Transmission Lines Complex Numbers, Phasors and Circuits Complex numbers are defined by points or vectors in the complex plane, and can be represented in Cartesian coordinates z = a + jb j = −1 or in polar (exponential) form z = A exp( jφ) = A cos ( φ ) + jA sin ( φ ) a = A cos ( φ ) real part b = A sin ( φ ) imaginary part where 2 A= a + b © Amanogawa, 2006 – Digital Maestro Series 2 φ = tan −1 b a 1 Transmission Lines Im z b A φ a Note : Re z = A exp( jφ) = A exp( jφ ± j 2 nπ) © Amanogawa, 2006 – Digital Maestro Series 2 Transmission Lines Every complex number has a complex conjugate z* = ( a + jb ) * = a − jb so that z ⋅ z* = ( a + jb) ⋅ ( a − jb) 2 2 =a +b = z 2 = A2 In polar form we have z* = ( A exp( jφ) ) * = A exp(− jφ) = A exp ( j 2 π − jφ ) = A cos ( φ ) − jA sin ( φ ) © Amanogawa, 2006 – Digital Maestro Series 3 Transmission Lines The polar form is more useful in some cases. For instance, when raising a complex number to a power, the Cartesian form z n = ( a + jb) ⋅ ( a + jb)… ( a + jb) is cumbersome, and impractical for non−integer exponents. polar form, instead, the result is immediate In n z = [ A exp( jφ)] = An exp ( jnφ ) n In the case of roots, one should remember to consider φ + 2kπ as argument of the exponential, with k = integer, otherwise possible roots are skipped: n z = n A exp jφ + j 2 kπ = n A exp j φ + j 2 kπ ( ) n n The results corresponding to angles up to 2π are solutions of the root operation. © Amanogawa, 2006 – Digital Maestro Series 4 Transmission Lines In electromagnetic problems it is often convenient to keep in mind the following simple identities π j = exp j 2 π − j = exp − j 2 It is also useful to remember the following expressions for trigonometric functions exp( jz) + exp(− jz) exp( jz) − exp(− jz) cos ( z ) = ; sin ( z ) = 2 2j resulting from Euler’s identity exp(± jz) = cos( z) ± j sin( z) © Amanogawa, 2006 – Digital Maestro Series 5 Transmission Lines Complex representation is very useful for time-harmonic functions of the form A cos ( ω t + φ ) = Re [ A exp ( jω t + jφ )] = Re [ A exp ( jφ ) exp ( jω t )] = Re [ A exp ( jω t )] The complex quantity A = A exp ( jφ ) contains all the information about amplitude and phase of the signal and is called the phasor of A cos ( ω t + φ ) If it is known that the signal is time-harmonic with frequency phasor completely characterizes its behavior. © Amanogawa, 2006 – Digital Maestro Series ω, the 6 Transmission Lines Often, a time-harmonic signal may be of the form: A sin ( ω t + φ ) and we have the following complex representation A sin ( ω t + φ ) = Re − jA ( cos ( ω t + φ ) + j sin ( ω t + φ ) ) = Re [ − jA exp ( jω t + jφ )] = Re [ A exp ( − jπ / 2 ) exp ( jφ ) exp ( jω t )] = Re A exp ( j ( φ − π / 2 ) ) exp ( jω t ) = Re [ A exp ( jω t )] with phasor A = A exp ( j ( φ − π / 2 ) ) This result is not surprising, since cos(ω t + φ − π / 2) = sin(ω t + φ) © Amanogawa, 2006 – Digital Maestro Series 7 Transmission Lines Time differentiation can be greatly simplified by the use of phasors. Consider for instance the signal V ( t ) = V0 cos ( ω t + φ ) with phasor V = V0 exp ( jφ ) The time derivative can be expressed as ∂ V ( t) = −ωV0 sin ( ω t + φ ) ∂t = Re { jωV0 exp ( jφ ) exp ( jω t )} ⇒ jωV0 exp ( jφ ) = jω V © Amanogawa, 2006 – Digital Maestro Series is the phasor of ∂ V ( t) ∂t 8 Transmission Lines With phasors, time-differential equations for time harmonic signals can be transformed into algebraic equations. Consider the simple circuit below, realized with lumped elements R L i (t) v (t) C This circuit is described by the integro-differential equation d i(t) 1 t v( t ) = L + R i + ∫ i( t ) dt dt C −∞ © Amanogawa, 2006 – Digital Maestro Series 9 Transmission Lines Upon time-differentiation we can eliminate the integral as d2 i( t) d v( t ) di 1 =L + R + i( t ) dt dt C dt 2 If we assume a time-harmonic excitation, we know that voltage and current should have the form v( t ) = V0 cos(ω t + αV ) phasor ⇒ V = V0 exp( jαV ) i( t ) = I0 cos(ω t + α I ) phasor ⇒ If I = I0 exp( jα I ) V0 and αV are given, ⇒ I0 and αI are the unknowns of the problem. © Amanogawa, 2006 – Digital Maestro Series 10 Transmission Lines The differential equation can be rewritten using phasors { } L Re −ω2 I exp ( jω t ) + R Re { jω I exp ( jω t )} 1 + Re { I exp ( jω t )} = Re { jωV exp ( jω t )} C Finally, the transform phasor equation is obtained as 1 V = R + jω L − j I=ZI ωC where Z = Impedance © Amanogawa, 2006 – Digital Maestro Series R Resistance + 1 j ωL − C ω Reactance 11 Transmission Lines The result for the phasor current is simply obtained as V V I= = = I0 exp ( jα I ) 1 Z R + jω L − j ω C which readily yields the unknowns I0 and αI . The time dependent current is then obtained from i( t ) = Re { I0 exp ( jα I ) exp ( jω t )} = I0 cos ( ω t + α I ) © Amanogawa, 2006 – Digital Maestro Series 12 Transmission Lines The phasor formalism provides a convenient way to solve timeharmonic problems in steady state, without having to solve directly a differential equation. The key to the success of phasors is that with the exponential representation one can immediately separate frequency and phase information. Direct solution of the timedependent differential equation is only necessary for transients. Integro-differential equations Transform Algebraic equations based on phasors I=? i(t)=? Direct Solution ( Transients ) i(t) © Amanogawa, 2006 – Digital Maestro Series Solution AntiTransform I 13 Transmission Lines The phasor representation of the circuit example above has introduced the concept of impedance. Note that the resistance is not explicitly a function of frequency. The reactance components are instead linear functions of frequency: Inductive component ⇒ proportional to ω Capacitive component ⇒ inversely proportional to ω Because of this frequency dependence, for specified values of L and C , one can always find a frequency at which the magnitudes of the inductive and capacitive terms are equal ωr L = 1 ωr C ⇒ ωr = 1 LC This is a resonance condition. The reactance cancels out and the impedance becomes purely resistive. © Amanogawa, 2006 – Digital Maestro Series 14 Transmission Lines The peak value of the current phasor is maximum at resonance I0 = | I0| V0 1 R + ωL − ω C 2 2 IM ωr © Amanogawa, 2006 – Digital Maestro Series ω 15 Transmission Lines Consider now the circuit below where an inductor and a capacitor are in parallel I L R C V The input impedance of the circuit is 1 Zin = R + + jω C jω L © Amanogawa, 2006 – Digital Maestro Series −1 = R+ jω L 1 − ω2 LC 16 Transmission Lines When ω=0 Zin = R 1 ω= LC ω→∞ Zin → ∞ Zin = R At the resonance condition ωr = 1 LC the part of the circuit containing the reactance components behaves like an open circuit, and no current can flow. The voltage at the terminals of the parallel circuit is the same as the input voltage V. © Amanogawa, 2006 – Digital Maestro Series 17 Transmission Lines Power in Circuits Consider the input impedance of a transmission line circuit, with an applied voltage v(t) inducing an input current i(t). i(t) v(t) Zin For sinusoidal excitation, we can write v (t ) = V0 cos(ω t + φ) φ ∈ [ −π/2 , π/2 ] i (t ) = I 0 cos(ω t ) where V0 and I0 are peak values and φ is the phase difference between voltage and current. Note that φ = 0 only when the input impedance is real (purely resistive). © Amanogawa, 2006 – Digital Maestro Series 18 Transmission Lines The time-dependent input power is given by P (t ) = v (t ) i (t ) = V0 I 0 cos(ω t + φ) cos(ω t ) V0 I 0 = [ cos (φ) + cos(2ω t + φ )] 2 The power has two (Fourier) components: (A) an average value V0 I 0 cos(φ) 2 (B) an oscillatory component with frequency 2f V0 I 0 cos(2ω t + φ) 2 © Amanogawa, 2006 – Digital Maestro Series 19 Transmission Lines The power flow changes periodically in time with an oscillation like (B) about the average value (A). Note that only when φ = 0 we have cos(φ) = 1, implying that for a resistive impedance the power is always positive (flowing from generator to load). When voltage and current are out of phase, the average value of the power has lower magnitude than the peak value of the oscillatory component. Therefore, during portions of the period of oscillation the power can be negative (flowing from load to generator). This means that when the power flow is positive, the reactive component of the input impedance stores energy, which is reflected back to the generator side when the power flow becomes negative. For an oscillatory excitation, we are interested in finding the behavior of the power