Unit Iv
Unit Iv
Unit Iv
r cylinders. Other cross-sectional shapes are possible, but in general these other shapes offer no electrical advantages over the simpler forms in use and are more expensive to manufacture. In order to determine the electromagnetic field configuration within the guide, Maxwells equations are solved subject to the appropriate boundary conditions at the walls of the guide.
Again assuming perfect conductivity for the walls of the guide, the boundary conditions are simply that Etan and Hnorm will be zero at the surface of the conductors.
H z + H y = j E x y
E z + E y = j x H y
H z + H x = j E y x
H y x H z = j E z y
E z + E x = j H y x
E y x
- - - - - - - - (1)
E x = j H z y
2 Ez 2 Ez + + E z = 2 z E 2 2 x y 2 H z 2 H z + + H z = 2 z H 2 2 x y
Equations (1) can be obtained into the form
- - - - - - - - (2)
H E Hx = 2 z + j 2 z h x h y
H E Hy = 2 z j 2 z h y h x
E H Ex = 2 z j 2 z h x h y
E H Ex = 2 z + j 2 z h y h x
Where h 2 = 2 + 2
- - - - - - - - (3)
For rectangular guides Maxwells equations and the solution follow almost exactly as for waves between parallel plans. Assuming that variation in the z direction may be expressed as, e-z, where = + j , Maxwells equation become (for the loss-free region within the guide)
_ _ _
These equations give the relationships among the fields within the guide. It will be noticed that, if Ez and Hz are both zero, all the fields within the guide will vanish. Therefore, for wave-guide transmission (no inner conductor) there must exist either an Ez or Hz component.
As in the case of waves between parallel planes, it is convenient to divide the possible field configurations within the guide into sets, transverse magnetic (TM) waves for which Hz = 0, and transverse electric (TE) waves for which E z = 0. For the rectangular guide shown and the boundary conditions are: Ex = Ez = 0 Ey = Ez = 0 at y = 0 and y = b at x = 0 and x = a
TRANSVERSE MAGNETIC WAVES IN RECTANGULAR GUIDES: The wave equation (2) is partial differential equations that can be solved by the usual technique of assuming a product solution. This procedure leads to two ordinary differential equations, the solutions of which are known. Nothing that
E z ( x, y, z ) = E z0 ( x, y )e z
Let
E z0 = XY
d2X d 2Y +X + 2 XY = XY dx 2 dy 2
- - - - - - - - (4)
Where X is a function of x alone and Y is a function of y alone. Inserting (4) in (2) gives
Divide by XY,
1 d2X 1 d 2Y + h2 = 2 X dx Y dy 2
- - - - - - - - (5)
Equation (5) equates a function of x alone to a function of y alone. The only way in which such a relation can hold for all values of x and y is to have each of these functions equal to some constant, say A2. Then
1 d2X + h 2 = A2 2 X dx
1 d 2Y = A 2 Y dy 2
- - - - - - - - (6) - - - - - - - - (7)
A solution of equation (6) is X = C1 cos Bx + C2 sin Bx B2 = h2 A2 Where the solution of equation (7) is Y = C3 cos Ay + C4 sin Ay This gives
0 E z = XY = C1C 3 cos Bx cos Ay + C1C 4 cos Bx sin Ay +
- - - - - - - - (8)
The constants c1, c2, c3, c4, A and B must now be selected to fit the boundary condition, viz.
E z0 = 0
When x = 0, x = a, y = 0, y = b
- - - - - - - - (9)
E z0 = C 2 C 3 sin Bx
For this to be zero for all values of x it is possible to have either c2 or c3 equal to zero
0 (assuming B 0) Putting c2 = 0 in (9) would make E z identically zero, so instead c3will be put
equal to zero.
0 Then the general expression (9) for E z reduces to
E z0 = C 2 C 4 sin Bx sin Ay
- - - - - - - - (10)
In addition to the amplitude constant C =c2c4, there are still two unknown constants, A and B. However, there are two more boundary conditions to be applied. If x = a
E z0 = C sin Ba sin Ay
In order for this to vanish for all values of y (and assuming A 0, because A = 0 would
0 make E z identically zero) the constant B must have the value
B = m / a Again if y = b, E z0 = C sin
and for this to vanish for all values of x, A must have the value A = n / b
Making use of eqs. (3) and putting = j (as in sec. 7.02) for frequencies above the cut-off frequency, the following expressions are obtained:
jC E = B cos Bx sin Ay h2
0 x
jC E = A sin Bx cos Ay h2
0 y
0 Hx =
- - - - - - - - (12)
0 Hy =
Where B = m / a
These expressions show how each of the components of electric and magnetic field strengths varies with x and y. The variation with time and along the axis of the guide, which is in the z direction, is shown by putting back into each of these expressions the factor and then taking the real part. In the derivation of the fields it was found necessary to restrict he constant A and B to the values given by expressions (12). In these expressions a and b are the width and height of the guide, and m and n are integers. Now, by definition, A2 + B2 = h2
and
h 2 = 2 + 2
Therefore, = h 2 2
= A 2 + B 2 2
- - - - - - - - (14)
m n 2 = + a b
Equation (14) defines the propagation constant for a rectangular guide for TM waves. For low frequencies, where 2 is small, will be a real number. The propagation constant met with in ordinary transmission-line theory is a complex number, that is = + j , where is the attenuation constant (attenuation per unit length) and is the phase shift constant (phase shift per unit length). If is real, must be zero, and there can be o phase shift along the tube. This means there can be no wave motion along the tube for low frequencies. However, as the frequency is increased, a value for will be reached that will make the expression under the radical in (14) equal to zero. If this value of is called c, then for all values of c, the propagation constant
_ _ _ _ _ _ _
greater than
will be imaginary and will have the form = j . For the case is zero for all
under consideration (perfectly conducting walls) the attenuation constant frequencies such that
2
m n = a b
- - - - - - - - (15)
m n + a b 1
2 2
- - - - - - - - (16)
The cut-off frequency, that is the frequency below which wave propagation will not occur, is
