noc19-ee21 Lecture 25-Rectangular Waveguides

IIT Kanpur July 201
20 followers
Dec 3, 2023
Rectangular waveguides, perfect conductors, longitudinal component, transverse component.
[Music]
thank you hello and welcome to nptel
mooc on electromagnetic waves in guided
and wireless media in the previous
module we derived an expression for the
angle Theta 1 at which you have to send
in the electromagnetic waves onto the
top plate right or into the plate at x
equal to zero and we found that this cos
Theta 1 was actually something like n
Lambda by 2 a right for a given Lambda
and a ah the corresponding Theta 1 which
would you know Theta 1 we will also make
it Theta 1 n because n you know can be
an integer which will take on different
values and for each n you will have
either a te 1 or tm1
T2 tm2 and so on as these waves okay
however which waves would press would be
present and which base will not be
present you have to actually go back and
find out from the electric field and
magnetic field Expressions themselves
okay so just because you have some t e
ten it does not mean that you will
always have a t 10 you have to also you
know find out whether you will have that
particular mode by looking at the
electric field and the corresponding
magnetic field expression of course in
this case there is only one in sorry
this is T one not t ten yeah t 10 not
yeah t e ten is a tenth mode right so
that is a tenth mode that we we thought
of for the transverse electric case so
whether this exists or not you have to
go back to Maxwell's equations and I
mean go by to the electric field
expressions and magnetic field
expressions and then determine that
indeed that particular mode can
propagate ok now there is a bit of a
curious curious thing that is going on
here
what we thought or what you would expect
is that if you have a plate at say x
equal to zero and then you send in this
electromagnetic wave at an angle Theta 1
right this wave gets reflected and then
you are going to get all those you know
things that we talked about
interestingly not all values of theta 1
are allowed to give you the guided
conditions right in other words when I
fix both x equal to 0 as well as x equal
to a that is if I fix a then not every
angle Theta 1 will be you know able to
generate this guided waves along the Z
Direction This of course is the Z
Direction not every possible value of
theta 1 is allowed why because Theta 1
has to satisfy this particular condition
so when you fix Lambda and you fix a
right only the multiples of this angle I
mean the cosine of the angle should be
such that only the multiples of this
should be allowed right so that multiple
value we have taken it to be as n right
there is also another Point here so yeah
sorry let us finish this first point the
first point means that if I change Theta
1 by a small value here and if this
angle the new one that we have written
in terms of the green line does not
satisfy this equation then that
particular Theta 1 or the
electromagnetic wave which will be
incident at that angle Theta 1 will not
be propagated it would be attenuated and
therefore there is no propagation for
that particular angle Theta which with
respect to the green one which does not
satisfy the condition that we have put
down for angle Theta one so it is clear
that not every value of theta 1 is
allowed there is a certain minimum value
to a maximum value of theta that is
allowed of course the simple case when
Theta 1 can be equal to zero so in that
case what would happen the K Vector
would go here the K vector would come
back the electric field incident and the
reflected wave would be bouncing back
and forth they would actually create a
nice standing wave pattern between the
plates but this standing wave pattern
does not really propagate ok so in other
words if you have Theta 1 equal to zero
it corresponds to shooting up the
electromagnetic wave to the top plate
and then getting reflected then this
would simply be going back and forth
without any propagation along the Z
Direction so this is no good for us what
you want is a sort of oblique incidence
with an angle Theta one such a way that
there will be propagation of course what
would happen when you try to take the
electric field and then just shoot it
straight across along the Z Direction so
if you shoot directly along the plate or
the Z axis it is possible for you to
have this propagation done but then
getting electromagnetic waves you know
is propagating in this straight line
with the electric field being in that
kind of a configuration the magnetic
field essentially becomes that of a
transverse electromagnetic wave that is
it essentially becomes a uniform plane
wave and you can show later on when we
talk of dispersion sorry diffraction
that it is not possible for us to
actually propagate this electromagnetic
waves in this direction given by the red
lines here ok so the
possible angles would lie from whatever
the angle Theta 1 that satisfies that
equation and the minimum value of that
angle will be Lambda by 2 a that is
Theta 1 will be equal to inverse COS of
Lambda by 2 a and to a certain maximum
value okay so what could be the maximum
value that you will have say for example
I start increasing the value of n here
ok so I go to so if I fix Lambda by two