during one full period, because from this we can easily obtain the average behavior in time. From the point of view of power consumption, we are also interested in knowing the power dissipated by the resistive component of the impedance. © Amanogawa, 2006 – Digital Maestro Series 20 Transmission Lines Using cos ( A + B ) = cos A cos B − sin A sin B one can write v (t ) = V0 cos(ω t + φ) = V0 cos φ cos ω t − V0 sin φ sin ω t in phase with current in quadrature with current This gives an alternative expression for power: P (t ) = V0 cos( ω t ) I 0 cos( ω t ) cos( φ ) − V0 cos( ω t ) I 0 sin( ω t ) sin( φ ) 2 = V0 I 0 cos( φ ) cos ( ω t ) − Real Power V0 I 0 sin( φ ) sin(2ω t ) 2 Reactive Power V0 I 0 V0 I 0 cos( φ ) + cos( φ ) cos(2ω t ) − = 2 2 V0 I 0 sin( φ ) sin(2ω t ) 2 Real Power Reactive Power © Amanogawa, 2006 – Digital Maestro Series 21 Transmission Lines The real power corresponds to the power dissipated by the resistive component of the impedance, and it is always positive. The reactive power corresponds to power stored and then reflected by the reactive component of the impedance. It oscillates from positive to negative during the period. Until now we have discussed properties of instantaneous power. Since we are considering time-harmonic periodic signals, it is very convenient to consider the time-average power 1 T 〈 P(t ) 〉 = ∫ P(t ) dt T 0 where T = 1 / f is the period of the oscillation. To determine the time-average power, we can use either the Fourier or the real/reactive power formulation. © Amanogawa, 2006 – Digital Maestro Series 22 Transmission Lines Fourier representation 1 T V0 I 0 1 T V0 I 0 cos(φ) dt + ∫ cos(2ω t − φ) dt 〈 P (t ) 〉 = ∫ 0 0 2 2 T T =0 V0 I 0 = cos(φ) 2 As one should expect, the time-average power flow is simply given by the Fourier component corresponding to the average of the original signal. © Amanogawa, 2006 – Digital Maestro Series 23 Transmission Lines Real/Reactive power representation 1 T T 〈 P (t ) 〉 = ( ∫ V0 I 0 cos(φ) dt + ∫ V0 I 0 cos(φ) cos(2ω t ) dt ) 0 2T 0 =0 1 T − V0 I 0 sin(φ) sin(2ω t ) dt ∫ 0 2T =0 = V0 I 0 cos(φ) 2 This result tells us that the time-average power flow is the average of the real power. The reactive power has zero time-average, since power is stored and completely reflected by the reactive component of the input impedance during the period of oscillation. © Amanogawa, 2006 – Digital Maestro Series 24 Transmission Lines The maximum of the reactive power is V0 I 0 V0 I 0 max{ Preac } = max{ sin ( φ ) sin ( 2ω t ) } = sin ( φ ) 2 2 Since the time-average of the reactive power is zero, we often use the maximum value above as an indication of the reactive power. The sign of the phase φ tells us about the imaginary part of the impedance or reactance: φ > 0 The reactance is inductive Current is lagging with respect to voltage Voltage is leading with respect to current φ < 0 The reactance is capacitive Voltage is lagging with respect to current Current is leading with respect to voltage © Amanogawa, 2006 – Digital Maestro Series 25 Transmission Lines If the total reactance is inductive V = Z I = R I + jω L I Im jωL I V Current lags φ>0 φ I © Amanogawa, 2006 – Digital Maestro Series RI Re 26 Transmission Lines If the total reactance is capacitive 1 V =ZI=RI− j I ωC Im I RI Re φ Voltage lags φ<0 V - j I /ω C © Amanogawa, 2006 – Digital Maestro Series 27 Transmission Lines In many situations, we may use the root-mean-square (r.m.s.) values of quantities, instead of the peak value. For a given signal v(t ) = V0 cos(ω t ) the r.m.s. value is defined as 1 T 2 1 2π 2 2 Vrms = V0 cos ( ω t ) dt = V0 cos ( ω t ) d ω t ∫ ∫ T 0 ωT 0 1 2π 2 1 = V0 V0 cos ( ϑ ) d ϑ = ∫ 2π 0 2 This result is valid for sinusoidal signals. Any given signal shape corresponds to a specific coefficient ( peak factor = V0 / Vrms ) that allows one to convert directly from peak value to r.m.s. value. © Amanogawa, 2006 – Digital Maestro Series 28 Transmission Lines The peak factor for sinusoidal signals is V0 = Vrms 2 ≈ 1.4142 For a symmetric triangular signal the peak factor is V0 = 3 ≈ 1.732 Vrms V0 t For a symmetric square signal the peak factor is simply V0 = 1 Vrms © Amanogawa, 2006 – Digital Maestro Series V0 t 29 Transmission Lines For a non-sinusoidal periodic signal, we can use a decomposition into orthogonal Fourier components to obtain the r.m.s. value: V (t ) = Vav + V1 (t ) + V2 (t ) + V3 (t ) … = [Vrms ] = 1 2 T ∫ T 0 = [∑ 1 T ∫ T 0 2 V (t ) dt = 2 (t )] V k k dt + 1 1 T ∫ T 0 ∑ k Vk ( t ) [ ∑ k Vk (t )] dt = 2 T [ ∑ i ≠ j 2 Vi (t ) V j (t )] dt = ∫ 0 T orthogonal 1 T 2 1 T 2 ( ) [ 2 ( ) ( ) ] = ( V ) = ∑k + V t dt V t V t dt ∑ k krms j T ∫0 k ∑ i ≠ j T ∫0 i ⇒ Vrms = ∑ k (Vkrms )2 = 2 Vav + (V1 rms 2 ) + (V2 2 rms ) +… The final result holds for any decomposition into orthogonal functions and it is known in mathemics as Parseval’s identity. © Amanogawa, 2006 – Digital Maestro Series 30 Transmission Lines In terms of r.m.s. values, the time-average power for a sinusoidal signal is then V0 I 0 〈 P (t ) 〉 = cos(φ) = Vrms I rms cos(φ) 2 2 Finally, we can relate the time-average power to the phasors of voltage and current. Since v (t ) = V0 cos(ω t ) = Re{ V0 exp( jω t ) } i (t ) = I 0 cos(ω t − φ) = Re{ I 0 exp(− jφ) exp( jω t ) } we have phasors V = V0 I = I 0 exp(− jφ) © Amanogawa, 2006 – Digital Maestro Series 31 Transmission Lines The time-average power in terms of phasors is given by 1 1 * 〈 P (t ) 〉 = Re{ V I } = Re{ V0 I 0 exp( jφ) } 2 2 V0 I 0 = cos(φ) 2 Note that one must always use the complex conjugate of the phasor current to obtain the time-average power. It is important to remember this when voltage and current are expressed as functions of each other. Only when the impedance is purely resistive, I = I* = I0 since φ = 0. Also, note that the time-average power is always a real positive quantity and that it is not the phasor of the time-dependent power. It is a common mistake to think so. © Amanogawa, 2006 – Digital Maestro Series 32 Transmission Lines Now we consider power flow including explicitly the generator, to understand in which conditions maximum power transfer to a load can take place. ZR Vin = VG ZG + Z R 1 Iin = VG ZG + Z R 1 * 〈 Pin 〉 = Re{ Vin Iin } 2 © Amanogawa, 2006 – Digital Maestro Series Iin ZG VG ZR Vin Generator Load 33 Transmission Lines As a first case, we examine resistive impedances ZG = RG Z R = RR Voltage and current are in phase at the input. The time-average power dissipated by the load is 1 RR 1 * 〈 P (t ) 〉 = VG VG 2 RG + RR RG + RR 1 RR 2 = VG 2 ( RG + RR )2 © Amanogawa, 2006 – Digital Maestro Series 34 Transmission Lines To find the load resistance that maximizes power transfer to the load for a given generator we impose d 〈 P (t )〉 =0 d RR from which we obtain d RR =0 2 dRR ( RG + RR ) ( RG + RR )2 − 2 RR ( RG + RR ) ( RG + RR ) 4 ( RG + RR ) − 2 RR = 0 ⇒ =0 RR = RG We conclude that for maximum power transfer the load resistance must be identical to the generator resistance. © Amanogawa, 2006 – Digital Maestro Series 35 Transmission Lines Let’s consider now complex impedances Z R = RR + jX R ZG = RG + jX G For maximum power transfer, generator and load impedances must be complex conjugate of each other: * Z R = ZG ⇒ RR = RG X R = − XG RG VG jXG −jXG RG This can be easily understood by considering that, to maximize the active power supplied to the load, voltage and current of the generator should remain in phase. If the reactances of generator and load are opposite and cancel each other along the path of the current, the generator will only see a resistance. Voltage and current will be in phase with maximum power delivered to the load. © Amanogawa, 2006 – Digital Maestro Series 36 Transmission Lines The total time-average power supplied by the generator in conditions of maximum power transfer is 1 1 1 2 1 2 1 * 〈 Ptot 〉 = Re{ VG Iin } = VG = VG RR 2 2 2 RR 4 The time-average power supplied to the load is * 1 1 ZR 1 * * VG 〈 Pin 〉 = Re{ Vin Iin } = Re VG 2 2 ZG + Z R Z Z + G R RR + jX R 1 1 2 2 1 = VG Re = VG 2 2 8 RR 4 RR © Amanogawa, 2006 – Digital Maestro Series 37 Transmission Lines The power dissipated by the internal generator impedance is 1 * } 〈 PG 〉 = Re { (VG − Vin ) Iin 2 1 1 1 2 1 2 1 2 1 = VG − VG = VG 4 8 8 RR RR RR We conclude that, in conditions of maximum power transfer, only half of the total active power supplied by the generator is actually used by the load. The generator impedance dissipates the remaining half of the available active power. This may seem a disappointing result, but it