定可以從兩個(gè)方面進(jìn)行：一是前端模擬預(yù)處理部分，從元器件選擇、參數(shù)及工作狀態(tài)要盡量做到兩通道間的匹配，但由于高精密元器件的參數(shù)不可能連續(xù)變化，加上各種誤差的存在，實(shí)際中可以做到兩通道的匹配程度優(yōu)于1%；其次是后端的數(shù)字標(biāo)定，即對(duì)兩通道的采集結(jié)果再進(jìn)行數(shù)字精確校正，使其保證有較高的一致性.經(jīng)過(guò)數(shù)字標(biāo)定后兩通道的匹配程度可以做到小于0.1%.

3.3噪聲控制

地震記錄系統(tǒng)的關(guān)鍵在于數(shù)據(jù)采集通道的低噪聲、高精度設(shè)計(jì)(陳祖斌等，2006).為降低數(shù)字控制系統(tǒng)部分的數(shù)字脈沖噪聲通過(guò)公共地線串入信號(hào)通道系統(tǒng)引起信噪比變壞，信號(hào)通道和控制系統(tǒng)部分采用了電源隔離和數(shù)字信號(hào)隔離.并且在信號(hào)通道、模擬電源和數(shù)字電源也是分開的，在電源引腳旁邊并聯(lián)高品質(zhì)的電容進(jìn)行退耦和濾波.模擬信號(hào)正負(fù)輸入端的電平變換網(wǎng)絡(luò)和走線則采用對(duì)稱設(shè)計(jì)，最大限度抑制外部共模噪聲輸入.AD轉(zhuǎn)換的基準(zhǔn)電壓電路也選用高精度、低噪聲、低溫漂的芯片，并合理設(shè)計(jì)濾波電路.此外因?yàn)閮赏ǖ繟D的匹配標(biāo)定是靜態(tài)的，只對(duì)固定的偏移和電壓靈敏度有效，而元件老化、溫度漂移等動(dòng)態(tài)變化量則無(wú)法消除.因而前端的模擬通道要采用低噪聲器件.如采用斬波放大器可以有效地降低直流偏移、溫漂、噪聲、1/f噪聲等.此外還要采用噪聲低、溫度系數(shù)小的精密金屬膜電阻.

4主要性能指標(biāo)測(cè)試

4.1動(dòng)態(tài)范圍

圖5　輸入短路噪聲Fig.5　Noise test

在輸入分別接入正弦電壓和直流電壓，測(cè)試滿幅輸入電壓范圍FSR(Full-Scale-Range)，獲得正弦記錄波形如圖6所示，可以看出其滿幅輸入電壓范圍FSR大于±40 V.于是計(jì)算采集器動(dòng)態(tài)范圍SNR為

≈157dB.

(8)

4.2線性測(cè)試

在多通道并行分級(jí)采集結(jié)構(gòu)中，盡管各通道AD本身具有很高的線性度，但由于存在通道間的失配現(xiàn)象，各通道間的增益或電壓靈敏度的不一致會(huì)導(dǎo)致采集器整體的線性變差，因而線性是一個(gè)重要的測(cè)試指標(biāo).在50 Hz采樣率下，根據(jù)《地震數(shù)據(jù)采集器質(zhì)量檢測(cè)技術(shù)規(guī)程(2007)》，測(cè)試幅值選擇為滿量程的0.1%、0.2%、0.4%、0.8%、1%，2%、4%、8%、10%、20%、40%、80%、100%，輸入信號(hào)為直流電壓，每個(gè)測(cè)試點(diǎn)記錄30 s長(zhǎng)的數(shù)據(jù)再取其平均值，測(cè)得數(shù)據(jù)結(jié)果如表1及圖7所示.