a I will fix that particular thing and
then the next one will be 2 times Lambda
by 2 a which is fine and next will be 3
times Lambda by two a and so on and so
forth right and please note that this
term in the bracket remains constant
whereas the term that are getting
multiplied is increasing so at some
particular Point OK you will have the
multiple Lambda by two a to be so high
so let us call that as some capital N
such that this cosine Theta or rather
this n Lambda by 2 a actually becomes
equal to one and cos Theta being 1 means
that Theta is either reducing to zero
right so Theta is actually reducing to
zero in which case you will simply have
plus you know just going back and forth
in this particular case and therefore
there won't be any ah propagation of
this one right and you cannot have n
plus 1 times Lambda by two a because n
Lambda by two a itself has become one so
cosine Theta has become equal to 1 and
Theta has become equal to 0 in this
particular case whereas when you go to
the next value available for n that
would actually make cosine Theta greater
than 1 in which case there is no
propagation right so you will actually
have from set of values of theta from
certain value which is at 0 all the way
up to the maximum value of theta which
is say or cosine Theta which would be
given by this particular value right so
these set of values that you are going
to have for Theta one they all can be
guided but they all correspond
correspond to different modes okay and
moreover because Lambda by two a is kind
of a minimum value at this point then
you will actually see that this fellow
will actually act as a filter right so
in terms of the frequency ah lower
wavelength would correspond to higher
frequency in terms of a wavelength that
would be the other way around ah in
terms of frequency when you consider
there will be a minimum frequency Beyond
which only the mode will actually be
propagating right so this this parallel
plate waveguide actually performs the
job as a high pass filter kind of a
thing it will only allow those
frequencies which are greater than this
minimum frequency as determined by
Lambda and a to propagate and moreover
not every frequency range I mean every
frequency greater than this minimum
frequency can propagate only those
frequencies which correspond to the
angles of theta one that are allowed
there is those discrete values of
frequencies are only allowed OK so you
will not get a continuous range of
allowed values of theta 1 or omegas you
will actually get this the discrete
values ok so this is very important to
note down OK all right so we have said
many things about the parallel plate
wave guide let us note one of the
problems with parallel plate waveguide
well the problem with parallel waveguide
is that you have to have these two
plates and these two plates have to be
quite wide apart right so even in the
more practical scenario where you will
actually take two big square plates or
maybe rectangular plates that you take
you still see that it is not closed at
any other Junction right there is no ah
you know either you have to fill in the
material in between and outside what
would happen is which we haven't
discussed but outside what would happen
is there will be electric field lines
just fringing off the edges ok so the
electric fields are confined but not
completely confined to our ah
satisfaction so we want to have coverage
or complete the ah plates you know
complete guidance what we can do is just
as we consider two plates here you can
consider two plates at the side as well
OK and then let the wave propagate in
between OK structures such as those are
called as rectangular wave guides in
general there will be one dimension
which is larger than the other and
please note that all these walls are
perfect conduct we are assuming that
these are all perfect conductors this is
simply a cube or a rectangular piece of
hollow metal which is protruding along
the Z direction or extending around the
Z Direction such that waves are
essentially propagating or modes are
propagating along that Z direction of
course this structure may look simple
from the point of view however the
analysis of this waveguide is slightly
more complicated than the panel plate
waveguide we will look at that analysis
it is interesting for couple of reasons
one it will tell us that ah there is
there is a one it will tell us how to
solve such equations when you have more
than one variables ah we haven't done
that in the optical fiber case where we
should have done it however in this case
we want to do it because this is much
more more this is simpler not as
complicated as the optical fiber and
this also has an additional thing to
tell us that the mode functions that you
look can actually be written after you
have solved the equations we will show
that the mode functions could have been
written down directly from the knowledge
of the boundary conditions itself OK
however the mathematical tricks that you
will be coming across when solving these
type of equations will be very useful
for later studies as well therefore we
will look at this particular wave but in
somewhat more detail than we have looked
at the parallel plate waveguide our
starting point will be the cross section
