D RADAR ENGINEE NG

92 MICROWAVE AN

z22,222222zzz;zz222222222222

===,"1 tr======i"1 (,=== 1

l
--, I I I I , - -1 I I I I ,--
1 I I 11 I 11 I 1111 I a
I I I I I I l__) I I I I I I
__ _) I I I I I._ __ _) I I I I L--
____ , ______
___ _J l I
/
L ____ _) I I L __ _
/ '---- b
222222z222zzz;;;z;z;zzzz222,
s
Electric field

:::[([[[[Illi(I::: f b
Magnetic fields ----

Fig. 4~29 Radiation pattern for TE10 mode.

4.3.6 Propagation of TM Waves in Rectangular Waveguide

For TM wave, Hz = 0 ; Ez =t- O
The wave equation of a TM wave is
2 2
a Ez + a Ez + h 2 Ez = O ... (4.38)
ax ay
2 2

This is a partial differential equation (p.d.e) which can be solved to

get the different field components Ex, Ey, Hx and Hy by 'separation of
variables method'.
Let us assume a solution
Ez = XY ... ( 4.39)
where, X.is a pure function of 'x' only.
Y is a pure function of 'y' only.
Since X and Y are independent variables
a2 (XY) - y a2X, '
a2 Ez = --.....-
·~
aX ax2 - , dx 2
a Ez
2
a2 (XY) a'V
a7 = <bi2 == Xa?
Using these two in Eq. 4.38, we get
d 2X d 2Y
Y-+X-+h2xY
2 2 0
dx dy ==
... (4.40)
Dividing throughout by XY, we get
2
1 dX 1 d 2Y
X dx2 + y dy2 + h2 == 0
... (4.41)
_!_ ~ is a pure function of x only
X dx~ ·
 MICROWAVE AND RADAR ENGINEERING

94
1 dzy . pure function of y only.
--;--:r1sa
Y dy . onstant Hence each term must be equal to
O f these 1s a c · . •
The sum • eX and Yare independent variables .
t arately sine
a constan sep . of variables method to solve the differential Eq.
We use separation
4.41.
.. .(4.42)
Let
... (4.43)
and
where:.... A2 and_ B 2 are constants.
Subtituting Eqs. 4.42 and 4.43 in Eq. 4.41, we get
- B2 -A2 + h2 = 0
h2 = A2 +B2 ... (4.44)
or
Equations 4.42 a.n d 4.43 are ordinary 2nd order differential
equations, the solutions of which are given by,
X = C1 cos Bx+ C2 sin Bx .. .(4.45)
Y = Ca cos Ay + C4 sinAy ... (4.46)
where C1, C2 , Ca and C4 are constants which can be evaluated by
applying the boundary conditions.
The complete solution is given by Eq. 4.39
i.e., Ez =X.Y
Substituting the values of X and Y from Eq. 4.45 and 4.46, we get
E = [C 1 cos Bx+ C2 sin Bx] [Ca cos Ay + C4 sin A~]
2 .. .(4.47)
Boundary Conditions
Since the entire surface of the rectangular waveguide acts as a short
circuit or ground for electric field, Ez = 0 all along the boundary walls
of the waveguide. Since there are four walls, as shown in Fig. 4.31 there
are four boundary conditionH.
1st boundary condition : f Bottom plane or bottom wall]
We know th1,d, I~:, 0, ulJ 11lon1{ ihc bottom wall.
i.e., (J :it y 0 v- :r - ~ 0 toa
stanch~for ''for all" and r - , 0 t.o '" means x varying between o to a .
--1

2nd boundary ,:,,n,lition : II i.,ft, t:-i idP plane or left side wall]
K O at, :) o v- y ~ o to b
MICROWAVE TRANSMISSION LINEs 95

