Problems 1. Rectangular Waveguide Fundamentals: C M N FC A B
Problems 1. Rectangular Waveguide Fundamentals: C M N FC A B
Problems 1. Rectangular Waveguide Fundamentals: C M N FC A B
Problems
1. Rectangular Waveguide Fundamentals
P7.1: Find the cutoff frequency for the first 8 modes of WR430.
P7.2: Calculate the cutoff frequency for the first 8 modes of a waveguide that has a =
0.900 inches and b = 0.600 inches.
P7.3: Calculate the cutoff frequency for the first 8 modes of a waveguide that has a =
0.900 inches and b = 0.300 inches.
7-2
P7.4: Calculate uG, the wavelength in the guide and the wave impedance at 10 GHz for
WR90.
P7.5: Consider WR975 is filled with polyethylene. Find (a) uu, (b) up and (c) uG at 600
MHz.
From Table 7.1 for WR975 we have a = 9.75 in and b = 4.875 in. Then
7-3
c 3 x108 m s 1 1in
fc10 403MHz
2 r a 2 2.26 9.75in 0.0254m
2 2
fc 403
F 1 1 0.741
f 600
Now,
c 3x108 m
uU 2 x108
r 2.26 s
uU m
uP 2.7 x108
F s
m
uG uU F 1.48 x108
s
P7.6: MATLAB: Plot up and wavelength in the guide as a function of frequency over the
cited useful frequency range for WR90.
% MLP0706
%
% Plot propagation velocity and guide wavelength
% over the cited useful freq range of WR90.
%
% 2/2/03 Wentworth
%
c=3e8;
a=0.900;b=0.450;
fc=(c/(2*.0254*a));
flo=8.2e9;
fhi=12.4e9;
N=100;
df=(fhi-flo)/N;
f=flo:df:fhi;
A=sqrt(1-(fc./f).^2);
Lu=c./f;
LG=Lu./A;
up=c./A;
fG=f./1e9;
subplot(2,1,1) Fig. P7.6
plot(fG,LG)
ylabel('guide wavelength (m)')
7-4
grid on
subplot(2,1,2)
plot(fG,up)
xlabel('frequency(GHz)')
ylabel('propagation velocity (m/s)')
grid on
P7.7: WR90 waveguide is to be operated at 16 GHz. Tabulate the values of the guide
wavelength, phase velocity, group velocity and impedance for each supported mode.
1 fc 16
2 2 m
uG uu 3x108 1 6.562 2.74 x108
f s
u 120
TE
Z mn 413
f
2 2
1 fc 1 6.562
16
Likewise values are found for the TE20 and TE11 mode. For the TM11 mode, a different
expression for impedance is used:
f
2
Z TM
mn u 1 fc
P7.8: MATLAB: Modify MATLAB 7.1 by plotting uG and up versus frequency for the
same guide over the same frequency range.
% M-File: MLP0708
%
% Waveguide Velocity Plot
% Plots uG and uP for TE11 and TM11
% vs freq. for air-filled waveguide
% (modifies ML0701)
%
% Wentworth, 11/26/02
%
clc %clears command window
clear%clears variables
% Initialize variables
c=2.998e8; %speed of light
Zo=120*pi;
ainches=0.900;
binches=0.450; Fig. P7.8
% convert to metric
a=ainches*0.0254;
b=binches*0.0254;
% calc fc11
fc=c*sqrt((1/a)^2+(1/b)^2)/2;
% Perform calculations
f=15e9:.1e9:25e9;
fghz=f/1e9;
Factor=sqrt(1-(fc./f).^2);
uu=c.*Factor./Factor; %just filling array with c
up=c./Factor;
ug=c.*Factor;
% Display results
plot(fghz,up,'-.k',fghz,uu,'--k',fghz,ug,'-k')
legend('up','c','uG')
xlabel('frequency, (GHz)')
ylabel('velocity (m/s)')
grid on
7-6
P7.9: MATLAB: Plot the TE10 wave impedance for WR430 waveguide versus frequency
if the guide is filled with Teflon. Choose a suitable frequency range for your plot.
% MLP0709
%
% Plot TE10 wave impedance for teflon filled
% WR430 guide over a suitable frequency range.
