Non-contact surface charge/voltage measurements

1 TrekApplicationNoteNumber3002 Non-contactsurfacecharge/ voltage measurementsFieldmeterandvoltmetermethod sDr. chargesandpotentialsusingelectrostatic voltmeteris alsodescribedasanuniquemethodthatcombine sadvantagesofboth broadvarietyofinstrumentsthatcanmeasurea nelectricchargeand/orvoltageona eldmetersandvoltmetersbelongtothecate-go ry eldmetermethods. Bothmeasurementtechniquescomewithmany variationsduetoanextraordinary effort putintodevelopingoflow-cost,accuratedevi ces[1 10].Whichmethodisbetter?Hopefully, theanswercanbefoundin thecomparisoneasier, considera voltmeteranda eldmeter, bothusingtheKelvinvibratingcapacitive parallel-plateca-pacitor.

2 Inthiscon gurationanelectric currentIis beinginducedin thesinusoidallyvibratingsen-sor[11,12].T hiscurrentisproportionaltothevalueofthee lectric potentialpresentonthesur-faceundertest[1 2]:I=U dCdt=U ddt 0AD0+D1 sin(!t) ==-U 0A D1!cos(!t)[D0+D1sin(!t)]2(1)Uisthediffer enceofpotentialsbetweenthetestedsurfacea ndthevibratingprobe, [V],D0isaconstantrepresentingtheseparati onbetweenelectrodeandthetestedsurfacewhe ntheelectrodeis notvibrating,[m],D1is theamplitudeof vibrations, [m],!is thecircularfrequencyofvibrations,!= 2 f[rad/s],wheref is a frequencyin [Hz],Aisthesurfaceareaofthesensingelectr ode,[m2], istherelative electricpermittivityofthemate-rialbetwee ntheelectrodeandthesurfaceundertest, 1 forair, 0is theelectric permittivityofvacuum, = 10-12[F/m].

3 ThecurrentsignalI isampli edanddemodu-latedusinga phase-sensitive demodulatorcircuit(Figures1 and2)toproducea voltageVpdi-rectlyproportionalto theamplitudeof eldmetersutilize theprocessedsignalVpis eldmetersFigure1 presentsanelectrostatic eldmeter. Afractionof thedetectedandprocessedvoltageVpis invertedandfedback toa uencedby two electric elds:onecreatedby thetestedsurfaceandonegeneratedby , thegreaterthesurfacevoltage, thesetwo voltagescanceleachother. Po-tentiometerP is usedto establisha constantratiobetweenVsandthemeasuredvolt ageVp.

4 WhenTREK,INC. 11601 MapleRidgeRoad Medina,NY14103 Tel:(585)798-3140 Call:1 800 FORTREK FAX:(585)798-3106 of 6 Copyright 2002 TREK, 0aTrekApplicationNoteNumber3002 Non-contactsurfacecharge/ voltage measurementsFieldmeterandvoltmetermethod sthesumofthetwo eldsequalszero, thestabil-ityofthesignaldetectedby , thepotentialdif-ferencebetweenthesurface andthesensorcanleadtothedischargeanddama geoftheequip-mentif measuredVsis alsosensitive to highvoltageampli ercircuitto replicatethevoltageonthetestedsurface. Theampli edvoltageis thenappliedtothesensorthusnullifyingthee lectric eldbetweenthetestedsurfaceandthesens-ing electrode.

5 Potentialontheelectrode"follows"thepoten tialonthesurface. Inthiscasethereisnothreatoftheeventualdi schargebetweentheprobeandthesurfaceunder test, atleastwithina certainrangeofD0. If thespanbetweenthesurfaceandsensoris toobig,theprobebe-comesin uencedby otherelectric eldspresentin voltmeterTheAC-feedback voltmeterusesa differenttech-niquetoachievespacingindep endentsurfacevoltage/chargemeasurements[ 6].RatherthancancellingtheKelvincurrentI by useofa feed-back DCvoltagewhichfollowsthesurfacetestvolta geto producezeroelectric eld,theAC feed-back methodutilizesa nullifyingcurrentI' tozerotheKelvincurrentI.

