MIT OpenCourseWare http://ocw.mit.

edu

6.013 Electromagnetics and Applications

For information about citing these materials or our Terms of Use, visit: http://ocw.mit.edu/terms.

Problem Set 7 Solutions Problem 7.1 (a) (i) For an air lled line we know that the propagation velocity is c . Looking at the gure we see the distance between adjacent voltage minima is A, so the wavelength is = 2A. The frequency of the wave is given by the ratio of the propagation velocity to the wavelength: f=
vp

3108 2A

1.5 A

108 [Hz]

(ii) The voltage standing wave ratio is the ratio of the maximum voltage on the line to the minimum voltage on the line: V SW R =
|V (z)|max | (z)|min V Zn 1 Zn +1 .

(iii) At a voltage minima (at the load in this case) we know that = || = Z(z = 0)N = RL = 1 = Z We can solve this for Z Z = (iv)
RL Z(z=0)n 1+ 1|| 1+||

So we can write:

1 V SW R

= RL V SW R = RL | (z)|max V min

|V (z)|

1 We know that the velocity of propagation on the line is Vp = LC , and that the characteristic L impedance of the line is Z = C , so the Capacitance per unit length of the line is

C=

1 Vp Z

1 (3108 )RL V SW R

[F/m]

(v) Using the same relations as in part (iv) we know that the inductance per unit length will be L=
Z Vp

RL V SW R 3108

(vi) We know that the magnitude of the reection coecient is given by || = V SW R1 V SW R+1 and that at z = 0 the reection coecient is = | So the reection coecient is (z) = V SW R1 ej2z/A V SW R+1 (vii) The fraction of the power reected by the load is the square of the magnitude of the reection coecient 2 P = V SW R1 P+ V SW R+1 (b) In this case, we can calculate VSWR and ||. V SW R = 2 = 2, || = V SW R1 = 21 = 1 1 V SW R+1 2+1 3 Again, at a voltage minimum the reection coecient is |V min = || = 1 . So at the voltage minimum, 3 the normalized impedance is
1 Z |V min = 1 = 11/3 = 2 1+1/3 From this point we can either solve the problem using a Smith Chart or by using equation 7.2.24. 1+

To solve using a Smith Chart rst nd the point corresponding to ZN = 0.5. Next draw the || = const circle. Finally, rotate from the ZN = 0.5 point toward the load by . The new point is the normalized load 8 impedance at the load. Based on the attached Smith Chart, ZL = Z (0.8 0.6j). To solve using equation 7.2.24, rst observe that when z = , tan(kz) = tan( 2 ) = tan(/4) = 1 8 8
+jZ Z( )N = 1 = ZL+jZL 2 Z 8 0.5Z + 0.5jZL = ZL + jZ

ZL (1 0.5j) = Z (0.5 j)
.5 ZL = Z 100jj = Z 122jj .5 2+j 2+j

= Z 43j = Z (0.8 0.6j) 5

Which is the same as the answer from the Smith Chart. (c) For shunt matching it is easier to work with admittances. The admittance of a capacitor is YC = jC > 0, so we need to nd the rst point on the line where the input admittance as a real part of 1 (matched) and negative imaginary part (so that the admittance of the capacitor can cancel it out). Again, its possible to solve this problem using a equation 7.2.24 or the Smith Chart. To solve using the Smith Chart, rst locate the load impedance. Next draw the circle of constant ||, and nd the point /2 from the load. This is the admittance of the load. Follow the constant || circle to the point it crosses the Re(Y ) = 1 circle, with a negative imaginary value. This is the point at which we can match the load with a shunt capacitance. From the Smith Chart, we can match at z = 0.2225 with a capacitance C = 0.7/(Z ). We can also solve this problem using equation 7.2.24. First we solve for the normalized admittance:
L L L YN = Z1 = Z jZL tan(kz) = Z jZL tan(kz) ZL +jZ tan(kz) = L |ZL |2 2ZLI Z tan(kz)+Z 2 tan2 (kz) N L L L Where ZLI is the imaginary part of the load impedance. To match the load, we need to nd a place where the real part of the complex admittance is 1.