so you mark two axis X and Y we will
assume that one plate here is at x equal
to zero and the width along X is x equal
to I mean is is a and you have y equal
to zero and y equal to B OK so you have
this four plates which or four lines you
can think of and these lines are
essentially the corners of the wall so
its basically a rectangular thing for
normal operation a is greater than b and
that is what we will assume when we are
writing this ah equations for this
please note that these are all perfect
electric conductor so I am going to just
indicate them by drawing this small
hatches around that ok inside we will
assume for now that there is nothing
filled but if you want you can also fill
up the waveguide with some material
homogeneous linear isotropic material is
what we will assume with a dielectric
constant Epsilon R of course this is the
cross section and this is how the
waveguide actually looks ok so if you
concentrate on this one this is how the
waveguide looks and this of course is
the walls that you have so these are the
walls that we have and the wave is
actually propagating inside through this
ok
so how do we go about writing Maxwell's
equations Well a true analysis of this
would actually be quite difficult for us
but nevertheless we will look at it the
there are a few steps that you need to
follow and these steps can be followed
for any wave guide as well even the
dielectric waveguide can be Revisited by
looking at the steps as well as the
optical fibers can be analyzed using
these same steps okay the first step is
to write down Maxwell's equations and
then Express the transverse components
which is e x t y h x and h y in terms of
e z and h z okay if the geometry were
not to be you know Cartesian or
rectangular geometry if it was a
cylinder that you are looking for so you
can think of a hollow cylinder right so
just a pipe and a pipe is being extended
along the Z Direction so that kind of ah
geometry would involve e r h r e Phi and
H Phi to be expressed again in terms of
e z and H set OK of course if you take a
round ball then you do not have e z and
H sir and the corresponding analysis
becomes very complicated luckily we do
not go into that one these two are more
common geometries that we will be
dealing with okay so it is important to
note that in the transverse plane
whatever the transverse electric field
and magnetic field components that are
present can be expressed in terms of the
longitudinal or the axial components e z
and H set OK the Second Step involves
solving Helm holds equation
for e z as well as for head set OK the
result of solving helmholtz equation for
e z and HZ will result in some
constants that must be chosen OK of
course you cannot go and choose them in
an arbitrary manner you want you will
have to apply boundary conditions and
then choose the constants okay so after
you apply the boundary conditions and
then choose this appropriate constants
you can then complete the picture by
calculating the propagation constant
remember that is one of the things that
you are actually looking at OK and the
field components that you want to find
out so you can do both you can calculate
the propagation constant which is what
would be very interesting and very
important for you ah which will tell you
what should be the values of Lambda what
should be the values of Omega for a
given structure for it to propagate now
from our knowledge of parallel plate
waveguide you do expect that you are
going to have discrete set of omegas
that are present and only those
frequencies which are greater than a
certain frequency called as the cutoff
frequency for a given mode would be
allowed to propagate OK our job is to
find out what the cutoff frequency is in
terms of A and B as well as to find out
what additional modes are going to be
present in this particular waveguide
structure okay so you have this four
step procedure I encourage you to follow
this four step procedure and apply this
step four step procedure to the case of
dielectric slab waveguide as well the
steps would remain the same except for
the boundary conditions in the metallic
case or the medium surrounded by the
perfect electric conductor boundary
condition tells you that total
tangential field on that boundary must
be equal to 0 whereas for the dielectric
to dielectric ah region what it means is
that e tan one must be equal to e tan
two that is to say the tangential
electric field must only exhibit
continuity and not really the ah going
to zero there okay in fact that is the
reason why you got this evanescent waves
in the dielectric case for the magnetic
field case the magnetic field continuity
would tell you that H tangential
component will not be zero in fact that
would actually be be equal to the
current sheet remember this is a
magnetic field here and the source of
the magnetic field can be the currents
of course I am assuming no free currents
but these are the induced currents
because of the ah you know field
structure that are present here so the
tangential magnetic field is not going
to zero at that at this condition for
the dielectric case the tangential
component is basically equal to the
tangential component on the other side
ok so this is for the metal case and
this is for the dielectric case ok so
