Rectangular Waveguides

Dr. Chandan Singh Rawat

Dr Chandan Singh Rawat

 Waveguide components

Rectangular waveguide Waveguide to coax adapter

Waveguide bends E-tee

Dr Chandan Singh Rawat

Figures from: www.microwaves101.com/encyclopedia/waveguide.cfm

 More waveguides

Dr Chandan Singh Rawat

http://www.tallguide.com/Waveguidelinearity.html
Uses

◼ To reduce attenuation loss

 High frequencies
 High power
◼ Can operate only above certain
frequencies
 Acts as a High-pass filter
◼ Normally circular or rectangular
 We will assume lossless rectangular
Dr Chandan Singh Rawat
 Rectangular WG
◼ Need to find the fields
components of the
em wave inside the
waveguide
 Ez Hz Ex Hx Ey Hy
◼ We’ll find that
waveguides don’t
support TEM waves
Dr Chandan Singh Rawat

http://www.ee.surrey.ac.uk/Personal/D.Jefferies/wguide.html
Maxwell’s Equations

Dr Chandan Singh Rawat

Field components
From Faraday and Ampere Laws we can find the
remaining four components:
 E z j H z
Ex = − −
h 2 x h 2 y *So once we know
 E z j H z Ez and Hz, we can
Ey = − 2 − 2 find all the other
h y h x
fields.
j E z  H z
Hx = 2 − 2
h y h x
j E z  H z
Hy = − 2 − 2
h x h y
where
Dr Chandan Singh Rawat

h =  + k = k + k y2
2 2 2 2
x
 Rectangular Waveguides:
Fields inside
Using phasors & assuming waveguide
filled with
◼ lossless dielectric material and

◼ walls of perfect conductor,

the wave inside should obey…

2 E + k 2 E = 0
2 H + k 2 H = 0
where k 2 =  2  c
Dr Chandan Singh Rawat
Then applying on the z-component…

 2 Ez + k 2 Ez = 0

 2 Ez  2 Ez  2 Ez
+ + + k Ez = 0
2

x 2
y 2
z 2

Solving by method of Separation of Variables :

E z ( x, y, z ) = X ( x)Y ( y ) Z ( z )
from where we obtain :
X '' Y '' Z ''
+ + = −k 2
X Y Z Dr Chandan Singh Rawat
 Fields inside the waveguide
'' '' ''
X Y Z
+ + = −k 2

X Y Z
− k x − k y +  = −k
2 2 2 2 h 2 =  2 + k 2 = k x2 + k y2

which results in the expressions :

X +k X =0
'' 2
x X(x) = c1 cos k x x + c2 sin k x x
Y '' + k y2Y = 0 Y(y) = c3 cos k y y + c4 sin k y y
Z '' −  2 Z = 0 Z ( z ) = c5e + c6 e
Dr Chandan Singh Rawat
z −z
 Substituting
X(x) = c1 cos k x x + c2 sin k x x
Y(y) = c3 cos k y y + c4 sin k y y
Ez ( x, y, z) = X ( x)Y ( y)Z ( z) Z ( z ) = c5ez + c6 e −z

(
E z = (c1 cos k x x + c2 sin k x x )(c3 cos k y y + c4 sin k y y ) c5ez + c6 e −z )
If only looking at the wave traveling in + z - direction :
E z = ( A1 cos k x x + A2 sin k x x )(A3 cos k y y + A4 sin k y y )e −z
Similarly for the magnetic field,
H z = (B1 cos k x x + B2 sin k x x )(B3 cos k y y + B4 sin k y y )e −z
Dr Chandan Singh Rawat
Modes of propagation
From these equations we can conclude:
◼ TEM (Ez=Hz=0) can’t propagate.

◼ TE (Ez=0) transverse electric

 In TE mode, the electric lines of flux are
perpendicular to the axis of the waveguide

◼ TM (Hz=0) transverse magnetic, Ez exists

 In TM mode, the magnetic lines of flux are
perpendicular to the axis of the waveguide.

