Lecture 21

Cavity Resonators

Cavity resonators are important components of microwave and optical systems. They work
by constructive and destructive interference of waves in an enclosed region. They can be
used as filters, or as devices to enhance certain physical interactions. These can be radiation
antennas or electromagnetic sources such as magnetrons or lasers. They can also be used to
enhance the sensitivity of sensors. We will study a number of them, and some of them, only
heuristically in this lecture.

21.1 Transmission Line Model of a Resonator

The simplest cavity resonator is formed by using a transmission line. The source end can be
terminated by ZS and the load end can be terminated by ZL . When ZS and ZL are non-
dissipative, such as when they are reactive loads (capacitive or inductive), then no energy is
dissipitated as a wave is reflected off them. Therefore, if the wave can bounce and interfere
constructively between the two ends, a coherent solution or a resonant solution can exist due
to constructive inference.

The resonant solution exists even when the source is turned off. In mathematical par-
lance, this is a homogeneous solution to a partial differential equation or ordinary differential
equation, since the right-hand side of the pertinent equation is zero. The right-hand side of
these equations usually corresponds to a source term or a driving term. In physics parlance,
this is a natural solution since it exists naturally without the need for a driving or exciting
source.

219
220 Electromagnetic Field Theory

Figure 21.1: A simple resonator can be made by terminating a transmission line with two
reactive loads at its two ends, the source end with ZS and the load end with ZL .

The transverse resonance condition for 1D problem can be used to derive the resonance
condition, namely that

1 = ΓS ΓL e−2jβz d (21.1.1)

where ΓS and ΓL are the reflection coefficients at the source and the load ends, respectively,
βz the the wave number of the wave traveling in the z direction, and d is the length of
the transmission line. For a TEM mode in the transmission line, as in a coax filled with
homogeneous medium, then βz = β, where β is the wavenumber for the homogeneous medium.
Otherwise, for a quasi-TEM mode, βz = βe where βe is some effective wavenumber for a z-
propagating wave in a mixed medium. In general,

βe = ω/ve (21.1.2)

where ve is the effective phase velocity of the wave in the heterogeneous structure.
When the source and load impedances are replaced by short or open circuits, then the
reflection coefficients are −1 for a short, and +1 for an open circuit. The (21.1.1) above then
becomes

±1 = e−2jβe d (21.1.3)

The ± sign corresponds to different combinations of open and short circuits at the two ends
of the transmission lines. When a “+” sign is chosen, which corresponds to either both ends
are short circuit, or are open circuit, the resonance condition is such that

βe d = pπ, p = 0, 1, 2, . . . , or integer (21.1.4)

For a TEM or a quasi-TEM mode in a transmission line, p = 0 is not allowed as the voltage
will be uniformly zero on the transmisson line or V (z) = 0 for all z implying a trivial solution.
The lowest mode then is when p = 1 corresponding to a half wavelength on the transmission
line.
When the line is open at one end, and shorted at the other end in (21.1.1), the resonance
condition corresponds to the “−” sign in (21.1.3), which gives rise to

βe d = pπ/2, p odd (21.1.5)

Cavity Resonators 221

The lowest mode is when p = 1 corresponding to a quarter wavelength on the transmission

line, which is smaller than that of a transmission line terminated with short or open at both
ends. Designing a small resonator is a prerogative in modern day electronic design. For
example, miniaturization in cell phones calls for smaller components that can be packed into
smaller spaces.
A quarter wavelength resonator made with a coax is shown in Figure 21.2. It is easier to
make a short indicated at the left end, but it is hard to make a true open circuit as shown at
the right end. A true open circuit means that the current has to be zero. But when a coax is
terminated with an open, the electric current does end abruptly. The fringing field at the right
end gives rise to stray capacitance through which displacement current can flow in accordance
to the generalized Ampere’s law. Hence, we have to model the right end termination with
a small stray or fringing field capacitance as shown in Figure 21.2. This indicates that the
current does not abruptly go to zero at the right-hand side due to the presence of fringing
field and hence, displacement current.

Figure 21.2: A short and open circuited transmission line can be a resonator, but the open
end has to be modeled with a fringing field capacitance Cf since there is no exact open circuit.

21.2 Cylindrical Waveguide Resonators

Since a cylindrical waveguide1 is homomorphic to a transmission line, we can model a mode

in this waveguide as a transmission line. Then the termination of the waveguide with either
a short or an open circuit at its end makes it into a resonator.
Again, there is no true open circuit in an open ended waveguide, as there will be fringing
fields at its open ends. If the aperture is large enough, the open end of the waveguide radiates
and may be used as an antenna as shown in Figure 21.3.

1 Both rectangular and circular waveguides are cylindrical waveguides.

222 Electromagnetic Field Theory

Figure 21.3: A rectangular waveguide terminated with a short at one end, and an open circuit
at the other end. The open end can also act as an antenna as it also radiates (courtesy of
RFcurrent.com).

