noc19-ee21 Lecture 26-Modes of Rectangular Waveguides

IIT Kanpur July 201
20 followers
1 Video View·Dec 3, 2023
Modes, rectangular waveguides, field distribution, TE mode, TM mode, medium wavelength, operating frequency, cutoff frequency.
[Music]
[Music]
hello and welcome to nptel's MOOC on
electromagnetic waves in guided and
wireless media and we will continue the
discussion of rectangular waveguides
that we began in the last module after
going through several steps I would like
you to recall that we stopped at this
particular equation in which you had
this part you know X double prime of X
by X as being a function only of X then
the next term y double prime of Y by Y
is a function only of Y just to remind
you X double prime of X actually means
that it is the second derivative of the
function X of X with respect to the
variable X itself okay and then you have
this other term k0 square epsilon r
minus beta square as a constant okay now
we said that one of the ways in which
this equation or rather the way in which
this equation can be true for all values
of X and for all values of Y is that
this separately be equal to some
constant which we will call as minus KX
Square and this we will call it as minus
KY Square and let this also be another
constant okay so what we now have is the
sum of three constants which will be
equal to 0
this actually imposes the limits on beta
ok of course we still haven't decided
what would be KX and KY they will be you
know obtained once we have applied the
boundary conditions ok so because we
haven't yet done that so for every term
here is unknown except that k0 Square
and epsilon R are known to us because
epsilon R is the material with which we
have filled the waveguide and k0 is
equal to 2 pi by lambda and lambda is
the operating wave length of the mode
that we are trying to propagate inside
the waveguide so to sum up you have K
minus KX square minus KY square minus
beta square plus k0 square epsilon R is
equal to 0 dear
changing the equation you can write for
beta as square root of k0 square epsilon
r minus KX square plus KY square okay so
the sum which is KX square plus KY
square can itself be redefined so we can
redefine this KX square plus KY square
as some KC Square okay if you wish and
then the equation for beta will be
square root of K 0 square epsilon R
minus KC square you can even simplify
this k 0 square epsilon R by writing
this as K square ok with the assumption
that the permittivity of the material
has been absorbed into that definition
of K itself okay so with that beta can
be simply written as the square root of
K square minus KC square but again
please remember we don't know what KC's
and what is beta of course to know that
we have to know what sort of fields can
we write for the electric field
component Z which is what we are solving
here and from the knowledge of ez plus
the knowledge of headset you should be
able to then write down the other field
components and then apply the boundary
condition and hope is that if everything
goes well then you will be able to
determine what is KX what is K Y and
hence determine what is KC
and therefore determine what would be
beta for a given wavelength or for a
given mode that is propagating okay so
we will first look at this easy we have
two equations to solve for you have X
double prime by x equals minus KX square
and y double prime by y is equal to
minus KY square from your classes on
differential equations you know that
these two are two partial differential
sorry two ordinary differential
equations of second order and the
solutions can therefore be written as
sum a cos KX X plus B sine K K xx where
a and B are constants which we do not
know yet but which is fine then you have
Y of Y which is obtained by solving this
second you know a differential equation
and that would have additional constants
so we will call this as say C cosine k
YY okay
because this is the solution of the
second equation and B sine K by Y okay
so together you can write for the Z
component of the electric field as a
function of all three components X Y and
said as a cos KX X plus B sine K Y sorry
K X X multiplied by C cos KY y plus D
sine K Y times y and what about the
propagation part yes you have to write
down the propagation part as a power
minus J beta Z it is interesting that
after we have solved these two partial I
mean these two differential equations we
end up having more unknowns than what we
actually started off with earlier
however boundary conditions will come to
our rescue we can apply one boundary
condition right away here okay
recall the cross section that we have we
have the cross section in terms of this
x-axis and y-axis right and you have two
walls one wall is X equal to zero wall
and the other wall that you have is X
equal to a wall okay and then you have a
bottom wall which is y equal to zero and
the top wall which is y equal to B okay
the waiver of course is of cross section
e cross B it usually a greater than B
okay now with this one and we are
looking at the easy component what kind
of a component will be this easy when it
you know it's at the two walls now we
know that the tangential components to
the X equal to a wall as well as for X
equal to zero wall will be the easy
component as well as ie y component you
