CAL POLY POMONA ELECTRICAL AND COMPUTER ENGINEERING DEPARTMENT INTRODUCTION TO ELECTRIC CIRCUITS NOTES AND LABS FOR ECE 207/207L A.P. FELZER AND L.D. GRAHAM AUGUST 2000The true measure of human existence is not "sccess” and the power it brings in this generation, but the perfecting of the art of living, which endures actoss many generations” e Mary Clark - A thought inspired by the writings of Martin Buber ‘Concem for humanity must always form the chief interest of all technical endeavors in order that the creations of our minds shall be a blessing, not a curse, to humankind. Never forget that in the midst of your diagrams and equations. Albert Einstein Men will do what is logical but only after they've exhausted all other possibilities Abba Eban It is important that students bring a certain ragamuffin, barefoot irreverence to their studies, they are not here to worship what is known but to question it Jacob Bronowski e If you're not rejecting 80% of your ideas you're not thinking very hard. Jacobus Ten Broek Talways looked forward to Monday Bill WoottonPREFACE TO THE STUDENTS ‘Two observations have guided the writing of these notes. The first is that most people study new technical subjects only when they have to — when they're stuck on a problem they don't have any other way to solve. And even then, they usually don't want to delve too deep — they only want enough information to solve their problem and nothing more. The second observation is that "education is what you remember after you've forgotten everything you've learned”. In other words it’s not really the specifics that make an education ~ they're quickly forgotten — but instead the general ideas that can be applied to solve many problems. So the student — like everybody else - needs a pressing problem to get him going. But he also needs the subject matter presented in a sufficiently general way if it's going to stick. ‘The modus operandi of these notes is for the weekly labs to supply the pressing problems and for the preceding chapters to provide the needed information — hopefully in a general enough way for the ideas to take hold.TABLE OF CONTENTS GETTING STARTED ‘Welcome To New ECE Majors ... 0-1 The ECE Labs... . 0-2 Overview of BCE 207 And Its Lab 252LL. . . 0-3 CHAPTER 1-DUAL TRACE SCOPES Transition 1-1 Review Of Basic Scope Operation . . . 1-1 Basic Operation Of Dual Trace Scopes . . . 1-4 Some Basic Measurements With Dual Trace Scopes . . . 1~6 Other Kinds Of Scopes. . . 1-9 BBLS ECE 252L Lab 1-Basic Operation Of The Dual Trace Scope ECE 252L Lab 2 - Review of the Basics . . . 2-1 CHAPTER 2 - NONLINEAR RESISTOR CIRCUITS - Not printed CHAPTER 3- CONTROLLED SOURCES 3.0 Transition . 3-1 3.1 Characterizing Controlled Sources . . . 3-2 3.2 General Analysis Of Circuits Containing Controlled Sources . . . 3-3 3.3. The Special Case Of Operational Amplifiers. 3-17 Computer Assignments. . . 3-51 Problems . . 3-52 ECE 252L Lab 3 - Controlled Voltage Sources 3-73 ECE 252L Lab 4 - Basic Resistor Op Amp Circuits 315 CHAPTER 4 - TWO-PORT RESISTOR CIRCUITS - Not printed CHAPTER 5- CAPACITORS AND INDUCTORS 5.0 Transition . . . 5-1 5.1 What Are Capacitors? - 5-1 5.2 What Are Inductors? . . . 5-9 5.3. Changes In Capacitor And Inductor Voltages And Currents ... 5—16 5.4 Equivalent Capacitances And Inductances . . . 5-20 5.5 Coupled Inductors . . . 5-26Problems . 3-29 e ECE 252L Lab 5-I-V Relations For Resistors, Capacitors And Inductors 5-36 CHAPTER 6- FIRST ORDER RC AND RL CIRCUITS 6.0 Transition . . . 6-1 6.1 Basic Analysis Of First Order RC Circuits. . . 6-1 62 First Order RL Circuits... 6-23 6.3 Adding More Resistors And Sources . . . 6-24 64 First Order Circuits With Switches 6-26 6.5 First Order RC-Active Circuits... 6-33 6.6 Approximating The Responses Of Ist Order Circuits 6-38 Computer Assignments. . . 6—41 Problems . 6-42 ECE 252L Lab 6-RC Time Constants . . . 6-58 CHAPTER 7- SECOND ORDER CIRCUITS 7.0 Transition 7-1 7.1 The Natural Responses Of 2nd Order RLC Circuits 1-2 72 Complete Responses Of 2nd Order RLC Circuits. . . 7-29 73 RLC Circuits With "Outputs" Larger Than Inputs . . . 7-35 @ 7.4 LC Circuits With Switches... 7-37 7.5 Pulling Some 2nd Order RLC Circuits From Under The Rug... 7-39 7.6 2nd Order RC Circuits... 7-40 7.7 Coupled Inductors And Transformers 7-44 7.8 Higher Order RLC Circuits 7-44 7.9 Approximating The Responses Of 2nd And Higher Order Circuits... 7-45 Writing Assignment... 7-51 Problems... . 7-50 ECE 252L Lab 7- Transient Responses Of 2nd Order Underdamped Ckts . . . 7-64 CHAPTER 8 - Convolution 8.0 Transistion .. . 8-1 8.1 The Modus Operandi of Convolution 8-2 ECE 252L Lab 8 - Overdamped And Critically Damped 2nd Order Ckis .. . 8-35WELCOME TO NEW ECE MAJORS Welcome to the ECE Department. We hope your program of study at Cal Poly will prepare you for a successful engineering career as well as give you the opportunity to develop other rewarding interests The main objective of our ECE Curriculum is to prepare our students for the engineering profession directly upon graduation. We do not emphasize the highly analystical skills typically prerequisite to graduate work at the University level. Our goal is to put emphasis on the fundamentals and their applications with much hands-on laboratory experience. To meet these goals, we require our students to take a large number of ECE classes and their associated labs - the capstone of which is the Senior Project. The Senior Project gives the student a complete engineering experience from conception to design to implementation. Its goal is to give the student the confidence that he can do engineering, Now for some words of advice, Don't despair if it tums out that you do not have the aptitude or interest in electrical and computer engineering that you thought you would have. ‘The world is full of opportunities in which any given individual can be successful and can make a contribution. And who is to say that any particular job is really better or more important than any other. You should choose a career that not only promises success but one in which you will be happy. If you would like to discuss your situation and find out the options, don't hesitate to see your advisor or any other instructor you feel comfortable with. We are sure they would be more than happy to talk with you. In any event, good luck in your studies at Cal Poly. AN OVERVIEW OF ECE 207 AND ECE 252L ‘The main goals of ECE 207 are: 1. To extend the results of linear resistor circuit analysis developed in ECE 109 to include a, Analysis of linear circuits containing controlled sources and op amps 2. To develop analaytical and computer methods for the analysis of first and second order RC, RL, RLC and RC circuits containing op amps to include ‘a. Finding the differential equations for circuit variables ’b. Finding the form of the natural responseof second order overdamped, critcally damped and underdamped circuits Finding the forced response for typical inputs including steps, pulses and sinusoids Finding the complete responses for arbitary initial conditions ao The main goals of ECE 252L are: 1, To develop procedures for the use of dual trace digital scopes to measure the time responses of first and second RC, RL, RLC and RC circuits containing op ampsINTRODUCTION TO THE ECE LABS LABORATORY EQUIPMENT — The ECE Department will in general supply equipment Tike power supplies, scopes, digital multimeters and so forth. Breadboards are supplied inthe Frochemen Lat ECE 129, but must be supplied by the student in subsequent labs. ‘The student is iso required to supply the components and connectors needed. In particular in the networks sequence the student is required to obtain Asetof atleast 12 clip leads Three 1-1/2 volt size D batteries Resistors, capacitors, diodes and op amps as needed Small screwdriver Four BNC to clip connectors (not required in ECE 129) wane PRE-LABS — Pre-labs are to be done before the beginning of the period that the lab is (0 be done. LAB REPORTS ~ The student is required is to write a lab report for each experiment. The report is due at the beginning of the next lab session after the experiment has been performed. No lab reports will be accepted late unless there are extenuating circumstances. Lab reports must be able to stand alone. In particular, @ reader should not have to refer back 10 the Experiment to figure out what measurement is being made or which question is being answered. In addition 1, All voltages and currents discussed in your report must be indicated ~ along with their reference directions - on corresponding circuit diagrams _ AI currents measured and calculated — arc to be conventional currents ‘All calculations must be shown ‘AML graphs must have titles and all axis labelled | Always show units Whenever possible, use Tables for data and for comparing calculated and measured values * The use of color is often a great aid to clarity. So is the use of lines, arrows, boxes and spaces, SAUELN In short, the overall goa! is to write reports in such a fashion that the reader can extract aS much information as much information as possible as efficiently as possible. The extra time an, engineer ‘kee in structuring and documenting his reports as well as his hardware and software will, in the Jong run, be well worth the effort.CHAPTER 1 DUAL TRACE SCOPES 1.0 TRANSITION The scopes we used in ECE 129 were dual trace scopes — scopes capable of displaying two signals at once — even though we only used them to display one signal at a time. The objective of this chapter is to take the big step and make use of our scope's dual trace capability. As we shall see, dual trace scopes are particularly convenient for making measurements of such quantities as gains, phase differences and floating voltages — voltages across circuit elements not connected to ground. 1.1 REVIEW OF BASIC ANALOG SCOPE OPERATION As we indicated in ECE 129 a scope is really nothing more than a cathode ray tube that traces out waveforms like 0 volts as its electron beam moves across the screen. As we also know from ECE 129, the movement of the beam is caused by horizontal and vertical deflection plates as indicated inVertical Deflection e Plates Horizontal Deflection a Plates with the horizontal deflection plates causing the beam to sweep across the screen as the probe voltage applied to the vertical deflection plates causes the beam to move vertically up and down, ‘Asa result we see on the screen the same curve we would have obtained if we had plotted the corresponding circuit voltage on a graph as a function of time. 1.11 THE HORIZONTAL OR TIME BASE CONTROLS ‘The time base control sets the rate the beam is being swept horizontally across the screen. If, for ‘example, the time base control is set at 1 ms/div e the beam will sweep across the ten division screen in 10 msec. So, for example, if a signal repeats say every 4 divisions, then it has a period of T=4 msec and a frequency f= IT = 250Hz Note that when we change the time-base control setting, all we're really doing is changing the voltage across the horizontal plates - thereby increasing or decreasing the speed the electron beam moves horizontally across the screen, Note also that as soon as the electron has been swept all the way across, it quickly returns to get ready for another sweep. 1.1.2 THE VERTICAL CONTROLS The vertical deflection of the beam is determined by the probe voltage — the voltage applied to the vertical deflection plates. ‘The scaling is determined by the VOLT/DIV setting of the horizontal controls. For example, if the vertical controls of our scope are set to 2 VOLTS/DIV then the sinusoid0 volts CEP is 6.cos 103. Remember that the zero reference can be set by setting the AC/DC/GND switch to GND and then adjusting the vertical position control knob. 1.1.3 THE TRIGGERING CONTROLS ‘The most common oscilliscopes are really only good for displaying periodic signals. The reason is that almost as soon as the electron beam traces out a waveform on the screen, the display will staft to fade. So the trace has to be "refreshed". But to be able to refresh a given trace, the signal has to repeat itself. And not only must the signal repeat itself but the sweep of the beam across the e screen must be synchornized with the waveform — the beam must start each sweep at the same location on the waveform. This synchomizing of scope and signal is done by the triggering circuitry. To get the scope, for example, to trigger at the point indicated on the following sinusoid T LI 0 volts it we would (1) Set the scope to trigger on CH 1 e 1-3(2) Set the Trigger Level so the triggering will occur at the voltage level indicated (3) Setthe Slope to (+) since the voltage is increasing at the point we wish to trigger @ Then we would see on the scope 0 volts 1.2 BASIC OPERATION OF DUAL TRACE SCOPES Dual trace scopes operate pretty much the same as discussed in our review except that they have two probes and as a result can display two signals at once. We refer to these two signals as the signal at Channel 1 and the signal at Channel 2. @ 1.2.1 DISPLAYING SIGNALS ON DUAL TRACE SCOPES To see the different signals that can be displayed on a dual trace scope let us take a look at chi ch2 Yin with the probes from CH 1 and CH 2 connected as indicated and the ground connections of each probe connected to the common ground. ‘Then we can display — 1. CHI alone = vin(t)| e 2. CH2 alone = volt) 3. CHI and CH2 simultaneously in the ‘a, Chop Mode — in which case the electron beam alternately displays part of Channel 1 and then part of Channel 2 as it moves across the screen. The Chop Mode works best for low frequencies corresponding to SEC/DIV settings slower than ms/div. b. Alternate Mode — In which case the electron beam altemately sweeps out a whole trace of Channel 1 and then a whole trace of Channel 2, The Alternate Mode is best for high frequencies corresponding to SEC/DIV settings of faster than 1 ms/div. 4. The difference CH 1 ~ CH2 = vin(t) - vott) = vil) Note that for this display to give valid results, both CH 1 and CH 2 must be set for the same vertical scale. 1.2.2 TRIGGERING OF DUAL TRACE SCOPES In ECE 129 we considered only the case of triggering by the one signal being displayed. But dual trace scopes can be triggered on either of the signals being displayed - the one on CH | or e CH 2. In addition the scope can be triggered by a. External Tiggering — in which case the trace is triggered by the signal we connect to the scope's External Trigger Input. In particular, for Tr Ext Trig ea Yn v0 the scope will display vo(t) but trigger on vin(t). The result is that the trace of vo(t) will start when vin(t) reaches the trigger level — independent of what vo(t) happens to be at that time. b. Line Triggering — in which case the trace is triggered by the 60 Hz power signal.We specify which signal is to do the triggering with the Triggering Source Switch. e 1.3 SOME BASIC MEASUREMENTS WITH THE DUAL TRACE SCOPES 1.3.1 MEASURING VOLTAGE GAIN All we have to do to measure the voltage gain G = Vo/Vin in a circuit like chi ch2 Yin® vo is connect CH 1 and CH2 as indicated and then simply take the ratio of the magnitudes of the displayed sinusoids vo(t) and vin(t). 1.3.2 MEASURING FLOATING VOLTAGES e Floating circuit elements are circuit elements that are not connected to ground. All we have to do to measure the floating voltage vj(t) ina circuit as above is switch the display mode to CH1 - CH2 Note that we can't in general use single trace scopes to measure such floating voltages. In particular if we try to use a scope like our Tektronix — with its ground internally connected to the ‘ground of the source as indicated in@ _ tren & is going to be shorted out. We call this very untenuous situation a ground loop. 1.3.3 MEASURING PHASES : ‘Suppose we would like to measure the phase difference @ between vin(t) = A cos ot and Vo(t) = Bcos (wt + 8) in the circuit Yin @) vo Then all we have to do is display both vin(t) and vo(t) on the scope at the same time as follows © volts t Sy and simply calculate the phase difference from = Number of divisions between crossover points ( 369° ‘Number of divisions per cycle For our example the phase difference is @ 1-71 @ = — (360 ) = 45 8 @ If, on the other hand, we want to measure the phase difference between vin(t) and the floating voltage vi(t) as in chi 10 Vin) then we would have to go through a procedure something like 1. Display CH 1 and adjust the zero crossover point to a convenient position 2. Without changing the triggering or horizontal controls, display v() = CHI - CH2 e 3. Then take the number of divisions between the zero crossover points of where vin(t) was displayed and where v(t) is now displayed and calculate the phase difference as before Note that when making phase measurements, you can take both the common time base and vertical scale controls out of calibration. In particular, you can adjust the time scale to obtain a nice convenient number such as 10 divicycles. But don't change the time scale once you have started making measurements. 1.4 BELLS AND WHISTLES More sophsticated scopes of course have many features. Some interesting such features are as follows. 1.4.1 SWEEP DELAY Scopes having sweep delay allow us to select a portion of a waveform as in 1s eand display it across the whole screen as follows ‘What the scope does is delay the trigger until the selected part of signal is reached and then speeds up the sweep so the selected part of the signal is displayed across the whole screen. 