圖6　正弦波滿幅輸入電壓范圍測(cè)試Fig.6　Full scale range test

表1　線性測(cè)試數(shù)據(jù)記錄表

圖7　輸入輸出線性測(cè)試Fig.7　Relation of input to output

測(cè)量示值Y和標(biāo)準(zhǔn)值X的線性擬合關(guān)系為

(9)

其中，

計(jì)算線性偏差Δyi：

式中，i=1，2，3，…，13；

Δyi集合中絕對(duì)值最大的值為Δymax，則線性度為

=4.37368×10-5

(10)

4.3輸入阻抗

輸入阻抗設(shè)計(jì)得小些，有利于減小系統(tǒng)零輸入噪聲，提高小信號(hào)分辨率.但輸入阻抗過(guò)小會(huì)導(dǎo)致對(duì)前端傳感器的負(fù)載過(guò)大，一般不宜小于20 kΩ.輸入阻抗測(cè)試信號(hào)連接如圖8所示，標(biāo)準(zhǔn)信號(hào)源采用Fluke-5500A Calibrator標(biāo)準(zhǔn)校準(zhǔn)器.設(shè)置數(shù)據(jù)采集器采樣率50 Hz、輸入正弦信號(hào)幅度為采集器滿量程的約50%，測(cè)得沒(méi)有串接標(biāo)準(zhǔn)電阻R時(shí)數(shù)據(jù)采集器的讀數(shù)U1=14.150VRMS，串接標(biāo)準(zhǔn)電阻R后數(shù)據(jù)采集器的讀數(shù)U2=4.748VRMS，按(11)式計(jì)算輸入電阻Ri為

(11)

圖8　輸入阻抗測(cè)試信號(hào)連接示意圖Fig.8　Connect diagram of input impedance test

5結(jié)論

本文提出的并行多通道AD轉(zhuǎn)換分級(jí)采集的方法，能夠提高對(duì)微小地震信號(hào)的分辨率，同時(shí)也能記錄大震信號(hào)而不會(huì)使采集器提前地震計(jì)進(jìn)入飽和產(chǎn)生限幅失真，有效地?cái)U(kuò)展了地震數(shù)據(jù)采集器的記錄動(dòng)態(tài)范圍，能夠滿足地震觀測(cè)中對(duì)數(shù)據(jù)采集器大動(dòng)態(tài)范圍的要求.而且技術(shù)上易于實(shí)現(xiàn)，成本也較低.

致謝在本文研究過(guò)程中，得到中國(guó)地震局地震預(yù)測(cè)研究所薛兵研究員的熱忱幫助和研究指導(dǎo)；同時(shí)評(píng)審專家也提出了很多寶貴的修改意見，在此謹(jǐn)表示作者誠(chéng)摯的感謝.

References

Black W C, Hodges D A. 1980. Time interleaved converter arrays.IEEEJ.Solid-StateCircuits, 15(6): 1022-1029.

Bogner P, Kuttner F, Kropf C, et al. 2006. A 14b 100MS/s digitally self-calibrated pipelined ADC in 0.13μm CMOS.∥ IEEE International Solid-State Circuits Conference, Digest of Technical Papers. San Francisco, CA: IEEE, 832-841.

Candy J C, Temes G C. 1987. Oversampling methods for A/D and D/A conversion.∥ IEEE Transactions on Circuits and Systems. IEEE.Chen Z B, Teng J W, Lin J, et al. 2006. Design of BSR-2 broad band seismic recorder.ChineseJ.Geophys. (in Chinese), 49(5): 1475-1481. Elbornsson J, Gustafsson F, Eklund J E. 2003. Analysis of mismatch noise in randomly interleaved ADC system.∥ Proceedings of the 2003 IEEE International Conference on Acoustics, Speech, and Signal Processing. Hong Kong, China: IEEE.El-Chammas M, Murmann B. 2009. General analysis on the impact of phase-skew in time-interleaved ADCs.IEEETrans.CircuitsSyst.I,Reg.Papers, 56(5): 902-910. Evans J R, Followill F, Hutt C R, et al. 2010. Method for calculating self-noise spectra and operating ranges for seismographic inertial sensors and recorders.SeismologicalResearchLetters, 81(4): 640-646. Gao S H, Xue B, Jiang J M, et al. 2012. Idea of high performance seismic data acquisition system.RecentDevelopmentsinWorldSeismology(in Chinese), (6): 173. Ge H K, Chen H C, Ouyang B, et al. 2013. Transportable seismometer response to seismic noise in vault.ChineseJ.Geophys. (in Chinese), 56(3): 857-868, doi: 10.6038/cjg20130315. Geerts Y, Steyaert M S J, Sansen W. 2000. A high-performance multibit ΔΣ CMOS ADC.IEEEJournalofSolid-StateCircuits, 35(12): 1829-1840. Harrison R R, Charles C. 2003. A low-power low-noise CMOS for amplifier neural recording applications.IEEEJ.Solid-StateCircuits, 38(6): 958-965.