have to imagine that I am stretching
this guide along the z axis in this
manner and therefore easy it is kind of
gliding along with this wall so if you
imagine that this is the wall then
either it is just gliding along the wall
and then e Y is increasing along this
particular direction so you have this
easy and ey both will be tangential to
this will X be tangential no e X will be
pointing this way right where the pen is
pointing obviously this will not be
tangential to the X equal to 0 wall
rather this will be
perpendicular to the wall right similar
arguments will tell you that e^x again
cannot be the tangential component
however Eve I component you can see that
this is e Y and that Eve I can be a
tangential component as well as ez is a
component tangential similarly for the y
equal to b wall the tangential
components will be e x as well as easy
okay so you have both TX and easy and
now we don't know yet what is Eve I and
II X because we don't know what is head
said at this point so we can apply
boundary condition for easy at these two
or rather at these four walls and see
whether we are able to get something out
of it right okay let's apply the
boundary condition at X equal to zero
wall write the boundary condition will
be that he said must be equal to zero
because this is a tangential electric
field component and we don't have you
know and it's a perfect electric
conducting wall
therefore the tangency electric field
component must go to zero and this is
the boundary condition right with EZ
equal to 0 at X equal to 0 let's go to
this expression that we have written and
then say at X equal to 0 what will
happen we won't touch the y part we
won't touch the Z that is e power minus
J beta Z but immediately when you put X
equal to zero this cosine of zero will
be one a will be present whereas sine KX
times zero which is sine of 0 will be
equal to zero right and because this
entire term must be equal to zero and
the conclusion is that a must be equal
to zero okay now we will also apply a
boundary condition at y equal to zero
wall which again says that easy must be
equal to zero and then following the
same arguments you can show that C will
be equal to zero these are not very
complicated to find you just have to
substitute the value of x and y in these
expressions and then see which component
has to be made equal to zero right so
when you do that you will easily find
out that a is equal to zero and C is
equal to zero and with that done right
you can eliminate these terms from the
possible solution set interesting right
and so that the overall solution that
you now have for easy will be easy of XY
and Z will be some B times
d which itself could be a constant so we
will have to call this some other
constant so what shall we do we will
call this as some easy row ok
so easy Rho is a constant which is
basically the product of B into D
because B and D both were constants we
call the new constant as easy Rho and
then what you have is sine K X x times
sine K Y Y right so if you take Y is
equal to constant plane and then
actually plot what you are going to get
for the electric field as a function of
Z you would see that this electric field
would actually go something like this it
would be sine so it would be zero at the
center and it would actually go to sorry
it would be zero at the edges and it
would go to a Maxima at the center if
this is the lowest half wavelength that
we can fit okay of course you can also
have additional type of solutions in
which case you will have this one you
can have even more number of points but
all the time please understand that as
you move along X the solution will be in
the form of a sinusoidal wave with it
going to 0 at the two edges and having
multiple Maxima in fact you know that
this is this kind of a mode function was
exactly what we had for a parallel plate
waveguide okay that we tackled in the
earlier module and there we gave
different mode numbers right so if you
had only one Maxima it was te one if you
had two Maxima it was te two and so on
and so forth right so in much of the
same way you can actually have a
numbering scheme for this one as well
except that you now have to field two
dimensional field distribution is not
just it is varying along X but because
there are top and bottom walls there
will be a sinusoidal variation along Y
as well right which is what you can
obtain or which you can see from this
sign K by Y term right so there will be
a sinusoidal variation in the X Y
direction there will be sinusoidal
variation in the X direction therefore
all modes whether we are dealing with
the transverse electric mode or
transverse magnetic mode they will come
with two numbers with them right these
numbers can refer to the variations
along X and this n will refer to the
variations along Y and in contrast to
that parallel plane waveguide you now
have to specify every mode by two
numbers this
not something new for the fiber we have
already done that so you had for the
fiber lp0 one more LP 1 1 mode LP 2 1
more and so on
so therefore this makes sense that you
actually have two different you know
indices to go with whether you are
dealing with the TE more or a TM mode ok
so anyway with that as easy overwrite
let's also put down E power minus J beta