1.4.2 STORAGE SCOPES Sometimes signals occur only once — they are not repetitive. As a result, the corresponding trace would quickly fade away on a regular scope. To overcome this problem, storage scopes have memories to store the signal and “refresh” the trace by "replaying" it over and over. 1.4.3 SPECTRUM ANALYZERS To obtain the "frequency response” of a circuit — the gain of the circuit as a function of frequency ~ we must calculate the gain Magnitude of the Output Magnitude of the Input Gain as we vary the frequency. Spectrum analyzers, however, have a built in frequency generator ‘whose frequency automatically increases through the range of interest. ‘The output of the spectrum analyzer is then a plot of the Gain versus frequency ©ECE 252L - EXPERIMENT 1 BASIC OPERATION OF THE DUAL TRACE SCOPE @ OBJECTIVES The student will be able to use a dual trace scopes to measure gains and phases MATERIALS REQUIRED BY STUDENT 1. 12 Clip leads 2, 5% 1/2 watt resistors of value 1K (two of these) and 2K 3. 3 BNC toclip connectors PRE-LAB — Due at the beginning of the lab 1, Given vO) = 5cos 2m1000t a. What is the frequency in Hz, in radians/sec b. What is the period c. How many divisions are required to display one complete cycle if the time base is set to 0.2 msecidiv 2. Given e v(t) = 2 + 3. cos 2m1000. Sketch what you would see on the scope if the input coupling switch was on a DC b. AC c. GND 3. Given 0 voltsFind v(t) if the INPUT-COUPLING Switch ison DC and the scope settings are 2 e volts/div and 1 usec/div 4. Given @ Ms @ 5 2 j 2nl0*t cos 2x ) vo = @ with CH 1 connected at 1, CH2 connected at 2, both their grounds connected at 0 and the following scope settings — Slope = + Coupling = AC Trigger Level = 2 volts Time Base = 2 msecidiv Vertical Deflection = 2 volts/div Trigger Source = CH 1 e ‘Sketch what you would see when you select cH I CH 2 Chop CHI - CH2 Be ge 5. Given (0) IK @ () 2K SY Scos 2n10 with the External Trigger at 1 and CHI at 2 and scope settings Slope = + Trigger Level = 2 volts ‘What will be the value of vo(t) displayed on CH 1 when the scope triggers é T-1l6. Given e S.cos 2103 t @ SCOPE 9 oO a, What is vj(t) before the scope is connected b. What is the magnitude of vy(1) displayed on the scope ¢c. Why did the student who made this measurement change majors — in other words, why did v1() change when the scope was connected 7. Given two sinusoids of the same frequency displayed on CH1 and CH2 as follows v0 10 divisions —] LS divisions between v0 SS crossover points find the phase of v2(t) if the phase of vi(t) is 0° 8. Explain in your own words what triggering is and why it is necessary LAB ~ Each student is to do his own experiment 1. Turn on the scope and obtain a horizontal trace. Set Trigger Source on CHI and Level on auto. Adjust the focus and intensity to give a good clear picture with minimum brightness. 1-122. Move the horizontal trace up and down with the vertical position control. Leave it set in the e center of the screen 3. Slow down the sweep rate until it appear that a spot of light is moving rapidly across the screen, Tum the red knob in the center of the SWEEP TIMEJ/DIV control a. Explain what effect, if any, the red knob has on the sweep speed . Explain what happens when the large switch is tumed to the SWEEP DISABLED position (fully counter-clockwise) c. Calculate the maximum and minimum frequencies for which the oscilloscope can display one complete cycle of a sine wave across the 10 division screen width using "calibrated" sweep Return the red knob to the calibrated position 4. Move the horizontal position control over its full range. What happens. Leave it set to make the sweep start at the left edge of the screen 5. Display vin(t) = 5ccos (210001) from the signal generator on CH 1. Adjust the controls so one cycle is displayed in 6 divisions, Be sure to return the red knob to the calibrated position if you took it out of calibration. 6. Given @ +. @ e 1 : Scos 20103 t 2K & YRO =©@ with CH1 at 1 and CH2 at 2 sketch and specify which voltage va, VB or Vin you see when you select a CHI b, CH2 c. CHOP d. CHI - CH2 7. Given YL 1K : Scos2nio> 2x = 20 1KDisplay and write an expressions for vj(t) and v(t). Demonstrate your measurements to Deine e 8. Given @ z Scos 2n103t 2x = Yo Display and sketch vo(t) with the scope triggered by the External Trigger at 1 9. Given ® IK @ é *in® *) 2x S YQ with CH1 at 1 and CH2at 2. describe the difference in the displays between CHOP and ALTERNATE when 00 Hz e a, b. c. 4. www a, Measure the magnitude of vo(t) (Beware of common grounds) b. Measure the phase difference between vo(t) and vin(t) cc. Write an expression for vo(t) 11, What is the input resistance of CH 1 and CH 2 as printed on the front of the scopee QUESTIONS AND PROBLEMS 1. About how large can the resistors in your circuit get before yout circuit would be loaded down by the scope 2. Given 0} 10 5 ‘Sketch what will be displayed if the trigger level is set at a. 7 volts b. O volts 3. Describe how you would measure voltages in a circuit if the all the sources have a common ground. e 4. What does a "times 10” probe do 5, Write and put in matrix form the node equations of 1K Yin = Ab Scos 103 3K 2KECE 252 - LAB 2 - REVIEW OF THE BASICS > SPRING 1999 AP. FELZER OBJECTIVE — The objective of this lab is to review node analysis, equivalent resistance, transfer functions and Thevenin equivalents of resistor circuits with sinusoidal inputs. 1. Given the following circuit i 4 RO - © do worsen le @ "xe § “7k PARTNER |: R=1K PARTNER 2: R= 2K Measure your resistor values. Compare with nominal values Measure the node voltage v(t) Use node equations to solve for vi(t). Then compare your calculated and measured amplitudes for vj(t) Measure vp(t) and then use it to calculate int) Make use of current division to calculate ii(®) Now make use of your vj(t) to calculate iy(¢) and then compare its amplitude with your result from part (e) 2. Given the following circuit e AW ir® 4 vRIO _ © ire + vRO_| @ IK vs(t) = 5 cos (2n10%1) 1K x PARTNER I: R=1K PARTNER 2: R=2K a. Measure your resistor values. Compare with nominal values . Display the node voltages on the scope. Make use of what you see to draw pictures of and write equations for vj(t) and vat) ¢, Make use of your node voltages to calculate the voltages vRi(t) and vR2(t) d. Measure vpi(t) and vpo(t) ¢. Compare the amplitudes of your calculated and measured vpi(t) and vRo(0) £. Make use of your values for vp)(t) and vpo(t) to calculate ig (t) and in2(®) 3. Given the following circuit im 1K @ v(t) = 5 cos (271031) 4.7K a: : € PARTNER I: R= 1K PARTNER 2: R= 4.7K4. Given the following circuit gaece Make use of your scope to determine i(t) Measure your resistor values. Compare with nominal values Make use of your result in part (b) to calculate the circuit's equivalent resistance @ Reg = vs/i Make use of your result in part (c) for Reg to predict v(t) inthe following eireuit IK 1K + ve(t) = 5 cos (271031) (F vit) 4.7K PARTNER I: R=1K PARTNER 2: R=4.7K Now measure vj(t) in your circuit and compare its amplitude to your prediction in part (d) e s(t) = 5 cos (21031) ® 1K : R ean PARTNER |: R=1K PARTNER 2: R=4.7K Measure your resistor values. Compare with nominal values e Measure Vo(t) Make use of your result in part (b) to find the voltage transfer function G = vo/vs Make use of your result for G to predict vo(t) when vs(t) = 7.5 cos (277103t) Measure vo(t) when vs(t) = 7.5 cos (27105t) and then compare the amplitudes of your measured and calculated results 5. Given the following circuit sacgp 1K —w— R 2K = s(t) = 5 cos (201031) PARTNER I: R=1K PARTNER 2: R=4.7K Measure your resistor values. Compare with nominal values Measure Vy, Use voltage division to calculate Vr. Compare your measured and calculated values for Vy Measure Rrq by measuring the equivalent resistance of the circuit with vs(t) set to zero (replaced by a short) @f. Now make use of the following scheme to find Rr without setting vs(t) to zero. Do this by measuring v;(t) with a 1K resistor connected to the circuit as follows 1K + R nog IK xg . vs(t) = 5 cos (271031) PARTNER I: R=1K PARTNER 2: R=4.7K and then analyzing the corresponding circuit Rry + v(t) 21K with vi(t) and Vy known g. Compare your values for Ryu in parts (e) and (£) h. Draw your Thevenin Equivalent circuit Make use of your Thevenin Equivalent to predict v(t) in the following circuit IK IK R vat) = 5 cos (27103) PARTNER 1: R=1K PARTNER 2: R=4.7K j. Then measure v)(t) in your circuit and compare its amplitude with your prediction in part (i)CHAPTER 3 CONTROLLED SOURCES 3.0 TRANSITION In the previous chapters we've looked at circuits that can attenuate signals and circuits that change signal's shape. The objective of this chapter is to start on the road to understanding circuits that can amplify — that can make signals bigger. Such circuits work basically like the following mechanical "amplifier" ‘where the amount of water flowing thru the pipe is controlled by the valve. We refer to this as a "amplifier" because the energy it takes to turn the valve is in general much less than the energy being controlled - the energy of the water, To illustrate this amplification, let's suppose that the distance of the valve stem from the middle of the pipe is varying sinusoidally with time as follows Distance of valve stem from center of pipe ‘Then the water will also flow sinusoidally as indicated in Rate water is flowing and so we've got ourselves an amplifier. Now this is exactly the same kind of thing that goes onin electronic amplifiers — in particular in transistors like coal except, of course, with voltages and currents rather than water. ‘The main objective of this chapter is to introduce controlled sources — the circuit elements that model the valve action in biased transitors — and to do the corresponding small signal analysis of circuits containing them 3.1 CHARACTERIZING CONTROLLED SOURCES Controlled sources — also referred to as dependent sources - such as vy 100V. have two main characteristics e (1) First of all they're multiterminal — they have more than two terminals (2) And second they are sources very much like our independent sources except that their values are controlled by — are dependent on — voltages and currents in other parts of the circuit. In particular, the value of the controlled voltage source shown above depends on the voltage V; in another part of the circuit ‘We draw controlled sources with diamonds to distinguish them from "regular" independent sources like batteries and power supplies. Altogether there are four different kinds of controlled sources — 1. Voltage Controlled Voltage Sources - VCVS ay wv The voltage across the dependent source is j1 times the voltage v. Note that 1 is dimensionless 3-2 @2. Voltage Controlled Current Source - VCCS The current through the dependent source is g times the voltage v. Note that g has the units mhos of conductance 3. Current Controlled Voltage Source - CCVS yi Oa The voltage across the dependent source is r times the current i, Note that r the units ohms of resistance. 4. Current Controlled Current Sources - CCCS The current through the dependent source is Bi times the current i, Note that f is dimensionless. We're now going to look at equivalent circuits for some typical amplifiers. As promised these equivalent circuits will contain controlled sources. 3.2 GENERAL ANALYSIS OF CIRCUITS CONTAINING CONTROLLED SOURCES We analyze resistor circuits with controlled source in basically the same way as we analyze resistor 3-3circuits without controlled sources. The only difference is that in general we have to add extra equations for the controlled sources as we'll see in some examples @ EXAMPLE 3.1 To find Vj, for ‘we proceed as usual and simply write the corresponding Node Equation 1 +e 103 2 o.iv, + YL + ML = 2x10? 108 put them in matrix form « 2 0y\)_(10" 200 3)y,) Lo and solve to obtain V, = -33mV EXAMPLE 3.2 To find the transfer function G= VL/Vin ofwwe again, as usual, write the Node Equations 1 Vi-Vin, Wi, Vi-W 108 2x10? 2x108 2 ¥2-Vi 4, oy, + 220 2x10* 3x103 put them in matrix form | 4 3] v Vin 507s |lv, 0 solve for V2 V2 = - 1.94Vin volts and then Vi. Vi = 2/3 0.65Vin volts from which we finally obtain the transfer function we're after = Sb. 4 Gey 0.65 AN IMPORTANT OBSERVATION It's interesting to note from our examples that we can calculate all the circuit variables without having to write node equations at grounded controlled sources. The situation is analogous to our not having to write node equations at grounded independent voltage sources. Writing node equations at such nodes is of course not forbidden but doing so would only require us to introduce as unknown variables the corresponding currents going through them — and so nothing would be gained. We'd get an extra equation but also an extra unknown. 3.2.1 EQUIVALENT RESISTANCES OF CIRCUITS CONTAINING CONTROLLED SOURCES Just as it was for the 2-terminal resistor networks of ECE 109, it tums out that two-terminal resistive networks with dependent sources like: 3 3 also have I-V characteristic equations of the form, V = Reql and so can be replaced — as far as the outside world is concemed - by equivalent resistances Reg. EXAMPLE 3.3. To find the equivalent resistance Req = Vin/lin of the following circuit lin Kk “IQ «x ”%@® 2K Vin 2K > 1K we (1) Write the node equations - which we already did in the last example 5975 4 ] yy, Vin 2 (2) Solve for the voltages we need to calculate lin ~ which in this case is V Vi = 0.016 Vin (3) And finally calculate vj y; Vis = Mi = = Vin Rin = Via in Kn (Vin = Vi) [108 (Vin ~ 0.016 Vig) [10° 3-6e which gives us EXAMPLE 3.4 One way to find the equivalent input resistance of 5 is to proceed as in the previous example and e (1) Write the node and controlled source equations Node Equation My ¥, Vv v 1 A+ te joo - Sl = 0 2x10" 10 2 sol + Y2 = 0 103 Controlled Source ‘Equation Vin - Vi 1 res 2x10° (2) Solve for the voltages we need to calculate Jin ~ which in this case we can do by substituting the controlled source equation into the first node equation to obtain 3-7(3) And finally calculate in ~ Vi) /2xt0° © (Vin ~ 699 Vin) /2510° which gives us But for this particular circuit it is probably easier to simply write the mesh equation at the input taking into account the fact that the current through the parallel combination of the 1K and 100 ohm resistor is S11. This gives Vin = (105jn + 103lin + S11(103 Il 100 ) = 6600Iin and so ‘All we have to do to show that all 2-terminal resistive networks containing dependent sources can be characterized by their equivalent resistances is simply put in general terms what we've done in the previous two examples. 3.2.2. SUPERPOSITION FOR RESISTIVE CIRCUITS CONTAINING CONTROLLED SOURCES We will not prove it but it's straightforward to show that a circuit containing dependent sources can be analyzed with superposition by (1) First finding the response to each independent source with all the others turned OFF (2) And then adding all the separately calculated responses to obtain the total response Vo=Vor+Va+ ° °° + Von 3.2.3 THEVENIN EQUIVALENTS OF RESISTOR CIRCUITS CONTAINING CONTROLLED SOURCES 38 eSince resistive circuits like the following ww < = ) 8] can be analyzed with superposition, then they ~ just like their cousins in ECE 109 ~ can be characterized by V = Regl + VEQ and therefore can be replaced by Thevenin Equivalents. But care must be taken when calculating Reg. In particular the dependent sources must be LEFT ON when Req = V/I is calculated. As long as some independent source is causing a given controlling variable to be nonzero the controlled source will be generating a real live signal! ‘This can be verified by looking at the form of the node equations of general resistor circuits containing dependent sources. EXAMPLE 3.5 Given 2K Ry=1K Find the Thevenin Equivalent Resistance Ryy as seen by the load Ry. With Vin set to zero the cireuit reduces to 1K + O1Vv,Let us take the easy way out. Since V, must equal zero, the controlled current source must be e equal to zero and so we are simply left with 2K for which EXAMPLE 3.6 To find the Thevenin Equivalent Resistance Rry as seen by Ry in 1K @ 2K @ 2K vin @) yoryim Sek e we turn off the independent source and then write the node equations for 3-10 eEXAMPLE 3.7 To find the Thevenin Equivalent Resistance Rrq as seen by the load Ry, in e the circuit of Example 3.4 we proceed as usual to tum off the independent source Vin, add a source Vos at the output in place of RL IK v,@ and then write the node and controlled source equations to obtain Node Equation 1 Vos 4 Vos _ 19, - sol = 0 e@ 2x1 108 Controlled Source — Equation I Tis =o 2x103 from which we obtain Rr = 298 = 380 los — avery small value. The reason is that the large current thru the controlled source causes Ips to be very large. This result is not only very interesting but also very useful. Note that in general we must add controlled source equations to our node equations when the controlling Variable is not a node voltage — in particular, when the controlling variable is a current ora floating voltage. Let us now illustrate another way to calculate the Thevenin Equivalent Resistance originally é 3-Hintroduced in ECE 109 which turns out to be very useful in the lab. First of all, given that the Thevenin Equivalent as seen by a resistor Ry, e is of the form RL we can e (1) Obtain Vu: by simply calculating or measuring the open circuit voltage Voc wv + vm &) (2) And obtain Rrx by connecting anice “healthy” load resistor of known value Ry to N N eR and