Huang S, Levy B C. 2006. Adaptive blind calibration of timing offset and gain mismatch for two-channel time-interleaved ADCs.IEEETrans.onCircuitsSyst.I:Reg.Papers, 53(6): 1278-1288.Johansson H, Lowenborg P. 2002. Reconstruction of nonuniformly sampled bandlimited signals by means of digital fractional delay filters.IEEETrans.SignalProcess., 50(11): 2757-2767.

Kurosawa N, Kobayashi H, Maruyama K, et al. 2001. Explicit analysis of channel mismatch effects in time-interleaved ADC systems.IEEETrans.CircuitsSyst.I:Fundam.TheoryAppl., 48(3): 261-271.Lee K, Temes G C. 2006. Enhanced split-architecture delta-sigma ADC.∥ Proceedings of the 13th IEEE International Conference on Electronics, Circuits and Systems. Nice: IEEE, 427-430.Lee Z M, Wang C Y, Wu J T. 2007. A CMOS 15-bit 125-MS/s time-interleaved ADC with digital background calibration.IEEEJournalofSolid-StateCircuits, 42(10): 2149-2160.

Leger G, Peralias E, Rueda A, et al. 2004. Impact of random channel mismatch on the SNR and SFDR of time-interleaved ADCs.IEEETrans.CircuitsSyst.I:Fundam.TheoryAppl., 51(1): 140-150.Li G. 2005. Study and implementation of improving ADC resolution based on over sampling technology.ComputerEngineering(in Chinese), 31(S1): 244-248.

Li J, Xue B, Zhu X Y, et al. 2013a. Key technology research of seismic data acquisition system.SeismologicalandGeomagneticObservationandResearch(in Chinese), 34(3-4): 179-185, doi: 10.3969/j.issn.1003-3246.2013.03/04.033. Li J, Liu Y, Liu H, et al. 2013b. A timing skew calibration scheme in time-interleaved ADC.JournalofComputerandCommunications, 1(6): 37-40. Li W, Chen Z B. 2006. Noise analysis in data acquisition channel of broad band seismic recorder.JournalofJilinUniversity(InformationScienceEdition) (in Chinese), 24(5): 462-465.

Liao S G. 2005. The limitation of 24-bits seismic data acquisition system for high precision petroleum seismic exploration.ScienceandTechnology(in Chinese), 15(2): 13-15.

Liu H C, Lee Z M, Wu J T. 2005. A 15-b 40-MS/s CMOS pipelined analog-to-digital converter with digital background calibration.IEEEJ.Solid-StateCircuits, 40(5): 1047-1056.

Luo Y X, Wu X Y, Zeng J, et al. 2006. Design of a 7-bands instantaneous floating point gain amplifier.JournalofEEE(in Chinese), 28(6): 39-41, 49.

Marelli D, Mahata K, Fu M Y. 2009. Linear LMS compensation for timing mismatch in time-interleaved ADCs.IEEETransactiononCircuitsandSystemsI:RegularPapers, 56(11): 2476-2486.