Dir~ction of
y propogation

Fig. 4.31
3rd boundary condition : [Top plane or top wall]
Ez = 0 at y = b Y x ---7 0 to a
4th boundary condition : [Right side plane or right side wall]
Ez = 0 at!, = a Y y ---7 0 to b
Substituting 1st boundary condition in Eq. 4.4 7, given by
(i) '
Ez = (C 1 cos Bx+ C2 sin Bx] [Ca cos Ay + C4 sin Ay]
We have, Ez = 0 aty = 0 Y X---7 0 to a.
or O = [C 1 cos Bx+ C2 sin Bx] Ca[·.· cos O = 1, sin O - O]
This is true for all (Y) x - 0 to a
C cos Bx+ C2 sin Bx -:;: O; :. Ca = 0
1
Using this in Eq. 4.47, the solution reduces to,
Ez = [C 1 cos Bx+ C2 sin Bx] (C4 sin Ay] ... (4.48)
(ii) Substituting 2nd boundary condition in Eq. 4.48 above, we get
Ez = 0 = C1 C4 sinAy Y y ~ 0 to b
[·: cos O = 1 and sin O = O]
Since sinAy -:;: 0 and C4 -:;: 0.
C1 = O
Now using this in Eq. 4.48, the solution further reduces to '
Ez = C2 C4 sin Bx sin Ay ... (4.49)
(iii) Substituting 3rd boundary condition in Eq. 4.49 above, we get
Ez = 0 = C2C4sinBxsinAb [aty = b, Yx~Otoa]
Since sin Bx * 0 C4 :t; 0, C2 =I= 0, otherwise there would be
no solution
 MJCWJWAVE AND RADAR ENGINEERING
96
sin.Ab= 0 (1'1, NA~ ~'\'\"(tr .f I ?1.
Ab = t multiple of 7t -- n7t
or
where ,1 is a constant, n = 0, 1, 2, ...
... (4.50)
A - ~
- b
(iu) Substituting 4th boundary condition in Eq. (4.49),
E, = 0 = C2 C, sin Ba sin Ay [at x ~ a, Y y ~ 0 to b1
Since sin Ay '# 0, C, -:t- 0, C2 '# 0
sin Ba= 0
or Ba =mrt
where m is another constant, m= 0, 1, 2, ...
B = m1t ...(4.51)
a
Now the complete solution is given by, Eq. 4.49
Ez = C2C• sin Bx sin Ay
where A and Bare as in Eqs. 4.50 and 4.51.

i.e., E, = C2C• sin (':'} sin (: } · e-r< · e•fe>t

where, e- ri = propagation along 'z' direction

e'IJX = sinusoidal variation w.r .t. 't'
Let C = C2C•, some other constant.

,f'E. = C sin (m : }~ sin (n : )y e'"' - 'P ... ( 4.52)

Since E, is known E.r, E.1, H:c and H, are given by the following
equations [from Eqs. 4.31 to 4.34) ~ - Y\1./'A

Ex = =-1 aE, _ ~ aH,

. h2 ax h2 ay
E _ .:1 aE,
1 - h2a;- (asforaTMwaveHz = O)

✓ E, = ~ :x)cos (";,x) xsin (~Jy ei"" -rz

C( ... (4.53)

and ,,,--
E, = =-t--r
- iJE .
+1~ a H,
h dy 1z2 dX

ET.c:: Ve
-y (nn) . (mn)xcos (nbn )
b sm-;;- ye'Ull - "F ...(4.54)

H - ::.l
J:
a H.r jwc ?[J E
2 -+
- l
h iJx h2 ,)y
 97
ANSMISSIO N LINES
MlCROWAVH TR
~ ... (4 .5 5)
j;:: c (: "}m (':"]x cos ( : " ] y ~ ,.
-
H, =
cJ Hz jw e aEz
Hy _ .::.1 ox
- h ay ,.. h
2
2