then from either calculations or measurements of V1, and Iz, proceed to calculate 3-12 ee Rru by analyzing the circuit Vv. + TH } Ry in which everthing is known except. Rrp which is given by Von - Vy Ry = = Ned EXAMPLE 3.8 - Using the above method to find the Thevenin Equivalent Resistance as seen by the load of we (1) First find Vir by writing and solving the node equations to obtain the open circuit voltage Voc= ~ 0.9785 volts (2) And then find Rrw by connecting a load Ry, e calculating Vi = -0.9765 volts and Ty = ~0.9765 ma and then analyzing Rqw TL = -0.9765 v # VqH =-0:9785 ¥ Vi, = -0.9765 v to obtain Rey = eT Ye) | 2 mv ae TI. ~ 0.9765 ma Rr = 29 ‘And finally one last but very interesting Thevenin Equivalent example. EXAMPLE 3.9 - The Thevenin Equivalent Resistance as seen by Ry in this circuitmust be zero since the transfer function we Vin S G= > 7 is independent of Ry! 3.2.4 VOLTAGE GAINS OF CIRCUITS CONTAINING CONTROLLED SOURCES One convenient way to calculate the voltage gain Vi Gare Vin of a general circuit like the following —wv— + Vin ® N Ru 3 is to make use of the Thevenin Equivalent as seen by the load. Then G is simply the gain of the following circuit where K is equal to the open circuit voltage gain of the circuit Rrwwhich is certainly easy to calculate. But this approach is more than computationally efficient. It e makes clear that the smaller Ry, the: smaller is going to be the gain G no matter what combination of resistors and dependent sources is inside N. We say Ry, is loading down the circuit, 3.2.5 POWER GAINS OF CIRCUITS CONTAINING CONTROLLED SOURCES As illustrated in the following example, an interesting thing happens when we innocently calculate the power gains of circuits containing dependent sources. EXAMPLE 3.10 _ Find the power gain of the following small signal equivalent cireuit of a simple transistor amplifier 1K VW a Ip o1v, ¢ og 3 yy 2K KS vy Writing and solving the node equations and then substituting into e VilL we obtain (-3.3x107? )(- 3.3x10-5) = 2178 (10 )(sx10-7) We know that we can't get power for free. So where does the power come from. As it turns out the power is coming from the DC supply that establishes the operating point of the transistor. So far so good. As a matter of fact things are pretty nifty. But as good as amplifiers in our examples are — they do have a problem. It's hard to build them with specific gains that won't vary with things like changes in temperature. But luckily there is a way to build more “robust” circuits as well see in the next section33 THE SPECIAL CASE OF OPERATIONAL AMPLIFIERS e Op amps — operational amplifiers — are simply high gain voltage amplifiers, They get their name from the fact that one of their mair applications is to perform basic mathematical operations such as addition, subtraction, multiplication, integration and so forth. They were originally made with tubes but did not gain widespread use until the advent of integrated circuits, Today they're ubiquitous. The goal of this section is to get a feel for how basic op amp circuits wotk and to develop effective ways to analyze them, 33.1 CHARACTERIZATION OF OP AMPS Op amps as symbolized by + Voc - Voc rr are simply transistor circuits which when biased by the dual DC supplies +Voc and —Ve¢ e have small signal signal characteristic curves (curves that specify their responses with respect to their operating points) of the form Vo +Voo Wy -Vv_) -Veo We refer to op amps as differential voltage amplifiers because in the linear part of the characteristic curve where IVol < Voo as indicated in 3-17Vo i Slope = A *Yee @ (W4-V_) - Voc the output is proportional to the difference between the inputs ~ in particular Vo = A(Vs - V-) where A = gain of the op amp = slope of the V/O characteristic curve in the linear region From the characteristic curves we see that the magnitude of Vg cannot get larger than Voc. When the op amp is "trying" to getlarger than Veq butcan't — when IVy = VI > Voo/A e we say the op amp is in saturation. ‘Typical op amps are characterized by a. Very large gains A (= 105) b. Very large input resistances at the V, and V_ terminals (~ 100 Meg) ¢. Relatively small output resistances Ry (= 75.2) As a result of these values we can very often get away with modelling op amps in garden variety applications with one of our controlled sources as follows v. + AV, - Via Vywhich has infinite input resistances at V4 and V_ and zero output resistance. As a matter of fact we can often go one step further and assume that the gain A is infinite in which case we then refer to our model as the ideal op amp model. (Our modus operandi in this chapter is to take the op amp as a magical circuit element that has simply been given to us. Our objective is to see how some basic op amp circuits work. 3.3.2 UNDERSTANDING BASIC OP AMP CIRCUITS THE BASIC NEGATIVE GAIN OP AMP CIRCUIT The basic negative gain op amp circuit is We refer to it as being a negative gain op amp circuit because the output R; Vo=- ra Vin is minus (Ro/Rj) times the input Vig. To get a handle on what's going on and to see where the expression for the gain comes from let us analyze the circuit with Vin=2volts Ry=1K R,=1K Assuming the op amp is characterized by a. Againof A= 105 b. Infinite input resistances at the + and — terminals c. Zero output resistance Ro we can first of all conclude that the currents I thru the two resistorsVin =2¥ must be the same since there can be no current flowing into the V_ terminal with its infinite input resistance. So the voltage drop from the input to the output is 2 Vin — Vo = 1031 + 1051 from which we have the important result Now let's suppose that Vin has just turned ON and see what happens — (1) The Source Turns ON ‘Assuming that the source is tumed ON at time t=0 and goes right to Vin =2 volts e as follows Vin then the current I at time t=0 will be Vin® - Yo _ 2-0 10) = 3g = 7 = hme as indicated in 3- 20I (ma) where we have assumed that Vo(0) =0 since the op amp is going to take at least some time to get going. @) The Op Amp Takes OFF Now with V0) = -1031 + Vin = -1+2 = I volt the input to the op amp is going to initially be V4(0) - V0) = 0 - =1 volt and so the op amp is going to take off as if it's going all the way to A(¥.(0) — V-(0)) = 105(-1) = -105 volts e as indicated in ; But as Vo starts on its way to —105 volts, the current thru the 1K resistors 2- Vo 2K will also start increasing | 3-21I (ma) causing - and this is what is significant — the op amp’s differential input Va. - VL =-V. = Vin - 1051 = 2 - 1031 to start decreasing in magnitude W,- Vv) So the increasing of the op amp output has the affect of decreasing the op amp's input. ‘The growth of the op amp output is self limiting. (3) The Op Amp Skids To A Halt So it's not going to be possible for Vo to make it all the way to -105. The growing output is choking off the input. As a matter of fact the output can't even make it all the way to the colossal value of —2 volts because a value of Vo =—2 volts would mean that which would mean that V. = 2 - 2x101 = and so V4-V.=0 which of course is impossible if the output is in fact 2 volts. So Vi must stop just before it reaches ~2 volts. In particular it must stop at a value which gives a really small V_ but one large enough that Vo = -105V_ is just short of —2 volts. 3-22Tose what Vo and V. actually end up being let us go back to basics and write the node ; Dp equations. In particular for 1K x Vin =2V + © Ye we have Controlled Source Equation. fo = +105 (V4 - V-) = -105V_ which gives us -105 ) Vo = = - 1.99996 volts ° Fe +2 v= a = +1,99996x10-5 volts 1 Now we need these precise values to calculate the op amp’s input voltage (V4. V_). But to calculate all the other voltages and currents in the circuit we can for all practical purposes use 2volis and V. Vo = O volts Generalizing on our results we have for 3 - 23Controlled Source — Equation Vo Vo = A(V+ - V.) = -AV. = -105V. which gives us 1/R, Vin Yer -Tig-g Ro AIR, Ra Now with A= 105 and typical values of R; and Rp around 1K the second term in the denominator is so small that for all practical purposes and so ‘Therefore is so small that for all practical purposes 3-24V,-V.=0 We say that the input to the op amp is a virtual short. We say it's a virtual short because no current is flowing even though the voltage is just about zero. THE BASIC POSITIVE GAIN OP AMP CIRCUIT The basic positive gain op amp circuit is We refer to it as a positive gain op amp circuit because the output Ry + R, Vo = am Vin is a positive number times the input Vin. To get a handle on how the circuit works and see where the expression for the gain comes from let us begin by analyzing the circuit with Vin =2volts Ry =IK Ry =1K Now assuming as before that the op amp is characterized by a Again of A = 105 b. Infinite input resistances at the + and — terminals c. Zero output resistance Ry we have first of all that V4. = Vin = 2volts and also that the currents I thru the two resistors 3-25Vin =2V must be the same since no current can flow into the V_ terminal with its infinite input resistance. Therefore we can conclude Vo = 2x11 and V. = 1031 Now let's proceed as we did for the negative gain op amp circuit and see what happens when the source is tuned ON ~ (1) The Source Turns ON Letus assume the source tums ON at time t=0 and goes right to Vin =2 volts as follows Vin e ‘Now as is to be expected the op amp isn't going to be able to respond instantaneously and so Vo(0) = 0 from which we can conclude V_(0) = 0 (2) The Op Amp Takes OFF ‘Now with V4(0) — V_(0) = 2 - 0 = 2volt the op amp is going to take off like crazy as if it's going all the way to 4 (V4) - V.@)) = 105(2) = 2x105 volts 3 - 266 as indicated in But as Vo starts increases then so must I and V. which causes — and this is what's significant — the op amp's differential input (V4 —V.) to start decreasing W,-Vv-) ) The Op Amp Skids To A HALT & ‘And so — just like for the negative gain op amr, circuit — the op amp is going thra a process of choking offs own input. As a matter of fact Vo cannot even make itt0 4volts because Vo=4 volts would mean Va -V.=2-2=0 which implies the contradictory result Vo =0 volts So Vo must stop just before it reaches 4 volts at a value which makes V_ justa little Jess than 2 volts but large enough that 105( V4 — V.) = 105(2 - V-) = Vo To see what V. and Vo actually end up being let us write the node equations of our circuit 3-27Controlled Source Equation Vo Vo = 105(V4 - V_) from which we have = 3.99992 volts 1 + 0.5x10° @ V. = 0.5 Vo = 1.99996 volts ° Again, as before, we need these precise values to calculate the op amp's differential input but not for calculating the other voltages and currents in the circuit for which we can — for all practical Purposes ~ use = 4vols = V_= 2volis Now for the general positive gain op amp circuit 3- 28Controlled Source Equation Vo Vo = ACVin - V-) from which we obtain Vo = ——1-_— \; cw Rk yi” R+kR*A So for most practical applications — which satisfy R+ RAR ¢ 100 and A> 10 RX e we have ~ for all practical purposes ~ that wei Ri + Ra in Rr and also # Athy R, in = Vin giving us ~ for all practical purposes 3- 29And so again we have (V,.— V_) = 0. and therefore a virtual short across the op amp. Note that this is the case even though the circuit does not satisfy V.= V_ = 0 butin fact satisfies e Vy = Vo = Vin. AN IMPORTANT OBSERVATION The voltage gain of our positive gain op amp circuit is always greater tian or equal to one and in fact will equal one for Ry=e and R2=0 in which case the circuit reduces to simply It may seem at first that such a circuit of gain one would be of no practical value. But in fact it turns out to be used all the time because of its large input resistance. It’s used as what is referred to.as a “buffer” — a circuit element that is able to “isolate” one part of a circuit from another. We'll illustrate this later. 3.3.3 SIMPLIFYING THE ANALYSIS OF OP AMP CIRCUITS OPERATING IN THE LINEAR ACTIVE REGION Let us consider our positive gain op amp circuit 3- 30We analyzed this circuit by solving the node and controlled source equations e vOVL-% — +— 2-0 R R AY, - V2 Vy Va = Vin Now if in fact the op amp is operating in its linear region as indicated in Vo Wy-Vv_) e then the controlled source equation I Vol Voc 4. Iv, -V.l = > < > = 15x10 implies that for all practical purposes V_-V.=0 which gives us Va = VL with no A's! So our circuit equations reduce to Ve, Ve-¥. Yi eSNG ss °* Ve Ve Va = aa Now to make the analysis even that much easier, we usually indicate V, = V_ directly on the circuit — in which case we don't have to write the corresponding equation. In particular, for our 3-31Positive gain op amp circuit with we have we have and so the single 3- 32from which we obtain our old familiar result Yo _ & Vin Rg This certainly makes the analysis easier !! But how are we to know — when handed an arbitrary op amp circuit — whether the op amp is really operating in its linear region with a virtual short actoss its differential input — that the op amp is not saturated. This sounds foreboding but really isn't. All we have to do is analyze our circuit as if in fact V}= V_ and see what happens to Vo. If IVol < Voc then we're safe — our assumption was valid. But if our calculations indicate that |Vol > Vcc then the op amp is really saturated and we have to go back and reanalyze the circuit with Vo = Voc depending on the sign of Vo, 3.3.4 WHAT MAKES OP AMP CIRCUITS SO NEAT At first blush it probably seems crazy to be using op amps with gains on the order of 105 to build amplifiers with gains of only 5 or 10. But what we then get — as we've seen — is amplifiers whose gains can be made to depend only on high precision resistors — resistors that cost only pennies. This is of mujor significance because it means we can produce inexpensive amplifiers without having to “worry” about the nominal gains of the op amps or even if they vary due to temperature or other conditions. The only requirement is that the op amp gains be "large". So our “crazy” circuits tum out to be elegant examples of superb design. ‘The underlying reason for our circuit's success is the way the resistors are connected from the output back to the input. For example, let us consider y, in with output "locked in” at Vo = 1.99996 volts = 2 volts. Any momentary increase in Vo will immediately cause a deere ng in the op amp's input and therefore “push” Vo back towards its original value, Similarly any momentary decrease in Vo will cause a self-correcting increase in the op amp input. The resistors determine the gain — the op amp does the work. 3 - 333.3.5 INPUT AND OUTPUT RESISTANCES OF OP AMP CIRCUITS Now that we have an idea of how individual op amp circuits behave, we're going to calculate e their input and Thevenin Equivalent output resistances s0 we can get an idea of what will happen when we connect them together. We begin with negative gain circuits. INPUT RESISTANCES OF NEGATIVE GAIN OP AMP CIRCUITS Directly from the circuit diagrams of negative gain op amp circuits Ry we have, =R i e OUTPUT RESISTANCES OF NEGATIVE GAIN OP AMP CIRCUITS - AN IRONIC TWIST Calculating the Thevenin Equivalent output resistance by first calculating and then calculating Vi, with aload Ry connectedRH RL from which we can conclude that Ray = Output Resistance = 00 since Ryp has no voltage across it. Now this result is in fact just about right. But it’s always somewhat disconcerting to end up with a value of zero for for something like Rryy — especially after making the kinds of assumptions Ro =0Q and Asoo => Vy=V. that we've been making. To reassure ourselves that we haven't thrown out the baby with the bathwater Iet us reanalyze our circuit with the op amp replaced with the more detailed modelwe can straightforwardly write and solve the node equations to show that the Thevenin Equivalent Resistance of the resulting circuit is = Bol Ri + Ra) Rra ARi which for typical circuit values is going to be pretty small. In particular if 3-36Ro=75Q Ry=R,= 1K A=105 then Ry = 75 (2x10) / (105x103) = 1.5x103.9 and so our previous result of Rrpy = 0.2 really was okay. But we're still left with the question of why Ryq isso small. Orequivalently, why Ry, must get so small before ~ at least ideally — before it has any real effect on the output voltage Vz. To see what's going on let us see what happens as Ry, decreases in ‘As Ry, decreases, the magnitude of Vi, will decrease and therefore so will the current v, I= But as I decreases the input to the op amp V4 — Vo = -V-= Vin - 1031 must increase and therefore so must the voltage -105V_ of the controlled source. Now this is the key result because as the controlled source voltage increases it will supply extra current to Ry, and so “prevent” it from increasing as “much as it would like". This is really ironic. The same “forces” that conspired to prevent Vz from getting larger thanare also preventing it from getting any smaller. The result is that the gain is "stablized”. ‘Now let us take a look at the basic positive gain op amp circuit rs INPUT RESISTANCES OF POSITIVE GAIN OP AMP CIRCUITS From the circuit diagram of a positive gain op amp circuit we see that at least ideally the input resistance ~ the resistance at the (+) terminal — is infinite. In fact, however, it is on the order of 500 Meg. OUTPUT RESISTANCES OF POSITIVE GAIN OP AMP CIRCUITS Proceeding analogously as above we can show that the output resistance of a positive gain op amp circuit and all such similar circuits with feedback resistors — resistors connected between the input and ourput — are all very small A WORD TO THE WISE When you actually go into the lab to measure Ry be sure to use the method described in this section. In particular, do not connect a source to or short the output of the op amp. If you do you'll be interferring with the biasing circuitry and get erroneous results. 