these differences subtle differences in
the boundary condition actually tell you
the mode functions going to be slightly
different in case of a metallic
waveguide metallic rectangular waveguide
as well as that of a parallel plate or a
dielectric waveguide as well ok so let
us begin with step ah one I will show
you how to solve ah you know for one
particular component I will leave
finding out for the other components as
an exercise to you we start with two
Maxwell's equation is written Side by so
one equation will be the
um Faraday's law so Del cross e is equal
to minus J Omega mu naught H I have
already gone into the phasor domain so I
hope this equation is you know you you
know this equation very well and then
you have Del cross H is equal to J Omega
Epsilon times e OK I will assume that
material inside is filled with air so
therefore I have J Omega Epsilon 0 e if
the metal is not filled with air then
you will have J Omega Epsilon times e
Epsilon will be equal to Epsilon R
Epsilon zero okay so that you keep in
mind ok so these are the two equations
that I have and now using rectangular
Cartesian in the rectangular Cartesian
coordinates I will expand this Del cross
e and Del cross H operators so I will
have X hat y hat Z hat Del by Del X Del
by Del y
Del by Del Z you have e x you have e y
and you have e z please note that in
general this X Y and Z can actually be
functions of x y z as well as on time t
or on the frequency Omega okay on the
right hand side of course you have minus
J Omega mu naught h x minus J Omega mu
naught h y and then minus Omega J Omega
mu naught h z okay so you have to equate
each of those X components to the
corresponding right hand side x
component y component to the Y component
Z component to the Z component
we also know that we want propagation of
waves along the Z Direction meaning that
in the Z direction we want them to be in
the form of E power minus J beta Z
therefore Del by Del Z operation can be
replaced by writing it as minus J beta
of course beta is unknown to me we will
have to find that out well that is where
the step number four will be since we
are in the step one we will simply put
this down as minus J beta and then find
out what beta will be in later on okay
so in this expression here instead of J
Del by Del Z I can go ahead and write
this as minus J beta okay so now let me
simplify this equation for the X case
you will have Del e z by Del y minus or
rather plus J beta e y this is for the X
ah component of the curl on the left
hand side and on the right hand side
this would be minus J Omega mu naught h
x ok
and the second one for the y y will be
minus Del e z by Del X
plus J beta so this is that ok let us
rewrite this correctly so I will have
for minus y hat so minus
ah Del e z by Del X Plus J beta e x okay
this would be equal to minus J Omega mu
naught h y cancel of minus from both
sides and remove it okay so that I have
done
and you can write down the third
equation I will leave it as a exercise
for you and the other set of equations
which come from curl of H can be written
in the same manner okay so you will have
Del h z by Del y ah plus J beta h y
because even the H component in the Del
by Del Z has to be replaced by minus J
beta right this would of course be equal
to J Omega Epsilon note that there is no
minus sign here so you have J Omega
Epsilon times e x OK the other equation
that you are going to get will be ah Del
h z by Del X Plus J beta h x should be
equal to minus J Omega Epsilon e y this
minus comes because in the previous
equation you could cancel out or the
corresponding equation you could cancel
out the minus sign on both sides here
you cannot cancel out the minus sign so
this is the equation that you get the
third equation also I will leave it as
an exercise for you to write
now notice this this expression ah in
this expression oh let us look at this
expression in this expression I have e z
by X which is ok Del e z by Del X I have
X Co e x component here and a h y
component here right however if I look
at this equation the cross equation here
this h y can be rearranged to write it
so we can recurring this equation in
this manner I can write this as J beta h
y being equal to J Omega Epsilon times e
x minus Del h z by Del y right and then
divide by J beta on both sides to get
divided by J beta 1 by J beta OK this
would be h y which I can substitute back
into this expression ok so now doing all
these tricks right what you will now get
is that Del e z by Del X Plus J beta e x
is equal to J Omega mu naught times h y
right you write that one J Omega Epsilon
by J beta
e x
OK you have to multiply this and then
you have minus J Omega mu naught by J
beta divided times Del h z by Del y
right so I can simplify this equation
slightly ah in fact what I want is I
want to combine these two equations
together so that I can express e x in
terms of e z and H set or in terms of
their derivatives OK so you combine
these two so before combination you know
you can ah yeah you can write it in this
manner so since writing this one will be
difficult if I go to the next slide what
I will do is I am going to write this
simplification in this present slide
itself I hope you have copied down all
this and what you will get from here is