is the best one can do for a real generator with a given internal impedance! © Amanogawa, 2006 – Digital Maestro Series 38 Transmission Lines Transmission Line Equations A typical engineering problem involves the transmission of a signal from a generator to a load. A transmission line is the part of the circuit that provides the direct link between generator and load. Transmission lines can be realized in a number of ways. Common examples are the parallel-wire line and the coaxial cable. For simplicity, we use in most diagrams the parallel-wire line to represent circuit connections, but the theory applies to all types of transmission lines. Load Generator ZG VG © Amanogawa, 2006 – Digital Maestro Series Transmission line ZR 39 Transmission Lines Examples of transmission lines d d D d Two-wire line D Coaxial cable w t h Microstrip © Amanogawa, 2006 – Digital Maestro Series 40 Transmission Lines If you are only familiar with low frequency circuits, you are used to treat all lines connecting the various circuit elements as perfect wires, with no voltage drop and no impedance associated to them (lumped impedance circuits). This is a reasonable procedure as long as the length of the wires is much smaller than the wavelength of the signal. At any given time, the measured voltage and current are the same for each location on the same wire. Load Generator ZG VG ZR VR = VG ZG + ZR © Amanogawa, 2006 – Digital Maestro Series VR VR VR ZR L << λ 41 Transmission Lines Let’s look at some examples. The electricity supplied to households consists of high power sinusoidal signals, with frequency of 60Hz or 50Hz, depending on the country. Assuming that the insulator between wires is air (ε ≈ ε0), the wavelength for 60Hz is: c 2.999 × 10 8 λ= = ≈ 5.0 × 106 m = 5, 000 km f 60 which is the about the distance between S. Francisco and Boston! Let’s compare to a frequency in the microwave range, for instance 60 GHz. The wavelength is given by c 2.999 × 10 8 λ= = ≈ 5.0 × 10 −3 m = 5.0 mm f 60 × 109 which is comparable to the size of a microprocessor chip. Which conclusions do you draw? © Amanogawa, 2006 – Digital Maestro Series 42 Transmission Lines For sufficiently high frequencies the wavelength is comparable with the length of conductors in a transmission line. The signal propagates as a wave of voltage and current along the line, because it cannot change instantaneously at all locations. Therefore, we cannot neglect the impedance properties of the wires (distributed impedance circuits). V (z) = V + e− jβ z + V − e jβ z Load Generator ZG VG V( 0 ) V( z ) V( L ) ZR L © Amanogawa, 2006 – Digital Maestro Series 43 Transmission Lines Note that the equivalent circuit of a generator consists of an ideal alternating voltage generator in series with its actual internal impedance. When the generator is open ( ZR → ∞ ) we have: Iin = 0 and Vin = VG If the generator is connected to a load ZR VG Iin = ( ZG + ZR ) VG ZR Vin = ( ZG + ZR ) If the load is a short ( ZR VG Iin = ZG ZG VG Vin ZR = 0) and Vin = 0 © Amanogawa, 2006 – Digital Maestro Series Iin Generator Load 44 Transmission Lines The simplest circuit problem that we can study consists of a voltage generator connected to a load through a uniform transmission line. In general, the impedance seen by the generator is not the same as the impedance of the load, because of the presence of the transmission line, except for some very particular cases: Zin = ZR Zin Transmission line only if λ L=n 2 [ n = integer ] ZR L Our first goal is to determine the equivalent impedance seen by the generator, that is, the input impedance of a line terminated by the load. Once that is known, standard circuit theory can be used. © Amanogawa, 2006 – Digital Maestro Series 45 Transmission Lines Generator Load ZG Transmission line VG Generator VG © Amanogawa, 2006 – Digital Maestro Series ZR Equivalent Load ZG Zin 46 Transmission Lines A uniform transmission line is a “distributed circuit” that we can describe as a cascade of identical cells with infinitesimal length. The conductors used to realize the line possess a certain series inductance and resistance. In addition, there is a shunt capacitance between the conductors, and even a shunt conductance if the medium insulating the wires is not perfect. We use the concept of shunt conductance, rather than resistance, because it is more convenient for adding the parallel elements of the shunt. We can represent the uniform transmission line with the distributed circuit below (general lossy line) L dz R dz C dz dz © Amanogawa, 2006 – Digital Maestro Series L dz G dz R dz C dz G dz dz 47 Transmission Lines The impedance parameters L, R, C, and G represent: L = series inductance per unit length R = series resistance per unit length C = shunt capacitance per unit length G = shunt conductance per unit length. Each cell of the distributed circuit will have impedance elements with values: Ldz, Rdz, Cdz, and Gdz, where dz is the infinitesimal length of the cells. If we can determine the differential behavior of an elementary cell of the distributed circuit, in terms of voltage and current, we can find a global differential equation that describes the entire transmission line. We can do so, because we assume the line to be uniform along its length. So, all we need to do is to study how voltage and current vary in a single elementary cell of the distributed circuit. © Amanogawa, 2006 – Digital Maestro Series 48 Transmission Lines Loss-less Transmission Line In many cases, it is possible to neglect resistive effects in the line. In this approximation there is no Joule effect loss because only reactive elements are present. The equivalent circuit for the elementary cell of a loss-less transmission line is shown in the figure below. L dz I (z)+dI I (z) V (z) V (z)+dV C dz dz © Amanogawa, 2006 – Digital Maestro Series 49 Transmission Lines The series inductance determines the variation of the voltage from input to output of the cell, according to the sub-circuit below L dz I (z) V (z) V (z)+dV dz The corresponding circuit equation is (V + dV ) − V = − jω L dz I which gives a first order differential equation for the voltage dV = − jω L I dz © Amanogawa, 2006 – Digital Maestro Series 50 Transmission Lines The current flowing through the shunt capacitance determines the variation of the current from input to output of the cell. I (z) dI I (z)+dI C dz V (z)+dV The circuit equation for the sub-circuit above is dI = − jω Cdz(V + dV ) = − jω CVdz − jω C dV dz The second term (including dV dz) tends to zero very rapidly in the limit of infinitesimal length dz leaving a first order differential equation for the current dI = − jω C V dz © Amanogawa, 2006 – Digital Maestro Series 51 Transmission Lines We have obtained a system of two coupled first order differential equations that describe the behavior of voltage and current on the uniform loss-less transmission line. The equations must be solved simultaneously. dV = − jω L I dz dI = − jω C V dz These are often called “telegraphers’ equations” of the loss-less transmission line. © Amanogawa, 2006 – Digital Maestro Series 52 Transmission Lines One can easily obtain a set of uncoupled equations by differentiating with respect to the space coordinate. The first order differential terms are eliminated by using the corresponding telegraphers’ equation dI = − jω C V dz d 2V dI = − jω L = jω L jω CV = − ω2 LC V dz dz2 d2 I dV = − jω C = jω C jω L I = − ω2 LC I dz dz2 dV = − jω L I dz These are often called “telephonists’ equations”. © Amanogawa, 2006 – Digital Maestro Series 53 Transmission Lines We have now two uncoupled second order differential equations for voltage and current, which give an equivalent description of the loss-less transmission line. Mathematically, these are wave equations and can be solved independently. The general solution for the voltage equation is V (z) = V + e− jβ z + V − e jβ z where the wave propagation constant is β = ω LC Note that the complex exponential terms including β have unitary magnitude and purely “imaginary” argument, therefore they only affect the “phase” of the wave in space. © Amanogawa, 2006 – Digital Maestro Series 54 Transmission Lines We have the following useful relations: 2π 2π f ω β= = = λ vp vp ω εr µr = = ω ε0 µ0 ε r µ r = ω ε µ c Here, λ = v p f is the wavelength of the dielectric medium surrounding the conductors of the transmission line and 1 1 vp = = ε0 ε r µ0 µ r εµ is the phase velocity of an electromagnetic wave in the dielectric. As you can see, the propagation