Z but because the C power minus J beta
there is a common thing I won't carry it
further ok so I will just remove this
one but please remember that that would
always be there ok so we have found easy
and we have seen that it would be in the
form of sine and sine can we try and
find out what would be the solution for
H said C the Helmholtz equation for head
said will be exactly the same equation
as that off for the easy component so
you have del square head said plus
k-square right h z is equal to 0 I can
write this del square as del square by
Del X square plus del square by Del Y
square minus beta square so the equation
essentially goes back to the same thing
so in following the same steps and
removing the Z dependence or suppressing
the Z dependence H said as a function of
x and y would be again some a cos KX x
plus b sine KX x times c cos KY y plus D
sine K Y Y right okay now you have to
apply boundary condition now here is
where it gets a little bit of a tricky
thing because there is no boundary
condition for the tangential H said when
there is a conductor right because you
have a conducting wall here right if you
consider H said component here then
there will be a current sheet that would
be present which would be along the Y
direction right which would be along the
Y direction in this particular case okay
so and you will of course have to also
imagine that there will be an edge wipe
along a particular wall and because of
that H Y there will be another current
sheet right so basically there will be
this surface currents okay
which would be propagating or which
would be present or which would be
indeed
on the walls and because those currents
are present you cannot take the time
mention magnetic field components to be
equal to zero in the case of a you know
perfect electric conductor to dielectric
interface something that we already have
seen in the equation when we read when
we wrote head and cross H 2 minus H 1 to
be equal to K where K was the current
sheet and hedge 2 minus H 1 was the
difference in the tangential electric
field of course you take H 2 equal to 0
but you still have to deal with the fact
that the boundary condition for the
tangential electric field is now
discontinuous or predicts that there is
a discontinuity in the magnetic field
and this discontinuity is actually taken
up by the current sheet K okay depending
on whether you are dealing with headset
or hy or headshakes the direction of the
current sheet will also differ because
there is a curling operation out there
right so will not worry about that but
the main point to note here is that I
can't make head set equal to 0 at any of
the two walls then what should I do
well it turns out that since H said is
or since the derivatives of headset del
headset by del X and del H Z by Del Y
are respectively proportional to some ie
Y as well as e X component depending on
the wall you have to invoke the
derivative being equal to 0 right
because e Y and DX are tangential
components vo is tangential at X equal
to 0 and at X equal to a II X is
tangential at y equal to 0 and y equal
to B so based on which wall you are
looking at you take the derivative of
headset whose expressions we have
already turned down and then set that to
0 okay to eliminate the constants a b c
and d and then to simplify the solution
as before i will leave this as an
exercise to you so i am going to not
deal with this but it is important that
you actually carry it out okay
so what I wanted to say is that compared
to how easily we were able to evaluate
he said you know and eliminate certain
constants it is not so simple to do that
for H Z keys okay however we are not
done with the easy yet I mean we did
apply the boundary conditions at X equal
to 0 wall and attic
at y equal to zero wall but we still
have two additional walls to work out
that is X equal to a 1 and y equal to B
1 right so let us apply the boundary
condition to X equal to a and y equal to
B for this expression okay so when you
write down that so let us go to a new
one so when you write down the boundary
condition at x equal to a wall and
demand that ez must also be equal to 0
what you are actually demanding is that
this sine KX x evaluated at sign K X X
evaluated at X equal to a be equal to 0
okay similarly when you write down
boundary condition at y equal to B wall
for easy and say that either it must be
equal to 0 the condition on the y
dependent function would be sine K Y B
must be equal to 0 now when can sign of
a function or sign of an argument go to
0 whenever this argument I mean whenever
this is a multiple of Pi sign will go to
0 because sine 0 is 0 sine PI is 0 sine
2 pi is 0 sine 3 pi 0 and so on right so
M is an integer of course M must not be
equal to 0 because when M is equal to 0
every field will be actually equal to 0
okay so that condition we don't allow
similarly sine K by B equal to 0 implies
that K Y B must be equal to n pi right
from which you can immediately write
down what is K X which is M PI by a and
K Y will be equal to n PI by B and
therefore beta can be obtained even
without finding the other field
components in a very simple manner as K
square minus M PI by a whole square plus
n PI by B whole square ok so this is
very important thing to note that and we
can also write down beta as 2 pi by
