Microwave Measurement and Beam Instrumentation Course

at Jefferson Laboratory, January 15-26th 2018

Lecture:
Transmission Lines and Waveguides

F. Marhauser
Wednesday, January 17, 2018
 This Lecture
－ Introduction to Various Transmission Lines
－ Coaxial Lines
• Wave Impedance
• Conditions for minimum Damping, maximum Voltage rating, and
maximum Power Transmission
• Attenuation and Power Capability, what are the Technical Limits?
• Bandwidth  Higher Order Mode TE11–mode Cutoff Frequency
－ Waveguides (Round and Rectangular)
• Most derivations are now in Appendix including full Set of RF field Components
• Cutoff Frequencies
• Group and Phase Velocity
• Examples of Mode Pattern
• Attenuation of Fields below Cutoff Frequency
• Poynting Vector
• Derivation of Transmitted Power in TE10 mode of Rectangular Waveguide
－ USAPS Experiment with Rectangular Waveguide

2
 Introduction
- Transmission lines and waveguides are utilized to transfer electromagnetic waves
carrying energy and information from a source to a receiver
- For an efficient transport one likes to guide the energy inside a line instead of
spreading it out in space
- Choice of the line technology depends on the purpose, e.g. operating frequency
range, the transmitted power level, and what power losses one can tolerate

Some typical transmission lines

Microstrip Co-planar waveguide Two-Wire Line Dielectric Waveguide

Coaxial Line Round Waveguide Rectangular Waveguide

3
Introduction

4
 Microstrip Line
ꟷ Microstrip lines are types of planar transmission lines widely used in printed circuit boards
(PCBs)
• Made by a strip conductor, dielectric substrate, and a ground plate
• Used in the microwave range with typical maximum frequency of 110 GHz
• Wave is confined mostly in dielectric, but is partially in upper substrate (usually air)
• The dielectric constant of the substrate usually decreases with frequency as dipolar polarization in the
material cannot follow anymore the oscillations of the electric field (starting around 10 GHz)
• The dielectric constant then approaches more and more that of air if the frequency increases
• At low frequencies, the fields resemble closely a TEM mode (v = 𝑐0 / 𝜀𝑟 ) with fields confined in the
dielectric, but at high frequency there are more non-negligible longitudinal components of both E or H
resulting in a ‘quasi’ TEM mode
• Comparably lossy
• Not shielded, may radiate parasitically and is vulnerable to cross-talk

𝜀𝑎𝑖𝑟

𝜀𝑑𝑖𝑒𝑙.

Frank Gustrau, “RF and Microwave Engineering: Fundamentals of

Wireless Communications’, ISBN: 9781118349571, 2012
5
 Coplanar Waveguide
ꟷ Coplanar waveguides (CPWs) are similar to microstrip lines and also used for PCBs
• Invented later than microstrips (1969 versus 1952)
• Easier to fabricate since having the return and main conductors in the same plane
• May or may not be grounded at the bottom
• Also operate in a quasi-TEM mode at a typical maximum frequency of 110 GHz.
𝜀𝑎𝑖𝑟

𝜀𝑑𝑖𝑒𝑙.

Frank Gustrau, “RF and Microwave Engineering: Fundamentals of

Wireless Communications’, ISBN: 9781118349571, 2012

Optical Fibers
ꟷ Dielectric waveguides can be optical fibers that have a circular cross-section
• Consist of a dielectric material surrounded by another dielectric material
• Allows transmitting optical and infrared signals with small losses (~0.2 dB per 1 km)
• Power transmitted is in the mW range.
6
 Two-Wire Line
ꟷ Two-wire (twin-lead) lines are used for telecommunication to transport RF wave
• Used e.g. for antenna lines to TV
• Separation of the wire is small compared to the wavelength (at 30 MHz wavelength is 100 m)
• Wave is transported in a TEM mode
• May offer smaller losses in the VHF band than miniature coaxial cables, e.g. 0.55 dB/100m versus 6.6
dB/100m for RG-58
• However, more vulnerable to interference even if shielded.