3.3.6 CASCADE CONNECTIONS OF OP AMP CIRCUITS In the previous sectir ~ we showed how to calculate the input and Thevenin Equivalent output resistances of op amp circuits. In this section we're going to use our results in calculating the gains of cascaded circuits — circuits made up of sections connected in "series" likeNi Now for such a circuit, the overall gain is where is the gain of Ny when it is loaded down by Nz and V2 Vy is the gain of Nz when itis loaded down by Ry. So in general we've got to take into account the loads when we calculate the gains of each individual section. But in the special case that Ny and No are op amp circuits like those discussed in the last section — circuits having input resistances much larger than their Thevenin Equivalent output resistances ~ then the loading will be minimal. And so for all practical purposes Ma Mw a and 7 equal the open circuit gains G, and Go of Rinand N2 So in general for a cascade of nop amps circuits R + Ny --+- Nn Ry LJ the overall gain G=G,-G.- will be the product of the individual open circuit gains. EXAMPLE 3.11 To find the overall voltage gain =< Vin of t = 2 + & + 7 © vy RLS YL ra Ra >e All we have to do is multiply the open circuit gain of > Ra Ri equal to times the open circuit gain of equal to R3 + Ry R3 to obtain y, +R,) 6-66, = gb - BSR in 1B 3.3.7 ANALYSIS OF GENERAL OP AMP CIRCUITS Suppose we want to analyze a more general op amp circuit likeAt first blush it looks like we have more constraints — the three node equations at the nodes 1 2 and 3 — than we have unknowns — the two unknown node voltages V4=V. and Vo But not really. The op amp output current Ip is also an unknown and so must be included in the node equation at node 3 as follows Vo-V. , Vo-¥. o=V. , Vo- Ve = mK * aK tb =° ‘The currents at the input terminals of an op amp may be zero, but the current Ip at the output is. definitely not. So the three node equations really do have three unknowns. But — be this as it may — we only actually have to write node equations at nodes 1 and 2. It's not necessary to write a node equation at node 3 because it alone contains the unknown Ip. Therefore all we have to dotofind V,=V_ and Vo is write the node equations atnodes 1 and 2 . Then we can calculate Ip from the results. Let us illustrate with an example. EXAMPLE 3.12 Writing the node equations for we havefrom which is obtained Vi=V_=3volts and Vo = 7 volts and so 7) val Oto 5K aK ‘The bottom line is that if you have an irresistible urge to write node equations at the outputs of ideal op amps then you must include the ourput currents Ig. There is nothing wrong with this except that nothing is really gained. You'll get an extra equation but also an extra unknown. =-3ma 3.3.8 SOME WORDS OF CAUTION BEWARE OF SATURATING THE OP AMP - ‘As we've mentionned before, too large an input will cause the op amp to saturate and as a result produce distorted ourputs like Vo 0+. -~N ~20 Ne ‘So whenever measuring voltage gains, input resistances, output resistances and so forth of op amp circuits in the lab make sure that the stays small enough that the op amp will not saturate. To guard against inadvertently making measurements on an op amp circuit that is going into saturation make sure that you always have the op amp hooked up to a scope to make sure the signal is clean ~ even if you are using a multimeter to make the actual measurements.BEWARE OF THE CURRENT LIMITATIONS OF OP AMPS. Not only is the output swing of an op amp limited but also is the maximum amount of current it can supply at its output. What this implies is that if we take a circuit like our negative gain op amp circuit but make Ry, too small, the op amp isn't going to be able to supply enough current Ip for Vi, to make it all the way to what it's “supposed” to be -RYRY Vin In particular once Ry, reaches QR) Vin Tomax further decreases will cause V1 to decrease. BEWARE OF SLEW RATES Besides all its other limitations it's not surprising to find that the op amp is limited in how fast its output can change. In particular if the input to an op amp circuit is a sinusoid that is changing faster than the op amp output can respond then the output will be distorted and look something like Yoo Slope = Max Rate At Which ge Output Can Change‘The maximum rate an op amp's output can change is specified by its slew rate. Slew rates e for typical op amps are on the order of 0.5 volts/isec. BEWARE OF HOOKING UP YOUR OP AMP BACKWARDS !! Suppose we take a positive gain op amp circuit that's supposed to be wired up like Vin = 2 and instead wire it with the op amp connected "backwards" as followsand so = 4.00008 for A= 105 which gives us Vs. = 2.00004 volts This looks exactly like the result when the op amp is wired correctly. But now suppose an innocent fly comes by and lands on the output lead — causing Vo to increase slightly. This may seem inconsequential at first but if Vo increases then so will V4 which will cause Vo=A (Vs - V-) to increase even more — the end result being that Vo will increase all the way to saturation at Vo= Vee: On the other hand if our fly had caused Vo to fall below 4 volts then V4. would be forced to fall below 2 volts and then Vor A (Vs - V-) e would tum negative and end up going all the way to saturation at Vo =~ Vcc. So Vo = 4.00008 volts is an equilibrium point for our circuit but it's an unstable equilibrium point. The circuit will work fine if nobody breathes, but as soon as somebody does it's going to go into saturation. This situation is similar to that of a ball balanced on top of a pin. Ill stay there as long as no one blinks. But as soon as someone does, then the ball is going to fall off !! Note that if Vin stans increasing from O volts then Vo will saturate at Vee. But if’ Vin starts decreasing from 0 volts then Vo will saturate at +Vgc. Going thru the same kind of analysis we can show that negative gain op amp circuits with their op amps wired backwards likealso have unstable equilibrium points. But not all is lost!! In fact oscillators turn out to be nothing more than circuits similar to ours with unstable equilibrium points, We'll take a look at one after we introduce capacitors. 3.3.9 NONLINEAR OP AMP CIRCUITS Up to now alll our op amp circuits have contained only linear resistors, but many useful op amp circuits such as contain nonlinear elements such as diodes in the above circuit. As usual all we have to do to analyze such circuits is to write and solve the corresponding node equations. For this circuit we have Node Equation k@-V,) + Ale -1)=0whose solution gives an output that depends on the log of the input. Such circuits are very useful in signal processing and in the design of circuits such as multipliers. @ 3.10 SOME NUTS AND BOLTS 1. Resistor Values To Use When Designing Op Amp Circuits Nice resistors to use when designing op amp circuits are roughly in the neighborhood of . 1K to 50K If the resistors used are too small, large currents may flow and the op amp will not operate as we've been assuming. On the other hand large resistors can add extraneous noise. 2. Wiring Up Op Amps The standard pinout for an op amp ina DIP package is NC (No Connetion) + Voc ‘There are two important points to remember when wiring up op amp circuits a. Most op amps require dual — that is, two — power supplies for setting their operating points. Typical values are’ +15 volts for the positive supply and —15 volts for the negative supply. The pinout diagram for your op amp will tell you where to make the connections. . Op.amps have no pin connection for ground. The ground side ofthe +15 volt and -15 volt power supplies are simply connected to the ground of the circuit. Including these connections, an op amp circuit looks like3.3.11 OP AMP EQUIVALENT CIRCUITS Tf we want to take into account the finite gains, finite input and output resistances and so forth of ‘op amps, we need corresponding models. Some basic models of increasing complexity are as, follows: 1. A simple model that takes into account finite op amp gain is 2. A model taking into account the op amp's finite input and output resistance is Ro Ve a 3. And finally a model that takes into account the fact that a real op amp’s performance e 3-49depends on the frequency of the inputis We include this model at this time even though we're not yet prepared to discuss it so that welll have all our op amp models in one place!COMPUTER ASSIGNMENTS 6 3.1 Given , 3K 2K Vin + Vo with the op amp v_ Vy é can be modeled by a Vis Ve 05 (V,-V_) a. Draw the circuit with the op amp replaced by its model b. Use SPICE to find the gain of the circuit — See Appendix C e 3- 51ECE 208 PROBLEMS - CHAPTER 3 REVIEW OF DIFFERENTIAL EQUATIONS 3.1 Sketch a v(t) = 2et — e-2t b. vit) = 2e2t — et cv) = 2+ Qe? et 3.2 Solve ¥+ 1 + lOv = 10 WO) = 0 v0) = -1 Use the method outlined in Chapter 7 on a cookbook method for solving 2nd order differential equations 33 Sketch av) = et + tet db. vit) = et ~ tet © v(t) = et + tet 3.4 Find and sketch the solution of VtWwe was vO) = 1 v(0) = -1 Use the method outlined in Chapter 7 on a cookbook method for solving 2nd order differential equations 3.5 Find and sketch the solution of V+ 4v + 4v=10 WO) =0 vO=1 ‘Use the method outlined in Chapter 7 on a cookbook method for: solving 2nd order differential equations 3.6 Find and sketch the solution of V+ 200¥ + 104y = 2x10¢ vO) =1 v(0) = 1 Use the method outlined in Chapter 7 on a cookbook method for solving 2nd order differential equations 3- 52: NODAL ANALYSIS OF CIRCUITS WITH CONTROLLED SOURCES @ 27 cee Vig = 10° cos 104 Find vo(t) 3.8 Given cos 103 ¢ Find vo(t) & 3.9 Given ik 10K Vi, 101 + n Cy 102 $ v, Find the transfer function G = Vo/Vin 3.10 Write and put in matrix form the node equatins ofcos 103 oe RESISTANCES OF CIRCUITS WITH CONTROLLED )URCE! 3.11 Find the input resistance Rig of 3.12 Find the input resistance of 1K O.1V, 3.13 Find the input resistance of 3-54 c1K 1K 101 1K 1K in 3.14 Find the input resistance of 1K 1K 3.15 Find the input resistance of I, 5K 2K 4K e 61 3.16 Find the input resistance of 3- 55" THEVENIN EQUIVALENTS OF CIRCUITS WITH CONTROLLED 3.17 Find the Thevenin Equivalent output resistance of 1K 2mv 3.19 Find and draw the Thevenin Equivalent of 3-56 ee Vin © gn IK 3.20 Find and draw the Thevenin Equivalent as seen by N 1K yy 1K cost 3.21 Given What can be said about Viy/Vin if v, G=%>1 1mv * N ©) KS, Find Vo if the open circuit voltage gain of N is 103 andthe Thevenin Equivalent 3.22 Givenoutput resistance seen by the 1K is 500 GAINS OF CIRCUITS WITH MULTI-TERMINAL ELEMENTS e 3.23 Given + + . > N IK with Vo/V; =5 and Rin = 2K a. Find the gain G = Vo/Vin of 1K VW b. What can we say about the gain of 1K v ® n | eB 3.24 GivenFind G = Vo/Vin if Vo/V1 = 5. and the input resistance of N with the 1K load is 2K 3.25 When is the gain of a cascade equal to the product of the open circuit gains 3.26 Given 3 10” cost cos ® Ny No ¥5 Find vo(t) if Ny and No both have infinite input resistance, zero output resistance and voltage gains of 100. What would happen to the magnitude of vo(t) if Ny and Np had finite input resistances and nonzero ourput resistances 3.27 Given 3-59502 + € Vin = Yo Find the overall gain G = Vo/Vin_ifteach section — when terminated with a $0 Q load as follows + + Ny = Si 502 S Vv, R jg = 502 has an inputresistance of Rpy = 500 anda gain of Vo/Vjn =5. Note that such a cascade of sections Nj, Np and N3 is said tobe "matched". é BASIC OP AMP CIRCUIT ANALYSIS 3.28 Find the gain of 3.29 Find the gain of 3- 60 e3.30 Find the gain of 3.31 Given a. Does V4=V. b. If not, what do they equal c. Draw an op amp circuit for which V4 = V_=0 3.32 Why is nothing gained by writing a node equation at the output of an ideal op amp 3.33 Find the transfer function G=Vo/Vin of 3- 6150002 3.34 Find the input resistance and output resistance of 1K —— 3- 62 €3.36 Find the transfer function G=Vo/Vin of 3.37 Given a. Find the transfer function G = Vo/Vin b. Find Ip if Vin = 1 volt 3.38 Find the transfer function G= Vo/Vin of 1K IK wm Ww Le3.39 Given Find vo(t) 3.40 Given Choose Ry and Ro so the overall gain is ~1/2 3.41 Choose resistor values in so the overal gain is G = Vo/Vin = 1/2. Redraw the circuit with the values specified 3- 64 €3.42 Given ? xe 1K “© © ve Find Vo 3.43 Find the transfer function G=Vo/Vin of10K" Find Vo 3.45 Find the transfer function G = Vo/Viq of Ps 1K Vin 7) 2K 1K W—— = © me oe 1K oa 2K 3.46 Given 3 - 66100K Find Vo interms of V; and V2 3.47 Draw the small signal equivalent circuits for each of the following circuits assuming the op can be modeled by “ Vy vig ® 3- 67Vin ® + ©) vo) NONIDEAL OP AMPS 3.48 Given - 2K i € ee + vo) a. Draw the equivalent circuit if the op amp can be modeled by 52 a —A— 10° (V,-V_) >. Find the Thevenin Equivalent output resistance of the complete circuit 3- 68 C33.49 Given the negative gain op amp circuit modelled by ‘How does the gain G=Vo/Vin vary as a function of R and A 3.50 Given the negative gein op amp circuit modelled by a. How does Ro affect the gain G= Vo/Vin b. Find the gain if Ry = 75.2 3.51 Given + v. Vo 1K IK s a. How does Ro affect the gain G= Vo/Vin b. Find the gain if Ry =75.2 3.52 Given 3- 69Vin® = 15 cos 107 t Sketch vo(t) if a. 3- 70 Yo3.53 Given in 7 cos 107 Sketch vo(t) if the op amp saturates at 10 volts. Indicate appropriate times 3.54 Given 1002 Vo) Sketch vo(t) if the output current of the op amp is limited by 25 mae ECE 252 - LAB 3 - CONTROLLED VOLTAGE SOURCES FALL 1998 A.P. FELZER OBJECTIVE — The objective of this lab is to make use of op amps — integrated circuits made from transistors ~ as follows Vee = 15 volts Ve vi We 0 Voc =~15 volts to build make use of voltage controlled voltage sources of the following form vd ¢ vat) =G vi) Note that it's not the objective of this lab to study the inner workings of op amps. Their inner workings are studied in the electronics classes. PRELAB - Find and draw the pinout diagram for a 741 op amp ~ there's a copy in the lab manual ® LAB - 1. Given the following op amp circuit (with the power supplies +Vce connected but not drawn in) > Vo(t) = B cos (27103t) vs(t) = A cos (272103t) PARTNER |: R=2K — PARTNER 2: R= 4.7K a. Redraw the circuit just like it is but with the op amp pin numbers shown on your diagram. Be sure to include the pin numbers for +Veo and -Vec b. Measure your resistor values. Compare with nominal values. ¢. Find and sketch vo(t) for inputs vs(t) with magnitudes A = 0, 1, 2, 3,4, 5. Make sure to lay out your circuit on your proto board to look exactly like your diagram d. Make use of your results in part (c) to plot B as a function of A. Make use of your results to come up with an equation for B as a function of A ¢. Make use of your result in part (d) to find vo(t) as a function of vs(t). ‘Then draw the controlled source realized by this circuit. 2. The objective of this problem is to build and analyze a simple resistor circuit containing 3} your voltage controlled voltage source from Problem (1) as followseg pee i 1K 10K v2(t) =G v(t) replaced by your op amp circuit. Show the pin numbers of the op amp. What is the G of your circuit. ‘Then measure your resistor values. Compare with nominal values Now build the circuit and measure v1(t), va(t) and v3(t). Again be sure to lay out ‘your circuit exactly like your diagram Calculate v2(t) and v3(t) and compare with your measured values Calculate i(t) from your measured values Use your measured data to calculate as v, and G, Use your results in part (f) to predict i(t) and v3(t) when vs(t) = 8 cos (271031) Measure i(t) and v3(t) when ve(t) = 8 cos (2110%t) and compare the magnitudes of the results with your predictions in part (g)ECE 252 - EXPERIMENT 8 4}- BASIC RESISTOR OP AMP CIRCUITS OBJECTIVES ‘The student will be able to calculate and measure the voltages and currents in basic resistor op amp circuits MATERIALS REQUIRED BY STUDENT 1. Clip leads 2 Seti ac resistors as needed 3. 307 op amp or equivalent — such as the 741 4, 3 BNC toclip connectors PRE-LAB — Due at the beginning of the lab 1. Given Yin (®) Calculate — assuming the op amp is ideal a. The gain G = voWVin b. The input resistance Rin as seen from the sourceRy =3K as discussed in Chapter 3 ¢. The Thevenin Equivalent output resistance as seen by the load 2K 1K bet Do not calculate Rry by actually shorting Vin and putting a source at the output. See Section 3.3.5 of Chapter 3 to see how tocalculate Rry with Vin present and aknown load Ry,“connected at the output Assume Rg =75Q and A = 105 2. Design a positive gain op amp circuit with gain isc ill = ( Birthday |! where by Birthday is meant the date of your birth (from 1 to 31). Assuming the op amp in your circuit is ideal calculate the a. Inputresistance Rin of your circuit as seen by the source b. Thevenin Equivalent output resistance Ro of your circuit as seen by the load 3- 73 €e 3. Given the three circuits qQ) 1k IK Vv. ¥ in ) 33K S Yor @ 1K 1K Vin cp ©) Voa (3) 1K vA 1K 1K 1K Vin + e) 33K SVo3 a, Find each circuit's gain G = voWin b. Why are the gains different? In particular, why does the second circuit have the gain with the largest magnitude 4, Obtain the pin diagram of your op amp. Draw the circuit of prelab Problem (1) with the pin connections specifed. Make a copy for the lab. LAB — Each student is to do his own experiment 1. Measure and record your resistor values 2. Given 3- 74Vin = Acos2nl03t Use your scope to a. Measure the gain G = Vo/vin with Vin increasing in steps of 1 volt until the output saturates . Measure the input resistance Rin as seen by the source ¢. Measure the Thevenin Equivalent ouput resistance Ro. Use the same method you used in your calculation of Ryy in the prelab. In particular, do not short Vin and put a source at the output 4. With A = 5 volts, makes measurements to verify that the voltage across the op amp input is a virtual short 3. For your positive gain op am, circuit you designed in the prelab é a. Measure the gain . Measure the input resistance ¢. Measure the output resistance ~ again do not short the input and put a source at the output d. Verify that the inputs to the op amps are virtually equal ¢. Meausre and sketch the response to vin(t) = 5 cos 2n105t 4, Measure the gains of the three circuits in Problem (3) of the prelab. Demonstrate these results to the instructor QUESTIONS AND PROBLEMS 1, For the negative gain op amp circuit you have built are the calculated and measured values b. How large can the input get before the op amp saturates c. Draw the circuit with the op amp replaced by its ideal ¢ ~trolled source model 3- 75Va es a ns 106(V,.