lambda G to denote ourselves that lambda
G is what is called as the guide
wavelength okay that is the wavelength
along the Z direction in the waveguide
whereas K which would be in general
given by 2 pi by lambda or 2 pi
by lambda where lambda can be called as
the medium wavelength right so if the
material is filled if the waveguide is
filled with air then lambda will be
equal to lambda 0 that is it's free
space value otherwise lambda will be
filled by lambda divided by epsilon R
our lambda 0 divided by square root
epsilon R we can kind of find this thing
out right so you have found beta but
again because you have these different
values of M and n possible you put a
subscript on em and you put a subscript
M N and for beta as well indicating that
not just one more is possible but there
are many modes which are actually
possible for us to find right there's
another relationship that one can
actually obtain because K is actually
equal to Omega square root mu Epsilon
okay
where Absalon will be equal to epsilon 0
epsilon R and this M pi by a square plus
n PI by B Square can also be rewritten
so we will do that one so beta can be
written as Omega square mu epsilon minus
Omega C square mu Epsilon okay so I can
write beta as Omega square mu epsilon
minus Omega C square mu epsilon where
Omega C square of mu epsilon will be
equal to M PI by a whole square plus n
PI by B whole square
and obviously beta will be positive you
want beta to be positive otherwise modes
will not propagate but they will simply
decay out so beta has to be greater than
zero indicates that Omega has to be
greater than Omega C okay and what is
the Omega C well again Omega C also
needs to be written with the subscript M
and n because you have two possible
values of M and n that would satisfy
this particular equation okay so this
expression Omega C square mu epsilon
which is equal to this thing can be
rearranged to find out what exactly is
Omega C Omega C will be 1 by square root
mu epsilon M PI by a whole square plus n
PI by B whole square under root and
because I don't like Omega I like F I
can even write this as FC as 1 by 2
by square root mu epsilon times square
root of this so given a given B right
you can keep changing m and n to
determine the cutoff frequencies for
different modes because the
corresponding frequency s at a
particular mode which you want to
propagate must always be greater than
this minimum value FC okay
because with F is less than FC the
component here or the term here Omega C
square mu epsilon will actually be more
than Omega square mu epsilon causing
beta to become complex but because your
solutions are X the form E power minus J
beta set and beta becoming complex means
that the waves actually decayed down as
you propagate along the waveguide you
don't want that you want ways to
actually propagate through the waveguide
so for that to happen beta has to be
positive and for positive value of beta
Omega must be greater than Omega C and
because Omega C itself is different for
different modes at a given frequency
only a certain number of modes can
actually propagate okay so you can
actually cut off many different
frequencies for example let's say we
find out what are the two fundamental
modes in this waveguide I haven't talked
about it but we talked about it later on
so I will find out two fundamental I
mean two modes of this one where the
first one is the first value the
smallest value of FC that you can get
for M and N and then there is a next
mode okay whose value FC also will find
out now if you want your wave way to
operate only in the single mode regime
that is it must carry only one
particular mode from point A to point B
then you can actually choose the
frequency F to be such that it is
greater than the first mode which you
want to propagate but it must be less
than the mode which you do not want to
propagate however if you want a lower
order mode to not propagate by lower
order we mean the mode which has a
lowest cutoff frequency if you don't
want that to propagate there is nothing
that you can do because as you increase
the frequency anything that would be you
know that mode would be present whose
cutoff frequency is actually less than
the operating frequency will begin to
propagate so in that case we say that
the waveguide is multi modal and multi
mode in many cases is not a very good
idea because the it will not
the energy and splitted according to
different modes in in certain manner and
the propagation also will be a little
bit of a complicated thing so we don't
want that but sometimes you can't help
it okay so most of the times you design
your waveguide in such a way that for a
given frequency range it should
hopefully operate only as a single mode
thing and you can design such wave
guides in the exercises that we will
give you okay
having said that whatever we have said
so far in terms of beta MN right in
terms of the cutoff frequency M and N
and even the waveguide lambda G which
you can find out by writing down so beta
was actually equal to 2 pi by lambda G
and since beta MN you know you can also
find out what is the corresponding guide
wavelengths these are all
characteristics of the waveguide
geometry they do not depend on whether
you are dealing with a TE or a TM mode