~ -~ ... (4 .5 6)
j;,e c (7'.:")cos (':"]x
~l
si n ( : " }
H, =
v eg u id es
in R ec ta n g u la r W a in TM
~TM M o d es
n , w e h av e v ar io u s m o d es
n d in g o n th e v al u es o f m an d
es as TM m n w h er e m an d n ar e
if ei fe th e m o d
av es . In g en er al w e re p re se n t
w
r.
as d ef in ed ea rl ie
M o d es
V a ri o u s TMmn
= 0 an d n = 0 x an d Hy, (E qs . 4.
53 to
T M o o m o d e: rn u te d in E x, Ey, H t.
If m = 0 an d n =
0 ar e su b st it
h en ce TM oo m o d e ca n n o t ex is
is h an d
56 ), w e se e th a t al l o f th em v an
4.
TMo1 m o d e : m
=0 an d n = 1 an d h en ce TM01 m o d e ca
nnot
an is h
co m p o n en ts v
A g ai n , al l field
ex is t.
T M1o m o d e : m
= 1 an d n = 0 h an d h en ce TM10 m o d
e ca n n o t
ts v an is
el d co m p o n en
E v en no w , al l fi
ex is t.
= 1 an d n = l Hx an d Hy, i.e., T 11t
M
TM11 m o d e : m p o n en ts E x, E y,
l th e fo u r co m n , th e co m p o n en
ts ex is
N o w w e h av e al al u es o f m an d
fo r al l h ig h er v
m o d e ex is ts an d
o d es do ex is t.
i.e., al l h ig h er m a ve g u id e a s a H ig h P a
ss
ve g u id e (W
u t- o ff F re q u en cy o f a W a
~
/2
\. )V " il te r )
o w th a t
an d 4. 51 , w e k n
F ro m E q s. 4 .5 0

h
2
= "{ + ro
2
µ£, = A
2
+ B
2
= ( ': " ] +(:"]
i. e. , "( =(': " ] + ( TJ-ro'µe
y = ✓(:"] +
(:"]-ro'µe = a + jp
or
nc ie s,
A t lo w er fr eq ue
 MICROWAVE ~'-.TI RADAR ENGINEERING
98
y then becomes real and positive and equal to the a~tenuation
constant 'a' i.e. the wave is completely attenuated and there 1s no phase
change. Hence the wave cannot propagate.
However, at higher frequencies,

co"µ£>(:n J+(n:J
y becomes imaginary, there will be phase change ~ and hence the
wave propagates. At the transition, y becomes zero and the propagation
just starts. The frequency at which yjust becomes zero is defined as the
cut-off frequency (or threshold frequency) '{c'.
i.e., At f = {c, y = 0 or ro = 2rtf = 2rt{c = We

0 =(:nJ +(nbnJ-~µ£
or
We= k[(:nJ +(~nJf
or
fc = 2n~[(:n] +(;nJf

or f, = :n [(:nJ+ (nnr 1
·.·c - - -
✓µ£

f, = ~[(:J +(~Jr d
... (4.57)

The cut-off wavelength (Ac) is

or ... (4.58)

All wavelengths greater than t-.e are attenuated and those less than
t-.e are allowed to propagate inside the waveguide.
4.3.8 Guide Wavelength, Group and Phase Velocity
Here we define the ~i~e wavelength, group velocity and phase velocity
relevant for transm1ss1on of a wave in a waveguide.
Guide Wavelength ()--g)
It is defined as the distance travelled by the wav · d
'ft f '> d" . . e m or
p ase s 1 o ...7t ra 1ans. This 1s shown by Fig. 4.32. er to undergo a
h h
MICROWAVE .
TRANS.MissroN LTh.TES 99

--1 19

. Fig. 4.32
I t 1s rela ted to the Ph ase con stan t by the rela tion

Ag =~
~ ... (4.59)

. a·1fferent from the wav elen gth

ngt h in the w ave gm·a e 1s
. f The wav ele e wav elen gth Ao and cut- off
spac
m ree spa ce. In fact it is rela ted to free
wav elen gth A., by ~ = ~ - ~ (to be pro ved
in sect ion 4.3.8.5)