%
% 2/2/03 Wentworth
%
c=3e8;
er=2.1;
uu=c/sqrt(er);
a=4.30;b=2.150;
fc=(uu/(2*.0254*a));
flo=1.7e9/sqrt(er);
fhi=2.6e9/sqrt(er);
N=100;
df=(fhi-flo)/N;
f=flo:df:fhi; Fig. P7.9
A=sqrt(1-(fc./f).^2);
ZTE=(120*pi/sqrt(er))./A;
fG=f./1e9;
plot(fG,ZTE)
xlabel('frequency(GHz)')
ylabel('TE10 mode impedance (ohms)')
grid on
From our study of T-Lines, we know that looking into a quarter guide-wavelength section
of waveguide terminated in a short circuit, the input impedance appears infinite. The
cutoff frequency for the TE10 mode is 4.29 GHz. Then,
U c f 3 x108 7 x109
0.0542m
2 2 2
1 fc 1 fc 1 4.29
f f 7
So the quarter wave length is 0.0542m/4 = 0.0136 m. Therefore a distance 1.4 cm away
from the short circuit, the input impedance appears infinite.
7-7
2
u 1 fc f .
So
2 2
u2 2 , u 2 1-1+ fc f =u 2 fc f
and
2
2 fc
m 2 n 2 2
u ,
f a b
2 2
2 fc m n 2
2 2
u f a b
2 2
2 fc m n
u f a b
Solving for fc, where we have uu = uf,
2 2 2 2
1 m n 1 m n
fc uu
2 a b 2 a b
P7.12: Find expressions for the phasor field components of the TE01 mode.
P7.13: Find an expression for the magnetic field of the TE11 mode.
7-8
j x y
H H o sin cos cos(t z )a x
u a
2 2
a b
j x y
H o cos sin cos(t z )a y
b
2
u
2
a b
x y
H o cos cos cos(t z )a z
a b
P7.14: Modify MATLAB 7.2 to look at the Hz field for the TE20 mode.
% M-File: MLP0714
%
% TE02 Hz Field Pattern
% Generates contour and surface plots
%
% Wentworth, 11/26/02
%
% Variables
% m,n mode indicators
% a,b unitless guide dimensions
% betax x component of phase constant
% betay y component of phase constant
% Ez Ez for contour plot
% Ezc Ez for conventional plot
% Ezs Ez for surface plot
% Initialize variables
m=0;n=2;
a=40;b=20;
betax=m*pi/a;
betay=n*pi/b;
end
subplot(3,1,1)
contour(x,y,Hz,4)
title('Hz')
ylabel('y')
axis('equal')
axis([0 a 0 b])
subplot(3,1,2)
plot(Hzc,yc)
ylabel('y')
xlabel('10*Hz, (at y=b/2)')
axis('equal')
axis([-20 20 0 b])
subplot(3,1,3)
surf(Hzs)
axis([0 40 0 20 0 10])
axis('equal')
xlabel('x')
ylabel('y') Fig. P7.14
zlabel('Hzs*10')
P7.15: MATLAB: You are to create a movie showing how a surface plot of Hz over a
cross section of rectangular waveguide changes with position. Use WR284 waveguide
operating at 5 GHz and animate Hz for the TE11 mode.
% M-File: MLP0715
%
% Movie of the Hz contour field pattern in TE11
% mode as a function of z-position along the line.
% Here we'll take a time snapshot at t = 0.
7-10
%
% Wentworth, 2/5/03
%
clc %clears the command window
clear %clears variables
% Initialize variables
m=1;n=1;
f=5e9;
c=3e8;
a=2.84*.0254;
b=1.34*.0254;
fc=(c/2)*sqrt((m/a)^2+(n/b)^2);
betau=2*pi*f/c;
betax=m*pi/a;
betay=n*pi/b;
beta=betau*sqrt(1-(fc/f)^2);
M=40;
N=20;
Lguide=2*pi/beta;
Hzs=10*cos(betax.*X).*cos(betay.*Y).*cos(-
beta.*z(n));
surf(Hzs);
axis([0 40 0 20 -20 20])
axis('equal')
xlabel('x')
ylabel('y')
zlabel('Hzs*10')
M(n)=getframe;
end
movie(M);
3. Dielectric Waveguide
P7.16: Start with (7.83) and derive (7.84) and (7.85).
