filters - university of colorado boulder

7
1 FILTERS LAB 3 INTRO: MEASURING THE FREQUENCY DEPENDENCE OF LOW PASS, HIGH PASS, AND BAND PASS FILTERS. GOALS In this lab, you will characterize the frequency dependence of three passive filters. You will gain more experience modeling both the response of the filters and how your measurement tools affect your measurements. Proficiency with new equipment: o Oscilloscope probe o Capacitors and inductors § Identify polarized capacitors and determine the correct installation orientation § Measure capacitance and inductance with an LCR meter. Modeling the physical system: o Develop mathematical models of frequency dependent voltage dividers o Determine the limitations of these models and range of applicability Modeling measurement systems: o Refine the model of scope measurement tool to include capacitance of the coax cable o Refine the measurement system to reduce the effect of the capacitance of a coax cable DEFINITIONS Scope probe – a test probe used to increase the resistive impedance and lower the capacitive impedance compared to a simple coax cable probe. Pass band – the range of frequencies that can pass through a filter without being attenuated. Attenuation band - the range of frequencies that a filter attenuates the signal. Cutoff frequency (or corner frequency or 3 dB frequency), f c – the frequency boundary between a pass band and an attenuation band. f c is the frequency at the half-power point or 3dB point, where the power transmitted is half the maximum power transmitted in the pass band. The output voltage amplitude at f = f c is 1/ 2 = 70.7% of the maximum amplitude. Low pass filter – a filter that passes low-frequency signals and attenuates (reduces the amplitude of) signals with frequencies higher than the cutoff frequency High pass filter – a filter that passes high-frequency signals and attenuates (reduces the amplitude of) signals with frequencies lower than the cutoff frequency Band pass filter – a device that passes frequencies within a certain range and rejects (attenuates) frequencies outside that range.

Upload: others

Post on 09-Feb-2022

4 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: FILTERS - University of Colorado Boulder

1

FILTERSLAB3INTRO:MEASURINGTHEFREQUENCYDEPENDENCEOFLOWPASS,HIGHPASS,ANDBANDPASSFILTERS.

GOALS

Inthislab,youwillcharacterizethefrequencydependenceofthreepassivefilters.Youwillgainmoreexperiencemodelingboththeresponseofthefiltersandhowyourmeasurementtoolsaffectyourmeasurements.

Proficiencywithnewequipment:

o Oscilloscopeprobeo Capacitorsandinductors

§ Identifypolarizedcapacitorsanddeterminethecorrectinstallationorientation§ MeasurecapacitanceandinductancewithanLCRmeter.

Modelingthephysicalsystem:

o Developmathematicalmodelsoffrequencydependentvoltagedividerso Determinethelimitationsofthesemodelsandrangeofapplicability

Modelingmeasurementsystems:

o Refinethemodelofscopemeasurementtooltoincludecapacitanceofthecoaxcableo Refinethemeasurementsystemtoreducetheeffectofthecapacitanceofacoaxcable

DEFINITIONSScopeprobe–atestprobeusedtoincreasetheresistiveimpedanceandlowerthecapacitiveimpedancecomparedtoasimplecoaxcableprobe.Passband–therangeoffrequenciesthatcanpassthroughafilterwithoutbeingattenuated.Attenuationband-therangeoffrequenciesthatafilterattenuatesthesignal.Cutofffrequency(orcornerfrequencyor3dBfrequency),fc–thefrequencyboundarybetweenapassbandandanattenuationband.fcisthefrequencyatthehalf-powerpointor3dBpoint,wherethepowertransmittedishalfthemaximumpowertransmittedinthepassband.Theoutputvoltageamplitudeatf=fcis1 / 2 =70.7%ofthemaximumamplitude.

Lowpassfilter–afilterthatpasseslow-frequencysignalsandattenuates(reducestheamplitudeof)signalswithfrequencieshigherthanthecutofffrequency

Highpassfilter–afilterthatpasseshigh-frequencysignalsandattenuates(reducestheamplitudeof)signalswithfrequencieslowerthanthecutofffrequency

Bandpassfilter–adevicethatpassesfrequencieswithinacertainrangeandrejects(attenuates)frequenciesoutsidethatrange.