Marques A, Peluso V, Steyaert M S, et al. 1998. Optimal parameters for ΔΣ modulator topologies. IEEE Trans.CircuitsSyst.II, 45(9): 1232-1241. McNeill J, Coln M C W, Larivee B J. 2005. ‘Split-ADC’ architecture for deterministic digital background calibration of a 16-bit 1-MS/s ADC.IEEEJ.Solid-StateCircuits, 40(12): 2437-2445.Medeiro F, Pérez-Verdú B, Rodríquez-Vázquez A. 1999. Top-Down Design of High-Performance Sigma-Delta Modulators. US: Springer.

Melton B S. 1976. The sensitivity and dynamic range of inertial seismographs.Rev.Geophys., 14(1): 93-116.

Mosheni P, Najafi K. 2004. A fully integrated neural recording amplifier with DC input stabilization.IEEETrans.Biomed.Eng., 51(5): 832-837. Mukhanov O A, Semenov V K, Li W Q, et al. 2001. A superconductor high-resolution ADC.IEEETransactionsonAppliedSuperconductivity, 11(1): 601-606. Mukhanov O A, Brock D K, Kirichenko A F, et al. 1999. Progress in the development of a superconductive high-resolution ADC.∥ Extended Abstracts of 7th ISEC′99. Berkeley, CA, 13-16.

Muramatu I. 1995. Development of a broadband velocity type strong motion seismometer and its recording range.JournaloftheSeismologicalSocietyofJapan, 48(2): 247-256.

Norsworthy S R, Schreier R, Temes G C. 1996. Delta-Sigma Data Converters: Theory, Design, and Simulation. NY: IEEE Press.Petraglia A, Mitra S K. 1991. Analysis of mismatch effects among A/D converters in a time-interleaved waveform digitizer.IEEETrans.Instrum.Meas., 40(5): 831-835.

Mendel S, Vogel C. 2006. A compensation method for magnitude response mismatches in two-channel time-interleaved analog-to-digital converters. ∥Electronics, Circuits and Systems, 2006. ICECS ′06. 13th IEEE International Conference on. 712-715.

Rao L P, Sitthimahachaikul N, Hurst P J. 2012. Correcting the effects of mismatches in time-interleaved analog adaptive FIR equalizers.IEEETransactionsonCircuitsandSystemI:RegularPapers, 59(11): 2529-2542.

Razavi B. 1995. Principles of Data Conversion System Design. New York: IEEE Press.

Ringler A T, Hutt C R, Evans J R, et al. 2011. A comparison of seismic instrument noise coherence analysis techniques.BulletinoftheSeismologicalsocietyofAmerica, 101(2): 558-567.Romeo G, Spinelli G. 2013. Extending a Lippmann style seismometer′s dynamic range by using a non-linear feedback circuit.Adv.Geosci., 36: 27-30, doi: 10.5194/adgeo-36-27-2013.Rylov S V, Robertazzi R P. 1995. Superconducting high-resolution A/D converter based on phase modulation and multichannel timing arbitration.IEEETrans.Appl.Supercond., 5(2): 2260-2263. Rylov S V, Brock D K, Gaidarenko D V, et al. 1999. High resolution ADC using phase modulation-demodulation architecture.IEEETrans.Appl.Supercond., 9(2): 3016-3019.Saleem S, Vogel C. 2010. Adaptive compensation of frequency response mismatches in high-resolution time-interleaved ADCs using a low-resolution ADC and a time-varying filter.∥ Proceedings of 2010 IEEE International Symposium on Circuits and Systems. Paris: IEEE, 561-564.

Saleem S, Vogel C. 2011. Adaptive blind background calibration of polynomial-represented frequency response mismatches in a two-channel time-interleaved ADC.IEEETransactionsonCircuitsandSystems-I:RegularPapers, 58(6): 1300-1310.

Satarzadeh P, Levy B C, Hurst P J. 2007. Bandwidth mismatch correction for a two-channel time-interleaved A/D converter.∥ Proceedings of the IEEE International Symposium on Circuits and Systems. New Orleans, LA: IEEE, 1705-1708.

Seo M, Rodwell M J W, Madhow U. 2005. Blind correction of gain and timing mismatches for a two-channel time-interleaved analog-to-digital converter.∥ Conference Record of the 39th Asilomar Conference on Signals, Systems and Computers. Pacific Grove, CA: IEEE, 1121-1125.