- V_) 2. Repeat question (1) for the positive gain op amp circuit you designed 3. How is the overall gain of a cascade like Rin 1 Vin Ny No N3 Ry L affected by the input and output resistances of the individual sections 4. Find the gain G = volvin of 10K 10K 5. Use SPICE to find the gain of the second circuit in Problem (3) of the prelab with ‘equivalent circuit 1K 1K 3.3K 2Y) & ‘a @ 3- 76CHAPTER 5 ® CAPACITORS AND INDUCTORS 5.0 TRANSITION As we've seen resistor circuits are great for attenuating signals, amplifying signals and changing the shapes of signals. But purely resistive circuits cannot separate signals from each other like, for example, those from different radio stations. In particular, it's not possible to build a resistor circuit with input Vin(t) = cos (10001) + cos (20000) and ourput Volt) = cos (1000t) To do this we need circuit elements whose responses depend on how fast the terms in the input are changing. As welll show in this chapter, this is exactly what capacitors and inductors do, 5.1 WHAT ARE CAPACITORS? Capacitors are simply circuit elements consisting of close but nontouching metal plates like e —Ub— 5.1.1 THE FUNDAMENTAL Q-V RELATION OF CAPACITORS If both plates of a capacitor are electrically neutral there will be no electric field between ther. But if we take Q coulombs of electrons from one of the plates and deposit them on the other plates as follows Q extra coulombs of electrons Oo 9 9 then the work we do will go into creating an electric field and so a potential difference beween the plates. Q fewer coulombs of electrons ‘What characterizes a given capacitor is how this voltage varies as a function of the amount of s s-1charge on the plates. A convenient way to indicate such a relationship is with a QV graph like Q 1 and V are proportional ~ as they often approximately are — then the Q-V characteristic Curve is a straight line as follows Q and we say the capacitor is a linear capacitor. Letting C be the slope of the Line we have c ig + N %_ energy will be transferred back and forth between N and the capacitor at the rate av (0 PO = ¥,OI0 = CHO This is interesting because unlike the case for resistors p(t) for a capacitor can be positive or negative. In particular 1. When PUD = velt)ig(t) > 0 in which case Volt) > O and icft) > 0 or v¢(t) < 0 and ip(t) < 0 the amount of charge on the plates is increasing and the capacitor receiving energy 2. But when PU) = velt)ig(t) < 0 in which case volt) > 0 and ig(t) <0 or —ve(t) < 0 and ig(t) > 0 then the amount of charge on the plates is decreasing and the capacitor wansferring energy back to N The capacitor, unlike the resistor, is able to transfer energy back to N_ because instead of dissipating the energy it receives, itis able to store itin the electric field between its plates. ‘The amount of energy a capacitor receives and stores can be determined as followst EW = Ett) + | pear € aS = Elto) + | velt) ig(t) dt ae = EG) + f Cvett MeO a dt to 1 t = Ete) + Lodo, 4 2 1 2 E) = E(t.) + Zev) - ZC% Cy) ‘Now assuming that our capacitor is brand spanking new ~ it just came off the assembly line at time tg and so has had no energy transferred to it and no voltage yet applied across it — we have Eto) = 0 and volta) = 0 and so the total energy delivered to the capacitor up to time t becomes simply Lain EQ = 5 CKO € which is always greater than or equal to zero. What this tells us is that a capacitor C cannot transfer back any more energy than it has received. We refer to such circuit elements — which themselves are not sources of new énergy — as being passive. To show that all the energy Lo? BO = sev 0) Gelivered to a capacitor C is in fact being stored, all we have to do is connect C to a resistor ig) 3 %o0 G R and show that all this energy will be dutifully delivered to and dissipated in the resistor, We will demonstrate this result in the next chapter.5.2 WHAT ARE INDUCTORS? Qevco: are circuit elements consisting simply of coils of wire — usually wound around iron cores. $.2.1 THE FUNDAMENTAL i-I RELATION OF INDUCTORS Whenever current flows through a wire, a magnetic field will form around it. When we wrap a long wire in a coil to form an inductor, the fields from the individual turns add and we are able to concentrate a relatively larger magnetic field in a small area. ‘What characterizes a given inductor then is how the strength of its magnetic field 2. varies as @ function of the current I through it ~ as indicated ina X-i graph like If 4 and I are proportional, then the A-I characteristic curve is a straight line _ ) i ind we ¢ , the inductor is a linear inductor. Letting L be the slope of the line we have ® 5-9L=4 € = inductance of the inductor The value of L typically depends on the number of tums and the material of the core. When the magnetic field strength 2 is in webers and I is in amps then L has theunits Henrys = webers/amp Note that the abbreviation for “Henrys” is "H", Inductors are harder to come by than capacitors but they are fairly readily available in the range from about 0.1 1H to 10 mh. The inductors we will be using in the lab will be in the mH range. §.2.2 THE I-V CHARACTERISTIC OF INDUCTORS Suppose we have a linear inductor | vy yo 4 “LO Then the strenth of the inductor’s magnetic field € XO = LiLo will vary in direct proportion to the current iy(t) thru it. The next question is how the voltage LC®) across the inductor io 4 vy . is related to it (1). To answer this we make use of the following basic observations 1 Ifthe magnitude of v_(t) is small then not much energy will be transferred between the charges flowing thru it and the magnetic field. Therefore the strength of the magnetic field as measured by X(t) and therefore also iz(t) will be changing slowly 2. If, on the other hand, the magnitude of vz(t) is large then a lot of energy will be tansferra between the charges flowing thra it and the magnetic field, Therefore the strength of the magnetic field as measured by 2(1) and therefore also iz(t) will be changing quickly 5-10 €3. And finally i¢(t) will be (@) Increasing when vit) is positive (©) Decreasing when ic (1) is negative Pulling all these observations together we that that v(t) is proprtional to the ir(t) is changing. Let us now illustrate with some examples. EXAMPLE 5.3 If ip(t) isa constant ipo t then A(t) = Liz (t) must also be a constant 2LO t and so v(t) must be zero 0 t since no change change in the strength of the magnetic field means there must not be any transfer of energy going on. EXAMPLE 5.4 If ip(t) isa sinusoid like i,@then A(t) =Lir(t) mustalso be a sinusoid ALO ©) get an idea of what vi(t) is going to look like let us look at the times tj, t, t3, \ and ts. Attime ty: v(t) is at its maximum positive value since A(t) is increasing at its fastest rate ty: vi() is zero since A(t) is not changing tg: vL() is atits maximum negative value since A(t) is decreasing at its fastest rate ty: v(t) is again zero since A(t) is not changing ts: vi) retums to its maximum value Therefore vi(t) should go thru points as shown below yo Looking at the intermediate points and doing some appropriate handwaving — as we did for the capacitor ~ itis not too hard to convince ourselves that vit) should look at least something like a sinewave, but 90° out of phase with iz(t) and Av. So what we would expect is 5-12THE ANALYTICAL APPROACH ‘We concluded from our observations at the beginning of this section that an inductor ie yo > — with associated reference directions has a voltage v(t) across it that is proportional to the rate the current thru it iz (t) is changing. All it takes to analytically obtain this result is to substitute into Faraday's Law y= Sr0 (which is simply a mathematical expression for what we've been saying all along — that the voltage, or difference in potential, across an inductor at time t is equal to the rate the strength of the magnetic field is changing at time t) our expression MQ) = Lip Doing this we obtain dO yO =L and therefore also t LOR) + Ef aoa 0 Let us now use these equations to confirm the results of our two previous ex” ples 5-13EXAMPLE 5.3 REVISITED With iL() = Ip = constant we have as before. EXAMPLE 5.4 REVISITED With i, = Ip sinax we have oy sin ot al yO = all, cos ot and so we see that v(t) is exactly a sinusoid just as we intimated. 5.2.3 INDUCTORS AND ENERGY ‘When it comes to energy transfer, the results for capacitors and inductors are pretty much the same. Summarizing the results for inductors we have the following. Given an inductor with associated reference directions ipo 4 “L So it is straightforward to see that whenever the power is greater than zero ao 0 iL = LO P= then the charged particles are delivering energy to increase the strength of the magnetic field. Conversely, when pp(t) < 0 then the collapsing magnetic field is delivering energy back to the charged particles. 5-14‘The amount of energy being stored in an inductor can be determined as follows t EQ) = Ey) + J ple dt ‘5 t = Ett.) + J Mig de ly : a0 = E(t.) + J Lio dt bo ‘ 1,2 = Egy + LA ol, 1,2 oe! E® = Ett.) *- zlipo ran ship) Now assuming that at time to the inductor is brand spanking new — there has not yet been any energy transfer — and so E(t) =0 and gt) = 0 we have that 1 Et) = zhao And so since E(t) > 0 forall timet ‘we see that inductors are passive just like capacitors and our garden variety positive valued resistors,$.3 CHANGES IN CAPACITOR AND INDUCTOR VOLTAGES AND CURRENTS 5.3.1 ABRUPT CHANGES IN CAPACITOR VOLTAGES € Suppose we are given a capacitor Yo) ig + dvo(t) at with ip@=c t i é VO = Vet) +E f ioe 0 that is in a circuit in which the current can only change gradually or at most by finite increments. If the current changes gradually like ip® «& then the voltage — given by ‘ 1 Ve = lt) +E f joae toy must increase smoothly like 5-16 €Yo t And even if ig(t) goes thru an abrupt step change like ig ® =u) ft t Ve(t) will still itself change continuously like Yel t In fact the only way for ve(t) to change abruptly — go thru a step change — is for a quantity of charge Q to be somehow magically and instantaneously added to or removed from the capacitor plates. Then vo(t) can jump like Yolo) But for a finite amount of charge to be delivered in an instant of time requires — at that time — an infinite current. This of course is not physically possible but it does turn out to be a 3 5-17mathematically useful idea. To represent such a current source that can deliver a finite amount of charge in “zero” time we introduce the Dirac Delta — or Impulse — Function 8(t) defined by @ @ b=0 if t#0 e Gi) f B@a=1 foray e>0 * Dirac Deltas are not of course functions in the regular sense in which we're used to thinking about functions but we can — "for all practical purposes” — think of them as very narrow and very tall pulses of area equal to one VA ‘We represent impulse functions by € bo So a capacitor connected to an impulse current source like ig . . + i, O= 8G Yo 5-18 ‘€‘would in fact go thru a step change in voltage, By the same kind of reasoning the “ideal” circuits (so resistance) . “1 L ‘will cause an impulse of current to flow when the switch closes. If we actually “build” such a circuit there will of course always be some series resistance — but you can still experience a preay hefty shock when the switch closes. $3.2 ABRUPT CHANGES IN INDUCTOR CURRENTS : ‘Suppose we are given an inductor ipa 4 “1 with ai, Li ‘ O-= Lz t ; 1 iLO sip) +o f yp dt 'o sex anslogously to the case for capacitors, the current thru an inductor cannot change ‘Ssumeancously like i®unless the charges flowing thru the inductor can somehow instantaneously deliver a finite amoun! of energy to the inductor ~ instantaneous cause the magnetic field to increase. The only way for such things to happen is for ideal switches to open in circuits like € ? + or close in circuits like §.4 EQUIVALENT CAPACITANCES AND INDUCTANCES, As usual we're always on the lookout for equivalent circuits — those for capacitors and inductors being no exception. @ 5.4.1 CAPACITORS IN PARALLEL Two capacitors in parallel in effect form one larger capacitor with a stored charge of Q=Aa+Q equal 10CV + QV e Q Therefore (Cy + CV Coq = WV = CL + In general, for n capacitors in parallel Ceq our above result easily generalizes 10 _. Le _t Coq = Cr + Cpt tC 5.4.2 CAPACITORS IN SERIES 72° 2 inany uncharged capaciors connected in series cy (f we apply a voltage V then a charge +Q will accumulate on the left plate of C; and a charge -Q on the lower plate of C2 as followsThe electrons between the two capacitors will then redistribute themselves so that ~Q is on the inside plate of Cy and +Q on the inside plate of Cp. My ‘Therefore and so In general for n capacitors in seriesVOLTAGE DIVISION IN A SERIES CONNECTION OF CAPACITORS Given a series connection of n capacitors oe with Q coulombs being stored on each capacitor we have WG, =2 = ‘e"G, 7 Tse I, “in INDUCTORS IN SERIES For two inductors in series 5-23‘ Ly i , 3 Ly Leg with 2y=Lyi and A, =Lpi, the total magnetic field 2 is equal to Re thy = Lit Li Gi + Li ‘Therefore Lg Bah + hy In general for n inductors in series Leg = Ly + ly tees + Lp 5.4.4 INDUCTORS IN PARALLEL ‘Suppose we have two inductors connected in parallelwith each one's magnetic field initially equal to zero. 41) = 2,0) = 0 ‘Then t t - dy = 2 + f vip dt = roa a a t t 40 = 2,0 + [una = una Oo 0 @ 2n8.s0 tre mapnesic feds of both inductors ae the same 20) = AO = AO Using this result we have ‘Therefore and in general for n_ inductors in parallel a 5-25CURRENT DIVISION IN A PARALLEL CONNECTION OF INDUCTORS Given a parallel connection of n inductors * 4 lin L L L 1 ke a aL My Ly” Wyss sy in i, k 5.5 COUPLED INDUCTORS Up to now we've been assuming that all of an inductor’s magnetic field is generated from the current flowing thru it alone. In particular for a given inductor i@ 7 —>—T__ we've been assuming that the total magnetic fields aw = Li@ 5-26 €But if two inductors are close enough to each other as in e io i20 + + y@ gli Leg 0 then the magnetic fields can overlap and we have AQ® = Lyi + Mint) AQ® = Mi) + Ly inh) and therefore the voltages vu = 220. 1) S10 2m S20 di vay = 220 sem +1, BO dt where L and Lz are referred to as the self-inductances and M the mutual inductance. We will not do it here but it can be shown that Ms i Now for the + signs. These are there because the magnetic fields from the two coils may be adding with or subtracting from each other depending on which way the coils are wound. To specify what's going on for a given pair of coils we draw dots like These dots are positioned so that (1)_ If we draw our reference directions with both arrows pointing from the dots to the inductors as followsi,@ Me i@ IG) M + .. ot +(e = yy@ BL L2g 20 vy@ A1jL2g 2% - ~ - e+ ig then the magnetic fields will add when both currents have the same sign and subtract when they have opposite signs. So this choice of reference directions comesponds to ‘our equations with the plus sign vy = tuo ot) SO. + M20. ao vat) = 20 = HO 5 1p a at (2) If, however, the reference directions are as follows i M i2® 5,0 M + |e oe +e + ¥y@ BLi Lrg 20 yy@ ArL2g 20 - + = | 1,0 with one arrow pointing from a dot into the inductor and the other pointing thru the inductor towards the dot then the magnetic fields will add when the currents are of opposite signs. And so in this case we have dd dij di; vy = SO = 1, HO ~ yy S20 vat) = 280 2 4 HHO. 4p 20 ‘Note that the equations in (1) and (2) are really uae ‘The difference comes about only because of the choice of different reference directions.ECE 208 PROBLEMS - CHAPTER 5 REVIEW OF DIFFERENTIAL EQUATIONS 5.1 Solve V+ v= cost v0) = W(0) = 1 Be very careful when finding the forced response CHARACTERIZING OF CAPACITORS 5.2 Find and sketch as a function of time the charge being stored on a capacitor's plates Ya c ig 7 _ lOne a. velt) = 10 volts b. volt) = cos 103 e ©. volt) = 5 cos 103 4. vc(t) as given by Yo t (msec) 5.3 Given a capacitorS with stored charge q(t) as given by a) ; & sketch a. volt) b. ic(t) 5.4 Givena 1 uf capacitor PO) i Lut Find ig(t) if « & Vc(t) = 10 volts b. volt) = 10.cos 100t ©. Volt) = 10 cos 104 5.5 Explain how the magnitude of the voltage across a capacitor can be increasing even though ‘the magnitude of the current through it is decreasing. 