Z j Z tan(kz)

Z j Z tan(kz) Z +jZ tan(kz)

Z Z +jZ 2 tan(kz)jZ Z tan(kz)+Z Z tan2 (kz)

Z ZLR Re(YN ) = 1 = |ZL |2 2Imag(Z+ZLR Z tan (kz)tan2 (kz) 2 L )Z tan(kz)+Z 2 We can plug in the known load impedance, and divide top and bottom by Z to get:

1=

0.8+0.8tan2 (kz) 1+1.2tan(kz)+tan2 (kz)

tan2 (kz) + 1.2tan(kz) + 1 = 0.8 + 0.8tan2 (kz) 0.2tan2 (kz) + 1.2tan(kz) + 0.2 = 0 tan2 (kz) + 6tan(kz) + 1 = 0 We can solve for tan(kz) using the quadratic formula 62 411 tan(kz) = 6 21 = 3 2 2 z = 2 arctan(3 2 2) Giving two possible solutions z = 0.02704, and z = 0.22296. Adding some multiple of 0.5 to either of these solutions will also satisify the Re(Y ) = 1 condition since tan(kz) is periodic, so we actually have an innite number of possible match points. For simplicity well take the rst point that meets our requirements. By plugging these distances into the impedance equation, we see that z = 0.22296 is the solution we want. At this point the input admittance is:

Yn = 1j(0.8j0.6)tan(20.22296) 1 j0.7071 (0.8j0.6)jtan(20.22296) So we can match the load by placing a shunt capacitance of C = 0.7071/(Z ) at position z = 0.22296

Problem 7.2 (a) From the VSWR we can determine the magnitude of the reection coecient. The point where the bulb is brightest will correspond to a current maximum ((z) = ||). || = V SW R1 = 31 = 2 = 1 V SW R+1 3+1 4 2 (/4) = 1 = L ej2k/4 = L ej2 4 2 1 j 4 L = 2 e 4 = 1 ej = 1 2 2 1+L ZLN = 1 L = 1.5 = 3 0.5
ZL = ZLN Z = 3 300 = 900 []
(b) Its possible to match the load using a short-circuited stub in series or in parallel with the antenna feed line. The position and length of the stub will likely be dierent in each case, so we need to test both methods and pick the one that results in the shortest stub length. We could work the problem exactly for each matching statagie, or we can use the Smith Chart to get the approximate solutions and then do the exact solution for the case predicted to be shorter. Since I like using the Smith Chart thats how well solve the problem. The Smith Chart solutions are attached at the end of this problem solution. The method for each solution is described below. Smith Chart Series Match To nd a match using the stub in series we work with impedances. (1) First locate the normalized load point on the Smith Chart. (2) Next draw the circle of constant ||. There are two possible match positions. These positions occur when the circle of constant || crosses the Re{Zn } = 1 circle. The imaginary component of the short-circuited stub impedance goes positive (from zero) before going negative, so we want the match point where the load has a negative imaginary component of impedance. (3) From the match point we can read o the value of the imaginary component of impedance (j1.16Z ), and get the position ((0.333 0.25) = /12 towards generator). Next we nd the length of the stub needed to match the load at this point. (4) First nd the stub impedance point (Zn = 0, position = 0 towards generator). Next nd the point corresponding to an impedance of j1.16Z (position = 0.136 towards generator). The dierence in positions is the length of the stub needed to match the load (0.136 = 0.544 [m]). Smith Chart Parallel Match To nd a match using the stub in parallel it is easier to work with admittances. (1) First locate the normalized load point on the Smith Chart. (2) Next draw the circle of constant ||. Rotate about this circle by /4 to nd the normalized load admittance. (3) Identify the two possible match positions. These positions occur when the circle of constant || crosses the Re{Yn } = 1 circle. The imaginary component of the short-circuited stub admittance goes negative (from innity) before going positive, so we want the match point where the load has a positive imaginary component of admittance. (4) From the match point we can read o the value of the imaginary component of admittance (+j1.155Z ), and get the position (0.1675 towards generator). Next we nd the length of the stub needed to match the
2

load at this point. (5) First nd the stub impedance point (Zn = 0, position = 0 towards generator). Now rotate by /4 to get the stub admittance (innity, position /4). Next nd the point corresponding to an admittance of j 1.155Z (position = 0.363 towards generator). The dierence in positions is the length of the stub needed to match the load (0.113 = 0.4544 [m]). From the Smith Chart results, we know that we can match the load using the shortest short-circuited stub when we place the stub in parallel with the feed line. Exact Solution to the Parallel Match Yn (z) = Yn (z) =
Z jZL tan(kz) ZL jZ tan(kz) Z jZL tan(kz) ZL jZ tan(kz)

ZL +jZ tan(kz) ZL +jZ tan(kz)

1 2 2 ZL +Z tan2 (kz)

2 2 Z ZL + jZ tan(kz) jZL tan(kz) + ZL Z tan2 (kz)

Re{Yn (z)} = 1 =

1 2 2 ZL +Z tan2 (kz)

Z ZL + ZL Z tan2 (kz)