Page 2: FILTERS - University of Colorado Boulder

2

BandpassfilterBandwidth–therangeoffrequenciesbetweentheupper(f+)andlower(f–)halfpower(3dB)points:

bandwidth ∆f=f+–f–.

APPLICATIONSOFFILTERS Afrequentprobleminphysicalexperimentsistodetectanelectronicsignalwhenitishiddeninabackgroundofnoiseandunwantedsignals.Thesignalofinterestmaybeataparticularfrequency,asinanNMRexperiment,oritmaybeanelectricalpulse,asfromanuclearparticledetector.Thebackgroundgenerallycontainsthermalnoisefromthetransducerandamplifier,60Hzpowerpickup,transientsfrommachinery,radiationfromradioandTVstations,cellphoneradiation,andsoforth.Thepurposeoffilteringistoenhancethesignalofinterestbyrecognizingitscharacteristictimedependenceandtoreducetheunwantedbackgroundtothelowestpossiblelevel.Aradiodoesthiswhenyoutunetoaparticularstation,usingaresonantcircuittorecognizethecharacteristicfrequency.Thesignalyouwantmaybelessthan10-6ofthetotalradiationpoweratyourantenna,yetyougetahighqualitysignalfromtheselectedstation. Manyexperimentsrequirespecificfiltersdesignedsothatthesignalfromthephenomenonofinterestliesinthepass-bandofthefilter,whiletheattenuationbandsarechosentosuppressthebackgroundandnoise. Thisexperimentintroducesyoutothefilteringpropertiesofsomewidelyusedbutsimplecircuits,employingonlyaresistorandcapacitorforhigh-andlow-passfiltersandanLCRcircuitforband-pass.

FILTERBASICSRCLow-andHigh-passfiltersTheresponseofRClow-passandhigh-passfilterstosinewavesisdiscussedinFCSections3.9&3.10.The3dBfrequencyis

fc =1

2πRC,

wherefcisthe3dBorhalf-powerpoint.

Theresponseofthefilterstoasquarewaveinthetimedomainisalsointeresting.

ParallelLCRBand-passfiltersSeeFCSection3.12(H&HSection1.22).TheresonantfrequencyandQaregivenby

f0 =1

2π LCQ =ω0RC =

f0Δf

whereω0=2πf0.Theresonantfrequency,f0,isthecenterfrequencyofthepassband,andtheQisequaltotheratioofthecenterfrequencytothebandwidth∆f.(ThesedefinitionsareexactlytrueonlyifQ>>1). ForaresonantLCRcircuitthecharacteristicimpedance,Z0,isthemagnitudeoftheimpedanceoftheinductororthecapacitorattheresonantfrequency:

CL

CLZ ===

000

ω

USEFULREADINGS1. FCSections3.4–3.18and10.1–10.62. H&HChapter1,especiallysections1.13-1.24.YouwillmakefrequentuseofthelasttopicinSection1.18,

“VoltageDividersGeneralized.”AppendixAonoscilloscopeprobes.

Page 3: FILTERS - University of Colorado Boulder

3

LABPREPACTIVITIES

AnswerthefollowingquestionsusingMathematica.SavethecompletenotebookasapdfandturnitintoCanvasbymidnightthedaybeforeyourlabsectionmeets.Bringanelectroniccopyofyournotebooktolab,preferablyonyourownlaptop.Youwilluseittoplotyourdataduringthelabsession.