2𝐷
𝑍 ≈ 276 Ω 𝐿𝑜𝑔 For D >> d
𝑑
d
D
Source: Electromagnetic Waves and Applications Part III, Y. MA

Source: Wikipedia

4
 Coaxial Line
ꟷ Coaxial cables are widely used in laboratories and carry signals in the TEM mode..
• At higher frequencies, the dimensions of the cables should be however limited as higher order modes
(with a cutoff) can propagates
• This in turn limits the power capability
• Coaxial cables are typically utilized below 3 GHz with attenuation losses of a few dB/100m in the UHF
range (around 100 MHz)
• Losses however quickly rise with frequency (for small cables to ~10 dB/100m at 1 GHz) with an average
power rating around just 1kW.
• The main losses arise due to the skin effect in the inner conductor, which is technically more difficult to
cool than the outer conductor
• At higher frequencies (around 10 GHz) the dielectric losses of the insulator can become dominant
• By enlarging the coaxial lines diameters (several inches for outer diameter), the power capability may
rise above 100 kW (at few hundred MHz) and into the MW regime (at few 10 MHz) with small
attenuation losses (< 1dB/100m)

http://www.winpoint.com.tw
8
 Coaxial Line – Wave Impedance
D
1 𝜇0 𝐷 60 Ω 𝐷
𝑍= ∙ ∙ 𝐿𝑛 𝑍≈ ∙ 𝐿𝑛 d
2𝜋 ∙ 𝜀𝑟 𝜀0 𝑑 𝜀𝑟 𝑑
𝜇0
; recall 𝑍0 = ≈ 120𝜋 Ω
𝜀0

ꟷ Z at some ratios D/d

Z (Ω)· 𝜀𝑟 D/d
41.56 2
50 2.302 TEM-mode
75 3.493
ꟷ Last 2 values are common cable impedances, why ?

9
 Coaxial Line – Minimum Damping
- There are attenuation losses along the coaxial line (conduction and dielectric losses)
- What is the optimum ratio D/d to minimize losses in the coaxial cable?
1 1 1 1 𝜇𝜔 𝜀𝑟
- Attenuation constant 𝛼 𝛼= ∙ ∙ + ∙ + 𝜋𝑓 ∙ 𝑡𝑎𝑛𝛿
2∙𝑍 𝜋 𝑑 𝐷 2𝜎 𝑐0
Resistive losses + dielectric losses
𝜇𝜔
; recall surface resistance 𝑅𝑆 = tan𝛿 loss tangent of dielectric material
2𝜎

- The 2nd term does not depend on the ratio D/d

- We then need to see for which ratio D/d the 1st term is at a minimum:
2𝜋 ∙ 𝜀𝑟 1 1 1 1 𝜇0 𝐷
𝛼= ∙ ∙ + ∙ 𝑅𝑆 ; 𝑍 = 2𝜋∙ ∙ ∙ 𝐿𝑛
𝐷 𝜋 𝑑 𝐷 𝜀𝑟 𝜀0 𝑑
2 ∙ 𝑍0 ∙ 𝐿𝑛
𝑑
𝜀𝑟 𝟏 𝑫 𝑫
𝛼= ∙ ∙ + 1 ∙ 𝑅𝑆 𝛼 ∙ 𝑍0 𝑫 +1
𝑫 𝑫 𝒅 or = 𝒅 ≡ 𝑓𝛼 (𝐷/𝑑)
𝑍0 ∙ 𝐿𝑛 𝜀𝑟 𝑅𝑆 𝑫
𝒅 𝐿𝑛
𝒅

- Note that there is D left, not only D/d

- One then may ask what is the optimum ratio D/d
to achieve the minimum attenuation at a given diameter of the cable D ?
 Coaxial Line – Minimum Damping
𝟕𝟔. 𝟖 𝛀 𝐷
- Minimum Attenuation: 𝑍𝑜𝑝𝑡. ≈ at ≈ 3.591
𝜖𝑟 𝑑
- E.g. PTFE with 𝜖𝑟 = 2.1 𝑍𝑜𝑝𝑡. ~ 53 Ω
- If inner and outer conductor are of different materials, this is not true anymore since
conductivity values are different, e.g. Al (D) and Cu (d) then Zopt, ~ 95 Ω/m

~76.8 Ω

D/d=3.591
 Coaxial Line
– Maximum Voltage and Maximum Power -
- However, sometimes one rather aims for the maximum voltage (Vmax) at a given D to
avoid a premature RF cable breakdown, which is given by the dielectric strength of the
material (at RF breakdown the dielectric fails to insulate)
- In that case, one wants to choose D/d to minimize the electrical field at given D and
given voltage V between the inner and outer conductor
- In a similar fashion, one finds that 𝑍opt. 𝑉 ~ 𝟔𝟎 𝛀 𝐷
𝑚𝑎𝑥
at = 2.718
𝜖𝑟 𝑑
- Moreover, in other cases one desires the maximum power transmission at a given D
- In this case 𝟑𝟎 𝛀 𝐷
𝑍opt. 𝑃𝑚𝑎𝑥 ~ at = 1.65
𝜖𝑟 𝑑
 Coaxial Line – Power Capability
- Losses quickly rise with frequency - To maximize power capability, use biggest
cables diameter
𝜀𝑟 𝟏 𝐷
RG=Radio Guide cables 𝛼= ∙ ∙ + 1 ∙ 𝑅𝑆
𝐷 𝑫 𝑑
𝑍0 ∙ 𝐿𝑛
𝑑
~ 3kW