constant β can be written in many different, equivalent ways. © Amanogawa, 2006 – Digital Maestro Series 55 Transmission Lines The current distribution on the transmission line can be readily obtained by differentiation of the result for the voltage dV = − jβ V + e− jβ z + jβ V − e jβ z = − jω L I dz which gives ( ) ( C + − jβ z 1 − jβ z −V e = I (z) = V e V + e− jβ z − V − e jβ z L Z0 ) The real quantity Z0 = L C is the “characteristic impedance” of the loss-less transmission line. © Amanogawa, 2006 – Digital Maestro Series 56 Transmission Lines Lossy Transmission Line The solution for a uniform lossy transmission line can be obtained with a very similar procedure, using the equivalent circuit for the elementary cell shown in the figure below. L dz R dz I (z)+dI I (z) V (z) C dz G dz V (z)+dV dz © Amanogawa, 2006 – Digital Maestro Series 57 Transmission Lines The series impedance determines the variation of the voltage from input to output of the cell, according to the sub-circuit L dz V (z) R dz I (z) V (z)+dV dz The corresponding circuit equation is (V + dV ) − V = − ( jω Ldz + Rdz) I from which we obtain a first order differential equation for the voltage dV = − ( jω L + R) I dz © Amanogawa, 2006 – Digital Maestro Series 58 Transmission Lines The current flowing through the shunt admittance determines the input-output variation of the current, according to the sub-circuit I (z) I (z)+dI dI C dz G dz V (z)+dV The corresponding circuit equation is dI = − ( jω Cdz + Gdz)(V + dV ) = − ( jω C + G)Vdz − ( jω C + G)dV dz The second term (including dV dz) can be ignored, giving a first order differential equation for the current dI = − ( jω C + G)V dz © Amanogawa, 2006 – Digital Maestro Series 59 Transmission Lines We have again a system of coupled first order differential equations that describe the behavior of voltage and current on the lossy transmission line dV = − ( jω L + R) I dz dI = − ( jω C + G)V dz These are the “telegraphers’ equations” for the lossy transmission line case. © Amanogawa, 2006 – Digital Maestro Series 60 Transmission Lines One can easily obtain a set of uncoupled equations differentiating with respect to the coordinate z as done earlier by dI = − ( jω C + G)V dz d 2V dI = − ( jω L + R) = ( jω L + R)( jω C + G)V 2 dz dz d2 I dV = − ( jω C + G) = ( jω C + G)( jω L + R) I dz dz2 dV = − ( jω L + R) I dz These are the “telephonists’ equations” for the lossy line. © Amanogawa, 2006 – Digital Maestro Series 61 Transmission Lines The telephonists’ equations for the lossy transmission line are uncoupled second order differential equations and are again wave equations. The general solution for the voltage equation is V (z) = V + e−γ z + V − eγ z = V + e−α z e− jβ z + V − eα z e jβ z where the wave propagation constant is now the complex quantity γ = ( jω L + R)( jω C + G) = α + jβ The real part α of the propagation constant γ describes the attenuation of the signal due to resistive losses. The imaginary part β describes the propagation properties of the signal waves as in loss-less lines. The exponential terms including α are “real”, therefore, they only affect the “magnitude” of the voltage phasor. The exponential terms including β have unitary magnitude and purely “imaginary” argument, affecting only the “phase” of the waves in space. © Amanogawa, 2006 – Digital Maestro Series 62 Transmission Lines The current distribution on a lossy transmission line can be readily obtained by differentiation of the result for the voltage dV = − ( jω L + R) I = −γ V + e−γ z + γ V − eγ z dz which gives ( jω C + G ) I (z) = (V + e−γz − V − e γ z ) ( jω L + R ) 1 (V + e−γz − V − e γ z ) = Z0 with the “characteristic impedance” of the lossy transmission line ( jω L + R) Z0 = ( jω C + G) © Amanogawa, 2006 – Digital Maestro Series Note: the characteristic impedance is now complex ! 63 Transmission Lines For both loss-less and lossy transmission lines the characteristic impedance does not depend on the line length but only on the metal of the conductors, the dielectric material surrounding the conductors and the geometry of the line crosssection, which determine L, R, C, and G. One must be careful not to interpret the characteristic impedance as some lumped impedance that can replace the transmission line in an equivalent circuit. This is a very common mistake! Z0 © Amanogawa, 2006 – Digital Maestro Series ZR Z0 ZR 64 Transmission Lines We have obtained the following solutions for the steady-state voltage and current phasors in a transmission line: Loss-less line + − jβ z − jβ z V (z) = V e Lossy line + −γz − γz +V e ( 1 I (z) = V + e − jβ z − V − e jβ z Z0 V (z) = V e ) ( +V e 1 I (z) = V + e− γ z − V −e γ z Z0 ) Since V (z) and I (z) are the solutions of second order differential + − (wave) equations, we must determine two unknowns, V and V , which represent the amplitudes of steady-state voltage waves, travelling in the positive and in the negative direction, respectively. Therefore, we need two boundary conditions to determine these unknowns, by considering the effect of the load and of the generator connected to the transmission line. © Amanogawa, 2006 – Digital Maestro Series 65 Transmission Lines Before we consider the boundary conditions, it is very convenient to shift the reference of the space coordinate so that the zero reference is at the location of the load instead of the generator. Since the analysis of the transmission line normally starts from the load itself, this will simplify considerably the problem later. ZR New Space Coordinate z d 0 We will also change the positive direction of the space coordinate, so that it increases when moving from load to generator along the transmission line. © Amanogawa, 2006 – Digital Maestro Series 66 Transmission Lines We adopt a new coordinate d = − z, with zero reference at the load location. The new equations for voltage and current along the lossy transmission line are Loss-less line + jβd − − jβ d V (d) = V e I (d) = 1 Z0 ( Lossy line + γd − −γ d +V e + jβd V e − − jβ d −V e V (d) = V e ) I (d) = 1 Z0 ( +V e + γd V e − −γd −V e ) At the load (d = 0) we have, for both cases, V (0) = V + + V − ( 1 I (0) = V + −V − Z0 © Amanogawa, 2006 – Digital Maestro Series ) 67 Transmission Lines For a given load impedance ZR , the load boundary condition is V (0) = Z R I (0) Therefore, we have ( ZR V +V = V + −V − Z0 + − ) from which we obtain the voltage load reflection coefficient V− Z R − Z0 ΓR = + = Z R + Z0 V © Amanogawa, 2006 – Digital Maestro Series 68 Transmission Lines We can introduce this result into the transmission line equations as Loss-less line Lossy line + jβ d V (d) = V e + jβ d I (d) = V e Z0 ( 1+ ΓR e ( 1 − Γ Re −2 jβ d −2 jβ d ) ) + γd V (d) = V e + γd I (d) = V e Z0 ( 1+ ΓR e ( 1 − Γ Re −2 γ d −2 γ d ) ) At each line location we define a Generalized Reflection Coefficient Γ (d) = Γ R e−2 jβ d Γ (d) = Γ R e−2 γ d and the line equations become V (d) = V + e jβ d (1 + Γ (d) ) V (d) = V + e γ d (1 + Γ (d) ) V + e jβ d I (d) = (1 − Γ (d) ) Z0 V +e γd I (d) = (1 − Γ (d) ) Z0 © Amanogawa, 2006 – Digital Maestro Series 69 Transmission Lines We define the line impedance as V (d) 1+ Γ (d) Z (d) = = I (d) 1 − Γ (d) A simple circuit diagram can illustrate the significance of line impedance and generalized reflection coefficient: ΓReq = Γ(d) d © Amanogawa, 2006 – Digital Maestro Series ZR Zeq=Z(d) 0 70 Transmission Lines If you imagine to cut the line at location d, the input impedance of the portion of line terminated by the load is the same as the line impedance at that location “before the cut”. The behavior of the line on the left of location d is the same if an equivalent impedance with value Z(d) replaces the cut out portion. The reflection coefficient of the new load is equal to Γ(d) Γ Req = Γ ( d ) = Z Req − Z 0 Z Req + Z 0 If the total length of the line is L, the input impedance is obtained from the formula for the line impedance as Vin V ( L ) 1 + Γ ( L ) Zin = = = Iin I ( L ) 1 − Γ ( L ) The input impedance is the equivalent impedance representing the entire line terminated by the load. © Amanogawa, 2006 – Digital Maestro Series 71 Transmission Lines An important practical case is the low-loss transmission line, where the reactive elements still dominate but R and G cannot be neglected as in a loss-less line. We have the following conditions: ω L >> R ω C >> G so that γ = ( jω L + R )( jω C + G ) = R G jω L jω C 1 + 1+ j ω L j ω C R G RG ≈ jω LC 1 + + − jω L jω C ω2 LC The last term under the square root can be neglected, because it is