◼ HE hybrid modes in which all components

exists Dr Chandan Singh Rawat
 TM Mode
E z = ( A1 cos k x x + A2 sin k x x )(A3 cos k y y + A4 sin k y y )e −z
◼ Boundary E z = 0 at y = 0,b
conditions: E = 0 at x = 0,a
z

From these, we conclude:

X(x) is in the form of sin kxx,
where kx=mp/a, m=1,2,3,…
Y(y) is in the form of sin kyy,
where ky=np/b, n=1,2,3,…
So the solution for Ez(x,y,z) is

E z = A2 A4 (sin k x x )(sin k y y )e − jz

Dr Chandan Singh Rawat

Figure from: www.ee.bilkent.edu.tr/~microwave/programs/magnetic/rect/info.htm

TM Mode

◼ Substituting

 mp   np  − jz
E z = Eo sin  x  sin  y e
 a   b 
where
 mp   np 
2 2

h =  +  = +k 2 2
2

 a   b 

Dr Chandan Singh Rawat

  mp   np  − jz
Ez = Eo sin  x  sin  y e
TMmn  a   b 
Hz = 0
◼ Other components are
 E z   mp   mpx   npy  −z
Ex = − Ex = − 2   Eo cos  sin  e
h 2 x h  a   a   b 
 E z   np   mpx   npy  −z
Ey = − 2 Ey = − 2   Eo sin   cos e
h y h  b   a   b 
j E z j  np   mpx   npy  −z
Hx = 2 Hx = 2   Eo sin   cos e
h y h  b   a   b 
j E z j  mp   mpx   npy  −z
Hy = − 2 Hy = − 2   Eo cos  sin  e
h x h  a   a   b 
Dr Chandan Singh Rawat
TM modes
◼ The m and n represent the mode of propagation
and indicates the number of variations of the
field in the x and y directions
◼ Note that for the TM mode, if n or m is zero, all
fields are zero.
◼ See applet by Paul Falstad

http://www.falstad.com/embox/guide.html

Dr Chandan Singh Rawat

 = (k 2
x )
+ k y2 − k 2
TM Cutoff  mp   np 
2 2

=   +  −  
2

 a   b 
◼ The cutoff frequency occurs when
 mp   np 
2 2

When c  =   +  then  =  + j = 0
2

 a   b 
 mp   np 
2 2
1 1
or f c =   + 
2p   a   b 
◼ Evanescent:

 mp   np 
2 2

When  2     +   =  and  = 0
 a   b 
 Means no propagation, everything is attenuated

 mp   np 
2 2

◼ Propagation: When    
2
 +   = j and  = 0
 a   b 
Dr Chandan Singh Rawat
 This is the case we are interested since is when the wave is allowed to
travel through the guide.
 attenuation Propagation

Cutoff of mode mn

fc,mn

◼ The cutoff frequency is the frequency

below which attenuation occurs and above
which propagation takes place. (High Pass)
2 2
u'  m   n 
f c mn =   + 
2  a  b
◼ The phase constant becomes
2
 mp   np   fc 
2 2

 =   − 
2
 −  =  ' 1 −  
 a   b   f 
Dr Chandan Singh Rawat
Phase velocity and impedance

◼ The phase velocity is defined as

 2p up
up = = =
'  f
◼ And the intrinsic impedance of the mode
is
2
Ex Ey  fc 
TM = =− = ' 1−  
Hy Hx f 
Dr Chandan Singh Rawat
 Summary of TM modes
Wave in the dielectric Inside the waveguide
medium
2
f 
 ' =  / u ' =    =  ' 1−  c 
f 

2
'=  /  TM
f 
= ' 1−  c 
f 

up = =/
u ' =  /  ' = f = 1 / 
2
f 
 ' 1−  c 
f 

'
 =
 '= u' / f  f 
2

1−  c 
Dr Chandan Singh Rawat  f 
 Related example of how fields look:
Parallel plate waveguide - TM modes
 mpx  e j(t − z )
Ez = A sin 
Ez  a 
m=1
0 a x

m=2

m=3
z a x
Dr Chandan Singh Rawat
 TE Mode
H z = (B1 cos k x x + B2 sin k x x )(B3 cos k y y + B4 sin k y y )e −z
◼ Boundary E x = 0 at y = 0,b
conditions: E y = 0 at x = 0,a
From these, we conclude:
X(x) is in the form of cos kxx,
where kx=mp/a, m=0,1,2,3,…
Y(y) is in the form of cos kyy,
where ky=np/b, n=0,1,2,3,…
So the solution for Ez(x,y,z) is