Question1

Low-andHigh-passfiltersa. DefinefunctionsinMathematicatocalculatethecut-offor3dBfrequency,fc,forthelow-andhigh-

passfiltersinFigure1(a)and(b).Theinputparameterstothisfunctionshouldbetheresistanceandcapacitanceofyourcircuit.Evaluatethefunctionsusingthenominalvaluesshownintheschematic.Duringthelab,youcaninputtheexactvaluesofyourcomponentsandthusquicklypredictthe3dByouexpectforyourcircuit.

b. CreatetwoBodeplots(oneforeachfilter)ofthefrequencyresponseofthelow-(1a)andhigh-pass(1b)filtersinFigure1.ABodeplotisalog-logplotof(Vout/Vin)versusfrequency.SeeH&HFig.4.31foranexample.Makesuretoincludealargeenoughrangeinfrequencytoseeboththepassandattenuationbands.HINT:DetailsaboutmakingplotsprettyareincludedinLabSkillActivity#2.

c. Duringthelabsection,youwillenteryourmeasurementsintoyourMathematicanotebookandplotthemwithyourmodelpredictions.Toprepareforthis,createalistof“fakedata”andplotitonyourBodeplots.Thiswillallowyoutocompareyourmodelandmeasurementsinrealtimeavoidinglosttimetakinglotsofdatawhensomethingiswrongwithyourcircuit.Thepointofthispartisjusttohaveyoucreateworkingcodetoenteralistofdataandplotitalongwiththefunction.Thenumericalvaluesofthefakedateareunimportant.HINT:ThereisahelpfulguideonourwebsiteundertheHINTSTabtitled“PlottingdataandtheorytogetherinMathematica.”

Question2

Band-passFiltersa. DefinefunctionsinMathematicatocalculatetheresonantfrequencyf0,thecharacteristic

impedanceZ0,andthequalityfactorQfortheband-passfilterinFigure1(c).Evaluatethefunctionsusingthenominalvaluesshownintheschematic.

b. CreateaBodeplotshowingthepredictedgain(|Vout/Vin|)versusfrequencyoftheband-passfilter.Makesuretoincludealargeenoughrangeinfrequencytoseeboththepassandattenuationbands.

c. Createalistof“fakedata”andplotitonyourBodeplots.Thepointofthispartisjusttohaveyoucreateworkingcodetoenteralistofdataandplotitalongwiththefunction.Thenumericalvaluesofthefakedateareunimportant.

Question3 Labactivitiesa. Readthroughallof the labstepsand identify thestep (or sub-step) thatyou thinkwillbe the

mostchallenging.b. Listatleastonequestionyouhaveaboutthelabactivity.

Figure1Filters.(a)low-pass,(b)high-pass,and(c)band-pass

(c)

Page 4: FILTERS - University of Colorado Boulder

4

SETTINGUPTHECIRCUITSANDPREDICTINGTHEBEHAVIOR

Figure2.GeneralVoltageDividers.(a)resistivedivider,(b)low-passfilter,(c)high-passfilter,and(d)band-passfilter.Step1 BuildingtheCircuits

a. GatherallthecomponentstobeabletobuildthefourcircuitsshowninFig.2Ifyou

cannotfindcomponentsinstockwiththespecifiedvalues,takethenearestinvaluethatyoucanfind,within30%ifpossible.

o Resistivedivider:R1=10kΩ,R2=6.8kΩo Low-passfilter:R=10kΩ,C=1000pFo High-passfilter:R=10kΩ,C=1000pFo Band-passfilter:R=10kΩ,C=.01µF,L=10mH

b. Measureallcomponentsbeforeplacingthemintothecircuit.Recordthevaluesinyourlabbook.Drawdiagramsofallthecircuits.Makesuretousethesamelabelsonthediagramsandforthevaluesofthecomponents.

c. Buildallfourcircuitsonyourproto-board(makesuretheyareallseparate)Step2 UsetheMathematicamodelstopredictthebehaviorofthefilters.

a. Calculatetheexpectedtransferfunctionofthedivider.b. Calculatetheexpectedvaluesofthecut-offfrequenciesforthehigh-andlow-passfilters

usingtheactualcomponentvalues.c. Calculatetheexpectedresonantfrequencyf0andqualityfactorQfortheband-passfilter

usingtheactualcomponentvalues.HINT:Youshouldhavealreadydonethesecalculationsinyourlabprepnotebook.Justentertheexactvaluesofyourcomponents.