6 ThecurrentI' is producedbyexternalgeneratorcircuittunedt othefrequencyof theKelvinsensoroscillations:I0=C dVtdt(2)Therefore, whencurrentsI andI' canceleachother,I0=IU dCdt=C dVtdt(3)AsbothI andI' currentsareinverselyproportionaltospacin gD0, theratiooftheamplitudeofVttoU (theDCtestsurfacevoltage)remainsconstant overthelargerangeofD0. AsshowninFigure3,theVtsignalis obtainedby ampli cationofthecurrentI convertedtoa voltageatthepreampli- er. AthighgainthecurrentI is beingcancelledto a very presentsa comparisonofmeasurementerrorsforastandar d eldmeterandtheTrekmodel520electrostaticv oltmeter.

7 Thedataindi-catethatit isimportanttokeeptheappropriatespacingbe tweenthe eldmetersensorandthetestedsurfaceinorder toconsiderthemeasure-mentreliable. Table showsa briefcomparisonbe-tween eldmeter, electrostaticvoltmeterandAC-feedback electrostaticvoltmeter. Becauseoftheirprincipleof operation,theelectrostatic eldmetersaresuitableformeasurementscondu ctedonrel-ativelylargeareas. They arealsonotasaccurateaselectrostaticvoltm eters. Sincetheresultspro-videdby the eldmetersdependstronglyontheprobe-to-sur facedistanceD0, it ismoreconve-nienttoreadthemastheelectric eldintensityvalues(thusthename, eldmeter).

8 Magnitudeof eldsmeasuredthisway is usuallyhigh,thereforethereisa riskofdischargesbetweentheprobeandthetes tedsurface. Fieldmetersarelessex-pensive thanelectrostaticvoltmeters, sincetheydonotrequirehighvoltagecircuitr y toproduceproperfeedback ,INC. 11601 MapleRidgeRoad Medina,NY14103 Tel:(585)798-3140 Call:1 800 FORTREK FAX:(585)798-3106 of 6 Copyright 2002 TREK, 0aTrekApplicationNoteNumber3002 Non-contactsurfacecharge/ voltage measurementsFieldmeterandvoltmetermethod svibratingKelvinsensorsurfaceunder testphasesensitivedemodulatorVPintegrato roscillatoramplifierpre-amplifierbootstr appedpower supplyFigure1:Electrostatic eldmeter[13].

9 VibratingKelvinsensorsurfaceunder testphasesensitivedemodulatorVintegrator oscillatoramplifierpre-amplifiercircuitc ommonbootstrappedpower supplyFigure2:Electrostaticvoltmeter(vol tagefollower)[13].vibratingKelvinsensors urfaceunder testV+-oscillatoramplifier 1pre-amplifieramplifier 2 VtFigure3:AC-feedback electrostaticvoltmeter[6,10].TREK,INC. 11601 MapleRidgeRoad Medina,NY14103 Tel:(585)798-3140 Call:1 800 FORTREK FAX:(585)798-3106 of 6 Copyright 2002 TREK, 0aTrekApplicationNoteNumber3002 Non-contactsurfacecharge/ voltage measurementsFieldmeterandvoltmetermethod s510152025020406080100120140160180200dis tance from the surface under test, [mm]error, [%]tested with 1 [kV]applied to the 310 [cm2] circular plateTrek Model 520 ESVM fieldmeterFigure4:Comparisontestbetweene lectrostaticvoltmeterand ,INC.

10 11601 MapleRidgeRoad Medina,NY14103 Tel:(585)798-3140 Call:1 800 FORTREK FAX:(585)798-3106 of 6 Copyright 2002 TREK, 0aTrekApplicationNoteNumber3002 Non-contactsurfacecharge/ voltage measurementsFieldmeterandvoltmetermethod sothertypesof eldmetersavailable, forexampleradioactive eldmeters, rotatingvaneunits. Eventhoughtheirconstructionandprincipleo fopera-tionarerelativelysimple, they sufferfromdisad-vantagessuchaspresenceof theradioactive ma-terial,pooraccuracyandhighpowerconsum ptionby thedrive motorof , particularlythevoltagefollowers, canbeemployedfortestsofrelativelysmallch argedareas- they have muchbetterres-olutionthan eldmeters.