5.6 What happens to the voltage across a capacitor as the frequency of the current through it increases — assuming the amplitude of the current does not change 5.7 Given a capacitor vo id OL ¢ 5- 30 «Express the energy E() being stored in erms of te expacitanee C and he charge Q() on e plates 5.8 What happens to the energy delivered to capacitors 5.9 Cana capacitor be storing energy at a time when the current through it is zero. Explain 5.10 Suppose the voltage across and current thru a capacitor were both sinusoids. Would the capacitor be storing more energy when the current was zero or when it was maximum, Justify your answer 5.11 How high would you have to lift a quarter to store as much potential energy as stored in a L uf capacitor with 1 volt across it CHARACTERIZING OF INDUCTORS 5.12 Find and sketch as a function of time the magnetic field strength A(t) of an inductor ipo + “LOL 2mh a. ip(t) = 10 volts b. p(t) = cos 10% ¢. ii, () = Scos 103 4d. ip(® is given by i,© (ma) t (msec) $5.13 Given an inductor ipo >——1-__—_with magnetic field strength (1) as given by © « Sketch a. iL) b. v(t) 5.14 Givena 1mh inductor ipo + “LO 1mh with « “0 t Sketch p(t) assuming that i(0) = 0 5.15 Given in + “LOL 10ma 5- 32 «Find v(t) if a.A() = 10ma b. i(t) = 10 cos 100t c. ip) = 10 cos 104r 5.16 Suppose the current through an inductor changes from —10ma to —15 ma. Has the inductor received energy or lost it. Justify your answer 5.17 What happens to the energy delivered to inductors EQUIVALENT CIRCUITS 5.18 Find the equivalent capacitor of Lut == 1st == And the equivalent inductance of SPARKS 5.19 Can the voltage across an inductor charge instantaneously even if the current thru it doesn't. Justify your answer 5.20 Under what circumstances will the current through an inductor change instantaneously 5.21 What does it take for the voltage across « -upacitor to change instaneously 5.22 GivenM i, i,® +L o+ 10 Zt le Fy@ a. Write the v-i equations if Ly = 10mh, L)=7mh, M=8mh — ’. Repeat the problem for the reference directions M io 10 + ipa v0 9t1 l2 Avie 7 + 5.23 Given 110 . io + [eo + vyy@® Oli Lo v0 a. Write the v-i equations if Ly = 10mh, Ly =7mbh, M=8 mh b. Repeat the problem for the reference directions M io i, + |e = vy Ali Ly 0 - , 5.24 Given i i,® draw in the dots if when ij >0 and i2>0 the increasing of iy causes ij to decrease S- 34ECE 252 - EXPERIMENT 5 3 I-V RELATIONSHIPS FOR RESISTORS, CAPACITORS AND INDUCTORS OBJECTIVES . Given the voltage across a resistor, capacitor or inductor the student will be able to calculate and measure the current through it and vice versa MATERIALS REQUIRED BY STUDENT 1, Clip leads . 2. 5% 1/2 watt resistor of valc)1K 3. 0.1 uf capacitor (not polarized) 4, Inductor box (check out from stockroom) 5. 4 BNC to clip connectors PRE-LAB - Due at the beginning of the lab 1. Given ip® @ Yin ® = + Scos 104 © vR® a. Find v(t) and ip(t) b. Sketch va(t) and ig(t) ¢. What is the phase difference between va(t) and ig(t) 2. Given igo Vi, = + .L et Scos 104¢ id vow a, Find ve(t) and ig(t) e 5-35b. Sketch vo(t) and ic(t) c. What is the phase difference between vc(t) and ict) en iro Vin) = Ae Sos 105+ S Lo ip = a. Find vp) and i(t) b. Sketch vp(0) and ip(t) c. Whats the phase difference between v(t) and ip(t) en v(t) = 10cos 104t = and v(t) = Scos (104 + 7/3) played on a scope with a Time Base setting of 0.1 msec/div a. By how many divisions will their traces differ b. Sketch vj(t) and v9(t) on the same graph zn two sinusoids v;(t) and v2(t) of the same frequency and magnitude v0 vo) ce an expression for v2(t) if v(t) = 10.cos 1000t and if v;(t) takes 10 sions/period and v2(t) crosses over 1 division after vy(t) n 5- 36CHI + “10 - cH2 vi, v2 Explain why the following procedure works for measuring the phase difference between ‘vj and v9(t) on a dual trace scope ~ a. Set Channel 1 and Channel 2 sensitivities to the same scale b. Set the scope to trigger on Channel } c Choose a reference point on the graticule and then observe with respect to that int @ First v(t) = (CH1 - CH2) Gi) And then (without changing any control settings) v2(t) = CH2 4. Finally calculate the phase difference from the relative positions of the waveforms as you did in Problem (5) 7. Why can we take the voltage scales out of calibration when making phases measurements 8. When can we take the time base out of calibration when making phase measurements 3 LAB — Each student is to do his own experiment 1. Measure and record your resistor and capacitor values. Do not test for capacitor leakage current, 2. Given IK i, vin 0 = + 5 cos 2n(1.5x103 )t O1pt vet where we have the assumed the source has zero phase. Use the frequency counter to set the frequency. Then a. Measure the phase difference between Yci(t) and ic(t) b. Measure and write an expression for vc(t). Note that we have assumed that the source has zero phasec, Measure and write an expression for ic(t). Note that the current through the capacitor equals the current through the resistor Demonstrate these measurements to your instructor € 3. Given 1K . iL@ Vp ® = bet 5 cos 2n(1.5x104)¢ () mh & YLO where we have again assumed the source has zero phase a. Measure the phase difference between vz (t) and iy(t) b. Measure and write expressions for v(t) and iz.(t) QUESTIONS AND PROBLEM 1, Do the currents and voltages measured in the lab satisfy the i-v relation a id = ot e ». vo = LoD 2. Given . ic + inf %oO Find vc(t) if ig(t) = 10cos (1000 + n/4) and v¢(0-) = 3 volts 3. Given 5- 38 €i@ 10 cos 104 N Find the circuit element N and its value if a. i(t) = 10-2 cos 104 b. it) = 10-3 cos (104 + 1/2) c. io = 10-2 cos (104t — n/2) 4. Given i R 10 cos 104 L Find R and L if i() = 0.707x10-? cos (104 — 1/4) 5. Given lnt ‘What capacitor should be added, and how should it be added, to obtain a total capacitance of a Lipe b. O9ufCHAPTER 6 a FIRST ORDER RC AND RL CIRCUITS 6.0 TRANSITION Now that we've introduced capacitors and inductors we want to start connecting them together with our other circuit elements — resistors, dependent sources and op amps — and see what happens. In this chapter we'll look at the simplest such circuits — those containing one capacitor or one inductor such as 3 © We call these Ist order circuits because to analyze them we're going to have to write and solve Ist order differential equations. 6.1 BASIC ANALYSIS OF FIRST ORDER RC CIRCUITS 6.1.1 BEGINNING AS SIMPLY AS POSSIBLE Let us begin by looking at the simplest possible Ist order RC circuit consisting of one resistor and one capacitor as follows ° Zig ‘c © € ‘x0 i “0 8 ve(0) = 2 volts EXAMPLE 6.1 Given suppose we want to find vo(t) for t > 0 given that the circuit starts off from the given initial condition of v¢(0) =2 voits. USING OUR INTUITION @ Since there is an initial voltage of 2 volts across the capacitor — corresponding to an initial charge being stored on its plates — va(0) is also going to be 2 volts. Asa result an initial current of value into) ~ RO - vo) 2- in@ = “RE 0) - 2 = 2ma is going to flow thru the resistor and and in so doing start discharging the capacitor as indicated in ce Now as time passes the voltage ve(t) will continue to get smaller but at a slower rate because @ 6-2 €smaller vc(t) means a smaller ig(t) as indicated in Yo As we will show in the next section vo(t) is in fact decaying exponentially and is therefore of the form Kerat ‘as we have implied in our sketch. We also expect that all the capacitor's initially stored energy is being dissipated in the resistor. VERIFYING OUR INTUITION Since Kirchhoff's Laws hold for RC circuits just like they do for all the other circuits we've been working with, we can write node equations for them in the usual way as follows — Node Equation vo 6 BV (0) 1 — + 19 =0 10 But now the result is a first order differential equation which — after multiplying by 106 - we can write in the standard form ve + 10 ve = 0 vo(0) = 2 Note that the initial condition ve(0) =2 has to be given to us ~ there is no way we can obtain it if we're not given the input before time t=0. The initial condition tells us the net result of everything that has happened to the circuit up to time t=0. So far so good. But now we have to figure out a way to solve our differential equation! There are several ways to go about this. What we're going to do here is take the easy way out and present a cookbook method that hopefully you've seen in your class on differential equations. Butif you haven't — don't despair — we'll be coming back and verifying our procedure when we develop LaPlace Transforms in ECE 307.COOKBOOK PROCEDURE FOR SOLVING FIRST ORDER a DIFFERENTIAL EQUATIONS WITH CONSTANT COEFFICIENTS ‘Typical examples of first order differential equations with constant coefficients that we are going to be running into are like ¥ + 103v = 3x10? (0) = 2 volts and ¥ + 2x103v = cos 10% = -v(0) = —2 volts ‘The functions 3x103 and cos (103t) on the right hand sides are referred to as forcing functions. They depend on the circuit's sources. To solve such equations — of the general form Vtav=f) v0) = Vo we proceed as follows — 1. We first set the right hand side of the differential equation to zero and solve the resulting homogeneous differential equation wh + avy = 0 tofind the form of the homogeneous part of the response vp(t) we a, Obtain what is referred to as the characteristic equation € D+az=0 by replacing vhoby =D ad vy by Di=D and then b. Solve the algebraic characteristic equation for its characteristic root D=-a which we then substitute into the equation for the homogeneous solution vptt) = Ke-at The value of K will be determined later. It depends on the initial conditions. 2. Once we've got the homogeneous solution we go back to the original differential equation to Obtain what we refer to-as the particular solution vp(t). The particular solution dependson the forcing function. Let us illustrate with an example, Suppose we want to find the particular solution of Yp + 1B vp = 4x10? The particular solution vp(t) is that function which when substirated into the left hand side of the equation yields the right hand side — in this case 2x103. To find the form of vp(0) swe again take the easy way out by looking in our “cookbook” which tells us that whenever the forcing function of a first order differential equation is a constant then so is the particular solution. In particular p(t) 2A To actually find the value of A we have to substitute into our differential equation as follows dA + 1034 = 4x103 a to obtain vp(t) = A= 4 3. The final step in solving our differential equation is to obtain the complete solution as the sum of the homogeneous and particular solutions v(t) = vate) + vp(t) = Kerl0000 + 4 ‘And then use the inital conditions to solve for K. Doing this we have vO) =2=K+4 which gives us K=—2. And so finally we have v(t) = -2e-1000t +4 ‘Now let us apply our cookbook procedure to our original differential equation Yo + Wve =0 = ve(0) = 2 1. The first thing we do is take the homogeneous differential equation wy + 18 vy = 0 and come up with the corresponding characteristic equation D+ 103 =0 som which we calculate the characteristic root 6-5D=-108 A From all this we have that the homogeneous solution is of the form € vat) = Ker2000t 2. Next we have that the particular solution is zero since that's the value of the function satisfying Vp + 103 vp = 0 3. Now combining the homogeneous and particular solutions we have the complete response ve(t) = vat) + vp(t) = Ken1000t + 0 = Ke-l000t Now using the initial condition to find K we have vel) = 2 = KeO = K and so finally the final solution volt) = 2¢71000t which as predicted looks like e , oO 2 6.1.2 THE TIME CONSTANT + Now let's see if we can get a handle on how fast we would expect the capacitor in a general RC circuitR c YQ to discharge. We would expect that the larger R. the smaller the current will be and therefore the longer it should take for the capacitor to discharge. And analogously the larger C the more charge the capacitor will have stored and so again the longer it should take the capacitor to discharge. In both cases this would make the coefficient a smaller. Writing and solving the node equation to verify our intuition we have the ‘The characteristic equation is then D+IRC=0 ‘with characteristic root D=-1RC As a result we have that vo(t) is of the general form Vo(t) = Ket /RC where as always K depends on the initial condition. We now define the RC product — which has units of time - to be the time constant 1 of the circuit. Specifically we define t= RCVelt) = Kert/t ‘What's particularly nice about knowing t is that every t seconds the response decreases by a factor of 1 = 0.3679 as illustrated by So + gives us a real handle on how fast the capacitor is discharging — how fast all the voltages and currents are decaying. In particular from the following Table t Ke t/RC = Ke-Vt € . 2n 3t at St we see that after 5 time constants t= St the response will for all practical purposes have died out — decayed to zero. ‘Note that just as we originally predicted the larger R and C, the larger t and therefore the longer it takes for the capacitor to discharge.EXAMPLE 6.2 Given ip) = 2ma find and sketch ig(t) and also find how long it will take the capacitor to discharge. From the Node = Equation 1 ne + 04x10), = 0 we obtain Ve + 5000 ve = 0 which implies that Volt) = Ke-5000t Now the initial condition vo(0) is hiding in the value of ip(0) but it’s easy to ferret out because vc(0) = Rig(0) = 500(2ma) = 1 volt and so we have ve) = 1 = KeO=K ‘Substituting in we have YVelt) = 75000 and soig(®) = veltVR = 2°5000t 500 = 0,5 -$000t ma which looks like in ® (aa) Now since the time constant = RC = 1K(4uF) = 4 msec the capacitor will forall practical purposes discharge in about T = St = 20msec Note also that we refer to Ut which has the units of one over time as the natural frequency of the circuit. 6.13 KEEPING TABS ON THE CAPACITOR'S STORED ENERGY ‘We would expect, of course, that all the capacitor's stored energy is going to be dissipated in the resistor. To verify this, let us suppose we have a "general" Ist order RC circuit like R c Vo with initial condition V¢(0). Then the capacitor will start off with an initial energy Log E=5CV.© 6-10Now the amount of energy dissipated in the resistor will be y @ 7 + vw E = [oa = apa 4 ey URC 2 veo BS “| yt c.2 -2uRC -FVjLO(e )I, 1.2 = “FZ CV¥.@ (0 - 1) 1 sieve which as we expected is exactly equal to the amount of energy originally stored in the capacitor. 6.14 ADDING A CONSTANT SOURCE : r ) ‘Now suppose we add a constant voltage source to our Ist order RC circuit R in c Yo and see how the circuit responds, USING OUR INTUITION Let us take a look at a specific example to see if we can get 2 handle on what should happen in such circuits, EXAMPLE 6.3 Given1K = 10volts * inf ¥o Vo(0) = 2 volts find vo(t) if The first thing that's going to start happening is that the voltage across the resistor will cause a current to start flowing thru it equal to ig@ = (10-2)/1K =8 ma thus causing the capacitor to begin charging ‘up as follows %oo 2 t As a result the voltage across and current thru the resistor will decrease YR ig (aa) ‘NO ; . AAS this happens the capacitor will continue charging and the voltage across it continue to increase but at a slower and slower rateYe Now the question is whether there's any limit to how large v¢(t) can get. The answeris yes and that the limit is 10 volts. The reason is that if vo(t) were to somehow or other reach 10 volts then the voltage across the resistor would be zero and so the charging current ig(t) would have to stop flowing. But in our circuit — at least the way it stands — the voltage can never actually reach 10 volts. The best it can do is get asymptotically close as indicated by %0 VERIFYING OUR INTUITION - Taking our circuit 1K ® = 10 volts ie Yo and writing the Node Equationvo 710 < 3 +10 v, = 0 10 we obtain after suitable manipulation Wg + 10°v, = 104 ‘Now going back to our cookbook procedure for solving first order differential equations we have 1. First the homogencous differential equation vy + 10°v, = 0 with solution vy(t) = Ke1000e Note that by setting the right hand side of the differential equation to zero we're in effect turning off the source. This gives us the same equation as in the previous section because with the source off both circuits are the same! Note also that we don’t solve for K until the last step when we have an expression for the complete solution. 