Step3 UsetheMathematicamodelstoplottheexpectedthebehaviorofthefilters.

a. Plotyourmathematicalmodelsofallthreefiltercircuits(threeindependentplots)usingyouractualcomponentvalues.Thefrequencyrangeshouldcoveratleastf=10-3fc(orf0)

tof=103fc(orf0)toshowthefullbehavior.b. Afterthe lab iscompletedandyouhaveyourmeasurementsontheseplotsaswell,you

willprintofftheplotsandtapethemintoyourlabbook.Makesuretoleaveroominyourlabbookfortheplots.

HINT:Youshouldhavealreadymade theseplots inyour labprepnotebook. Justenter theexactvaluesofyourcomponents.

Page 5: FILTERS - University of Colorado Boulder

5

SETTINGUPTESTANDMEASUREMENTEQUIPMENT

Step4

Preparetotestthecircuits a. ConnectthecircuitboardtothefunctiongeneratorandtheoscilloscopeasshowninFig.

3.Itisalwayshelpfultodisplayboththeinputvoltageaswellastheoutputvoltageonthescopeatthesametime.

b. Testyoursetupbycreatinga1kHzsinewaveat1voltp-pusingthefunctiongeneratorandconfirmthewaveformfrequencyandamplitudebymeasuringthesignalonthescope.TriggerthescopeontheSync.outputofthefunctiongenerator.

Figure3.TestandMeasurementSet-up.Channel1will“pickoff”thefunctiongeneratorsignalonitswaytothecircuitboard.YoucandothisusingaBNC“T”connectormounteddirectlyontheoscilloscopeinput.

RESISTIVEVOLTAGEDIVIDER

Step5

a. Measurethefrequencydependenceofthevoltagedividera. Connectthesignalfromthefunctiongeneratortotheinputofthevoltagedivider.

Measurethetransferfunction(=Vout/Vin)overalargerangeinfrequency(100Hzto1MHzinapproximatelydecade(X10)steps).Recordyourmeasurementsinyourlabbook.

b. Atlowfrequencies(1kHz),compareyourmeasuredvalueofthetransferfunctiontowhatyourmodelpredictedusingyouractualcomponentvalues.Doesyourmeasurementagreewithyourprediction?Explicitlyrecordwhatcriteriayouusedtodeterminewhetherornotthemodelandmeasurementsagree.Ifthereisahighfrequencycut-off(3dBfrequency),measureitsvalue(wherethevoltageisreducedto0.7ofthelowfrequencyvalue).Recordthecut-offfrequency.

c. Holdontotheresistorsfromthisvoltagedivider.You’llbereturningtothiscircuitattheendofthelab.

LOWANDHIGHPASSFILTERS

Page 6: FILTERS - University of Colorado Boulder

6

Step6

b. Measurethefrequencydependenceofthefilters.a. Connectthesignalfromthefunctiongeneratortotheinputofthelow-passfilter.

Measurethetransferfunction(=Vout/Vin)overalargerangeinfrequency(100Hzto1MHz)inatleastonestepperdecade,withseveralextrastepswithinthedecadearoundyourexpectedcutofffrequency.Recordyourmeasurementsinyourlabbook.Determineandrecordthecut-offfrequencyforthelow-passfilter.Compareyourmeasuredhalfpowerpoint(Vout/Vin=0.707)withthecut-offfrequencycomputedfromtheactualcomponentvaluesused.Includeyourcomparisoninyourlabbook.Thendothesameforthehigh-passfilter.

b. TestthepredictedfrequencyresponsebyplottingyourdatapointsdirectlyonyourtwoBodeplots.Doesthemodelagreewithyourdata?Explicitlyrecordwhatcriteriayouusedtodeterminewhetherornotthemodelandmeasurementsagree.