As previously shown, single-section waveguide resonators can be modeled p

with a transmis-
sion line using homomorphism with the appropriately chosen βz . Then, βz = β 2 − βs2 where
βs can be found by first solving a 2D waveguide problem corresponding to the reduced-wave
equation.
For a rectangular waveguide, for example, from previous lecture,
r  mπ 2  nπ 2
βz = β2 − − (21.2.1)
a b

for both TEmn and TMmn modes.2 If the waveguide is terminated with two shorts (which is
easy to make) at its ends, then the resonance condition is that

βz = pπ/d, p integer (21.2.2)

Together, using (21.2.1), we have the condition that

ω2  mπ 2  nπ 2  pπ 2
β2 = = + + (21.2.3)
c2 a b d
The above can only be satisfied by certain select frequencies, and these frequencies are the
resonant frequencies of the rectangular cavity. The corresponding mode is called the TEmnp
mode or the TMmnp mode depending on if these modes are TE to z or TM to z.
2 It is noted that for a certain mn mode, with a choice of frequency, β = 0 which does not happen in a
z
transmission line.
Cavity Resonators 223

The entire electromagnetic fields of the cavity can be found from the scalar potentials
previously defined, namely that
E = ∇ × ẑΨh , H = ∇ × E/(−jω) (21.2.4)
H = ∇ × ẑΨe , E = ∇ × H/(jωε) (21.2.5)

Figure 21.4: A waveguide filled with layered dielectrics can also become a resonator. The
transverse resonance condition can be used to find the resonant modes.

Since the layered medium problem in a waveguide is the same as the layered medium
problem in open space, we can use the generalized transverse resonance condition to find the
resonant modes of a waveguide cavity loaded with layered medium as shown in Figure 21.4.
This condition is repeated below as:
R̃− R̃+ e−2jβz d = 1 (21.2.6)
where d is the length of the waveguide section where the above is applied, and R̃− and R̃+
are the generalized reflection coefficient to the left and right of the center waveguide section.
The above is similar to the resonant condition using the transmission line model in (21.1.1),
except that now, we have replaced the transmission line reflection coefficient with TE or TM
generalized reflection coefficients.

21.2.1 βz = 0 Case
In this case, we can still look at the TE and the TM modes in the waveguide. This corresponds
to a waveguide mode that bounces off the waveguide wall, but make no progress in the z
direction. The modes are independent of z since βz = 0. It is quite easy to show that for the
TE case, a z-independent H = ẑH0 , and E = Es exist inside the waveguide, and for the TM
case, a z-independent E = ẑE0 , and H = Hs being the only components in the waveguide.
Consider now a single section waveguide. For the TE mode, if either one of the ends of
the waveguide is terminated with a PEC wall, then n̂ · H = 0 at the end. This will force
the z-independent H field to be zero in the entire waveguide. Thus for the TE mode, it can
only exist if both ends are terminated with open, but this mode is not trapped inside since it
easily leaks energy to the outside via the ends of the waveguide.
For the TM mode, since E = ẑE0 , it easily satisfy the boundary condition if both ends are
terminated with PEC walls since the boundary condition is that n̂ × E = 0. The wonderful
part about this mode is that the length or d of the cavity can be as short as possible.
224 Electromagnetic Field Theory

21.2.2 Lowest Mode of a Rectangular Cavity

The lowest TM mode in a rectanglar waveguide is the TM11 mode. At the cutoff of this mode,
the βz = 0 or p = 0, implying no variation of the field in the z direction. When the two ends
are terminated with metallic shorts, the tangential magnetic field is not shorted out. But the
tangential electric field is shorted to zero in the entire cavity, or that the TE mode cannot
exist. Therefore, the longitudinal electric field of the TM mode still exists (see Figures 21.5
and 21.6). As such, for the TM mode, m = 1, n = 1 and p = 0 is possible giving a non-zero
field in the cavity. This is the TM110 mode of the resonant cavity, which is the lowest mode in
the cavity if a > b > d. The top and side views of the fields of this mode is shown in Figures
21.5 and 21.6. The corresponding resonant frequency of this mode satisfies the equation
2  π 2  π 2
ω110
= + (21.2.7)
c2 a b

Figure 21.5: The top view of the E and H fields of a rectangular resonant cavity.

Figure 21.6: The side view of the E and H fields of a rectangular resonant cavity (courtesy
of J.A. Kong [32]).

For the TE modes, it is required that p 6= 0, otherwise, the field is zero in the cavity. For
example, it is possible to have the TE101 mode with nonzero E field. The resonant frequency
Cavity Resonators 225

of this mode is
2  π 2  π 2
ω101
= + (21.2.8)
c2 a d
Clearly, this mode has a higher resonant frequency compared to the TM110 mode if d < b.
The above analysis can be applied to circular and other cylindrical waveguides with βs
determined differently. For instance, for a circular waveguide, βs is determined differently
using Bessel functions, and for a general arbitrarily shaped waveguide, βs may have to be
determined numerically.

Figure 21.7: A circular resonant cavity made by terminating a circular waveguide (courtesy
of Kong [32]).

For a spherical cavity, one would have to analyze the problem in spherical coordinates.
The equations will have to be solved by the separation of variables using spherical harmonics.
Details are given on p. 468 of Kong [32].