fc = m n + 2 a b 1
2 2
- - - - - - - - (17)
c =
2 m n + a b
2 2
- - - - - - - - (18)
From which it is clear that fcc = v0 The velocity of wave propagation will be given by
_
v=
m n 2 a b
2 2
- - - - - - - - (19)
This last expression indicates that the velocity of propagation of the wave in the guide is greater than the phase velocity in free space. As the frequency is increased above cut-off, the
phase velocity decreases from an infinitely large value and approaches c, the velocity in free space, as the frequency increases without limit. Since the wavelength in the guide is given by
_
2 m n 2 a b
2 2
- - - - - - - - (20)
In the above expressions the only restriction on m and n is that they be integers. However from eqs. (12) and (13) it is seen that if either m possible value for either m or n (for TM waves) is unity. From eq. (17)it is evident that the lowest cut-off frequency will occur for m = n =1. Substituting these values in equation (13) gives the fields for the lowest frequency TM wave which can be propagated through the guide. This particular wave is called the TM wave for obvious reasons. Higher-order waves (larger values of m and n) require higher frequencies in order to be propagated along a guide of given dimensions. TRANSVERSE ELECTRIC WAVES IN RECTANGULAR GUIDES: The equations for transverse electric waves (Ez = 0) can be derived in a manner similar
0 to that for transverse magnetic waves. This is left as an exercise for the student. H z will be
found to have the same general form as eq. (8). This is differentiated with respect to x and y
0 0 0 0 find E x , E y , H x and .H y . The boundary conditions are then applied to 0 0 E x .and .E y to give
H z0 = C cos Bx cos Ay
- - - - - - - - (21)
0 Ey =
B=
m a
and
TM waves. However, in equation (21) for TE waves it will be found possible to make either m or n (but not both) equal to zero without causing all the fields to vanish. That is, a lower order is possible than in the TM wave case. The lowest-order TE wave in rectangular guides is therefore the which has the lowest cut-off frequency is called the dominant wave. It is seen that the subscripts m and n represent the number of half-period variations of the field along the x and y co-ordinates respectively. By convention,* the x co-ordinate is assumed to coincide with the larger transverse dimension, so the off frequency than the . wave has a lower cut wave. This wave
For practical reasons in most experimental work with rectangular guides the dominant wave is used. For this wave, substituting m = 1. and n = 0, the fields are H z0 = C cos
0 x
x a
j aC x H = sin a
0 Ey =
jaC
sin
x
a
- - - - - - - - (22)
0 Ex = H y = 0
= 2 a
fc = c / 2a For the
c = 2a
h=/a
wave the cut-off frequency is that frequency for which the corresponding
For the
The field configurations for the lower-order TE and TM waves in rectangular guides.
Possible methods for feeding rectangular guides so that these waves may be initiated are shown. In order to launch a particular mode, a type of probe is chosen which will produce lines of E and H that are roughly parallel to the lines of E and H for that mode.
The probe is parallel to the y axis and so produces lines of E in the y direction and lines of H which lie in the x-z plane. This is the correct field configuration for the probes fed with opposite phase tend to set up the modes. mode. In (b), the parallel mode. In (d) the probe parallel to
the z axis produces magnetic field lines in the x-y plane, which is correct for the TM It is possible for several modes to exist simultaneously in a guide if the frequency is above cut-off for those particular modes. However the guide dimensions are often chosen so that only the dominant mode can exist. IMPOSSIBILITY OF TM WAVE IN WAVE GUIDES ANDEXCITATIONS OF MODES IN RECTANGULAR WAVEGUIDES The waves that will propagate inside hollow rectangular cylinders have been divided into two sets: the transverse magnetic waves of equations (11) and (12) which have no z component of H, and the transverse electric waves of equation (21) that have no z component of E. It will be found that corresponding sets of TM and TE waves can also propagate within circular wave guides, or indeed, in cylindrical guides of any cross-sectional shape. guide. It is easily shown, however, that the familiar TEM wave, for which there is no axial component of either E or H, cannot possibly propagate within a single conductor wave
Suppose a TEM wave is assumed to exist within a hollow guide of any shape. Then
lines of H must lie entirely in the transverse plane. Also in a nonmagnetic material,
This requires that the lines of H be closed loops. Therefore, if a TEM exists inside the guide, the lines of H will be closed loops in plane perpendicular to the axis. equal the axial current (conduction or displacement) through the loop. o In the case of a guide with an inner conductor, e.g., a coaxial transmission line, this axial current through the H loops is the conduction current in the inner conductor. However, for a hollow wave guide having no inner conductor, this axial current must be a displacement current. o But an axial-displacement current requires an axial component of E, something not present in a TEM wave. Therefore the TEM wave cannot exist in a single-conductor wave guide. WAVE IMPEDANCES Wave impedance are defined as the ratio of electric and magnetic field strength Wave impedance for the positive coordinate system are
+ Z xy =
Now by
Maxwells first equation the Magneto Motive Force around each of these closed loops must
Ey + Ex + E Z yz = Z zx = z Hy Hz Hx
Z xy =
For TE wave impedance is Ey j + Z yx = = = Hx j After simplifying the above equation we get the final equation of the wave impedance for TE waves + Z yx = 2 fc 1 f For TM waves the wave impedance are
+ Z yx =
Ey Hx
j = j
After simplifying the above equation we get the final equation of the wave impedance for TE waves
+ Z yx = 0