HELIFLEX are air-dielectric cables. The

inner conductor is centered by using a
dielectric helix made from high density
polyethylene)

Z = 50 Ω

D = 24.77 cm (corrugated Aluminium)

d = 9.94 cm (corrugated Copper tube)
Max. peak power rating: 5.8 MW at 0.5 MHz
but only 236 W at 560 MHz (fmax)
 High Power Coaxial Lines for Cavities
- Consider: 3rd generation storage ring light sources can store few hundreds of mA

- Powered by coaxial coupler feeding cavity via

water-cooled loop coupler
- Typical effective operating voltage is 1 MV
- At Ib = 100 mA, the forward power required for the
beam (beam loading) is 100 kW (CW) at 500 MHz
- We need more power than shown so far!

500 MHz BESSY (European)

HOM-Damped Cavity

14
High Power Coaxial Lines for Cavities
EiA coaxial transmission lines (50, 75, or 100 Ω)

Vendor for instance Mega Industries: In high

radiation areas where Teflon is not suitable as
supports, special polymer insulators or ceramic
supports are available for sizes up to 14” in diameter.

500 MHz

Horizontal region
indicates peak
power value

https://www.megaind.com
15
Attenuation in such Coaxial Lines

500 MHz

https://www.megaind.com
16
 What Defines Bandwidth of Transmission Lines?
- To maximize power capability, use biggest cables diameter AND avoid exciting the next mode
- One does not want any higher order mode to propagate beside the lowest (dominant) mode
- In coaxial line the TEM-mode is the dominant mode
- Such modes have no cutoff frequency (transmission line works all the way down to DC)
- Example: 2nd mode in coaxial cable is a dipole TE11-mode
- This dipole mode changes polarity twice around cable circumference
- The corresponding wavelength equals the cable circumference C, but at which radius?
𝑑+𝐷
- Approximation: Use average circumference 𝜆𝑐 𝑇𝐸11 = 𝐶 = 𝜋 ∙
2

Cutoff wavelength below which TE11 mode can propagate

E-field H-field
- Corresponding cutoff frequency 𝑇𝐸11 v 𝑐0 1
𝑓𝑐 = = ∙
𝜆𝑐 𝑇𝐸11 𝜇𝑟 ∙ 𝜀𝑟 𝜋 ∙ 𝑑 + 𝐷
2

Cutoff frequency above which TE11 mode can propagate

- Example: d = 1cm, D/d = 2.302 for Z = 50 Ω (in vacuum) 𝑓𝑐 𝑇𝐸11 = 5.78 𝐺𝐻𝑧
- Numerical solution: 𝑓𝑐 𝑇𝐸11 = 5.9 𝐺𝐻𝑧
 Round Waveguide
- Derivation of field components in round waveguide follows the same method covered
in lecture about resonators
- This time however we have no reflection plate, and wave can propagate freely
- We thus lose one constraint (index p) in longitudinal direction that we had derived for the
cylindrical resonator, and only have to deal with two integers m and n
- Appendix covers derivations
- For a perfect conductor (no resistive attenuation) we obtain
the propagation constants for TE- and TM-modes z
2
𝑥𝑚𝑛 𝑜𝑟 𝑥′𝑚𝑛 𝑥𝑚𝑛 for TM-modes
𝛾 = 𝑖𝛽 = ± 𝑖 𝜇𝜖𝜔 2 −
𝑅 𝑥′𝑚𝑛 for TE-modes

- We see that only if β is real, the wave can propagate without decay by means of 𝑒 −𝑖𝛽𝑧
- This leads to so-called cutoff frequencies, above which wave may propagate for given m, n
2
𝑥𝑚𝑛 𝑜𝑟 𝑥′𝑚𝑛 1 𝑥𝑚𝑛 𝑜𝑟 𝑥′𝑚𝑛
2 𝜔𝑐 = ∙
𝜇𝜖𝜔 ≥
𝑅 𝜇𝜖 𝑅
- The spectrum of possible modes above cutoff frequencies is continuous
- Inserting last in first equation on this page yields: 𝛽 2 = 𝜇𝜖 𝜔2 − 𝜔𝑐2 ; TE or TM modes
 Phase and Group Velocity in Waveguide
𝜔
𝘷𝑝ℎ = Phase velocity ; in free space 𝘷𝑝ℎ = 𝑐0 due to ; 𝑘 2 = 𝜇𝜖𝜔2
𝑘
- In a waveguide the wavenumber is constrained compared to free space (𝑘 = 𝜇𝜖𝜔)
due to 𝛽2 = 𝜇𝜖 𝜔2 − 𝜔𝑐2
2𝜋 2 𝑘𝑧 2
𝛽 = 𝑘𝑧 = = 𝑘2 − 𝑘𝑐 = 𝜇𝜖 𝜔2 − 𝜔𝑐2 𝜔= + 𝜔𝑐2
Λ 𝜇𝜖
-  is the wavelength of the waveguide (‘guide length’)
𝜔 𝜔 v
𝘷𝑝ℎ = =𝑓∙Λ= =
𝑘𝑧 2 2
𝜇𝜖 𝜔 − 𝜔𝑐 𝑓𝑐2 Phase velocity is always > speed of light
1− 2
𝑓