2, The next step is to retum to our original differential equation and solve for the particular solution satisfying Vp + 10%vp = 104 Since our cookbook tells us that the Particular solutions in such situations are constants we have — after substituting in SA + 084 = 108 the result vp) = A = 10 3. Now putting everyting together we have the expression for the complete response Volt) = va) + vp(t) = Ke™1000t + 10 from which we can obtain K by substituting in the initial conditionsvo0) = 2 =Ke0 + 10 =K + 10 . and solving for K to obtain K=+8 ‘Therefore — finally we have which looks just like our prediction YQ) 2 6.15 SOME TERMINOLOGY 1. THE NATURAL AND FORCED RESPONSES ‘We as engineers usually refer to the homogeneous part of the solution vp(t) asthe natural response yn(t) and the particular part of the solution vp(t) as the forced response v(t) We refer to the homogeneous part of the complete solution as the natural response because its form — which for our first onder circuits is a decaying exponential Ket and the value of t are completely independent of the magnitude of the source as well as whether it happens to be a constant, a sinusoid, or whatever. (On the other hand, we refer to the particular solution as the forced response because its form ) 6-15docs depend on the form of the input —"'the forcing function. So we can now write ve(t) as — velt) = val) + vat) € where for our example va(t) = —8e1000t and vet) = 10 Now plotting vo(t) as the sum of these two terms we have that Vo(t) equals the sum of the namural response v0 plus the forced response ¥,@) 10 t and so equals * ‘ofIt should be clear that the responses of first order RC circuits with constant inputs are always es going to be of the form v(t) = 2671000 + 5 asin ourexample, An interesting characteristic of such responses is that the natural parts responses are decaying to zero while the forced responses are staying with us. When the natural responses of our circuits do in fact decay to zero as in our examples, we refer to them as transient responses. On the other hand, we refer to that part of the response that does not decay to zero — that stays around — as the steady state response. In particular, for v(t) above Transient Response = 2¢7100t Steady State Response = 5 Note that the steady state response does not depend on the initial conditions. Note also that once an RC. circuit with a constant input like ig is in the steady state then iC(t) will be zero and the corresponding steady state values can be calculated by replacing the capacitor by an open circuit Vig =A and then analyzing the comesponding resistor circuit. 6.1.6 FIRST ORDER RC CIRCUITS WITH SINUSOIDAL INPUTS Sinusoids are magical functions. The reason we say they are magical is that the steady state response of an RLC circuit to a sinusoidal input vin(t) = A cos(at +8) is a sinusoid at the same frequencyBeos(at + $) Now the magnitude and phase are in general different — but if they were the same that would take all the fun out of it and leave us nothing to figure out. Other common functions like square waves, sawtooth waves and so forth definitely do not have this property. The underlying reason that sinusoids have this magical propery is that the derivative of a sinusoid is again a sinusoid at the same frequency. To see how to analyze a 1st order circuit with a sinusoid input let us do an example. EXAMPLE 6.4 Given 1K Yin® = = Scos 103+ tHe ¥e0 find v¢(t) given thar ve(0) = 1 volt Writing the node equation as we have done in our other examples Ve + 10° vq = Sx10% cos 10°¢ and then following our cookbook procedure we have 1. A natural response equal to the solution of Vat 10°v, = 0 which is vq(t) = Ker1000e 2. And a forced response equal to the function which when substituted into the left hand side of Vg+ 10° vp = Sx10°cos 10°t yields the right hand side. From our cookbook we have that if the forcing function for such a 6-18 € €e differential equation is a sinusoid of a given frequency then so is the forced response is also a sinusoid at the same frequency. In particular for our example we have vet) = A cos (10%t +8) ‘The problem is to find A and 0. As before we substutite into the differential equation 4 (40s (10° + @)) + 10°(A.cos (107 + 8)]= Sos (10°) to obtain — 103A sin ( 103t+6) + 103A cos (103t +) = 5x103 cos 103r Now one way to find A and @ is to make use of the trig identity Boosx + Dsinx = VB? + D* cos + unt EP} B Substituting in the values from our example we obtain 2 2 V co%a) + (107A) cos [o + an which reduces to = 5x10° cos 10° ~10°A) (107A) A107 YZcos (10% + @ + m/4) = Sx10°cos 10% So AlO*{Z = 5x10? and 0 + m4 = 0 and therefore which gives us 5 3 ve = Fy e0s (10"t - 7/4) 3. Now finally putting everything togethervolt) = vplt) + vet) e vel = KEM % cos (10% - wi) and using the initial condition 5 ve(0) = 1 = K + cos (-n/4) we have K=-15 and so Ye () = = 15 71000 4. $ cos (10% = /4) Plotting vo(t) as the sum of its namural and forced responses we obtain ¥,0 plus %@equals Yo) Note that we refer to the sinusoid remaining after the transient has died out as the sinusoidal steady state response. AN IMPORTANT OBSERVATION Tf we take our first order RC circuit + c v0 Vin O= Acos at and increase the frequency of the sinusoidal input then the voltage ve(t) across the capacitor is going to decrease. The reason is that as the frequency increases the charge is going to have less and less time to accumulate on the plates before the sinusoidal source changes sign and the capacitor has to start discharging, Of particular interest is when the frequency gets “really large” and the voltage across the capacitor is just about zero. We can then find the moi steady state response of such a circuit by replacing the capacitor by a short and analyzing the resulting resistor circuit. EXAMPLE 65 To find i(t) in ig) 1K Scos otat "very high" frequencies we replace the capacitor by a short as follows if) IK Scos wt and analyze the corresponding resistor circuit to obtain i) = Scos (@t) ma‘eo 62 FIRST ORDER RL CIRCUITS Fist order RL: circuits behave basically the same as first order RC circuits, In particular if we take a look at just one R and one L connected as follows ipo + R SJ v0 then we would expect the current thru the inductor to decay exponentially ipo as the resistor dissipates the inductor’s stored energy. The only real question is how the time constant t depends on R and L. What we would expect is that 1. The larger L the more energy the inductor is storing and so the longer it should take the resistor to dissipate it and 2. The larger R the faster the resistor will dissipate the inductor's energy As a result we would expect the time constant of a first order RL circuit to be directly Proportional to L and inversely proportional to Ras indicated in as we can easily verify by Writing and solving the differential equation for any of the variables in the circuit, In particular writing the node equation for vz(t) in 6-23ipo € R vo we obtain Node Equation 10 R t +i,@ + tf vWd=0 a ‘Taking the derivative of this integrodifferential equation to put it in our familiar form we obtain . R ‘+p w=0 which has the solution vit) = Ke “(RL with time constant t= LR as predicted. Note that the procedure for finding the response of RL circuits to constant and sinusoidal inputs is analogous to what we did for RC circuits, One particularly interesting difference, however, is that we can the steady state response of an ‘RL circuit wo a constant input by replacing the inductor by a short and to 2 sinusoidal input of "very high” frequency (> R/L) by replacing the inductor by an open. This is just the opposite of what we do for RC circuits. 63 ADDING MORE RESISTORS AND SOURCES ‘The easiest way to analyze circuits like 6 - 24is usually to find the Thevenin Equivalient as seen by C Roy ¥ TH cS and then analyze this simplified circuit. EXAMPLE 6.6 Given 1K pf = lov © i IK 1. The time constant t and Find 2. The steady state value of vo(t) First of all the Thevenin Equivalent of the resistive part of the circuit as seen by C 6-25V;, = 10v m™O® os 1K is simply Rry = 15K 9 Va = S volts and so our circuit reduces to Tat 15K + Vig = 5 TH inf Yo for which Lt = Ray = 1.5x10%x10- = 1.5 msec 2. The steady state value of volt) is vr = 5 volts 6.4 FIRST ORDER CIRCUITS WITH SWITCHES 6.4.1 RC CIRCUITS WITH SWITCHES Let's begin by looking at 1K =i J: Ov pf Yo with the switch closing at time t= 0. The central question is how all the voltages and currents in € 6 - 26the circuit are going to change from what they are just before the switch closes — their values attime t=0- to what they will be just after the switch closes — their values at time t= 0+ Clearly there can be no currents flowing in our circuit at time t=0- no matter what ve(0-) is and so ig(O-) = ig(0-) = Oma Now let's suppose that yc(0-) = 2 volts and see what happens when the switch is closed. From what we've observed in the previous chapter on capacitors we know that the voltage across Fs capacitor can't change instantaneously unles there isan impulse of current. What this ly means is that the voltage across the capaictor is not going to change instantaneously unless the capacitor gets shored out oo Lt or switched in parallel with a voltage source Since nothing like this is happening in our circuit — the IK resistor is “in the way" — it must be that the capacitor is staying put at 2 volts — ve(0+) = vc(0-) = 2 volts Now to calculate all the other voltages and currents at time t = 0+ let us replace the capacitor 6-27by an equivalent, "voltage source” ip) 1K ® € if c (0+) @) (3 Yo(+H)=2v Vig(0+) = 10¥ and analyze the resulting circuit just as if it were a regular linear resistor circuit. In fact since we have the node voltage vc(0+) = 2volts we have in(0+) = (10-2)1K = 8ma ic(O+) = ig(OH) = 8ma ‘Be sure to note that the currents — in particular ig(t) — did change instantaneously even though c(t) did not. The problem of finding v¢(t) for t> 0 therefore reduces to analyzing 1K . Vin =10v Ine re) ye(0+) = 2volts 6 - 28find eS 1. The resistor currents before the switch closes 2. The resistor currents and capacitor voltage after the switch closes at t=. given that vo(0-) = 0 volts Now a. Before the switch closes we clearly have igy(0+) = igg(0-) = 25ma and igg(0-) = Oma b. And since the capacitor is being neither shored our nor. ending up in parallel with a voltage source when the switch closes its voltage must be staying put at vo(0+) = vo(0-) = Ovolts ‘Therefore we have at time t= 0+ ©) 1K Vo(0+) =0v with node voltage v4(0+) =2.5 volts and so ig O+)=3.75 ma —igg(0+)=1.25 ma —_igag(0+) = 2.5 ma One particularly important observation to make is that the calculation of the voltages and currents in such circuits at time t= 0+ does not invoive the writing or solving of any differential equation. These problems are basically simple resistor analysis problems. 6.4.2 FIRST ORDER RC CIRCUITS WITH STEP INPUTS The unit step function u(t) is by defined by wo= {2 ESB 6-29Graphically the unit step looks like ue) € ‘What it really corresponds to is a source of value one that is being "switched" into the circuit at time t=0. The unit step response s(t) is then the response of the circuit to a unit step assuming all initial conditions are zero, EXAMPLE 6.8 The unit step response s(t) of 1K + Vig O = WO) ibe YQ) is simply a. Equal to 0 for t<0 b, And equal to the response of 1K 7 Yin ®= 1 nt Yo ve(0-) = 0 for t 2 0 — whichis simply - 1000: ve = 1-e Combining our results for t 1 msec if for t< 1 msec the circuit is in the steady state when the 3° input changes value at t= 1 msec as indicated in Vin © 10 5 t (msec) 1 6.34 Given R + Yin c vo @ with vin(t) equal to the pulse train Vin t 12 T Sketch v(t) if 6.35 Given 3 6 - 531K Vin © : lyf == ¥o® t (msec) t (msec) Find Vinin and Vmax FIRST ORDER RC-ACTIVE CIRCUITS 6.36 Find the step response of 6.37 Find the step response of 6- 54 «10K 6.38 Given vo(0-) = 2 volts a. Find the steady state response », Find the complete response 6.39 Find the complete response of & 6-5510K cos 103 © = vo) assuming zero initial conditions COMPUTER-AIDED ANALYSIS OF FIRST ORDER CIRCUITS 6.40 Given that the responsé of a particular first order circuit is 4 v() = 10027 a. Sketch v(t) 'b. Where is the derivative of v(t) maximum c. Where is the derivative of v(t) changing the fastest @ @. Approximating the derivative by av _ Mtg + Ad) - vUto) at) = Find for some interesting value of fy how small t has to be for the approximation to be within 10% of the actual value - within 1% of the actual value e. Are the results what you expected. £. How do your results affect a computer approximation to the solution of the differential equation for v(t) 6.41 Given 3 cos 1035 © pt Yoo 6 - 56Draw the companion model for approximating volt) at t = 0.1 msec given that vc(0) = © 5 volts. Then find the corresponding approximation at t= 1 sec 6.42 Draw the companion model for calculating the response of 1K cos 105 ¢ & 1K e 6 - 57ECE 252 - EXPERIMENT 6 RC TIME CONSTANTS OBJECTIVES ‘The student will be able to measure and calculate the transient responses of a first order RC circuit MATERIALS REQUIRED BY STUDENT L 2 Oth expack 3. Two IK and one 10K 5% 1/2 watt resistors 4, 10K 5. 4 BNC to clip connectors PRE-LAB — Due at the beginning of the lab 1. Given | Vi, 0 =5¥ vc(0-) = -Svolts a. Write the differential equation for v¢(t) b. Solve for v(t) c. Calculate igft) 4. Use the program PLOT — oran equivalent — to obtain separate graphs of vc(t) and ig(). Make an extra copy for the lab . How long does it take the transient response to for all practical purposes to decay £. Redo the above steps for vin = —Svolts, vc(0-) = +5 volts g. Explain why, and under what conditions, the results for parts (a) thm (f) can be used to find the response of vc(t) toa square wave of 5 volts amplimde (10 volts peak to peak) 2._"Now suppose we replace the $ volt source in our crit withthe squerewave of magninue volts 6- 58‘Then what should we make the squarewave's frequency in order to alternately see the capacitor "charge up" for 7 time constants to +5 volts and then “charge down” for 7 time constants to ~5 volts LAB — Each student is to do his own experiment 1. Measure and record your resistor and capacitor values. Do not check for leakage currents 2. Build the circuit analyzed in the prelab 1K na a Yin ope Yo with vin(t) equal to a squarewave of magnitude 5 volts and the 1K resistor equal to your resistor plus the Thevenin Equivalent of the source a. Using the frequency counter, set the frequency of the squarewave at the value calculated in the prelab. Put data points from your trace onto your plots of vc(t) and ig(t) from the prelab. b. Replace the 1K resistor by your pot and see what happens asthe resistor value is vari Demonstrate these results to your instructor 3. Sketch the response of Y¢(t) e 6- 591K WwW + vin IK O1pe Yo to a squarewave of judicious frequency that gives us a good picture of the circuit's capacitor reaching steady state 4, Given the following op amp circuit 10K Vin ‘Adjust the frequency of the squarewave input until you get a nice trace of the transient response, Then make an accurate enough sketch of your result so that you will be able to estimate + QUESTIONS AND PROBLEMS 1. Compare the calculated and measured responses of your RC circuit at the first 5 time constants 2. Given the RC circuit 1K + Vin ® O1pe vo® ‘What happens to the RC time constant t and the time it takes the transient response to 6- 60decay as a R decreases e b. R increases ©. C decreases dC increases 3, Estimate the time constant of the op amp circuit you builtin the lab from your sketch of its frequency response 4, Given vi, =5v¥ mM O1pe Yow) ve) = a. How much charge is stored in the capacitor when it is fully charged b. How much energy has been dissipated in the resistor after the capacitor is fully charged ¢. How much energy is supplied by the source during charging e 5. Whats the time constant of v0 6. Given 1K Vin® = Sos 103 fant - Yc a. Write the differential equation for vc(t) ». Find the forced response 6 6- 61CHAPTER 7 SECOND ORDER CIRCUITS 7.0 TRANSITION Up to now we've only been investigating first order circuits — circuits with one L or one C and therefore characterized by first order differential equations. Now we will consider circuits with two energy storage elements — two capacitors, two inductors or one capacitor and onr inductor like a Yin ® =— 6 RI cl Yin ® C2 ad P>‘We call these second order circuits since we must solve 2nd order differential equations to analyze them. € ‘The solutions of these 2nd order differential equations pretty much follows the same pattern as the solution to Ist order differential equations. In particular the solutions can be found as the sum of a natural plus a forced response. But the form of the natural responses do turn out to be tantalizingly different. ‘We will start our discussion with an investigation of the natural responses of 2nd order LC circuits, move on to the natural responses of series and parallel RLC circuits and then put everything to find the complete responses of general RLC circuits. Finally we will Took at the interesting case of RC-Active circuits. 