2 2 ZL + Z tan2 (kz) = Z ZL + ZL Z tan2 (kz) 2 2 ZL Z ZL = tan2 (kz)(ZL Z Z ) 2 2 Z Z Z tan(kz) = ZL Z ZL = 900 300900 = 3 2 9003003002 L

kz =

, 3

z = (this is the sign that will work for us). 6 Plugging this back into the expression for impedance we get Y (/6) = Y (1 + j 3 2 ) 3
So our stub has to have an admittance of j 3 2 Y .
3 2 jY YS = tan(kz) = j 3 3 Y

tan(kz) =
2 z

3 2

= tan1 (

z= So,

3 2 ) 1 ( 23 ) 2 tan

= 0.1136 = 0.4544 [m]

(i) The stub should be 0.4544 meters long (ii) The stub should be in parallel with the feed line (iii) The stub should be placed
6

2 3

meters from the antenna port

Problem Set 7 Draft Solution Problem 7.3 Problem 7.3 (a) For a half-wave line, the input impedance is just the load impedance. (b) For a 3 -wave line, the input impedance is the inverse of the load impedance. 4
1 1 1 1 Zin = Z jZ = Z 1j 1+j = Z 1+j = 0.5 + j 0.5 1+j 2 Z Z At the input, the impedance looks like an inductor in series with a resistor.

(c) In this case the problem is a little more complicated. The Thevenin equivalent resistance will be Z . To nd the open circuit voltage, we need to think about the standing wave pattern that results in the steady state. At the input (from the sources point of view) of the transmission line, the lines equivalent impedance is zero V (the /4 transform of an open circuit). So the current has to be IT = S . This gives us two equations: Z VT = 0 = V+ + V = (I+ + I )Z IT =
VS Z

= I+ I
V VS 2Z .

The rst equation tells us that I = I+ , and when we put this into the second equation we get I+ =

S Using this we can get V+ = 2 . At the open circuit the voltage will be twice the forward going voltage (since the reection coecient is 1 at an open circuit). Finally, we know that the open circuit voltage will lag behind the source voltage by a quarter wave ( = kz = 2 = ). 4 2

So the Thevenin voltage will be VT hevenin = VS ej 2 . (d) No solution, skip this problem

10

Problem Set 7 Draft Solution Problem 7.4 Problem 7.4 (a) Assume we want to match at normal incidence, and at the center of the wavelength band given (0.5m). Further, assume that we can neglect reections from the back of the solar cell. We can approximate this problem as a quarter wave line matching problem. The impedance of one line will be Z = and the impedance of the second line will be ZL = L . We know from the notes that the impedance of the line used to match the two materials should be Za = (Z ZL )0.5 , so we have the following relation a = a = ( L )0.5 = ( L )0.5 So we need to choose a = L = 2 For this value of , a free-space wavelength of 0.5m has wavelength = / 2 = thickness will be one quarter of this value. d=
1 8 2 0.5 2

[m]. The coating

[m] 88 [nm]
8

(b) At one micron, the layer will be coating will be

thick, so we know the apparent impedance at the surface of the

+jZa a L L Zin = Za ZL+jZL = 0.5 0.5 +j 0.5 = 0.5 a +j0.5 0.5 a a Za Zin Z

0.5 +j 0.5

2+j2
1 2 2+j 2
1 +1

Zin Z Zin Z

The fraction of the power reected is ||2 , which is ||2 = 0.059, at one micron.

11

Problem Set 7 Draft Solution Problem 7.5 Problem 7.5 (a) For a series RLC resonance, we have the following relations 1 L L Q = R C which tells us C = RQ = 2000. 1 = LC = 2 106 which tells us LC = (2 106 )1 . L 1 L = LC C = (2 106 )1 2000 = 103 1 1 C = LC C = (2 106 )1 2000 = 4 109 L (b) For a parallel RLC resonance, we have the following relations Q = R C which tells us C = Q/R = 0.2 L L 1 = LC = 2 106 which tells us LC = (2 106 )1 . L 1 1 L = LC C = (2 106 )1 0.2 = 4 105 1 C = LC C = (2 106 )1 0.2 = 107 L (c) From the problem statement we know we are looking for a parallel resonator (short circuit far from resonance), with the input to the amplier serving as the load resistor. We also know that QLoaded = 21106 250106 = 0.02. QLoaded = 0.02 = CRth R , which tells us that Rth +R C= = L=
Q Rth +R Rth R 1 LC

1 200 250106 1002

1010

which tells us L = =
1 16

1 2 C .

1 4 (2106 )2 1010

102

12