... (4.60)
or

a wav egu ide of any cross

Thi s equ atio n is true for any mode in
e and the cros s sect ion of the
sec tion , pro vid ed Ac cor resp ond s to the mod
wav egu ide .
Fro m the abo ve rela tion , it is clea r tha t if
Ao < < Ac, the den omi nato r
As Ao sym bol Ac, Ag incr eas es
is app rox ima tely equ al to '1' and Ag= Ao.
and rea che s infi nity whe n Ao=~- Wh en
Ao> Ac, it is evid ent tha t 'Ag is
atio n in the wav egu ide.
ima gin ary whi ch is not hin g but no pro pag

Ph ase Vel oci ty (Vp )

in the wav egu ide whe n guid e
We hav e jus t see n tha t wav e pro pag ates
ce wav elen gth Ao. Sin ce the
wav elen ght Ag is gre ater tpa n the free spa
A and f, it follows tha t in a
velo city of pro pag atio n is the pro duc t of
se velocity. But the spe ed of
wav egu ide, Vp = Ag/ whe re VP is the pha
Vp is gre ater thRn the spe ed of
ligh t is equ al to pro duc t of Ao and f. Thi s
e no sign al can trav el fast er
ligh t sinc e Ag > Ao, Thi s is con trad icti ng sinc
elen gth in the guid e is the
tha n the spe ed of ligh t. How eve r, the wav
velocity of the pha se. In fact
len gth of the cycle and Vp rep rese nts the
it is def ined as the rate at whi ch the wav
e cha nge s its pha se in term s of
the gui de wav elen gth.
v, Ag = Ag . f = 2nf · Ag _ 2nf
i.e., P = uni t tim e
21t - 21t/Ag
... (4.61)
i.e.,
 MICROWAVE AND RAUAR ENGJNHERl NC
100
. 27t
where, ro = 21tf, ~ = Ag
Since no intelligence or modulation travel at this velocity, VP is
termed as phase velocity.
Group Velocity (V1 )
If there is modulation in the carrier, the modulation envelope actually
travels at velo'city slower-than that of carrier alone and of course slower
than speed of light. The velocity of modulation envelope is called the
group velocity V,. This happens when a modulated signal travels in a
waveguide, the modulation goes on slipping backward with respect to
the carrier.
It is defined as the rate at which the wave propagates through the
waveguide and is given by
V, = dro
dp ...(4.62)

Expression for Phase Velocity and Group Velocity

/j Expression for V, : From Eq. 4.61 we know that V, = ~

Also, h2 = '( +ro'µt = A'+B' = (:"J +(~"J

and y = ex+ jp
For wave propagation, y=jP (·.· attenuation, ex = O)

"( = U~)2 = (':" J+(n:J- 2

00 µ£
... (4.63)

At f = {c, ro = <.t>c, y = O
~ µ£ = (:"J +(nn ... (4.64)

Using Eq. 4.64 in Eq. 4.63, we get

y = U~)2 = ro~µe - ro2µe
... . i' 2
= ~2 = .ro µe - ro~µE

or ~ = ✓ ro2µE - ro~µE
p = ✓µe (ro2 _ ~) = ~ ✓ 002 _ co; ... (4.65)

ro ro 1 1
VP = ~ = ✓µe ✓co2 - CO: = ✓µe ✓1 - (coJco)2

V, - C
i.e., p - ✓ 1 - ({c I /) 2
MICROW A\'I· fRANSMlSSlON l INFS
101
We also know thnt , f( nny frequoncy) :::i c~ where~ 1s
. free space 6
wave I ength nnd 1r/.' lcu
• t off-frequency) = c/Ac where A.t, ts
. cut~off
.
wavel ength

Vp == C ... (4.66)
✓1 - (~)2
2. Expression for V, : From Eq. 4.62 we know that, V6 = ~;

But from Eq. 4.65, ~ = ✓µe (ro2- ro;)

Now differenting ~ w.r.t. 'ro', we get

~ =
1
· 2 roµe
dro 2 ✓(ci-~) µe
~-
dro -
¥
✓ 1 - (rojro)2 - ✓1- (fjrf
- ¥
V _ dro _ ✓1 - (fjfJ2
or 6 - d~ - {µe

or V, = c✓ 1 -(~ J ... (4.67)

Consider the product of Vp and V1

i.e.,
2
Vp V6 = c ... (4.68)

ALITER : Now we give another method for proving VP V6 = c 2

As per earlier discussion we know that Vp is greater than the speed
of light by the ratio V,'A.o

V.p -- Ao~c
The group velocity V8 is shortened by the same ratio.