okay of course once you fix a value of M
and n you have to go back and evaluate
whether that particular te field
components are nonzero or not for
example the lowest order mode in this
particular waveguide is what is called
as te 1 0 which we will study in detail
later on
ok there is no corresponding TM 1 0
because when you look at what would be
the expressions for transverse magnetic
1 & 0 you will find that electric field
component is it will be 0 okay and
because of that all the other components
will also be equal to 0 so this
condition does not really occur in the
sense that although the cutoff
frequencies for TE 1 0 and TM 1 0 are
the same the fact that there are no
field components which are nonzero for
the TM 1 0 means that that more would
not exist
however this p 1 0 mode would exist and
in fact it has the lowest cutoff
frequency that is this would be the mode
that would propagate right after the
frequency F is greater than F c10 okay
so this is the mode that would always
begin propagating first before the other
modes can begin to propagate and as such
this mode is called as dominant mode
okay
we will see what the field expressions
for this dominant mode are but what I
wanted to point out is that all the
other characteristics that we saw but
all
independent of whether you are dealing
with the te mode or a TM mode
okay now without too much of a
derivation I would simply like to give
you the field expressions in the
transverse magnetic mode okay when I say
transverse magnetic mode okay I am going
to assume that H Z will be equal to zero
okay causing only easy to be nonzero and
because of that you can actually still
see what would be e Y E and H Y and H X
easy we have already returned down easy
will be sine and sine and then what
would be e Y well I had asked you to
look at the expressions for this you
know
ey e X and other things right so ey will
be equal to or will be proportional to
del e Z by Del Y similarly e X will be
proportional to del Y Z by Del X okay so
the field components for Eve I will be
sine K X X but it would be cosine K Y Y
whereas for the X component it would be
cos K X X and sine k YY okay please
verify this this is a big exercise okay
which follows the results of which
follows from earlier solutions of the
earlier exercises in the assignment or
in the handout we will actually give you
the complete derivation of this equation
so that you don't have to you know run
around finding the muddy revelations but
I strongly suggest that you carry out
exercises as I have suggested then you
will be able to find these expressions
for yourself okay what I want to tell
you at this point is that well hy + HX I
will again leave it as an exercise for
you what I want to tell you here is that
are these solutions sine KX x and cos K
YY plausible from a boundary condition
point of view right to do that let us
write down our cross section again so I
have X equal to 0 wall X equal to a-1
right now look at ey ey is a tangential
component at X equal to 0 it is also a
tangential component at X equal to
meaning that it has to go to zero at
these two points the function that would
make it go to zero at these two points
would be a sine KX X that is as a
function of X the ey component must go
to zero at X equal to zero and X equal
to ay consequently it should behave as a
sine thing however at y equal to 0 and
at y equal to a eva is not tangential
here is normal okay and this normal
means they need not necessarily go to 0
which is captured by this cos K by Y
well the argument is very strong for the
sine part not so much for the cosine
thing but anyway we'll take it to the
case that when the boundary conditions
are not asking specifically the fields
to go to zero then the possible solution
is a cosine wave not a sine wave right
and you can show again by following this
kind of a logic that e^x should go to
zero at both y equal to 0 as well as at
y equal to B and therefore its solution
in terms of sine k YY is valid right
because e x must also go to 0 at the two
walls bottom and the top ones okay so
these are the field equations for TM
mode we will have to write down the
field equations for TE mode but for
Timor the solution is slightly
complicated or longer it's not
complicated I'm sorry it is basically
just longer because I have to first find
out H said set the derivatives of
headset with appropriate x and y
derivatives to 0 at the appropriate
walls and then obtain the other field
components and then go ahead and find
out the corresponding field solutions or
the form of the fields ok you can do
that in the next model I will give you
the expressions and you can verify that
you have actually done it correctly by
looking at your expressions and
verifying it with the expressions that
we will give but what is important in
these two aspects is that most of the
times you can actually write down the
solution by looking at the boundary
conditions or at least after you obtain
the solutions you can verify that they
actually satisfy boundary conditions and
therefore the solutions are at least
correct in terms of boundary conditions
ok so we will continue the discussion of
te modes and as well as point out couple
of things in terms of T and TM mode
in the next module thank you very much
[Music]
[Music]
[Music]