𝑑𝜔
𝘷𝑔𝑟 = Group velocity (energy flows at group velocity)
𝑑𝑘

𝑑𝜔 𝑘𝑧 1 𝜔 2 − 𝜔𝑐2 𝑓𝑐2 Group velocity is < speed of light

𝘷𝑔𝑟 = = = =v∙ 1− 2
𝑑𝑘 𝜇𝜖𝜔 𝜇𝜖 𝜔 𝑓 Group velocity is not constant
with frequency (dispersion)
Phase and Group Velocity in Waveguide

Phase velocity is > speed of light

Group velocity is < speed of light

 Roots of Bessel Function and its Derivative
Lowest Cutoff Frequencies and Degeneracy
1 𝑥𝑚𝑛 1 𝑥′𝑚𝑛
𝜔𝑐𝑇𝑀 = ∙ 𝜔𝑐𝑇𝐸 = ∙
𝜇𝜖 𝑅 𝜇𝜖 𝑅

- 1st cutoff is dipole mode TE11 (x’11 = 1.84118)

- 2nd cutoff is monopole mode TM01 (x01 = 2.40483)
TM TE
- Degeneracy (differing modes, but same cutoff frequency) occurs when x1n = x’0n  𝑓1𝑛 = 𝑓0𝑛
- Recommended operating bandwidth (single-mode operation) is from slightly above TE11-mode
cutoff to maximally TM01-mode cutoff (factor ~1.31 higher than TE11-mode cutoff )
bandwidth

Cutoff frequencies in a round waveguide in sequential order for the first

TE and TM modes (normalized to first TE11 cutoff frequency)
21
 TE-Mode Pattern
TE01

TE11

22
 Waveguide and Frequencies below Cutoff
- We recall that for the propagation constant 𝛾 = 𝑖𝛽 and 𝛽 = 𝜇𝜖 𝜔 2 − 𝜔𝑐2

- Specific modes (with indices m, n) propagate undamped (perfect conductor)

according to 𝑒 −𝑖𝛽𝑧 and only above the cutoff frequency
- What if mode frequency is smaller than its cutoff frequency ?

𝛽 = 𝑖 ∙ 𝜇𝜖 𝜔𝑐2 − 𝜔 2 ; for𝜔 < 𝜔𝑐

- β becomes imaginary itself and  becomes real

- The wave is therefore damped even for the loss-less case ( =0) according to 𝒆−𝜸𝒛
 Example: Single-Cell (TESLA) Cavity

TM010 mode f = 1.302 GHz, Rsh = 105.6 Ω

R = 39 mm

- First monopole harmonic of fundamental mode (TM011)

- First relevant monopole cutoff frequency is TM01 (not TE11)

TM011 mode f = 2.451 GHz, Rsh = 28.4 Ω c xmn

𝑓𝑐,𝑇𝑀01 = ∙ = 2.94 𝐺𝐻𝑧 > 𝑓𝑇𝑀011
2π R
 TM011 resonator mode is trapped
inside cavity  we need HOM-
couplers to suppress this field
 Even More Power: Hollow Waveguides
ꟷ Hollow waveguides can transmit very high average power signals in the microwave spectrum
• Rectangular and round waveguides are commonly employed.
• Without an inner conductor, they can sustain much higher power levels than coaxial lines
• Metal walls can be readily cooled
• Recommended bandwidth is limited – as for coaxial line – by preventing the next waveguide mode to
co-exist with the dominant mode
• The inner surface of the waveguides can be plated with high conductive material
(e.g. copper, silver, gold) to reduce losses due to the skin effect
• Average power levels into the MW range can be achieved with rather small attenuation losses (few
dB/100 m) with large scale waveguides
• Standard rectangular waveguides (WR) sizes are available up to WR2300 (0.584 m (23”) x 0.2921 m)
covering 320-450 MHz and down to WR3 (0.864 mm x 0.432 mm) covering 220-330 GHz.