H z = B1 B3 (cos k x x )(cos k y y )e − jz

Dr Chandan Singh Rawat

Figure from: www.ee.bilkent.edu.tr/~microwave/programs/magnetic/rect/info.htm

TE Mode
◼ Substituting

 mpx   np  − jz
H z = H o cos  cos y e
 a   b 
where again
 mp   np 
2 2

h2 =   + 
 a   b 

◼ Note that n and m cannot be both zero

because the fields will all be zero.
Dr Chandan Singh Rawat
  mp   np  − jz
H z = H o cos x  cos y e
TEmn  a   b 
Ez = 0
◼ Other components are
j H z j  np   mpx   npy  −z
Ex = − 2 Ex = 2   H o cos  sin  e
h y h  b   a   b 
j H z j  mp   mpx   npy  −z
Ey = − 2 Ey = − 2   H o sin   cos e
h x h  a   a   b 
 H z j  mp   mpx   npy  −z
Hx = − 2 Hx = 2   H o sin   cos e
h x h  a   a   b 
 H z
Hy = − 2 j  np   mpx   npy  −z
h y Hy = 2   H o cos  sin  e
h  b   a   b 
Dr Chandan Singh Rawat
Various TE modes

Dr Chandan Singh Rawat

 attenuation Propagation

Cutoff of mode mn

fc,mn

◼ The cutoff frequency is the same

expression as for the TM mode
2 2
u'  m   n 
f c mn =   + 
2  a  b

◼ But the lowest attainable frequencies are

lowest because here n or m can be zero.

Dr Chandan Singh Rawat

Dominant Mode

◼ The dominant mode is the mode with

lowest cutoff frequency.
◼ It’s always TE10
◼ The order of the next modes change
depending on the dimensions of the
guide.

Dr Chandan Singh Rawat

 Summary of TE modes
Wave in the dielectric Inside the waveguide
medium
2
f 
 ' =  / u ' =    =  ' 1−  c 
f 

'
TE =
'=  /  f 
1−  c 
2

f 


up = =/
u ' =  /  ' = f = 1 / 
2
f 
 ' 1−  c 
f 

'
 '= u' / f  =
 f 
2

1−  c 
Dr Chandan Singh Rawat  f 
Variation of wave impedance

◼ Wave impedance varies with

frequency and mode

TE

’
TM

0 fc,mn Dr Chandan Singh Rawat

Example:
Consider a length of air-filled copper X-band
waveguide, with dimensions a=2.286cm,
b=1.016cm operating at 10GHz. Find the
cutoff frequencies of all possible propagating
modes.
Solution:
◼ From the formula for the cut-off frequency
2 2
u'  m   n 
f c mn =   + 
2  a  b
Dr Chandan Singh Rawat
Example

An air-filled 5-by 2-cm waveguide has

Ez = 20 sin (40px )sin (50py )e− jz V/m
at 15GHz
◼ What mode is being propagated?

◼ Find 

◼ Determine Ey/Ex

Dr Chandan Singh Rawat

 Group velocity, ug

◼ Is the velocity at which

the energy travels.
2 j  p   mpx  −z
 fc  Ey = −    e
1  rad/s   m  h2  a 
H o sin
 a 
ug = = u' 1 −    rad/m  =  s 
 /  f 

◼ It is always less than u’

u p u g = (u ')
2

Dr Chandan Singh Rawat

http://www.tpub.com/content/et/14092/css/14092_71.htm
 Group Velocity

◼ As frequency is increased,
the group velocity increases.

Dr Chandan Singh Rawat

 Power transmission
◼ The average Poynting vector for the waveguide
fields is
Pave = ReE  H  = Re E x H y − E y H x
1
2
* 1 * *

2
 
2
Ex + E y
2
[W/m2]
= zˆ
2
◼ where  = TE or TM depending on the mode
2
Ex + E y
2
a b
Pave =  Pave  dS =   dy dx [W]
x =0 y =0
2
Dr Chandan Singh Rawat
Attenuation in Lossy
waveguide
◼ When dielectric inside guide is lossy, and walls
are not perfect conductors, power is lost as it
travels along guide.
Pave = Po e −2z
dPave
◼ The loss power is PL = − = 2Pave
dz
◼ Where =c+d are the attenuation due to ohmic
(conduction) and dielectric losses
◼ Usually c >> d
Dr Chandan Singh Rawat
Attenuation for TE10