the product of two very small quantities. © Amanogawa, 2006 – Digital Maestro Series 72 Transmission Lines What remains of the square root can be expanded into a truncated Taylor series 1 R G + γ ≈ jω LC 1 + 2 j L j C ω ω 1 C L = R +G + jω LC 2 L C so that 1 C L α = R +G L C 2 © Amanogawa, 2006 – Digital Maestro Series β = ω LC 73 Transmission Lines The characteristic impedance of the low-loss line is a real quantity for all practical purposes and it is approximately the same as in a corresponding loss-less line R + jω L L Z0 = ≈ G + jωC C and the phase velocity associated to the wave propagation is ω vp = ≈ β 1 LC BUT NOTE: In the case of the low-loss line, the equations for voltage and current retain the same form obtained for general lossy lines. © Amanogawa, 2006 – Digital Maestro Series 74 Transmission Lines Again, we obtain the loss-less transmission line if we assume R=0 G=0 This is often acceptable in relatively short transmission lines, where the overall attenuation is small. As shown earlier, the characteristic impedance in a loss-less line is exactly real L Z0 = C while the propagation constant has no attenuation term γ = ( jω L)( jω C ) = jω LC = jβ The loss-less line does not dissipate power, because α = 0. © Amanogawa, 2006 – Digital Maestro Series 75 Transmission Lines For all cases, the line impedance was defined as V (d) 1 + Γ (d) Z (d) = = Z0 I (d) 1 − Γ (d) By including the appropriate generalized reflection coefficient, we can derive alternative expressions of the line impedance: A) Loss-less line 1 + Γ R e−2 jβd Z R + jZ 0 tan(β d) Z (d) = Z 0 = Z0 −2 jβd jZ R tan(β d) + Z 0 1 − Γ Re B) Lossy line (including low-loss) 1 + Γ R e −2 γ d Z R + Z 0 tanh( γ d) Z (d) = Z 0 = Z0 −2 γ d Z R tanh( γ d) + Z 0 1 − Γ Re © Amanogawa, 2006 – Digital Maestro Series 76 Transmission Lines Let’s now consider power flow in a transmission line, limiting the discussion to the time-average power, which accounts for the active power dissipated by the resistive elements in the circuit. The time-average power at any transmission line location is { 1 〈 P(d , t ) 〉 = Re V (d) I * (d) 2 } This quantity indicates the time-average power that flows through the line cross-section at location d. In other words, this is the power that, given a certain input, is able to reach location d and then flows into the remaining portion of the line beyond this point. It is a common mistake to think that the quantity above is the power dissipated at location d ! © Amanogawa, 2006 – Digital Maestro Series 77 Transmission Lines The generator, the input impedance, the input voltage and the input current determine the power injected at the transmission line input. Iin ZG VG Vin Generator © Amanogawa, 2006 – Digital Maestro Series Zin Line Zin Vin = VG ZG + Zin 1 Iin = VG ZG + Zin 1 * 〈 Pin 〉 = Re Vin Iin 2 { } 78 Transmission Lines The time-average power reaching the load of the transmission line is given by the general expression { } 1 〈 P(d=0 , t ) 〉 = Re V (0) I * (0) 2 + 1 1 = Re V (1 + Γ R ) * V + (1 − Γ R ) 2 Z0 ( * ) This represents the power dissipated by the load. The time-average power absorbed by the line is simply the difference between the input power and the power absorbed by the load 〈 P line 〉 = 〈 P in 〉 − 〈 P (d = 0 , t ) 〉 In a loss-less transmission line no power is absorbed by the line, so the input time-average power is the same as the time-average power absorbed by the load. Remember that the internal impedance of the generator dissipates part of the total power generated. © Amanogawa, 2006 – Digital Maestro Series 79 Transmission Lines It is instructive to develop further the general expression for the time-average power at the load, using Z0=R0+jX0 for the characteristic impedance, so that R0 + jX 0 R0 + jX 0 = * = = 2 * 2 Z0 Z0 Z0 R0 + X 02 Z0 1 Z0 Alternatively, one may simplify the analysis by introducing the line characteristic admittance 1 Y0 = = G0 + jB0 Z0 It may be more convenient to deal with the complex admittance at the numerator of the power expression, rather than the complex characteristic impedance at the denominator. © Amanogawa, 2006 – Digital Maestro Series 80 Transmission Lines * 0 〈 P(d=0, t ) 〉 = 1 Re V + 1+Γ R 1 V + 1−Γ Z 2 = V+ 2 Z0 = V+ 2 Z0 = V+ 2 Z0 = V+ 2 Z0 2 2 Re R0 + jX 2 0 R + jX Re 0 0 2 2 1+ Re Γ + j Im Γ R 2 2 2 2 R0 − R0 Γ R − 2 X 0 Im Γ © Amanogawa, 2006 – Digital Maestro Series R = 1− Re Γ 2 2 1− Re Γ + Im Γ + R R Re + + 2Im R jX j 1− Γ Γ R 0 0 2 * R R R + j Im Γ R j 2Im Γ R = R 81 Transmission Lines Equivalently, using the complex characteristic admittance: * 0 〈 P(d=0, t ) 〉 = 1 Re V + 1+Γ R Y V + 1−Γ 2 = = = = V+ 2 2 2 2 1+ Re Γ + j Im Γ R 2 R = 1− Re Γ 2 2 1− Re Γ R + Im Γ R + 2 Re G0 − jB0 1− Γ R + j 2Im Γ 2 V+ Re G0 − jB0 2 V+ 0 Re G0 − jB 2 V+ * R 2 G0 − G0 Γ R + 2B0 Im Γ © Amanogawa, 2006 – Digital Maestro Series R R + j Im Γ R j 2Im Γ R = R 82 Transmission Lines The time-average power, injected into the input of the transmission line, is maximized when the input impedance of the transmission line and the internal generator impedance are complex conjugate of each other. Zin Generator Load ZG VG ZG = Z*in © Amanogawa, 2006 – Digital Maestro Series Transmission line ZR for maximum power transfer 83 Transmission Lines The characteristic impedance of the loss-less line is real and we can express the power flow, anywhere on the line, as 1 〈 P (d , t ) 〉 = Re{ V (d) I * (d) } 2 1 + jβ d 1 + Γ R e− j 2β d = Re V e 2 ( ) ( 1 (V + )* e− jβ d 1 − Γ R e− j 2β d Z0 1 +2 = V 2Z0 Incident wave ) * 1 +2 − V ΓR 2 2Z0 Reflected wave This result is valid for any location, including the input and the load, since the transmission line does not absorb any power. © Amanogawa, 2006 – Digital Maestro Series 84 Transmission Lines In the case of low-loss lines, the characteristic impedance is again real, but the time-average power flow is position dependent because the line absorbs power. { { } ( 1 〈 P (d , t ) 〉 = Re V (d) I * (d) 2 1 = Re V + eα d e jβ d 1 + Γ R e−2 γ d 2 1 (V + )* eα d e− jβ d 1 − Γ R e−2 γ d Z0 ) ( ) * 1 1 + 2 2α d + 2 −2α d = V e − V e ΓR 2 2Z0 2Z0 Incident wave © Amanogawa, 2006 – Digital Maestro Series Reflected wave 85 Transmission Lines Note that in a lossy line the reference for the amplitude of the incident voltage wave is at the load and that the amplitude grows exponentially moving towards the input. The amplitude of the incident wave behaves in the following way V + eα L ⇔ input V + eα d inside the line ⇔ V+ load The reflected voltage wave has maximum amplitude at the load, and it decays exponentially moving back towards the generator. The amplitude of the reflected wave behaves in the following way V + Γ R e−α L ⇔ input © Amanogawa, 2006 – Digital Maestro Series V + Γ R e−α d ⇔ inside the line V +ΓR load 86 Transmission Lines For a general lossy line the power flow is again position dependent. Since the characteristic impedance is complex, the result has an additional term involving the imaginary part of the characteristic admittance, B0, as { { } 1 〈 P (d , t ) 〉 = Re V (d) I * (d) 2 1 = Re V + eα d e jβ d (1 + Γ (d) ) 2 } Y0* (V + )* eα d e− jβ d (1 − Γ (d) )* G0 + 2 2α d G0 + 2 −2α d = V e − V e ΓR 2 2 2 + B0 V © Amanogawa, 2006 – Digital Maestro Series +2 e2α d Im(Γ (d)) 87 Transmission Lines For the general lossy line, keep in mind that Z 0* R0 − jX 0 R0 − jX 0 1 Y0 = = = = 2 = G0 + jB0 * 2 2 Z0 Z0 Z0 R0 + X 0 Z0 −X0 R0 = G0 = 2 B 0 R0 + X 02 R02 + X 02 Recall that for a low-loss transmission line the characteristic impedance is approximately real, so that B0 ≈ 0 and Z 0 ≈ 1 G0 ≈ R0 . The previous result for the low-loss line can be readily recovered from the time-average power for the general lossy line. © Amanogawa, 2006 – Digital Maestro Series 88 Transmission Lines To completely specify the transmission line problem, we still have to determine the value of V+ from the input boundary condition. ¾ The load boundary condition imposes the shape of the interference pattern of voltage and current along the line. ¾ The input boundary condition, linked to the generator, imposes the scaling for the interference patterns. We have Zin Vin = V (L) = VG ZG + Zin or with 1 + Γ (L) Zin = Z 0 1 − Γ (L) Z R + jZ 0 tan(β L) Zin = Z 0 jZ R tan(β L) + Z 0 loss - less line Z R + Z 0 tanh( γ L) Zin = Z 0 Z R tanh( γ L) + Z 0 lossy line © Amanogawa, 2006 – Digital Maestro Series 89 Transmission Lines For a loss-less transmission line: V (L) = V + e jβ L [1 + Γ (L) ] = V + e jβ L (1 + Γ R e− j 2β L ) ⇒ Zin 1 V = VG ZG + Zin e jβ L (1 + Γ R e− j 2β L ) + For