n1 cos i n2 cos t
(7.83)
n1 cos i n2 cos t
Now, from the definition of index of refraction, we have
n1 sin t n
, or sin t 1 sin i ,
n2 sin i n2
2
n
cos t 1 1 sin i
n2
Substituting this into (7.83) we have
2 2
n n n
n1 cos i n2 1 1 sin i cos i 2 1 1 sin i
n2 n1 n2
2 2
n n n
n1 cos i n2 1 1 sin i cos i 2 1 1 sin i
n2 n1 n2
n2 n1 sin i n2 n1 sin i
2 2 2 2
cos i cos i j 2
n2 n1 sin i n2 n1 sin i
2 2 2 2
cos i cos i j 2
7-12
sin i n2 n1
2 2
cos i j
(7.84)
sin i n2 n1
2 2
cos i j
To arrive at the phase, we must break (7.84) into its real and imaginary parts.
Multiplying both numerator and denominator by the conjugate of the denominator results
in:
2
sin i n2 n1
2 2
cos i j
,
cos 2 i sin 2 i n2 n1
2
or
sin i n2 n1
2 2
cos i j
,
1 n2 n1
2
cos i
P7.17: MATLAB: Compose a program that will plot the left and right side of (7.86)
versus all possible values of i for m = 0. Test the program using the following values: a
= 1 mm, f = 100 GHz, n1 = 3 and n2 = 1.
% MLP0717
% plot the left and right sides of (7.86)
% vs incident angle.
clc
clear
n1=3;
n2=1;
m=0;
a=.001;
f=100e9;
c=3e8;
N=100;
b1=2*pi*f*n1/c;
7-13
dth=round(90-thcd)/N;
thi1=thmind+dth:dth:90;
thir1=thi1*pi/180;
phi=a*b1*cos(thir1)/2;
left=tan(phi);
thi2=thcd+dth:dth:90-2*dth;
% Note that we avoid thi2 getting too close to 90 degrees
% where the function goes to infinity.
thir2=thi2*pi/180;
right=sqrt(sin(thir2).^2-(n2/n1)^2)./cos(thir2);
plot(thi1,left,'-*',thi2,right,'-o')
xlabel('incident angle (degrees)')
ylabel('equation (7.86)')
legend('left','right')
grid on
Fig. P7.17
7-14
P7.18: MATLAB: Compose a program that will plot the left and right side of (7.89)
versus all possible values of i for m = 0. Test the program using the following values: a
= 1 mm, f = 100 GHz, n1 = 3 and n2 = 1.
% MLP0718
% plot the left and right sides of (7.89)
% vs incident angle.
clc
clear
n1=3;
n2=1;
m=0;
a=.001;
f=100e9;
c=3e8;
N=100;
b1=2*pi*f*n1/c;
thc=asin(n2/n1); %critical angle in radians
thcd=thc*180/pi; %critical angle in degrees
dth=round(90-thcd)/N;
thmin=acos(pi/(a*b1));
thmind=thmin*180/pi;
thi1=thmind+dth:dth:90;
thir1=thi1*pi/180;
phi=a*b1*cos(thir1)/2;
left=tan(phi);
thi2=thcd+dth:dth:90-10*dth;
thir2=thi2*pi/180;
rightnum=sqrt(sin(thir2).^2-(n2/n1)^2);
rightden=((n2/n1)^2)*cos(thir2);
right=rightnum./rightden;
plot(thi1,left,'-*',thi2,right,'-o')
xlabel('incident angle (degrees)')
ylabel('equation (7.89)')
legend('left','right')
grid on
7-15
F xi
We then calculate F and F’ at this better point for x in order to arrive at a better still value
for x. Such an iterative procedure can usually quickly arrive at a solution.
We can also consider that the minimum possible i occurs where tan(u) = ∞, at cos u = 0.
And the maximum possible i occurs where tan(u) = 0, at sin u = 0. Solving, we find
(m 1) 1 m
i min cos 1 , and i max cos .
a 1 a 1
% MLP0719
%
% Use a Newton-Raphson iterative approach to solve
% for data in Equation 7.86.