- In SRF accelerators we may run 20 MV/m in CW in Energy Recovery Linacs with large currents
(up to 1A machines have been proposed in the past)
- The injector of an ERL is not energy-recovered  If structure is 1m long, we have a voltage of 20 MV (CW)
 beam loading at 100 mA is then already 2 MW.
- This is the power we would need to deliver into cavity (wall losses are negligible)
- Otherwise or we need to split powerand/or make cavities shorter to reduce require power per cavity

25
 Rectangular Waveguide
Cutoff Frequency, Phase and Group Velocity
- The cutoff frequency can be calculated by (no full derivation here, refer to textbooks):
y
v 𝑚 2 𝑛 2
𝑓𝑐 = ∙ + ; v is speed of light
2 𝑎 𝑏 in linear medium
z
𝑚∙𝜋 2 𝑛∙𝜋 2
𝜔𝑐 = v ∙ + a = long side
𝑎 𝑏 x
b = short side

- We obtain the same relations for the phase and group velocity with the given cutoff
frequency in the rectangular waveguide as for round waveguides and all other related
consequences apply similarly
𝜔 v
𝘷𝑝ℎ = = 𝑑𝜔 𝑓𝑐2
𝑘
𝑓2 𝘷𝑔𝑟 = =v 1− 2
1− 𝑐2 𝑑𝑘 𝑓
𝑓

Phase velocity Group velocity

 Rectangular Waveguide
Cutoff Frequencies of Lowest Modes

v 𝑚 2 𝑛 2
𝑓𝑐 = ∙ +
2 𝑎 𝑏
- Note: For TE-modes: m,n ≥ 0, but m=n=0 is not allowed since only trivial solution exist
- Note: For TM-modes: m,n ≥ 1
-  1st dominant mode is dipole TE10 -mode (m = 1, n= 0)
- Cutoff frequency can be easily derived graphically

𝐸
a = λ/2
𝑇𝐸10
TE10 b 𝜆𝑐 =2 ∙𝑎
𝑇𝐸 𝘷
𝑓𝑐 10 =
a 2 ∙𝑎

half wavelength
 Rectangular Waveguide
Cutoff Frequencies of Lowest Modes

v 𝑚 2 𝑛 2
𝑓𝑐 = ∙ +
2 𝑎 𝑏
- Note: For TE modes: m,n ≥ 0, but m=n=0 is not allowed since only trivial solution exist
- Note: For TM modes: m,n ≥ 1

- Ditto for the TE20-mode

a=λ
TE20 𝑇𝐸20
𝜆𝑐 =𝑎
𝑇𝐸 𝘷
𝑓𝑐 20 =
𝑎

1 wavelength
 Rectangular Waveguide
Cutoff Frequencies of Lowest Modes

v 𝑚 2 𝑛 2
𝑓𝑐 = ∙ +
2 𝑎 𝑏
- Note: For TE modes: m,n ≥ 0, but m=n=0 is not allowed since only trivial solution exist
- Note: For TM modes: m,n ≥ 1

- Ditto for the TE01-mode

half wavelength
b = λ/2
TE01 𝑇𝐸01
𝜆𝑐 =2 ∙𝑏
𝑇𝐸 𝘷
𝑓𝑐 01 =
2 ∙𝑏
 Lowest Cutoff Frequencies and Degeneracy
- Various mode degeneracies may occur depending on ratio of waveguide height (b) to width (a)
and for all TE and TM modes with same indices m and n
- Again: Recommended operating bandwidth (single-mode operation) is from slightly above
TE10-mode cutoff to maximally TE20-mode cutoff for b ≤ a/2 (larger bandwidth than for round
waveguides). For b > a/2 the bandwidth is reduced as TE01-mode becomes 2nd mode

bandwidth

bandwidth

bandwidth

Cutoff frequencies in a rectangular waveguide for the first TE and TM modes (normalized
to first TE10 cutoff frequency) depending on the ratio of the waveguide height to the
waveguide width b. Mode degeneracies may occur.
30
Power Transmission in a Rectangular Waveguide
 Power Transmission along a Waveguide
- Power transmitted along a waveguide can be generally by integrating the Poynting vector over
the-cross section of the waveguide 𝑆 = 𝐸×𝐻 Poynting vector
- Poynting vector points in the direction of the wave propagation and is the energy transferred
per unit area and per unit time (units are V/m· A/m = W/m2)
- For harmonic signals, the time-averaged power is given by real part of cross product integrated
over the transverse cross-section of the guide normal to the propagation