a lossy transmission line: ( V (L) = V + e γ L [1 + Γ (L) ] = V + e γ L 1 + Γ R e−2 γ d ⇒ ) Zin 1 V = VG ZG + Zin e γ L (1 + Γ R e−2 γ L ) + © Amanogawa, 2006 – Digital Maestro Series 90 Transmission Lines In order to have good control on the behavior of a high frequency circuit, it is very important to realize transmission lines as uniform as possible along their length, so that the impedance behavior of the line does not vary and can be easily characterized. A change in transmission line properties, wanted or unwanted, entails a change in the characteristic impedance, which causes a reflection. Example: Z01 Z02 ZR Γ1 Z01 © Amanogawa, 2006 – Digital Maestro Series Zin Zin − Z01 Γ1 = Zin + Z01 91 Transmission Lines Special Cases ZR → 0 (SHORT CIRCUIT) ZR = 0 Z0 The load boundary condition due to the short circuit is V (0) = 0 ⇒ V (d = 0) = V + e jβ 0 (1 + Γ R e− j 2β 0 ) = V + (1 + Γ R ) = 0 ⇒ © Amanogawa, 2006 – Digital Maestro Series Γ R = −1 92 Transmission Lines Since ΓR = V− V+ ⇒ V − = −V + We can write the line voltage phasor as V (d) = V + e jβ d + V − e− jβ d = V + e jβ d − V + e− jβ d = V + ( e jβ d − e− jβ d ) = 2 jV + sin(β d) © Amanogawa, 2006 – Digital Maestro Series 93 Transmission Lines For the line current phasor we have 1 (V + e jβ d − V − e− jβ d ) I (d) = Z0 1 (V + e jβ d + V + e− jβ d ) = Z0 V + jβ d (e = + e− jβ d ) Z0 2V + cos(β d) = Z0 The line impedance is given by V (d) 2 jV + sin(β d) = = jZ0 tan(β d) Z(d) = + I (d) 2V cos(β d) / Z0 © Amanogawa, 2006 – Digital Maestro Series 94 Transmission Lines The time-dependent values of voltage and current are obtained as V (d, t ) = Re[V (d) e jω t ] = Re[2 j | V + | e jθ sin(β d) e jω t ] = 2| V + |sin(β d) ⋅ Re[ j e j (ω t +θ) ] = 2| V + |sin(β d) ⋅ Re[ j cos(ω t + θ) − sin(ω t + θ)] = −2| V + |sin(β d) sin(ω t + θ) I (d, t ) = Re[ I (d) e jω t ] = Re[2 | V + | e jθ cos(β d) e jω t ] / Z0 = 2| V + |cos(β d) ⋅ Re[ e j (ω t +θ) ] / Z0 = 2| V + |cos(β d) ⋅ Re[ (cos(ω t + θ) + j sin(ω t + θ)] / Z0 | V+ | =2 cos(β d) cos(ω t + θ) Z0 © Amanogawa, 2006 – Digital Maestro Series 95 Transmission Lines The time-dependent power is given by P(d, t ) = V (d, t ) ⋅ I (d, t ) | V + |2 =−4 sin(β d) cos(β d) sin(ω t + θ) cos(ω t + θ) Z0 | V + |2 =− sin(2β d) sin (2ω t + 2θ) Z0 and the corresponding time-average power is 1 T < P(d, t ) > = ∫ P(d, t ) dt T 0 | V + |2 1 T =− sin(2β d) ∫ sin (2ω t + 2θ) = 0 Z0 T 0 © Amanogawa, 2006 – Digital Maestro Series 96 Transmission Lines ZR → ∞ (OPEN CIRCUIT) ZR → ∞ Z0 The load boundary condition due to the open circuit is I (0) = 0 V + jβ 0 ⇒ I (d = 0) = e (1 − Γ R e− j 2β 0 ) Z0 V+ = (1 − Γ R ) = 0 Z0 ⇒ © Amanogawa, 2006 – Digital Maestro Series ΓR = 1 97 Transmission Lines Since ΓR = V− V+ ⇒ V− = V+ We can write the line current phasor as 1 I (d) = (V + e jβ d − V − e− jβ d ) Z0 1 = (V + e jβ d − V + e− jβ d ) Z0 V + jβ d − jβ d 2 jV + = −e (e )= sin(β d) Z0 Z0 © Amanogawa, 2006 – Digital Maestro Series 98 Transmission Lines For the line voltage phasor we have V (d) = (V + e jβ d + V − e− jβ d ) = (V + e jβ d + V + e− jβ d ) = V + ( e jβ d + e− jβ d ) = 2V + cos(β d) The line impedance is given by V (d) 2V + cos(β d) Z0 Z(d) = = =−j I (d) 2 jV + sin(β d) / Z0 tan(β d) © Amanogawa, 2006 – Digital Maestro Series 99 Transmission Lines The time-dependent values of voltage and current are obtained as V (d, t ) = Re[V (d) e jω t ] = Re[2 | V + | e jθ cos(β d) e jω t ] = 2| V + |cos(β d) ⋅ Re[ e j(ω t +θ) ] = 2| V + |cos(β d) ⋅ Re[ (cos(ω t + θ) + j sin(ω t + θ)] = 2| V + |cos(β d) cos(ω t + θ) I (d, t ) = Re[ I (d) e jω t ] = Re[2 j | V + | e jθ sin(β d) e jω t ] / Z0 = 2| V + |sin(β d) ⋅ Re[ j e j(ω t +θ) ] / Z0 = 2| V + |sin(β d) ⋅ Re[ j cos(ω t + θ) − sin(ω t + θ)] / Z0 | V+ | = −2 sin(β d) sin(ω t + θ) Z0 © Amanogawa, 2006 – Digital Maestro Series 100 Transmission Lines The time-dependent power is given by P(d, t ) = V (d, t ) ⋅ I (d, t ) = | V + |2 =−4 cos(β d) sin(β d) cos(ω t + θ) sin(ω t + θ) Z0 | V + |2 =− sin(2β d) sin (2ω t + 2θ) Z0 and the corresponding time-average power is 1 T < P(d, t ) > = ∫ P(d, t ) dt T 0 | V + |2 1 T =− sin(2β d) ∫ sin (2ω t + 2θ) = 0 Z0 T 0 © Amanogawa, 2006 – Digital Maestro Series 101 Transmission Lines ZR = Z0 (MATCHED LOAD) Z0 ZR = Z0 The reflection coefficient for a matched load is ZR − Z0 Z0 − Z0 ΓR = = =0 ZR + Z0 Z0 + Z0 no reflection! The line voltage and line current phasors are V (d) = V + e jβ d (1 + Γ R e−2 jβ d ) = V + e jβ d + V + jβ d V (1 − Γ R e−2 jβ d ) = I( d ) = e e jβ d Z0 Z0 © Amanogawa, 2006 – Digital Maestro Series 102 Transmission Lines The line impedance is independent of position and equal to the characteristic impedance of the line V (d) V + e jβ d Z(d) = = = Z0 + I (d) V jβ d e Z0 The time-dependent voltage and current are V (d, t ) = Re[| V + | e jθ e jβ d e jω t ] = | V + | ⋅ Re[e j(ω t +β d +θ) ] =| V + | cos(ω t + β d + θ) I (d, t ) = Re[| V + | e jθ e jβ d e jω t ] / Z0 + | V+ | | V | j (ω t +β d +θ) = ⋅ Re[e ]= cos(ω t + β d + θ) Z0 Z0 © Amanogawa, 2006 – Digital Maestro Series 103 Transmission Lines The time-dependent power is + | | V P(d, t ) = | V + | cos(ω t + β d + θ) cos(ω t + β d + θ) Z0 | V + |2 cos2 (ω t + β d + θ) = Z0 and the time average power absorbed by the load is 1 t | V + |2 < P(d) > = ∫ cos2 (ω t + β d + θ) dt T 0 Z0 | V + |2 = 2 Z0 © Amanogawa, 2006 – Digital Maestro Series 104 Transmission Lines ZR = jX (PURE REACTANCE) ZR = j X Z0 The reflection coefficient for a purely reactive load is ZR − Z0 jX − Z0 ΓR = = = ZR + Z0 jX + Z0 ( jX − Z0 )( jX − Z0 ) = = ( jX + Z0 )( jX − Z0 ) © Amanogawa, 2006 – Digital Maestro Series X 2 − Z02 Z02 + X 2 +2j XZ0 Z02 + X 2 105 Transmission Lines In polar form Γ R = Γ R exp( jθ) where ΓR = ( ( ) ) ( 2 2 X − Z0 4 X 2 Z02 + = 2 2 Z02 + X 2 Z02 + X 2 2 ) ( ( ) ) 2 2 2 Z0 + X =1 2 Z02 + X 2 −1 2 XZ0 θ = tan X 2 − Z2 0 The reflection coefficient has unitary magnitude, as in the case of short and open circuit load, with zero time average power absorbed by the load. Both voltage and current are finite at the load, and the time-dependent power oscillates between positive and negative values. This means that the load periodically stores power and then returns it to the line without dissipation. © Amanogawa, 2006 – Digital Maestro Series 106 Transmission Lines Reactive impedances can be realized with transmission lines terminated by a short or by an open circuit. The input impedance of a loss-less transmission line of length L terminated by a short circuit is purely imaginary 2π f 2π Zin = j Z0 tan ( β L ) = j Z0 tan L = j Z0 tan L v λ p For a specified frequency f, any reactance value (positive or negative!) can be obtained by changing the length of the line from 0 to λ/2. An inductance is realized for L < λ/4 (positive tangent) while a capacitance is realized for λ/4 < L < λ/2 (negative tangent). When L = 0 and L = λ/2 the tangent is zero, and the input impedance corresponds to a short circuit. However, when L = λ/4 the tangent is infinite and the input impedance corresponds to an open circuit. © Amarcord, 2006 – Digital Maestro Series 107 Transmission Lines Since the tangent function is periodic, the same impedance behavior of the impedance will repeat identically for each additional line increment of length λ/2. A similar periodic behavior is also obtained when the length of the line is fixed and the frequency of operation is changed. At zero frequency (infinite wavelength), the short circuited line behaves as a short circuit for any line length. When the frequency is increased, the wavelength shortens and one obtains an inductance for L < λ/4 and a capacitance for λ/4 < L < λ/2, with an open circuit at L = λ/4 and a short circuit again at L = λ/2. Note that the frequency behavior of lumped elements is very different. Consider an ideal inductor with inductance L assumed to be constant with frequency, for simplicity. At zero frequency the inductor also behaves as a short circuit, but the reactance varies monotonically and linearly with frequency as X = ωL (always an inductance) © Amarcord, 2006 – Digital Maestro Series 108 Transmission Lines Short circuited transmission line – Fixed frequency L Z in = 0 L= 0 0< L< L= λ 4 λ 2 4 λ 2 λ k p Im Z in > 0 in d u c ta n c e Z in → ∞ o p e n c irc u it k p c a p a c ita n c e Im Z in < 0 Z in = 0 2 < L< L= 4 λ < L< L= λ 3λ 4 3λ 4 3λ < L< λ 4 s h o rt c irc u it k p s h o rt c irc u it Im Z in > 0 in d u c ta n c e Z in → ∞ o p e n c irc u it k p Im Z in < 0 c a p a c ita n c e … © Amarcord, 2006 – Digital Maestro Series 109 Z(L)/Zo = j tan(β