%
% 2/3/03 Wentworth
%
% Variables
% a dielectric thickness (m)
% er relative permittivity
% f operating frequency (Hz)
% m mode
% n1,n2 indices of refraction
% n21 ratio of n2 to n1
% qi angle theta (rad)
% qimin minimum qi (rad)
% qimax maxiumum qi (rad)
% qicrit critical angle for qi (rad)
% b1 beta (rad/m)
% N number of iterations
% F,Fp function and 1st derivative
clc
clear
% Initialize Variables
a=0.050;
er=4;
f=4.5e9;
m=0;
n1=sqrt(er);
n2=1;
n21=n2/n1;
c=3e8;
b1=2*pi*f*sqrt(er)/c;
N=100;
7-17
for n=1:N
u=(a*b1*cos(qi(n))-m*pi)/2;
du=-a*b1*sin(qi(n))/2;
v=sqrt(sin(qi(n))^2-(n21)^2);
F(n)=tan(u)-v/cos(qi(n));
Fp(n)=(du/cos(qi(n))^2)-(sin(qi(n))*cos(qi(n))^2/v)
+v*sin(qi(n));
qi(n+1)=qi(n)-F(n)/Fp(n);
end
qidegrees=180*qi(N+1)/pi
qidegrees =
74.5224
>>
The program also works with m = 1 and m = 2. For m = 3, where there is no solution, a
complex number is calculated.
P7.20: Find (i)critical for a wave incident from distilled water into air.
P7.21: Suppose a Teflon slab of thickness 60 mm exists in air. What is the maximum
frequency at which this slab will support only one mode?
7-18
a 1 1 1
, so a o .
o 2 n12 n22 2 n12 n22
1c 1
a
2 f n n22
2
1
1 3x108 m s 1
a 9
0.273m
2 1x10 / s 1.6 1.5032
2
P7.23: MATLAB: Modify MATLAB 7.3 to find the odd TE mode field patterns. Use the
example information and duplicate the m = 1 plot for Figure 7.19.
% M-File: MLP0723
%
% This modifies ML0703 to find the odd TE mode
% field patterns at z = 0.
% The theta angle must be
% entered for a particular mode. The
% "hold on" function allows the results of
% multiple runs to be placed on one plot.
%
% Wentworth, 2/2/03
7-19
%
% Variables
% a dielectric thickness (m)
% b phase constant
% thdeg angle theta in degrees
% th angle theta in radians
% n1,n2 indices of refraction
% n21 the ratio n2/n1
% f frequency (Hz)
% c speed of light (m/s)
% w radian frequency (rad/s)
% Eo initial amplitude (V/m)
% alpha Fig P7.23
% initialize variables
a=50e-3;
thdeg=57.9; %corresponds to m = 1
th=pi*thdeg/180;
n2=1;
n1=2;
n21=n2/n1;
f=4.5e9;
w=2*pi*f;
c=2.998e8;
Eo=1;
b=(w/c)*n1;
alpha=b*sqrt(sin(th)^2-n21^2);
%m=0
x=-a/2:a/40:a/2;
Ey=Eo*sin(b*cos(th)*x);
hold on
plot(x,Ey,'k')
grid on
xlow=-2*a/2:a/40:-a/2;
Eylow=-Eo*sin(b*cos(th)*a/2)*exp(alpha*(xlow+a/2));
plot(xlow,Eylow,'k')
7-20
xhi=a/2:a/40:2*a/2;
Eyhi=Eo*sin(b*cos(th)*a/2)*exp(-alpha*(xhi-a/2));
plot(xhi,Eyhi,'k')
P7.24: Generate a figure similar to Fig. 7.17(b) for the m=0 mode of 10 mm thick r1 = 9
dielectric for n1/n2 = 1.5, 2, and 3.
P07.24
50
m=0
40
n1/n2=3
30
20
n1/n2=2
10
n1/n2=1.5
0
90 80 70 60 50 40 30 20
incidence angle (degrees)
Fig. P7.24
4. Optical Fiber
P7.25: A 100/240 silica optical fiber has a core index of 1.460 and a cladding index of
1.450. Estimate the number of propagating modes for a source wavelength (a) 850 nm,
(b) 1300 nm and (c) 1550 nm.