7.1 THE NATURAL RESPONSES OF 2ND ORDER RLC CIRCUITS 7.1.1 THE NATURAL RESPONSE OF THE BASIC LC TANK Let us begin with the basic LC “tank” circuit ~ ‘We call these tank circuits because charge “sloshes” back and forth from “one side" to the other (and so does the energy) just like water in a tank. One particularly interesting property of this circuit is that it is one of those rare circuits for which the natural response does not die away with time, To get a handle on such a circuit let us assign the values ipo ig + + vo 10mah ine Yo) ip(Q) =5ma and —_v¢(0) = Ovolts and see what happens. USING OUR INTUITION ; From the initial conditions, as shown iniL (aa) e *y]- Yet t— we see that at time ¢=0 the indcutor is storing E,) = ZLEO = 5 (10 (5x10)? = 1.255107 joules ® of energy and the capacitor none since there is no voltage across it. Now as current flows thru the inductor, it's pulling electrons off the lower plate of the capacitor as it pushes them onto the upper plate as indicated in O~7-oOs> <=~O<0< ‘This results in an increase in the magnitude of ve(t) But all this takes work on the part of the inductor ~ the elecuons are not doing this of their own free will. The source of this energy, of course, is the energy being stored in the inductor's magnetic field. So what we have is a transfer of energy from the inductor’s magnetic field to the capacitor’s electric field.In particular, as, Lat Li? = 5 = joules 1 Ee and therefore i, decreases, ve and are going to be increasing as indicated in i, @ Ga) %oO Nore, in particular, that vc(t) is increasing its fastest attime t=0 when it, is maximum. ‘Now as time continues to pass and more and more charge is squeezed into the capacitor, the harder and harder i's going to push back and so the more rapidly the current ir(t) is going to decrease. But the smaller the current, the slower the voltage is going to change and so we've got something likei ja) ee ° \ : th and Yo "1 Ve where at time ty the push back of the charges squeezed into the capacitor has finally stopped the current and the capacitor is no longer charging. At t the inductor has transfered all of its @ energy to the capacitor. We notice, by the way, that at this point in time the curent through the inductor is changing (decreasing) very rapidly. We know that it takes a strong push to do this to the current in the inductor, but since we have so much charge squeezed into the capacitor this is just what we have got. ‘What happens next? Well, the charge squeezed into the capacitor is still nying to get out, itis still pushing hard and so it keeps changing the inductor rapidly as shown in () (ma) ig 5 \ "1But this means the current is now flowing the other way through the inductor — which means that some of the squeezed in charge is finally able to start flowing out of the capacitor — as e indicated in %o® . t Sr, -Vo ‘Nove that the rate of change of vc is small because ic =—iy_ is still small in magninude. But as Jong as the capacitor continues to push on iy in the direction iy is flowing, the magnitude of iy is going to get larger and larger. But the larger the magnimde of ‘ir, gets the faster the capacitor will discharge and so the left charge there will be left to push on iy!! Therefore the magnitude of iz will keep increasing but not as rapidly and so we have i, © (mm) Yow t 1, Finally the capacitor is totally discharged, the magnitude of vc is down to 2er0, the capacitor stops pushing to increasing the magnitude of i, and so the magnitude of iy. has gomten as large as it's going to get as indicated in =i aaand e@ Since vc is zero again — and the magnitude of iz, is large again — all the energy has sloshed bback from the capacitor to the inductor and we are in effect back where we started except that now the current thru the inductor is going in the opposite direction. In particular iy = ~Sma instead of +5 ma. So everything should repeat itself except that all the voltages and currents will now have the opposite signs as follows i L© jm)Now we're back where we started — but without any loss of energy — and so the cycle must repeat itself forever. As our plots indicate — and as we will in fact show when we write and solve our circuit comesponding 2nd order differential equation — the signals ij(t) and vc(t) are both sinusoids bbut 90° out of phase. In particular p(t) = 5x10°3 cos aot Velt) = Vo cos ((ogt + 7/2) where «& ©) =2n/T rad/sec and T= period in see To calculate Vo we can make use of the fact that at those times when the inductor has ansferred all its energy to the capacitor, the capacitor will be storing 1 Lig S v2 a 30 = FCO) V5 = 1.252107 joules = Vo = 0.5 volts and so we have Volt) = 0.5 cos (gt + 1/2) Now let's write and solve the differential equations to verify our intuition and in the process calculate @, for our example.VERIFYING OUR INTUITION ‘Writing the node equation for our circuit @ ip@ * 10mh ine Yo ip(0)=Sma_—_ ve(0) = 0 volts we have ihe Node — Equation t 1 1 + (SG | wee s ED <0 ‘Taking the derivatives of both sides we obtain fis Wye +0¥, = 0 and so with initial conditions[Cc A COOKBOOK PROCEDURE FOR SOLVING 2ND ORDER @ DIFFERENTIAL EQUATIONS WITH CONSTANT COEFFICIENTS Our cookbook approach for solving 2nd order differential equations with constant coefficients is basically the same as the one for solving their 1st order cousins — 1, The first step is to find the form of the natural response. We do this by a. First setting up the characteristic equation — which we do in the same way as in the first order case but now with the addition that the 2nd derivative is replaced by D2, b. And then using the result from differential equation theory that the characteristic roots — the natural frequencies of the circuit — correspond to the natural responses as follows CHARACTERISTIC FORM OF THE DAMPING TYPE ROOTS NATURAL RESPONSE a,b (both real) Kye! + Kyebt Overdamped a Ket! + Kpteat Critically Damped @ a+jb, a—jb Kea cos (bt + 0) Underdamped jb, 3 K cos (bt + 8) No Damping Note that the characteristic roots depend only on the coefficients of the differential equations — which depend on how the circuit elements are connected and together and what their values are. ‘They do not depend on initial conditions or the values of any inputs. 2. The forced response is then found in the same way as it found when we follow our cookbook procedure for solving ist order differential equations 3. And finally we add the narural and forced responses together to form the complete response Complete Response = Natural Response + Forced Response and then use the initial conditions to obtain the parameters K and 6. This procedure is relatively straightforward for inputs like constants and sinusoids but can cerainly be tedious. 7-10 «e Now letus use our cookbook procedure to solve our differential equation Wo + 18 ve = 0 1, First we take the differential equation for the natural response Yn + 108, = 0 obtain the characteristic equation D2 + 108 =0 and solve forthe characteristic roots D == j10+ ‘We then use our Table to find the form of the namural response va(t) = K cos (104t + @) 2. Now since our circuit has no input the forced response is zero vat) = 0 @ 3. And so the complete response equals Vet) = valt) + vat) = Kos (104 + 6) Now we can use our initial conditions to find the values of K and @. In particular since vc(0) = Keos@ = 0 4) = -10*Ksin@ =-Ssx10? = Ksin@=0.5 we have tan@ = Ksin@ =< K cos @ =>and so finally volt) = 0.5 cos (104t + 2) and 5x10 cos 10°t which are of the form and magnitude we predicted. HOW THE NATURAL FREQUENCY @, OF AN LC TANK DEPENDS ON L AND C ‘As we've seen the capacitor and inductor in an LC tank 1g Fe transfer the circuit's energy back and forth at a frequency we refer to as @g. Let's now take a Jock at how the values of L and C affect Wo (1) The Affect OF C — The larger a capacitor’s plates the more charge that has to be delivered before any “appreciable work — any real "squeezing in” of the charge has tobe done. So the larger C the longer the inductor's current is going to have to be flowing before it's done enough work to transfer all its energy to the Capacitor. Therefore the larger C the lower ag. (2) The Affect OF L - The larger an inductor the longer it takes to get its current going. ‘Asaresult the larger L the longer it's going to take the capacitor to discharge and so transfer its energy to the inductor. So just like for the capacitor, the larger L the lower O. Putting our two observations together we have that @, decreases as L and C increase. In fact itrums out that maki Som which we can easily show by writing and solving the node equation 7-12for a general LC tank Doing this we obtain the characteristic equation ae Dt + a the characteristic roots D=sj and therefore the claimed result that Vell) = Vo cos (gt + 8) with es ® = te ‘This is a fundamental result. We refer to aL Gis ° "We asthe frequency of oscillation of the LC tank. 7-137.1.2 THE NATURAL RESPONSE OF THE BASIC 2ND ORDER RLC SERIES CIRCUIT Now suppose we put a resistor in series with our L and C as follows 6 ipo 2 COL L c R Clearly iz(t) and vo(t) are going going to have to decay to zero as the resistor R dissipates the circuit's energy. But what is interesting — and at least at first somewhat surprising — is the fact that the form of the decay is diferent for different values of the R, L and C. In onder to get a better feel for what's going on let us take a look at a mechanical analog consisting of a ball rolling in a frictionless trough that is filled with a fluid whose viscosity may be very high, medium, very low or zero. A cross-section of such a set up is shown below Ee ‘Now let's suppose the ball starts from rest high up on the side of the trough as indicated and see if we can get a handle on what should happen. If the fluid is engine sludge the ball will take a long time to get to the bottom of the trough. And when it gets there it won't overshoot. In fact, it barely gets there. If the fluid is heavy oil, the ball will roll toward the bottom faster, but there may still be just enough viscuous friction to keep it from overshooting and rolling up the other side. If the fluid is air, the ball will certainly roll faster and overshoot the bottom and start rolling up the other side. The ball will roll back and forth — oscillate — from one side of the trough to the other. But sooner or later friction with the air wil bring the ball toa stop at the bortom of the rough. Finally in a vacuum the ball would just oscillate back and forth from side to side forever. These are just the cases of overdamped, critically damped and not damped behavior tat we see in the esponses of RLC circus. So let's go back over this again — ina litle more detail! If the fluid is “infinitely thin" — that is, a vacuum, no fluid at all — the ball will oscillate back and forth forever. 7-14If the fluid is very thin, like air, the ball will not be going quite as fast at the bottom of the hill as it was when it was in a vacuum. In particular it will have less kinetic energy, having lost some by friction to the air. But it will stil certainly be moving and so will head up the other side of the trough. But now two things should be clear — “1, After each oscillation cycle, the ball will climb a litle bit less up the slope. ” 2. The period of oscillation will be slightly longer than when the ball was in a vacuum. Now suppose we make the fluid thicker (more viscous) stil Atfirst — since the grade is steep — the ball will start rolling as indicated in we Now the question is, as the ball goes on down toward the bottom, what happens to its speed, that is, to its energy. As the ball approaches the bottom and the slope of the trough is less, the ‘component of gravity tending to speed up disappears. But the frictional force of the sludge, tending to slow it down remains. Thus, even before the ball reaches the bottom of the hill it might start slowing down; if the sludge is thick enough it will certainly start slowing down before it gets to the bottom. Now the curical question becomes — will the ball's speed have decreased to zero by the time it gets to the bottom, or will it still have some speed left even if only a lite. If it does have some speed left at the bottom, then it will pass the bottom and head up the other side, even though it won't go up as high as it was when it started. When it comes 10 a stop it will start back down again .... andso on. Thus we will still have oscillations, even if they'are of a very low frequency. On the other hand, if the ball has no speed left at the bortom, and it stops at the bottom, then thar's it. The oscillations have disappeared — all we have is an exponential decrease of speed and position toward zero. It would seem reasonable that there is some degree of sludge thickness (as the ball approaches the bottom and the slope of the hill disappears) that will be enough to bring the ball to a stop by the time it reaches the bottom. The ball will simply start rolling down from above, where the slope is steep, pick up some speed, but then slow down due to the sludge where the slope is shallower and finally come to rest at the bottom without ever overshooting. This is what we refer to as critical damping. If still thicker sludge is used the ball will still reach the bottom but will take longer still and of course will not overshoot. In this case we say the system is overdamped. Going back now to our series RLC circuit, electric charge is that system is the analog of the ball in our example. Charge starts built up on the capacitor, just as the ball started high on the side of the rough. In the electrical system charge “sloshes” back and forth from one side of the capacitor to the other. 7-15‘The capacitor's push is the electrical analog of gravity tending to pull the ball down the walls of the trough. The inductor’s tendency to keep current flowing is the electrical analog of the momentum of the e ball which carries it past the bottom of the rough, And electrical friction between the charges and the resistor is the electrical analog of the friction ‘between the ball and fluid in the rough. Not surprisingly, if you take a course in dynamics in the physics department, the equations you derive to describe the motion of the ball are very much the same as our equations for electrical circuits. Now let's go back and actually analyze our series RLC circuit. THE UNDERDAMPED CASE - SMALL R Assuming that R_is small enough, itis certainly reasonable to expect that our series RLC circuit would oscillate like the LC tank — but have oscillations that smaller and smaller as the resistor dissipates the circuit's stored energy as follows This of course is just like a ball in a trough filled with a not very viscous fluid, As indicated in our graph and as in fact we will show when we write and solve the corresponding differential equation the natural responses of the variables under such circumstances are damped sinusoids of the form Kel cos (apt + 8) ‘We refer to such natural responses as being underdamped. The parameters a and depend on the values of R, Land C while K and @ depend on the circuit's initial conditions. ‘Note that a circuit being underdamped corresponds to the characteristic roots being complex conjugates. Note also that 7-16forms an envelope which determines how fast the sinusoid. decays as indicated in In particular, the larger a the faster the decay. EXAMPLE 7.1 Find the response of ip(t) in ip@ 2 COL =. TH soa ve(0) = -5 volts and ip@) = 2ma This corresponds to a position in our mechanical analogy with initial conditions which have the ball on one side of the trough at t=0 and rolling up that side of the rough. For this circuit, we can, as always write the node equations. But our circuit would have two nods equations while only one mesh equation so let us make the prudent choice and write the one mesh equation! Tn particular for our circuit we have the Mesh — Equation din) 1 + . 1 1072! + vo(0) + af Ode + 50iL) =0 a 10° Jy ‘Taking the derivatives of both sides of our equation and multiplying thru by 100 we have 2; as PO 5 52103 HO 5 08s: = a? &tNow going thru the steps of our cookbook procedure for solving 2nd order differential equations we 1. First take the differential equation for the natural response @ i . Hin 5 55103 Hal 5 108 ince) = 0 at? dt Form the characteristic equation D2 + 5x103D + 108 = 0 and solve for the characteristic roots Dy, Dy = -2.5x103 + j 9.7x103 Now since the characteristic roots — the natural frequencies — of our cireuit are complex conjugates we have from our Table that the natural response is of the form in() = Keat cos (mat +8) = Ke-2500t cos (97001 + 6) 2. Since there is no input ig) = 0 3, And so the complete response is ip (t) = ig() + ig(t) = Ke~2500t cos (9.71051 + 8) Now using the initial conditions | i) = 2ma EO = bo = 45(-01 + 5) = 490 tofind K and @ we have from iL) = Keos@ = 2x10-3 SL ( = ~25x10° K cos - 9.74103 K sin = 490 Now substituting the result from the first equation Kcos@ = 2x10-3 into the second equation we obtain Ksin@ = -0.051 and so@=-153 rd and so finally we have ip (t) = 51 ¢°2500t cos (9.7x105 - 1.53) ma ‘Now with the time constant for the envelope °2500t being t= 1/2500 = 4x10“ sec and the period of the sinewave being T = 2nlag = 21/9700 = 0.6 msec the response of iz(t) should look like ipo which is of the form we predicted. Note that the response will for all practical purposes decay to zero in about St = 5(0.4 msec) = 2 msec THE OVERDAMPED CASE = LARGE R Now suppose we increase R. At first we might suspect that increasing R simply causes @, to get smaller and smaller until R reaches infinity when the circuit opens up and the oscillations are 7-19forced to stop. But surprisingly every RLC circuit has some value of R beyond which oscillations stop altogether and the natural response goes through a metamorphasis to the form e Kjeat + Kyebt where the characteristic roots or natural frequencies a and b are real. This comesponds to our mechanical analogy where the fluid is so viscous that the ball can hardly make it to the bottom of the trough — let alone do any oscillating. ‘What is going on is very much analogous to the tortoise who say every second is making it half way from where he is to the finish line of the race. The poor guy is just never going to make it. ‘The closer he gets the more he runs out of steam. For our RLC series circuit ip® 2 CO _ L € R the more the capacitor discharges, the slower it discharges so it can never reach the point where it start building up charge on the other plate. Let us now illustrate this situation with an example. EXAMPLE 7.2 Find the response of iz(t) iL@ vow 4100 _ | @ 10 mh 1H ong vc(0) = -Svolts and i, (0) = 2ma With these circuit element values we have for iy(t) the differential equation os ; FLO 4 gx19¢ BLY 4 108 Gy = 0 dr? da iL) = sh ged a. 7 ip(0) = 2ma ZO=2u0 rm) (-0.8 + 5) = 420 Now going thru our cookbook procedure 1. We first find the characteristic equation D2 + 4x104D + 108 = 0 and characteristic roots 7-20