L) Transmission Lines Impedance of a short circuited transmission line (fixed frequency, variable length) 40 30 20 inductive Normalized Input Impedance 10 inductive inductive 0 -10 capacitive cap. -20 -30 -40 0 100 π/(2β)= λ/4 200 π/β =λ/2 300 3π/(2β)= 3λ/4 400 2π/β= λ 500 5π/(2β)= 5λ/4 θ L [deg] Line Length L © Amarcord, 2006 – Digital Maestro Series 110 Z(L)/Zo = j tan(β L) Transmission Lines Impedance of a short circuited transmission line (fixed length, variable frequency) 40 30 20 inductive Normalized Input Impedance 10 inductive inductive 0 -10 capacitive cap. -20 -30 -40 0 100 vp / (4L) 200 vp / (2L) 300 3vp / (4L) 400 vp / L 500 5vp / (4L) θ [deg] f Frequency of operation © Amarcord, 2006 – Digital Maestro Series 111 Transmission Lines For a transmission line of length L terminated by an open circuit, the input impedance is again purely imaginary Z0 Zin = − j =−j tan ( β L ) Z0 Z0 =−j 2π 2π f tan L tan L λ v p We can also use the open circuited line to realize any reactance, but starting from a capacitive value when the line length is very short. Note once again that the frequency behavior of a corresponding lumped element is different. Consider an ideal capacitor with capacitance C assumed to be constant with frequency. At zero frequency the capacitor behaves as an open circuit, but the reactance varies monotonically and linearly with frequency as 1 X= (always a capacitance) ωC © Amarcord, 2006 – Digital Maestro Series 112 Transmission Lines Open circuit transmission line – Fixed frequency L L= 0 0< L< L= λ 4 λ 2 4 λ 2 λ 2 < L< L= 4 λ < L< L= λ 3λ 4 3λ 4 3λ < L< λ 4 Z in → ∞ o p e n c irc u it Im Z in < 0 k p c a p a c ita n c e Z in = 0 s h o rt c irc u it k p Im Z in > 0 in d u c ta n c e Z in → ∞ o p e n c irc u it Im Z in < 0 k p c a p a c ita n c e Z in = 0 s h o rt c irc u it k p Im Z in > 0 in d u c ta n c e … © Amarcord, 2006 – Digital Maestro Series 113 Normalized Input Impedance Z(L)/ Zo = - j cotan(β L) Transmission Lines Impedance of an open circuited transmission line (fixed frequency, variable length) 40 30 20 inductive 10 inductive inductive 0 -10 capacitive capacitive capacitive -20 -30 -40 0 0 100 π/(2β) = λ/4 200 π/β = λ/2 300 3π/(2β) = 3λ/4 400 2π/β = 2λ 500 5π/(2β) = 5λ/4 θ [deg] L Line Length L © Amarcord, 2006 – Digital Maestro Series 114 Normalized Input Impedance Z(L)/ Zo = - j cotan(β L) Transmission Lines Impedance of an open circuited transmission line (fixed length, variable frequency) 40 30 20 inductive 10 inductive inductive 0 -10 capacitive capacitive capacitive -20 -30 -40 0 0 100 vp / (4L) 200 vp / (2L) 300 3vp / (4L) 400 vp / L 500 5vp / (4L) θ [deg] f Frequency of operation © Amarcord, 2006 – Digital Maestro Series 115 Transmission Lines It is possible to realize resonant circuits by using transmission lines as reactive elements. For instance, consider the circuit below realized with lines having the same characteristic impedance: I L1 short circuit Z0 V Zin1 Zin1 = j Z0 tan ( β L 1 ) © Amarcord, 2006 – Digital Maestro Series L2 Z0 short circuit Zin2 Zin2 = j Z0 tan ( β L 2 ) 116 Transmission Lines The circuit is resonant if L1 and L2 are chosen such that an inductance and a capacitance are realized. A resonance condition is established when the total input impedance of the parallel circuit is infinite or, equivalently, when the input admittance of the parallel circuit is zero 1 1 + =0 j Z0 tan ( β r L1 ) j Z0 tan ( β r L 2 ) or ωr ωr tan L1 = − tan L 2 vp vp with 2π ωr βr = = λ r vp Since the tangent is a periodic function, there is a multiplicity of possible resonant angular frequencies ωr that satisfy the condition above. The values can be found by using a numerical procedure to solve the trascendental equation above. © Amarcord, 2006 – Digital Maestro Series 117 Transmission Lines Transient and Steady-State on a Transmission Line We need to give now a physical interpretation of the mathematical results obtained for transmission lines. First of all, note that we are considering a steady-state regime where the wave propagation along the transmission line is perfectly periodic in time. This means that all the transient phenomena have already decayed. To give a feeling of what the steady-state regime is, consider a transmission line that is connected to the generator by closing a switch at a reference time t = 0. For simplicity we assume that all impedances, including the line characteristic impedance, are real. Generator Switch Load RG t=0 VG © Amanogawa, 2006 – Digital Maestro Series Transmission line RR 118 Transmission Lines After the switch is closed, the voltage at the input of the transmission line will vary nearly instantaneously from the open voltage of the generator VG to a value V+ , with a current I+ beginning to flow into the line. A transient takes place in the transmission line, as charges in the conductors move, transporting the current towards the load. Until the front first reaches the end of the transmission line, the load voltage remains zero. Initially, the input impedance of the transmission line appears to be the same as the line characteristic impedance, because the current cannot yet sense the value of the load impedance. Therefore, a voltage front V+ propagates with the current front I+ , where Z0 V = I Z0 = VG RG + Z0 + + + V V0 + I = = Z0 RG + Z0 © Amanogawa, 2006 – Digital Maestro Series 119 Transmission Lines If the load does not match exactly the characteristic impedance of the line, the voltage V+ and the current I+ cannot be established across the load RR , when the front reaches the end of the line, because V + ≠ I + RR Therefore, voltage and current adjust themselves at the load by – reflecting back a wave front with voltage V and current I such that V + + V − = ( I + + I − ) RR © Amanogawa, 2006 – Digital Maestro Series 120 Transmission Lines Since also the reflected front will see an impedance Z0 we have V + = Z0 I + ; ⇒ V − V − = − Z0 I − + RR − Z0 =V RR + Z0 The quantity V− RR − Z0 ΓR = = + R +Z V R 0 is the load reflection coefficient. © Amanogawa, 2006 – Digital Maestro Series 121 Transmission Lines The wave reflected by the load propagates in the negative direction and interferes with the oscillating values of voltage and current found along the transmission line, which continue to be injected by the generator. When the reflected wave reaches the input of the transmission line, it terminates on the generator impedance RG and if this does not match the line characteristic impedance, reflection back into the line again occurs, generating now an additional forward wave with associated voltage + − RG − Z0 V2 = V RG + Z0 Remember that the ideal voltage source part of the generator will behave simply as a short for the reflected wave attempting to exit the line from the input. © Amanogawa, 2006 – Digital Maestro Series 122 Transmission Lines The front reflected by the generator side will again reach the load, and waves of ever decreasing amplitude will keep bouncing back and forth along the line until the process associated to that initial wave front dies out. Every subsequent wave front injected over time by the oscillating generator undergoes an identical phenomenon of multiple reflections. If we assume a sinusoidal generator, voltage and current injected into the line repeat periodically, according to the period of the oscillation. Therefore, successive reflections at the ends of the line obey the same reflection coefficients, but involving in time different values of amplitude and phase. If the generator continues to supply the line with a stable oscillation, after a sufficient time the combined interference of forward and backward waves becomes stable, and one can identify two well-defined incoming and reflected steady-state waves, arising from the superposition of the infinite transient components travelling on the line. © Amanogawa, 2006 – Digital Maestro Series 123 Transmission Lines Note that the wave fronts travel with a phase velocity equal to the speed of light in the medium surrounding the wires. Also, note that the length of the line will affect the interference pattern of the wave superposition, so that lines of different length will result in different voltage and current distributions along the line. When we study the steady-state voltages and currents in a transmission line, we only need to know the phasors that represent the stable steady-state oscillations at each line location. The phasors provide a snapshot of how values of voltage and current relate to each other in space, at a reference