2 2
a 2 (50 x106 )
N 2 n f nc 2 (1.46) 2 (1.45)2
2
9
1.436 x10
N .
2
(a) for 850 nm we have
1.436 x109
N 1990
(850 x109 ) 2
Likewise, (b) N = 850 and (c) N = 598.
P7.26: Given a fiber of core index 1.478 and cladding index 1.445, find the numerical
aperture for a source of light incident from (a) air, (b) distilled water.
P7.27: Suppose nf = 1.475 and nc = 1.470. Determine the numerical aperture for a source
of light incident from air. What is the maximum core diameter allowed to support only a
single propagating mode if the source wavelength is 1300 nm?
amax
k01
1300 x10 2.405 4.1 m,
9
2 n 2f nc2 2 0.121
diameter 2amax 8.2 m.
P7.28: Given a step-index fiber with nf = 1.480 and a source wavelength of 1550 nm,
determine the minimum value of nc that will allow only one propagating mode for a core
radius of 2 m.
2 a n 2f nc2 k01
, or n 2f nc2 .
k01 2 a
Rearranging,
2 2
k k
nc n 01 , so nc min n 2f 01
2
f
2 a 2 a
P7.29: At a source wavelength of 1550 nm for a 5/125 silica fiber with nf = 1.470, what is
the minimum value nc can be and only allow one propagating mode?
Power budget:
Source 0dBm
Source-to-fiber -12dB (from Table 7.5)
Connector -0.7dB (from Table 7.5)
Fiber (1dB/km * 10km) -10dB (from Table 7.2)
4 splices (.05dB each) -0.2dB (from Table 7.5)
Connector -0.7dB (from Table 7.5)
Fiber-to-detector -1.5dB (from Table 7.5)
-25.1dB
Power Margin=-25.1dB-(-35dBm)= 9.9dB
7-23
Frequency budget:
For an LED with = 50 nm,
ns nm
tchrom 0.003 50nm 10km 1.5ns (see Tables 7.2, 7.3)
km
also
ns
tinterm 0.5 10km 5ns (see Table 7.2)
km
Combining,
t f tchrom
2
tinterm
2
1.5ns 2 0.5ns 2 5.22ns
and using LED risetime of 10 ns from Table 7.3, and PIN diode risetime of 0.3 ns from
Table 7.4,
10ns 5.22ns 0.3ns 11.28ns.
2 2 2
ts tt2 t 2f tr2
The rise-time budget is satisfied for t < T/2, so T > 2t, or Tmin = 22.6 ns. Since T=1/f,
we have fmax =1/Tmin=44 MHz.
Using the data available in Tables 7.2 – 7.5, and with source-to-fiber loss = 10log(NA 2),
we find:
850nm GRIN MMF 1300nm GRIN MMF
NA 0.24 0.20
Atten 4dB/km 1dB/km
P7.32: Calculate the maximum data rate, in bits per second, that could be supported for
each case of the previous problem.
For a typical LED we have (from Table 7.3) = 50 nm and tt = 10 ns. The detector
has (from Table 7.4) tr = 0.3 ns. Data from Table 7.2 is also used.
ns nm
tchrom 0.10 50nm 2km 10ns
km
ns
tintermodal 3 2km 6ns
km
(3) t f tchrom
2
tintermodal
2
11.7 ns
P7.33: A 10. km optical link is established between a typical laser diode and a typical
PIN photodiode using a 1300 nm step-index single-mode fiber. Find the power margin
and the maximum frequency analog signal that can be supported by this link. Assume 2
connectors and 4 splices. Compare your answer with that of problem P7.30.
-11.1dB
Power Margin=-11.1dB-(-35dBm)= 23.9dB
Frequency budget:
For an LED with = 50 nm,
ns nm
tchrom 0.003 3nm 10km .09ns
km
and
0.4ns 0.09ns 0.3ns 0.508ns.
2 2 2
ts tt2 tchrom
2
tr2
Tmin = 2ts = 1.016 ns, and fmax=1/Tmin=984MHz.
So, the laser diode source and the SMF:SI results in a much larger power margin and
much higher maximum frequency.