◼ Dielectric attenuation, Np/m Dielectric

conductivity!
 '
d = − 2
f 
2 1 −  c 
 f 

◼ Conductor attenuation, Np/m

2 Rs  b  f 
2

c = −  0.5 +  c ,10  
2   f  
 f c ,10   a   
b ' 1 −  
 f 
Dr Chandan Singh Rawat
Waveguide Cavities
◼ Cavities, or resonators, are
used for storing energy
◼ Used in klystron tubes,
band-pass filters and
frequency meters
◼ It’s equivalent to a RLC
circuit at high frequency
◼ Their shape is that of a
cavity, either cylindrical or
cubical.
Dr Chandan Singh Rawat
Cavity TM Mode to z
Solving by Separation of Variables :
E z ( x, y, z ) = X ( x)Y ( y ) Z ( z )
from where we obtain :
X(x) = c1 cos k x x + c2 sin k x x
Y(y) = c3 cos k y y + c4 sin k y y
Z ( z ) = c5 cos k z z + c6 sin k z z

k 2 = k x2 + k y2 + k z
2
where
Dr Chandan Singh Rawat
 TMmnp Boundary Conditions
E z = 0 at y = 0,b
From these, we conclude: E z = 0 at x = 0,a
kx=mp/a
E y = E x = 0, at z = 0,c
ky=np/b
kz=pp/c
where c is the dimension in z-axis

 mpx   npy   ppz 

E z = Eo sin   sin   sin   c
 a   b   c 
where
 mp   np   pp 
2 2 2

k2 =   +  +  =  
2

 a   b   c 
Dr Chandan Singh Rawat
Resonant frequency

◼ The resonant frequency is the same

for TM or TE modes, except that the
lowest-order TM is TM111 and the
lowest-order in TE is TE101.
2 2 2
u'  m   n   p 
fr =   +  + 
2  a  b  c 

Dr Chandan Singh Rawat

 Cavity TE Mode to z
Solving by Separation of Variables :
H z ( x, y, z ) = X ( x)Y ( y ) Z ( z )
from where we obtain :
X(x) = c1 cos k x x + c2 sin k x x
Y(y) = c3 cos k y y + c4 sin k y y
Z ( z ) = c5 cos k z z + c6 sin k z z

k = k + k + kz
2 2 2 2
where x y

Dr Chandan Singh Rawat

 TEmnp Boundary Conditions
H z = 0 at z = 0,c
From these, we conclude: E y = 0 at x = 0,a
kx=mp/a
ky=np/b Ex = 0, at y = 0,b
kz=pp/c
where c is the dimension in z-axis

 mpx   npy   ppy  c

H z = H o cos  cos  sin  
 a   b   c 

Dr Chandan Singh Rawat

Quality Factor, Q

◼ The cavity has walls with finite

conductivity and is therefore losing
stored energy.
◼ The quality factor, Q, characterized the
loss and also the bandwidth of the
cavity resonator.
◼ Dielectric cavities are used for
resonators, amplifiers and oscillators at
microwave frequencies.
Dr Chandan Singh Rawat
A dielectric resonator antenna
with a cap for measuring the
radiation efficiency

Dr Chandan Singh Rawat

Univ. of Mississippi
 Quality Factor, Q

◼ Is defined as
Time average energy stored
Q = 2π
loss energy per cycle of oscillation
W
= 2p
PL

For the dominant mode TE101 where

=
(a + c 2 abc
2
) =
1
QTE101
 ( ) (
 2b a 3 + c 3 + ac a 2 + c 2 )
Dr Chandan Singh Rawat
pf101o c
 Example
For a cavity of dimensions; 3cm x 2cm x 7cm filled with
air and made of copper (c=5.8 x 107)
◼ Find the resonant frequency and the quality factor
for the dominant mode. 3 1010  1   1   0 
2 2 2

f r110 =   +   +   = 9GHz
Answer: 2  3  2  7 
3 1010  1   0   1 
2 2 2

fr =   +   +   = 5.44GHz
2  3  2  7 
1
= = 1.6 10−6
(5.44 109 ) o c

=
(3
+ 7 2 )3  2  7
2
= 568,378
QTE101

 2  2(3 + 7 ) + 3  7(3 + 7 )
3 3 2 2

Dr Chandan Singh Rawat