Siragusa E, Galton I. 2004. A digitally enhanced 1.8-V 15-bit 40-MSample/s CMOS pipelined ADC.IEEEJ.Solid-StateCircuits, 39(12): 2126-2138.

Sun X, Luo G E. 2008. Design and implementation of a high precision seismic data acquisition system.ComputerMeasurement&Control(in Chinese), 16(6): 881-883.Tsai T H, Hurst P J, Lewis S H. 2006. Bandwidth mismatch and its correction in time-interleaved analog-to-digital converters.IEEETrans.CircuitsSyst.II:Exp.Briefs, 53(10): 1133-1137.

Vogel C. 2005. The impact of combined channel mismatch effects in time-interleaved ADCs.IEEETrans.Instrum.Meas., 54(1): 415-427.

Vogel C, Saleem S, Mendel S. 2008. Adaptive blind compensation of gain and timing mismatches in m-channel time-interleaved ADCs.∥ Proceedings of the 15th IEEE International Conference on Electronics, Circuits and Systems. St. Julien′s: IEEE, 49-52.Wang P, Li X J. 2002. Improving ADC resolution and signal-noise ratio by oversampling.ChineseJournalofScientificInstrument(in Chinese), 23(3S): 1-2.Wang C F, Yang X Y, Song C, et al. 2010. The Principal shock waves of the 2008 Wenchuan 8.0 Earthquake Recorded by two types of Acquisition device.EarthquakeResearchinSichuan(in Chinese), (3): 19-22.

Xu Q, Duan Z M. 2014. A signal reconstruction method for time-interleaved ADC converters.JournalofInformation&ComputationalScience, 11(1): 123-132.Yin M, Ghovanloo M. 2007. A low-noise preamplifier with adjustable gain and bandwidth for biopotential recording applications.∥ IEEE International Symposium on Circuits and Systems. New Orleans, LA: IEEE, 321-324.

Yuan Z L, Cao G H. 2000. Improvement on seismograph data collection system.JournalofDaqingPetroleumInstitute(in Chinese), 24(4): 15-17.

附中文參考文獻(xiàn)

陳祖斌, 滕吉文, 林君等. 2006. BSR-2寬頻帶地震記錄儀的研制. 地球物理學(xué)報(bào), 49(5): 1475-1481.

高尚華, 薛兵, 蔣久明等. 2012. 高性能地震數(shù)據(jù)采集方案構(gòu)想. 國(guó)際地震動(dòng)態(tài), (6): 173.

葛洪魁, 陳海潮, 歐陽(yáng)飚等. 2013. 流動(dòng)地震觀測(cè)背景噪聲的臺(tái)基響應(yīng). 地球物理學(xué)報(bào), 56(3): 857-868, doi: 10.6038/cjg20130315.

李國(guó). 2005. 基于過(guò)采樣技術(shù)提高ADC分辨率的研究與實(shí)現(xiàn). 計(jì)算機(jī)工程, 31(增刊): 244-248.

李江, 薛兵, 朱小毅等. 2013. 地震數(shù)據(jù)采集器關(guān)鍵技術(shù)研究. 地震地磁觀測(cè)與研究, 34(3-4): 179-185, doi: 10.3969/j.issn.1003-3246.2013.03/04.033.

李威, 陳祖斌. 2006. 寬頻地震儀數(shù)據(jù)采集通道的噪聲分析. 吉林大學(xué)學(xué)報(bào)(信息科學(xué)版), 24(5): 462-465.

廖聲剛. 2005. 24位地震數(shù)據(jù)采集系統(tǒng)進(jìn)行高精度石油地震勘探的局限性. 江漢石油科技, 15(2): 13-15.

羅運(yùn)先, 吳雄英, 曾敬等. 2006. 七階型瞬時(shí)浮點(diǎn)增益放大器設(shè)計(jì)探討. 電氣電子教學(xué)學(xué)報(bào), 28(6): 39-41, 49.