21.3 Some Applications of Resonators

Resonators in microwaves and optics can be used for designing filters, energy trapping devices,
and antennas. As filters, they are used like LC resonators in circuit theory. A concatenation
of them can be used to narrow or broaden the bandwidth of a filter. As an energy trapping
device, a resonator can build up a strong field inside the cavity if it is excited with energy
close to its resonance frequency. They can be used in klystrons and magnetrons as microwave
sources, a laser cavity for optical sources, or as a wavemeter to measure the frequency of
226 Electromagnetic Field Theory

the electromagnetic field at microwave frequencies. An antenna is a radiator that we will

discuss more fully later. The use of a resonator can help in resonance tunneling to enhance
the radiation efficiency of an antenna.

21.3.1 Filters

Microstrip line resonators are often used to make filters. Transmission lines are often used to
model microstrip lines in a microwave integrated circuits (MIC)or monolithic MIC (MMIC).
In these circuits, due to the etching process, it is a lot easier to make an open circuit rather
than a short circuit. But a true open circuit is hard to make as an open ended microstrip line
has fringing field at its end as shown in Figure 21.8 [124, 125]. The fringing field gives rise to
fringing field capacitance as shown in Figure 21.2. Then the appropriate ΓS and ΓL can be
used to model the effect of fringing field capacitance. Figure 21.9 shows a concatenation of
several microstrip resonators to make a microstrip filter. This is like using a concatenation
of LC tank circuits to design filters in circuit theory.
Optical filters can be made with optical etalon as in a Fabry-Perot resonator, or concate-
nation of them. This is shown in Figure 21.10.

Figure 21.8: End effects and junction effects in a microwave integrated circuit [124, 125]
(courtesy of Microwave Journal).
Cavity Resonators 227

Figure 21.9: A microstrip filter designed using concatenated resonators. The connectors to
the coax cable are the SMA (sub-miniature type A) connectors (courtesy of aginas.fe.up.pt).

Figure 21.10: Design of a Fabry-Perot resonator [56, 83, 126, 127].

21.3.2 Electromagnetic Sources

Microwave sources are often made by transferring kinetic energy from an electron beam
to microwave energy. Klystrons, magnetrons, and traveling wave tubes are such devices.
228 Electromagnetic Field Theory

However, the cavity resonator in a klystron enhances the interaction of the electrons with the
microwave field allowing for such energy transfer, causing the field to grow in amplitude as
shown in Figure 21.11.

Magnetron cavity works also by transferring the kinetic energy of the electron into the
microwave energy. By injecting hot electrons into the magnetron cavity, the electromagnetic
cavity resonance is magnified by the absorbing kinetic energy from the hot electrons, giving
rise to microwave energy.

Figure 21.13 shows laser cavity resonator to enhance of light wave interaction with material
media. By using stimulated emission of electronic transition, light energy can be produced.

Figure 21.11: A klystron works by converting the kinetic energy of an electron beam into the
energy of a traveling microwave next to the beam (courtesy of Wiki [128]).
Cavity Resonators 229

Figure 21.12: A magnetron works by having a high-Q microwave cavity resonator. When the
cavity is injected with energetic electrons from the cathode to the anode, the kinetic energy
of the electron feeds into the energy of the microwave (courtesy of Wiki [129]).

Figure 21.13: A simple view of the physical principle behind the working of the laser (courtesy
of www.optique-ingenieur.org).

Energy trapping of a waveguide or a resonator can be used to enhance the efficiency of

a semiconductor laser as shown in Figure 21.14. The trapping of the light energy by the
heterojunctions as well as the index profile allows the light to interact more strongly with
230 Electromagnetic Field Theory

the lasing medium or the active medium of the laser. This enables a semiconductor laser
to work at room temperature. In 2000, Z. I. Alferov and H. Kroemer, together with J.S.
Kilby, were awarded the Nobel Prize for information and communication technology. Alferov
and Kroemer for the invention of room-temperature semiconductor laser, and Kilby for the
invention of electronic integrated circuit (IC) or the chip.

Figure 21.14: A semiconductor laser at work. Room temperature lasing is possible due to
both the tight confinement of light and carriers (courtesy of Photonics.com).

21.3.3 Frequency Sensor

A cavity resonator can be used as a frequency sensor. It acts as an energy trap, because it
will siphon off energy from a microwave when the microwave frequency hits the resonance
frequency of the cavity resonator. This can be used to determine the frequency of the passing
wave. Wavemeters are shown in Figures 21.15 and 21.16. As seen in the picture, there is
an entry microwave port for injecting microwave into the cavity, and another exit port for
the microwave to leave the cavity sensor. The passing microwave, when it hits the resonance
frequency of the cavity, will create a large field inside it. The larger field will dissipate more
energy on the cavity metallic wall, and gives rise to less energy leaving the cavity. This dip
in energy transmission reveals the frequency of the microwave.
Cavity Resonators 231

Figure 21.15: An absorption wave meter can be used to measure the frequency of microwave
(courtesy of Wiki [130]).

Figure 21.16: The innards of a wavemeter (courtesy of eeeguide.com).

232 Electromagnetic Field Theory