1
𝑃𝑎𝑣𝑔 = 𝑅𝑒 𝐸𝑡𝑟𝑎𝑛𝑠 × 𝐻 ∗ 𝑡𝑟𝑎𝑛𝑠 ∙ 𝑛 ⋅ 𝑑𝑆
2
𝑑𝑆

- Note that ratio of the transverse components Etrans/Htrans determines the wave impedance Z
∗ ∗ 2
1 2
𝐸𝑡𝑟𝑎𝑛𝑠 × 𝐻 𝑡𝑟𝑎𝑛𝑠 ∙ 𝑛 = 𝐸𝑡𝑟𝑎𝑛𝑠 ⋅ 𝐻 𝑡𝑟𝑎𝑛𝑠 = 𝑍 ⋅ 𝐻𝑡𝑟𝑎𝑛𝑠 = ⋅ 𝐸𝑡𝑟𝑎𝑛𝑠
𝑍
- The transverse components of E and H are all normal to each other for TEM, TE, and TM waves
and cross product points in direction of 𝑛 ⋅ 𝑑𝑆
- The transmitted power through the waveguide cross-section is thus:

1 2
1 2
𝑃𝑎𝑣𝑔 = ∙𝑍 𝐻𝑡𝑟𝑎𝑛𝑠 ∙ 𝑑𝑆 = 𝐸𝑡𝑟𝑎𝑛𝑠 ∙ 𝑑𝑆
2 𝑍
𝑑𝑆 𝑑𝑆
 Example: Power Transmitted in TE10 mode
- For dominant TE10 mode, the non-vanishing transverse electric field component is:
𝜔∙𝜇 𝜋 𝜋∙𝑥
𝐸𝑦 𝑧 = 0 = −𝑖𝐻0 ∙ ∙ ∙ 𝑠𝑖𝑛
𝑘𝑐2 𝑎 𝑎

- Consequently:

1 1 2
𝑃𝑎𝑣𝑔 = ∙ 𝐸𝑡𝑟𝑎𝑛𝑠 ∙ 𝑑𝑆 =
2 𝑍𝑤𝑎𝑣𝑒
𝑑𝑆
2
1 1 𝑏 𝑎 𝜔∙𝜇 𝜋 𝜋∙𝑥
= ∙ ∙ 𝑑𝑦 𝑑𝑥 ∙ 𝐻0 ∙ ∙ ∙ 𝑠𝑖𝑛 𝑎
2 𝑍𝑤𝑎𝑣𝑒 𝑦=0 𝑥=0 𝑘𝑐2 𝑎
1 1 𝜔∙𝜇 2 𝑎 𝜋∙𝑥 2
2 𝜋 2
=2 ∙𝑍 ∙ 𝑏 ∙ 𝐻0 ∙ ∙𝑎 𝑥=0
𝑑𝑥 ∙ 𝑠𝑖𝑛 ; 𝑘𝑐2 =
𝑤𝑎𝑣𝑒 𝜋 𝑎 𝑎
2𝜋∙𝑥 𝑎 2 𝑎∙sin(2𝜋∙𝑥
𝜋∙𝑥 𝑎 )
1 1 𝜔∙𝜇 2 𝑥 𝑎∙sin( 𝑎 ) ; 𝑑𝑥 𝑠𝑖𝑛 = 𝑥2 −
=2∙𝑍 ∙ 𝑏 ∙ 𝐻0 2 ∙ ∙𝑎 ∙ − 𝑎 4𝜋
𝑤𝑎𝑣𝑒 𝜋 2 4𝜋 0
1 1 𝜔∙𝜇 2 𝑎
𝑃𝑎𝑣𝑔 = 2 ∙ 𝑍 ∙ 𝑏 ∙ 𝐻0 2 ∙ ∙𝑎 ∙2
𝑤𝑎𝑣𝑒 𝜋
 Example: Wave Impedance TE10 mode
- What is yet missing is an expression for the wave impedance Z
- We thus need Etrans/Htrans : The only non-vanishing transverse electric and magnetic field
components for TE10 mode are:
𝜔 ∙ 𝜇 𝑚𝜋 𝑚∙𝜋∙𝑥 𝛽 𝑚𝜋 𝑚∙𝜋∙𝑥
𝐸𝑦 𝜔 = −𝑖𝐻0 ∙ ∙ ∙ 𝑠𝑖𝑛 ∙ 𝑒 𝑖𝜔𝑡 𝐻𝑥 𝜔 = 𝑖𝐻0 ∙ ∙ ∙ 𝑠𝑖𝑛 ∙ 𝑒 𝑖𝜔𝑡
𝑘𝑐2 𝑎 𝑎 2
𝑘𝑐 𝑎 𝑎

𝜇0 𝜇𝑟 1 𝑓𝑐 2
𝐸𝑦 𝑧 = 0 𝜔∙𝜇 𝑍𝑇𝐸10 = ∙ ∙ ;𝛽 = 𝜇𝜖𝜔 ∙ 1 −
𝑓2
𝑍𝑇𝐸10 =− = 𝜖0 𝜖𝑟 2
𝐻𝑥 𝑧 = 0 𝛽 𝑓𝑐 v
1−
𝑓2
with 𝑓𝑐,𝑇𝐸10 = 2𝑎