time, in terms of amplitude and phase. The actual time oscillation can be easily recovered, because in steady-state we know that voltage and current are perfectly periodic at each line location, according to the period of the generator. If the generator provides more than one frequency of oscillation, at steady-state the behavior of each frequency in the spectrum can be studied independently and the total result is obtained by superposition. © Amanogawa, 2006 – Digital Maestro Series 124 Transmission Lines Standing Wave Patterns In practical applications it is very convenient to plot the magnitude of phasor voltage and phasor current along the transmission line. These are the standing wave patterns: V (d) = V + ⋅ (1 + Γ(d) ) Loss - less line V+ ⋅ (1 − Γ(d) ) I (d) = Z0 V (d) = V + eα d ⋅ (1 + Γ ( d ) ) Lossy line V + eα d ⋅ (1 − Γ ( d ) ) I (d) = Z0 © Amanogawa, 2006 -–Digital Maestro Series 125 Transmission Lines The standing wave patterns provide the top envelopes that bound the time-oscillations of voltage and current along the line. In other words, the standing wave patterns provide the maximum values that voltage and current can ever establish at each location of the transmission line for given load and generator, due to the interference of incident and refelected wave. The patterns present a succession of maxima and minima which repeat in space with a period of length λ/2, due to constructive or destructive interference between forward and reflected waves. The patterns for a loss-less line are exactly periodic in space, repeating with a λ/2 period. Again, note that although we talk about maxima and minima of the standing wave pattern we are always examining a maximum of voltage or current that can be achieved at a transmission line location during any period of oscillation. © Amanogawa, 2006 -–Digital Maestro Series 126 Transmission Lines We limit now our discussion to the loss-less transmission line case where the generalized reflection coefficient varies as Γ(d) = Γ R exp ( − j 2β d ) = Γ R exp ( jφ ) exp ( − j 2β d ) Note that the magnitude of an exponential with imaginary argument is always unity exp ( jφ ) exp ( − j 2β d ) = 1 In a loss-less line it is always true that, for any line location, Γ(d) = Γ R When d increases, moving from load to generator, the generalized reflection coefficient on the complex plane moves clockwise on a circle with radius |ΓR| and is identified by the angle φ - 2β d . © Amanogawa, 2006 -–Digital Maestro Series 127 Transmission Lines The voltage standing wave pattern has a maximum at locations where the generalized reflection coefficient is real and positive Γ(d) = Γ R exp ( jφ ) exp ( − j 2β d ) = 1 ⇒ φ − 2β d = 2 nπ At these locations we have 1 + Γ(d) = 1 + Γ R ⇒ Vmax = V (d max ) = V + ⋅ (1 + Γ R ) The phase angle φ - 2β d changes by an amount 2π, when moving from one maximum to the next. This corresponds to a distance between successive maxima of λ/2. © Amanogawa, 2006 -–Digital Maestro Series 128 Transmission Lines The voltage standing wave pattern has a minimum at locations where the generalized reflection coefficient is real and negative Γ(d) = − Γ R exp ( jφ ) exp ( − j 2β d ) = −1 ⇒ φ − 2β d = ( 2 n + 1 ) π At these locations we have 1 + Γ(d) = 1 − Γ R ⇒ Vmin = V (d min ) = V + ⋅ ( 1 − Γ R ) Also when moving from one minimum to the next, the phase angle φ - 2β d changes by an amount 2π. This again corresponds to a distance between successive minima of λ/2. © Amanogawa, 2006 -–Digital Maestro Series 129 Transmission Lines The voltage standing wave pattern provides immediate information on the transmission line circuit If the load is matched to the transmission line ( ZR = Z0 ) the voltage standing wave pattern is flat, with value | V+ |. If the load is real and ZR > Z0 , the voltage standing wave pattern starts with a maximum at the load. If the load is real and ZR < Z0 , the voltage standing wave pattern starts with a minimum at the load. If the load is complex and Im(ZR ) > 0 (inductive reactance), the voltage standing wave pattern initially increases when moving from load to generator and reaches a maximum first. If the load is complex and Im(ZR ) < 0 (capacitive reactance), the voltage standing wave pattern initially decreases when moving from load to generator and reaches a minimum first. © Amanogawa, 2006 -–Digital Maestro Series 130 Transmission Lines Since in all possible cases Γ (d) ≤ 1 the voltage standing wave pattern V (d) = V + ⋅ (1 + Γ(d) ) cannot exceed the value 2 | V+ | in a loss-less transmission line. If the load is a short circuit, an open circuit, or a pure reactance, there is total reflection with Γ (d) = 1 since the load cannot consume any power. The voltage standing wave pattern in these cases is characterized by Vmax = 2 V + © Amanogawa, 2006 -–Digital Maestro Series and Vmin = 0 . 131 Transmission Lines The quantity 1 + Γ(d) is in general a complex number, that can be constructed as a vector on the complex plane. The number 1 is represented as 1 + j0 on the complex plane, and it is just a vector with coordinates (1,0) positioned on the Real axis. The reflection coefficient Γ(d) is a complex number such that |Γ(d)| ≤ 1. Im Γ 1+Γ 1 © Amanogawa, 2006 -–Digital Maestro Series Re 132 Transmission Lines We can use a geometric construction to visualize the behavior of the voltage standing wave pattern V (d) = V + ⋅ (1 + Γ(d) ) simply by looking at a vector plot of |(1 + Γ (d))| . |V+| is just a scaling factor, fixed by the generator. For convenience, we place the reference of the complex plane representing the reflection coefficient in correspondence of the tip of the vector (1, 0). Example: Load with inductive reactance Im( Γ ) 1+ΓR φ ΓR 1 © Amanogawa, 2006 -–Digital Maestro Series Re ( Γ ) 133 Transmission Lines Im( Γ ) Maximum of voltage standing wave pattern φ 1+Γ(d) ΓR 1 Γ(d) 2 β dmax Re ( Γ ) ∠ Γ ( d ) = φ − 2β d max = 0 Im( Γ ) Minimum of voltage standing wave pattern φ 1+Γ(d) 1 ∠ Γ ( d ) = φ − 2β d min = −π © Amanogawa, 2006 -–Digital Maestro Series Γ(d) ΓR Re ( Γ ) 2 β dmin 134 Transmission Lines The voltage standing wave ratio (VSWR) is an indicator of load matching which is widely used in engineering applications Vmax 1 + Γ R VSWR = = Vmin 1 − Γ R When the load is perfectly matched to the transmission line ΓR = 0 ⇒ VSWR = 1 When the load is a short circuit, an open circuit or a pure reactance ΓR = 1 ⇒ VSWR → ∞ We have the following useful relation VSWR − 1 ΓR = VSWR + 1 © Amanogawa, 2006 -–Digital Maestro Series 135 Transmission Lines Maxima and minima of the voltage standing wave pattern. Load with inductive reactance Im ( ZR ) > 0 ⇒ The load reflection coefficient is in this part of the domain ZR − Z0 Im ( Γ R ) = Im >0 ZR + Z0 Im Γ Re Γ 1 The first maximum of the voltage standing wave pattern is closest to the load, at location ∠ Γ ( d ) = φ − 2β d max = 0 © Amanogawa, 2006 -–Digital Maestro Series ⇒ φ d max = λ 4π 136 Transmission Lines Load with capacitive reactance Im ( ZR ) < 0 ⇒ ZR − Z0 Im ( Γ R ) = Im <0 ZR + Z0 The load reflection coefficient is in this part of the domain Im(Γ) 1 Re(Γ) The first minimum of the voltage standing wave pattern is closest to the load, at location ∠ Γ ( d ) = φ − 2β d min = π © Amanogawa, 2006 -–Digital Maestro Series ⇒ π−φ λ d min = 4π 137 Transmission Lines A measurement of the voltage standing wave pattern provides the locations of the first voltage maximum and of the first voltage minimum with respect to the load. The ratio of the voltage magnitude at these points gives directly the voltage standing wave ratio (VSWR). This information is sufficient to determine the load impedance ZR , if the characteristic impedance of the transmission line Z0 is known. STEP 1: The VSWR provides the magnitude of the load reflection coefficient VSWR − 1 ΓR = VSWR + 1 © Amanogawa, 2006 -–Digital Maestro Series 138 Transmission Lines STEP 2: The distance from the load of the first maximum or minimum gives the phase φ of the load reflection coefficient. Vmax |V| For an inductive reactance, a voltage maximum is closest to the load and 4π d max φ = 2β d max = λ Vmin dmax 0 |V| Vmax For a capacitive reactance, a voltage minimum is closest to the load and 4π φ = −π + 2β d min = −π + d min λ Vmin dmin 0 © Amanogawa, 2006 -–Digital Maestro Series 139 Transmission Lines STEP 3: The load impedance is obtained by inverting the expression for the reflection coefficient ZR − Z0 Γ R = Γ R exp ( jφ ) = ZR + Z0 ⇒ © Amanogawa, 2006 -–Digital Maestro Series 1 + Γ R exp ( jφ ) ZR = Z0 1 − Γ R exp ( jφ ) 140