孫嫻, 羅桂娥. 2008. 一種高精度地震勘探數(shù)據(jù)采集系統(tǒng)的設(shè)計(jì)與實(shí)現(xiàn). 計(jì)算機(jī)測(cè)量與控制, 16(6): 881-883.

王萍, 李小京. 2002. 利用過(guò)采樣法提高A/D分辨率和信噪比. 儀器儀表學(xué)報(bào), 23(3增刊): 1-2.

王翠芳, 楊曉源, 宋澄等. 2010. 兩種采集器記錄的汶川8.0級(jí)主震波形. 四川地震, (3): 19-22.

袁子龍, 曹廣華. 2000. 地震儀數(shù)據(jù)采集系統(tǒng)的改進(jìn). 大慶石油學(xué)院學(xué)報(bào), 24(4): 15-17.

(本文編輯汪海英)

Extending dynamic range of the seismic data acquisition system by using multi-channel ADC

TENG Yun-Tian, HU Xing-Xing*, WANG Xi-Zhen, WANG Xiao-Mei,LU Hong-Ya, WANG Zhe, ZHANG Yang

InstituteofGeophysics,ChineseEarthquakeAdministration,Beijing100081,China

AbstractThe seismic data acquisition system is an indispensable process of the digitalization of seismic signals. The quality of this system has a direct effect on the quality of the seismic signal recorded and the final data processing results. There is a wide dynamic range for the observed seismic signal, the total signal amplitude of which could even surpass 160 dB. So it is a dynamic range of broadband seismometers, which is more than 140 dB. However, the widely used 24-bit high-quality seismic data acquisition system in recent times, which is produced based on ΔΣAD conventional technology, has a dynamic range of only 135 dB@50SPS. As a result, seismic information cannot be extracted effectively in practical earthquake monitoring for both minor and strong signals. The loss of minor signal is due to the fact that data acquisition with low SNR (signal to noise ratio) has no high resolution to identify minor signals, and that the amplitude of the signal to record is at the same magnitude order as the resolution of the data acquisition system. The loss of large signal is caused by a saturated clipping distortion that appears in violent earthquakes. This makes local seismograph stations almost lose the ability to record data at the time that firsthand observation resources are needed for earthquake prevention and hazard reduction, and leads to lose precious opportunities to record strong seismic signals for seismic study. Therefore, to meet the application requirements, a multi-channel AD converter sample grading method for extending the dynamic range of seismic data acquisition is put forward in this paper.

KeywordsDynamic range; Seismic data acquisition system; High resolution; Channel mismatch; Calibration

基金項(xiàng)目國(guó)家自然科學(xué)基金(41304142)，中國(guó)地震局地球物理研究所中央級(jí)公益性科研院所基本科研業(yè)務(wù)(DQJB-13B10)資助.

作者簡(jiǎn)介滕云田，男，1966年生，中國(guó)地震局地球物理研究所研究員，博士生導(dǎo)師，主要從事地球物理觀測(cè)技術(shù)研究.E-mail：tyt1966@sohu.com*通訊作者胡星星，男，1978年生，中國(guó)地震局地球物理研究所副研究員，博士.研究方向?yàn)榈厍蛭锢碛^測(cè)技術(shù). E-mail：huxx@cea-igp.ac.cn

doi:10.6038/cjg20160424 中圖分類號(hào)P315

收稿日期2014-10-17，2016-02-25收修定稿

滕云田， 胡星星， 王喜珍等. 2016. 用多通道AD分級(jí)采集擴(kuò)展地震數(shù)據(jù)采集器的動(dòng)態(tài)范圍.地球物理學(xué)報(bào)，59(4):1435-1445,doi:10.6038/cjg20160424.

Teng Y T, Hu X X, Wang X Z, et al. 2016. Extending dynamic range of the seismic data acquisition system by using multi-channel ADC.ChineseJ.Geophys. (in Chinese),59(4):1435-1445,doi:10.6038/cjg20160424.