- From lecture on Maxwell’s equation we remember that vacuum impedance

𝜇0 𝜇𝑟 𝑍0
𝑍0 = ≈ 120𝜋 ≈ 376.73 [Ω] 𝑍𝑇𝐸10 = ∙
𝜀0 𝜖𝑟
𝑓𝑐 2
1− 2
𝑓

- The TE10 waveguide impedance is therefore always greater than the free space wave
impedance  Reflection occurs if wave would propagate out into free space
 Example: Power Transmitted in TE10 mode
- By inserting the wave impedance we eventually obtain:

1 𝛽 𝜔∙𝜇 2 𝑎 1 𝛽∙𝜔∙𝜇
2
∙ 𝑎3 ∙ 𝑏 ∙ 𝐻0 2
𝜔∙𝜇
𝑃𝑎𝑣𝑔 =2 ∙ ∙ 𝑏 ∙ 𝐻0 ∙ ∙𝑎 ∙2 =4∙ ; 𝑍𝑇𝐸10 = 𝛽
𝜔∙𝜇 𝜋 𝜋2

𝜔∙𝜇∙ 𝜇𝜖𝜔 3 2 𝑓𝑐 2 𝑓𝑐 2
= ∙ 𝑎 ∙ 𝑏 ∙ 𝐻0 ∙ 1 − ;𝛽 = 𝜇𝜖𝜔 ∙ 1 − 𝑓2
4𝜋2 𝑓2

𝑓𝑐 2
=𝑓 2 ∙ 𝜇 ∙ 𝜇𝜖 ∙ 𝑎3 ∙ 𝑏 ∙ 𝐻0 2 ∙ 1 −
𝑓2

1 𝑓𝑐 2
= 𝑓 2 ∙ 𝜇 ∙ ∙ 𝑎3 ∙ 𝑏 ∙ 𝐻0 2 ∙ 1 − 1
v 𝑓2 ;v= 𝜇𝜖

𝜇 𝑓2 3 2 𝑓𝑐 2 1
= ∙ 2 ∙ 𝑎 ∙ 𝑏 ∙ 𝐻0 ∙ 1− ;v=
𝜀 v 𝑓2 𝜇𝜖

1 𝜇 𝑓2 2 𝑓𝑐 2 v
𝑃𝑎𝑣𝑔 = 4 ∙ ∙ ∙ 𝑎 ∙ 𝑏 ∙ 𝐻0 ∙ 1 − 𝑓2 ; 𝑓𝑐,𝑇𝐸10 = 2𝑎
𝜀 𝑓𝑐 2

with H0 the peak field amplitude

 Example: Power Limit in TE10 mode (air-filled)
- The electrical field amplitude of TE10 mode (see appendix) is
𝜔∙𝜇 𝜋 𝜋∙𝑥
|𝐸𝑦 | = 𝐻0 ∙ 2 ∙ ∙ 𝑠𝑖𝑛
𝑘𝑐 𝑎 𝑎

- Maximum electric field amplitude is in the center of the waveguide (x = a/2)

𝜔∙𝜇 𝜋 𝜔∙𝜇 𝜋 2
𝐸𝑦 = 𝐻0 ∙ 2 ∙ = 𝐻0 ∙ ∙𝑎 ; 𝑘𝑐2 =
𝑘𝑐 𝑎 𝜋 𝑎

- If the waveguide is filled with air (1 bar), the dielectric strength is 3 MV/m (Emax)
- The electric field shall not exceed the dielectric strength, thus:
𝜋
𝐻0,𝑚𝑎𝑥 = ∙ 𝐸𝑚𝑎𝑥
𝜔∙𝜇∙𝑎
- Inserting into transmitted power formula yields:
1 𝜇 𝑓2 𝜋 2 𝑓𝑐 2
2
𝑃𝑎𝑣𝑔,𝑚𝑎𝑥 = ∙ ∙ ∙𝑎∙𝑏∙ ∙ 𝐸𝑚𝑎𝑥 ∙ 1 −
4 𝜀 𝑓𝑐 2 𝜔∙𝜇∙𝑎 𝑓2