V,g = ''11
Ao
i.e., 1 C

Ag ' Ao 2
Vp Vg = Ao . C • Ag . C = C

Relation between 'Ag, Ao, an·d A.c

We know that Vp = Ag · f = fA · c
--- ----- ------- --~--- ------ ------- -- ·------.:.

MI CROW AVE AND RADAR ENGINEERING

102

C
and also VP = -✓-;:::=====2
1- (Ao I~)
_______c__ = ~ .C
✓ 1 - (Ao I 'Ac) 2 'A;,

or Ag = --;::===== ...(4.69)
✓ 1 - (Ao/ Ac) 2 .....--
Physical Interpretatio n of Phase and Group Velocity
The wave propagating between conducting parallel planes will
encounter two different velocities viz. phase velocity and group velocity.
It so happens that any electromagneti c wave has two velocities, one with
which it propagates and the other with which it changes its phase.
However in free space these are naturally the same and equal to velocity
of light.
Consider a simple, single frequency wave which has the
mathematical form
V = Vm cos (wt - ~z) ... (4. 70)
The wave is shown in Fig. 4.33 as a function of t, for
t = O and t = Af.
The wave propagates through a distance of Ag !'1t in a
time !'1t or l3 Af 1n a t'1me
(J.) • A~
UL.

J f=o

-t

Fig. 4.33

Velocity = wl~ · M
!1t
= wfA.
fJ·

or Velocity = Ag = \ f = Ag 2 ref =
w = ~ (as before)
t IV 2 re 2 re/'A.g fJ

The velocity, defined as the phase velocity, Vp is the velocity at which

the whole wavefront of equiphase moves in the z-direction.
Now consider a wave made up of two frequency components of equal
amplitude one with a frequency less than Wa and other at a frequency
slightly greater than Wo with p~ase constant~ as shown in Fig. 4.34.
 · V, = Ve sin 105
sur f ace 1s 8 d
'l 'l I . an the wav 1
''11 = '¼> sin 8. If the frequenc is ' . e ength in this direction is
velocity) with which the wave y h r it follows that the velocity (phase
the conducting surface is give~ :nges phase in a direction parallel to
wavelength 'A-p. Y th e product of the frequency and

i.e., VP= {Apor{'A.g = f ~ = l

sin 8 sine ... (4.73)
Combining Eqs. 4. 72 and 4 _73 ,

Vg = Ve sin e and V. Ve
P = sin 0
This is the same as Eq. 4.GS.
Phase velocity is the velocity with
which the wave changes phase at a
pl~ne bou":1-dary and not the velocity CX)

with which it travels along the

boundary. Another example is the
phase velocity of a wave train
I
corresponding to sea waves
approaching a beach at an angle
rather than straight in. The
phenomenon which accompanies this
is that the edge of the wave appears Q.______ _ _ _ _ _ _ ___:,i

to sweep along the beach much faster o· 30' 60' 90'

than the wave is really travelling. It ---e
is the phase velocity that provides Fig. 4.38
this effect.
A plot of VP and Vg is shown in Fig. 4.38 as a function of wave angle.
It is seen from the plot that as 8 decreases,
I. VP increase and becomes equal to oo at 8 = 90°
2. Vg decreases and becomes equal to zero at 0 = 90°.
Thus the group velocity with which the energy is actually
transmitted can almost be equal to 'c' the velocity of light.
At cut off frequencies, VP becomes infinite and Vg becomes zero. This
means that there is no propagation of energy along the waveguide. The
wave simply bounces back and forth between parallel plates and normal
to them.
TM Modes in Rectangular Waveguides
Consider the various TMmn modes for various values of m and n.
1. TMn mode : Minimum possible mode m = 1 , n = 1.