BANDPASSFILTER

Step7

a. Measurethefrequencydependenceoftheband-passfilter.a. Onresonance,Voutwillbeamaximumandthephaseshiftbetweentheinputandoutput

waveformswillbezero.Findtheresonantfrequencyfobothways.Adjustthefrequencysothat(1)theoutputhasmaximumamplitude(Vout/Vin=max),(2)thereiszerophasedifferencebetweenVoutandVin.Recordbothmeasurements.Whichmethodismoreprecise?

b. TheLCRmetermeasuredtheinductanceofyourinductorataparticularfrequency.Yourinductor’sinductancechangesslightlyatdifferentfrequencies.UseyourmeasurementsoffotogetamoreaccuratemeasureofLonresonancebydoingthefollowing.Comparethemeasuredfowiththeexpectedvalue .Refinethemodeloftheinductorby

calculatingacorrectedvalueofLfromthemeasuredvaluesoffoandC,andusethisrefinedvaluebelow.ComparethisvalueofLtothevalueyoumeasureusingtheLCRmeterinthelab.

c. DeterminethequalityfactorQbymeasuringthefrequenciesatthetwohalf-powerpointsf+andf–aboveandbelowtheresonanceatfo.Recordyourmeasurements.Recallthat

Q =Resonant frequency f 0

Bandwidth ΔfwhereΔf=f+–f–.

HINT:Thehalf-powerpointsarewhereVout=𝑽𝒐𝒖𝒕(𝒎𝒂𝒙)/ 𝟐not𝑽𝒊𝒏/ 𝟐.d. ComparethemeasuredvalueofQwiththatpredictedfrommeasurementsofcomponent

values.Dotheyagree?e. ItiscommoninallelectricalcircuitstofindQvaluesthataresomewhatlowerthanvalues

youpredictusingmeasuredcomponentvalues.Thisisduetoadditionallossesinthecircuit,inthiscaselossesareintheinductor.Measuretheinductor’s“equivalentseriesresistance”(ESR)usingaDMM.Youcanrefineyourmodelbyincludingthisresistanceinyourcircuit.Drawaschematicthatincludesthisresistor.WhatisthepredictedQwhenyouincludethisresistanceinyourmodel?HINT:Seehintssectionbelow.DoesthishavebetteragreementwithyourmeasuredQ?

f. Measurethetransferfunction(=Vout/Vin)asfunctionoffrequency.Useyourmodelpredictiontodecidewhatvaluesoffrequencytotakedata.Plotyourmeasurementsonthesamegraphasyourmodel.Note,yourtransferfunctiondidnotincludetherefinedvalueofQ.

Step8

b. Explorehigh-frequencybehaviorwiththescopeprobe.a. Now,returntothevoltagedividerfromStep5.Measurethetransferfunctionofthe

voltagedividerinafewstepsbetween1MHzandthemaximumfrequencyofyour

1 / 2π LC( )

Page 7: FILTERS - University of Colorado Boulder

7

functiongenerator.NotewhetherVinchangesaswellasVout.b. Avoltagedividercontainingonlyresistorsshouldnothaveanyfrequencydependence.

However,acoaxcablehasacapacitanceof~25pF/foot.Youcouldrefineyourmodeltoincludethiscapacitance.However,inthiscase,refineyourphysicalsysteminsteadbyusingascopeprobe(seedefinitions)inplaceofthecoaxcabletoreducethecapacitanceofthemeasurementprobe.Repeatthemeasurements(andrecordtheminyourlabbook)ofQuestion5part(a)usingthe10xprobetomeasuretheoutputofthecircuit.Thisscopemodeldoesnotautomaticallydetectthepresenceofourscopeprobes,soyouwillhavetogointothescopesettingsandchangetheprobesettingto“10x.”Besuretoputitbackto“1x”whenyouremovethescopeprobe.

c. Doesyouoriginalmodelofjusttworesistorsbetterpredictthebehaviorofthecircuitwhenyouusea10Xprobe?

HINTS:REFINEDLCRBAND-PASSFILTERMODEL

Inductorsoftenhaveconsiderableresistanceastheyarejustwireswrappedaroundaferritecore.Onecanincludethisresistanceasaresistorinserieswiththeinductor.TherefinedmodeloftheQofthissystemis

𝑄0123415 =𝑅𝑅8

𝑅 𝐶𝐿 +

1𝑅8

𝐿𝐶

whereRListheequivalentseriesresistanceoftheinductor.Thisisnon-trivialtoderive.