that you have Del e z by Del X and then
you pull this y component onto the ah
well let us not will put the Y component
part here so you have omega mu naught by
Beta the JS here I have canceled and
then you have Del H set by Del y this
should be equal to
you will have minus Omega Square mu
naught Epsilon divided by Beta
OK times e x minus J beta times e x so
this I have just moved it on to left
hand side right hand side just to get
this equation ah in a simplified manner
so what you get here is that minus Omega
Square mu naught Epsilon minus J beta
Square OK divided by Beta times e x
right so I have this equation now now I
can write down what would be e x by
dividing this entire thing by this
particular ah expression ok so I hope I
got this equation correct ah if I missed
perhaps I missed a j somewhere so if
that is the case then I would like to
put on the J so I will leave this as an
exercise to you ok so find out which one
should be the correct ah term that you
will have and then write e x as 1 by
whatever this term that you have right
times Del e z by Del X Plus Omega mu
naught by Beta times Del h z by Del y
okay and then you can further simplify
the equations in case you want but the
important point to note is that you
could express you can express all the
longitudinal components that is sorry
all the transverse components in any
uniform guided structure where the
guiding structure will have a certain
longitudinal component all the
transverse components can be expressed
in terms of the longitudinal components
ok so that you can always do and you can
do it for all the different transverse
components
now I will leave that as a first part of
that one for you as an exercise you know
just getting the numbers or just getting
the terms right here and there will look
at step two we want to solve the
helmholtz equation for e z component
right the helmholtz equation for e z
component would tell you that you will
have Del square e z plus K 0 square OK
ah times Epsilon r e z equal to zero
this assuming that e z is actually a
phasor right just like every other
component e z also must be along as
minus J beta therefore Del Square by Del
Z square e z can be replaced as minus
beta square e z ok so then what you will
have is only the transverse component
which we will label it as Del
perpendicular Square for e z plus you
have K 0 square Epsilon R minus beta
Square I hope you know what is K 0 so I
am not going to repeat that one this is
the simplified Helm holds equation or a
reduced Helm holds equation that you are
going to get OK of course beta is still
unknown we have to find this beta out
but what is interesting is that then
this Del Square now in the helmholtz
equation that I have written I am only
now concerned with the transverse Delta
or transverse ah Del Square ok so this
fellow for the case of rectangular
waveguard is very simple you will have
Del Square by Del x square and Del by
Del Square by Del y Square so you can
substitute that into the expression
above and what you get is an interesting
thing you will get this expression this
expression plus K 0 square Epsilon R
minus beta Square Times e z must be
equal to zero ok
of course e z is a function of x y and z
now here is a trick which I will tell
you in the next two minutes ah what we
do is e z can be written in terms of Z
dependence we already know that is E
power minus J beta Z so we are not going
to worry about that in this equation we
can simply take e z to be a function of
X and Y only so I can write e z as a
function only of X and as a function
only of Y that is I can write this in
the product form as X of X and Y of Y
and therefore when you substitute this
back into the Expressions what you will
get is X double Prime of X meaning that
this derivative is only with respect to
x times y of Y plus y double Prime of y
x of X OK plus K 0 square Epsilon R
minus beta Square Times X of X and Y of
y equal to 0. now I can divide the whole
thing by X Y OK the product X Y and
assuming that that product is not 0 what
you get is an interesting thing so you
are going to get this one divided by X
Plus y double Prime by y this one is
equal to zero now look at this right
this term is basically a function only
of x
this one is a function only of Y
and this one is a constant right but the
sum of these three is actually going to
be equal to zero the only way it can
happen is when this function of X is a
constant of its own the function of Y is
a constant of its own and when you add
these two constants along with K 0
square Epsilon R minus beta Square the
sum should actually be equal to zero OK
because this is true for all values of X
and Y right so therefore X double Prime
of X by X will be a constant y double
Prime of Y will be another constant plus
that K 0 square Epsilon R minus beta
Square the sum of these three constants
will be equal to zero okay and this
method wherein we wrote e z as a product
of two functions and now we are saying
that each of those functions of X and Y
alone have to be constants or satisfy
the second order differential equation
is called as variable separable method
or separation of variables method and we
will have to say more about this ah
solution in the next module thank you
very much
foreign
[Music]
foreign
[Music]
