版权说明:本文档由用户提供并上传,收益归属内容提供方,若内容存在侵权,请进行举报或认领
文档简介
ALow-Cost,SmartCapacitivePositionSensorAbstractAnewhigh-performance,low-cost,capacitiveposition-measuringsystemisdescribed.Byusingahighlylinearoscillator,shieldingandathree-signalapproach,mostoftheerrorsareeliminated.Theaccuracyamountsto1μmovera1mmrange.Sincetheoutputoftheoscillatorcandirectlybeconnectedtoamicrocontroller,anA/Dconverterisnotneeded.I.INTRODUCTIONThispaperdescribesanovelhigh-performance,low-cost,capacitivedisplacementmeasuringsystemfeaturing:1mmmeasuringrange,1μmaccuracy,0.1stotalmeasuringtime.Translatedtothecapacitivedomain,thespecificationscorrespondto:apossiblerangeof1pF;only50fFofthisrangeisusedforthedisplacementtransducer;50aFabsolutecapacitance-measuringinaccuracy.MeijerandSchrier[l]andmorerecentlyVanDrecht,Meijer,andDeJong[2]haveproposedadisplacement-measuringsystem,usingaPSD(PositionSensitiveDetector)assensingelement.SomedisadvantagesofusingaPSDarethehighercostsandthehigherpowerconsumptionofthePSDandLED(Light-EmittingDiode)ascomparedtothecapacitivesensorelementsdescribedinthispaper.Thesignalprocessorusestheconceptspresentedin[2],butisadoptedfortheuseofcapacitiveelements.Bytheextensiveuseofshielding,guardingandsmartA/Dconversion,thesystemisabletocombineahighaccuracywithaverylowcost-price.Thetransducerproducesthree-period-modulatedsignalswhichcanbeselectedanddirectlyreadoutbyamicrocontroller.Themicrocontroller,inreturn,calculatesthedisplacementandcansendthisvaluetoahostcomputer(Fig.1)oradisplayordriveanactuator.ElectronicCircuitElectronicCircuitPersonalComputer ActuatorDisplayFig.1.BlockdiagramofthesystemFig.2.PerspectiveanddimensionsoftheelectrodestructureⅡ.THEELECTRODESTRUCTUREThebasicsensingelementconsistsoftwosimpleelectrodeswithcapacitanceCx,(Fig.2).Thesmallerone(E2)issurroundedbyaguardelectrode.Thankstotheuseoftheguardelectrode,thecapacitanceCxbetweenthetwoelectrodesisindependentofmovements(lateraldisplacementsaswellasrotations)paralleltotheelectrodesurface.TheinfluenceoftheparasiticcapacitancesCpwillbeeliminatedaswillbediscussedinSectionⅢ.AccordingtoHeerens[3],therelativedeviationinthecapacitanceCxbetweenthetwoelectrodescausedbythefiniteguardelectrodesizeissmallerthan:δ<e-π(x/d)(1)wherexisthewidthoftheguardanddthedistancebetweentheelectrodes.Thisdeviationintroducesanonlinearity.Thereforewerequirethatδislessthan100ppm.Alsothegapbetweenthesmallelectrodeandthesurroundingguardcausesadeviation:δ<e-π(d/s)(2)withsthewidthofthegap.Thisdeviationisnegligiblecomparedto(l),whenthegapwidthislessthan1/3ofthedistancebetweentheelectrodes.Anothercauseoferrorsoriginatesfromapossiblefiniteskewangleαbetweenthetwoelectrodes(Fig.3).Assumingthefollowingconditions:thepotentialsonthesmallelectrodeandtheguardelectrodeareequalto0V,thepotentialonthelargeelectrodeisequaltoVvolt,theguardelectrodeislargeenough,itcanbeseenthattheelectricfieldwillbeconcentric.ddl/2l/2Fig.3.Electrodeswithangleα.Tokeepthecalculationssimple,wewillassumetheelectrodestobeinfinitelylargeinonedirection.Nowtheproblemisatwo-dimensionalonethatcanbesolvedbyusingpolar-coordinates(r,φ).Inthiscasetheelectricalfieldcanbedescribedby:(3)Tocalculatethechargeonthesmallelectrode,wesetφto0andintegrateoverr:(4)withBltheleftborderofthesmallelectrode:(5)andBrtherightborder:(6)Solving(4)resultsin:(7)Forsmallα'sthiscanbeapproximatedby:(8)Itappearstobedesirabletochooselsmallerthand,sotheerrorwilldependonlyontheangleα.Inourcase,achangeintheangleof0.6°willcauseanerrorlessthan100ppm.Withaproperdesigntheparametersεoandlareconstant,andthenthecapacitancebetweenthetwoelectrodeswilldependonlyonthedistancedbetweentheelectrodes.Ⅲ.ELIMINATIONOFPARASITICCAPACITANCESBesidesthedesiredsensorcapacitanceC,therearealsomanyparasiticcapacitancesintheactualstructure(Fig.2).ThesecapacitancescanbemodeledasshowninFig.4.HereCplrepresentstheparasiticcapacitancesfromtheelectrodeE1andCp2fromtheelectrodeE2totheguardelectrodesandtheshielding.ParasiticcapacitanceCp3resultsfromimperfectshieldingandformsanoffsetcapacitance.WhenthetransducercapacitanceCxisconnectedtoanACvoltagesourceandthecurrentthroughtheelectrodeismeasured,CplandCp2willbeeliminated.Cp3canbeeliminatedbyperforminganoffsetmeasurement.Fig.4.EliminationofparasiticcapacitancesThecurrentismeasuredbytheamplifierwithshuntfeedback,whichhasaverylowinputimpedance.Toobtaintherequiredlinearity,theunity-gainbandwidthfToftheamplifierhastosatisfythefollowingcondition:(9)whereTistheperiodoftheinputsignal.SinceCp2consistsofcablecapacitancesandtheinputcapacitanceoftheopamp,itmayindeedbelargerthanCfandcannotbeneglected.IV.THECONCEPTOFTHESYSTEMThesystemusesthethree-signalconceptpresentedin[2],whichisbasedonthefollowingprinciples.WhenwemeasureacapacitorCxwithalinearsystem,weobtainavalue:(10)wheremistheunknowngainandMoff,theunknownoffset.ByperformingthemeasurementofareferencequantityCref,inanidenticalwayandbymeasuringtheoffset,Moff,bymakingm=0,theparametersmandMoffareeliminated.ThefinalmeasurementresultPisdefinedas:(11)Inourcase,forthesensorcapacitanceC,itholdsthat:(12)whereAxistheareaoftheelectrode,doistheinitialdistancebetweenthem,εisthedielectricconstantand△disthedisplacementtobemeasured.Forthereferenceelectrodesitholdsthat:(13)withAreftheareaanddrefthedistance.Substitutionof(12)and(13)into(10)andtheninto(11)yields:(14)Here,Pisavaluerepresentingthepositionwhilea1anda0areunknown,butstableconstants.Theconstanta1=Aref/Axisastableconstantprovidedthereisagoodmechanicalmatchingbetweentheelectrodeareas.Theconstantao=(Arefd0/(Axdref)willalsobeastableconstantprovidedthatdoanddrefareconstant.Theseconstantscanbedeterminedbyaone-timecalibration.Inmanyapplicationsthiscalibrationcanbeomitted;whenthedisplacementsensorispartofalargersystem,anoverallcalibrationisrequiredanyway.Thisoverallcalibrationeliminatestherequirementforaseparatedeterminationofa1anda0.V.THECAPACITANCE-TO-PERIODCONVERSIONThesignalswhichareproportionaltothecapacitorvaluesareconvertedintoaperiod,usingamodifiedMartinoscillator[4](Fig.5j.WhenthevoltageswingacrossthecapacitorisequaltothatacrosstheresistorandtheNANDgatesareswitchedoff,thisoscillatorhasaperiodToff:Toff=4RCoff.(15)Sincethevalueoftheresistoriskeptconstant,theperiodvariesonlywiththecapacitorvalue.Now,byswitchingontherightNANDport,thecapacitanceCXcanbeconnectedinparalleltoCoff.Thentheperiodbecomes:Tx=4R(Coff+Cx)=4RCx+Toff(16)TheconstantsRandToffareeliminatedinthewaydescribedinSectionIV.In[2]itisshownthatthesystemisimmuneformostofthenonidealitiesoftheopampandthecomparator,likeslewing,limitationsofbandwidthandgain,offsetvoltages,andinputbiascurrents.Thesenonidealitiesonlycauseadditiveormultiplicativeerrorswhichareeliminatedbythethree-signalapproach.VI.PERIODMEASUREMENTWITHAMICROCONTROLLERPerformingperiodmeasurementwithamicrocontrollerisaneasytask.Inourcase,anINTEL87C51FAisused,whichhas8kByteROM,256ByteRAM,andUARTforserialcommunication,andthecapabilitytomeasureperiodswitha333nsresolution.Eventhoughthecountersare16bwide,theycaneasilybeextendedinthesoftwareto24bormore.Theperiodmeasurementtakesplacemostlyinthehardwareofthemicrocontroller.Therefore,itispossibletolettheCPUofthemicrocontrollerperformothertasksatthesametime(Fig.6).Forinstance,simultaneouslywiththemeasurementofperiodTx,periodTrefandperiodToff,therelativecapacitancewithrespecttoCrefiscalculatedaccordingto(11),andtheresultistransferredthroughtheUARTtoapersonalcomputer.Fig.5.ModifiedMartinoscillatorwithmicrocontrollerandelectrodes.Fig.6.Periodmeasurementasbackgroundprocess.Fig.7.Positionerrorasfunctionofthepositionandestimateofthenonlinearity.VII.EXPERIMENTALRESULTSThesensorisnotsensitivetofabricationtolerancesoftheelectrodes.Thereforeinourexperimentalsetupweusedsimpleprintedcircuitboardtechnologytofabricatetheelectrodes,whichhaveaneffectiveareaof12mm×12mm.Theguardelectrodehasawidthof15mm,whilethedistancebetweentheelectrodesisabout5mm.Whenthedistancebetweentheelectrodesisvariedovera1mmrange,thecapacitancechangesfrom0.25pFto0.3pF.Thankstothechosenconcept,evenasimpledualopamp(TLC272AC)andCMOSNAND’scouldbeused,allowingasingle5Vsupplyvoltage.Thetotalmeasurementtimeamountstoonly100ms,wheretheoscillatorwasrunningatabout10kHz.Thesystemwastestedinafullyautomatedsetup,usinganelectricalXYtable,thedescribedsensorandapersonalcomputer.Toachievetherequiredmeasurementaccuracythesetupwasautozeroedeveryminute.Inthiswaythenonlinearity,long-termstabilityandrepeatabilityhavebeenfoundtobetterthan1μmoverarangeof1mm(Fig.7).ThisiscomparabletotheaccuracyandrangeofthesystembasedonaPSDasdescribedin[2].Asaresultoftheseexperiments,itwasfoundthattheresolutionamountstoapproximately20aF.Thisresultwasachievedbyaveragingover256oscillatorperiods.Afurtherincreaseoftheresolutionbylengtheningthemeasurementtimeisnotpossibleduetothel/fnoiseproducedbythefirststagesinboththeintegratorandtheComparator.Theabsoluteaccuracycanbederivedfromthepositionaccuracy.Sincea1mmdisplacementcorrespondstoachangeincapacitanceof50fF,theabsoluteaccuracyof1μminthepositionamountstoanabsoluteaccuracyof50aF.CONCLUSIONAlow-cost,high-performancedisplacementsensorhasbeenpresented.Thesystemisimplementedwithsimpleelectrodes,aninexpensivemicrocontrollerandalinearcapacitance-to-periodconverter.Whenthecircuitryisprovidedwithanaccuratereferencecapacitor,thecircuitcanalsobeusedtoreplaceexpensivecapacity-measuringsystems.REFERENCES[1]G.C.M.MeijerandR.Schner,“Alinearhigh-performancePSDdisplacementtransducerwithamicrocontrollerinterfacing,”SensorsandActuators,A21-A23,pp.538-543,1990.[2]J.vanDrecht,G.C.M.Meijer,andP.C.deJong,“ConceptsforthedesignofsmartsensorsandsmartsignalprocessorsandtheirapplicationtoPSDdisplacementtransducers,”DigesrofTechnicalPapers,Transducers’91.[3]W.C.Heerens,“Applicationofcapacitancetechniquesinsensordesign,”Phys.E:Sci.Insfrum.,vol.19,pp.897-906,1986.[4]K.Martin,‘‘Avoltage-controlledswitched-capacitorrelaxationoscillator,”IEEEJ.,vol.SC-16,pp.412-413,1981.一种低成本智能式电容位置传感器摘要本文描述了一种新的高性能,低成本电容位置测量系统。通过使用高线性振荡器,屏蔽和三信号通道,大部分误差被消除。其精确度在1毫米范围内达1微米。由于振荡器的输出可直接连接到微控制器,所以无需用A/D转换器。Ⅰ.导言本文介绍了一种新型高性能,低成本的电容位移测量系统,特点如下:1毫米测量范围1微米精确度0.1s总测量时间相应到电容域,规格相称于:1皮法的变化范围;只有这个范围的50fF(fF是法拉乘以10的负15次方。f是femto的缩写)用于位移传感器。50aF绝对电容测量误差。梅耶尔和施里尔[1]以及最近的范德雷赫特河,梅耶尔,和德容[2]提出了位移测量系统,采用一个PSD(位置敏感探测器)作为传感元件。和本文描述的电容传感器元件相比,使用PSD的缺陷是,PSD和LED(发光二极管)有更高的成本和功率消耗。使用[2]中所提概念的信号解决器,被采用到电容元件的使用中。通过广泛使用屏蔽,智能A/D转换,该系统可以将高精确度和低成本结合。换能器产生可以选择和直接由微控制器读出的三段调制信号。微控制器,相应的,计算位移及发送此值到主机电脑(图1)或显示或驱动执行器。电子电路电子电路上位机 执行器演示图1该系统的框图金属屏蔽电极屏蔽金属屏蔽电极屏蔽图2电极结构的尺寸和透视图Ⅱ.电极结构基本传感元件包含电容为Cx的两个简朴电极(图2)。较小的一个(E2)是由屏蔽电极包围。由于使用屏蔽电极,两电极间的电容Cx可平行于电极表面独立运动(横向平移以及旋转)。寄生电容Cp的影响可被消除,将在第3节讨论。据Heerens[3],由有限屏蔽电极大小导致的两个电极之间电容Cx的相对偏差小于:δ<e-π(x/d)(1)其中x是屏蔽的宽度,d是电极之间的距离。这种偏差引入了非线性。因此,我们规定δ小于100ppm。此外小电极和周边屏蔽之间的间距产生一个偏差:δ<e-π(d/s)(2)S是间距的宽度。当间距宽度小于电极之间距离的1/3时,这偏差和(1)相比是微局限性道的。另一个误差的因素也许源自两个电极之间的有限倾斜角α(图3)。假设符合下列条件:小电极和屏蔽电极上的电势等于0V大型电极电势等于V伏屏蔽电极足够大可以看出,电场将同心。ddl/2l/2图3倾斜角度α的电极为了使计算简朴,我们将假设电极在一个方向无限大。问题就成为一个二维问题,可以用极坐标(Υ,φ)方法解决。在这种情况下,电场可以表述为:(3)为了计算小电极的损耗,我们设定φ为0,整定Υ:(4)Bl是小电极的左侧边界:(5)Br是右边界:(6)求解(4)结果:(7)对小α的近似:(8)选择比d小的l似乎是可行的,因此该误差将只决定于角度α。在这种情况下,0.6°的角度变化,将产生小于100ppm的误差。对参数εo和l是常数的设计,两个电极之间的电容将仅仅取决于电极之间的距离d。Ⅲ.寄生电容的消除除了抱负传感器电容Cx,在实际结构中尚有许多寄生电容(图2)。这些电容可以建模,如图4所示。这里Cpl代表电极El的寄生电容,Cp2是从电极E2到屏蔽电极和屏蔽层的。寄生电容Cp3导致不完善屏蔽,形成一个偏移电容。当传感器电容Cx连接到AC电压源,通过电极的电流可测,Cpl和Cp2,将被消除。Cp3可通过偏移测量消除。图4消除寄生电容电流通过并联反馈放大器测量,它具有非常低的输入阻抗。要获取所需的线性度,放大器的单位增益带宽fT必须符合下列条件:(9)T是在此期间的输入信号。由于Cp2涉及电缆电容和运算放大器的输入电容,它很也许大于Cf而不可忽略。Ⅳ.本系统的概念该系统采用了[2]提出的三信号的概念,它是基于以下原则。当我们用线性系统测量电容Cx,得到一个值:(10)其中m是未知的增益,Moff是未知偏移。以相同的方式,通过测量参考量Cref,测量偏移Moff,使m=0,参数m和Moff被抵消。最后的测量结果P定义为:(11)在我们的例子中,传感器的电容Cx为:(12)其中Ax,是电极面积,do是它们之间最初的距离,ε是介电常数,△d是要测量的位移。对于参考电极,它为:
温馨提示
- 1. 本站所有资源如无特殊说明,都需要本地电脑安装OFFICE2007和PDF阅读器。图纸软件为CAD,CAXA,PROE,UG,SolidWorks等.压缩文件请下载最新的WinRAR软件解压。
- 2. 本站的文档不包含任何第三方提供的附件图纸等,如果需要附件,请联系上传者。文件的所有权益归上传用户所有。
- 3. 本站RAR压缩包中若带图纸,网页内容里面会有图纸预览,若没有图纸预览就没有图纸。
- 4. 未经权益所有人同意不得将文件中的内容挪作商业或盈利用途。
- 5. 人人文库网仅提供信息存储空间,仅对用户上传内容的表现方式做保护处理,对用户上传分享的文档内容本身不做任何修改或编辑,并不能对任何下载内容负责。
- 6. 下载文件中如有侵权或不适当内容,请与我们联系,我们立即纠正。
- 7. 本站不保证下载资源的准确性、安全性和完整性, 同时也不承担用户因使用这些下载资源对自己和他人造成任何形式的伤害或损失。
最新文档
- 朱自清春教案课件
- 医疗纠纷应对
- 室外定位技术智慧养老技术概论
- 树立正确职业价值观
- 《选房方式小结》课件
- 《光学工艺与测量》课件
- 系统详细设计流程及范畴
- 微课人力资源规划的程序财经管理人力资源管理系副教
- 外科手术饮食
- 《汽车驾驶虚拟现实》课件
- 民族团结主题班会教学课件
- 国开成本会计第14章综合练习试题及答案
- 幼儿园教育活动设计与指导(第二版)教案第二章第二节幼儿园语言教育活动设计二
- 外观检查记录表
- GB∕T 13171.1-2022 洗衣粉 第1部分:技术要求
- 气温的变化与分布 完整版课件
- 现在完成时的用法 完整版课件
- 中小学古诗词首
- DB11T 1411-2017 节能监测服务平台建设规范
- 外科学教案-心脏疾病
- 白内障手术流程
评论
0/150
提交评论