1 𝜇 1 𝑏 1 2 𝑓𝑐 2
𝑃𝑎𝑣𝑔,𝑚𝑎𝑥 = ∙ ∙ ∙ ∙ ∙ 𝐸𝑚𝑎𝑥 ∙ 1 −
16 𝜀 𝑓𝑐 2 𝑎 𝜇2 𝑓2
 Example: WR650 Waveguide
- Standard WR650 rectangular waveguides (6.5” x 3.25”) with TE10 cutoff at ~908 MHz are often
used to power L-band accelerators
- For JLab’s cavities (fRF = 1.497 MHz) the maximally transmitted power in air-filled waveguide
WR650 waveguides is ~65 MW (far above the requirements of ~13 kW (forward power) for
upgrade cavities)
- What about attenuation?
Attenuation of Various Transmission Lines

http://www.rfcafe.com/
 Example: WR650 Waveguide
- Standard WR650 rectangular waveguides (6.5” x 3.25”) with TE10 cutoff at ~908 MHz are often
used to power L-band accelerators
- For JLab’s cavities (fRF = 1.497 MHz) the maximally transmitted power in air-filled waveguide
WR650 waveguides is ~65 MW (far above the requirements of ~13 kW (forward power) for
upgrade cavities)
- What about attenuation?
- Everything OK?
- In reality: Reflections in waveguides such as arising from adapters, directional coupler, and
waveguide bends can reduce the breakdown field
- Other insertion devices can significantly reduce the breakdown field to a fraction of the
theoretical waveguide limit
- Waveguides and coaxial lines are sometimes filled with dielectric sulfur hexafluoride (SF6) gas
to increase the RF breakdown field limit (1 bar SF6 is equivalent to ~3 bar air), however SF6 is a
potential greenhouse gas

https://www.megaind.com
 Experiment with Rectangular Waveguide
- Rectangular waveguide with various terminations
1) Measure the reflection response S11 using coaxial-to-waveguide adapter using VNA
2) Make use of calibration kit (1-port calibration)
3) Learn how to de-embed a device under test (DUT) utilizing the Time Domain
Reflectometry (TDR) option of the VNA
4) Measure the reflection response of the adapter only by setting appropriate time gates
5) What is the useful range for the measurements? Note: All adapters are bandwidth-
limited and allow only for a certain frequency range to be transmitted efficiently
6) Measure the reflection response of the termination (HOM-loads) by re-adjusting the time
gates
7) Record the reflection response and determine the characteristics for each HOM-load
8) How does using the TDR option compare to regular results?
9) What is the best HOM-load at room temperature?

HOM damping waveguide dimensions:

H x B = 0.71” x 6.3” = 18 mm x 160 mm
Standard WR430 adapter:
H x B = 1.875” x 4.3”
http://www.agilent.com/

40
Appendix
 Round Waveguide
- We assume an infinitely long waveguide in z-direction
- An existing wave inside (previously launched from on side of the guide)
can be propagating only in one direction (no reflection plane)
- Instead of an infinitely long waveguide one can also assume that the
waveguide is perfectly matched on one end such that no reflection occurs z
- Derivation of fields is analogous to cylindrical resonator, except for z-direction
R = inner radius
𝐴=𝑁 𝑟 𝑀 𝜑 𝑃 𝑧 𝑃 𝑧 = 𝐵𝑒 ±𝛾𝑧 =+i𝛽
forwards or backwards travelling wave
- Longitudinal index p: Since there is no boundary in z-direction, a dependency on a number ‘p’ is
undetermined

- Azimuthal index m: 𝑀 𝜑 = 𝐶1 𝑒 −𝑖𝑚𝜑 + 𝐶2 𝑒 +𝑖𝑚𝜑 𝑀 𝜑 = 𝑀 𝜑 + 2𝜋 ; m ∈ ℕ0

- Radial index n: 𝑁 𝑟 = 𝐷1 𝐽𝑚 𝑟 𝑘 2 + 𝛾 2
- TM-modes: 𝑥𝑚𝑛 2
𝐸𝑧 (𝑟) ∝ 𝑁 𝑅 = 𝐷1 𝐽𝑚 𝑅 𝑘 2 + 𝜸2 = 0 𝑥𝑚𝑛 = 𝑅 𝑘2 + 𝛾2 𝛾 = ± 𝑖 𝜇𝜖𝜔 2 −
𝑅
Tangential fields of Ez(r=R) must vanish: ;𝑘 2 = 𝜇𝜖𝜔2

𝛿𝐻𝑧 𝑟 = 𝑅 𝜕𝑁 𝑟 = 𝑅 𝜕
- TE-modes: =0 = 𝐷 𝐽 𝑅 𝑘2 + 𝛾2 = 0
1
𝛿𝑟 𝜕𝑟 𝜕𝑟 𝑚
Ez = 0, azimuthal fields E(r=R) must vanish at cavity perimeter
𝑥′𝑚𝑛 2
𝑥′𝑚𝑛 = 𝑅 𝑘 2 + 𝛾 2 𝛾 = ± 𝑖 𝜇𝜖𝜔 2 −
𝑅
 Round Waveguide
- Loss-less case ( = 0)

2 2
𝑥𝑚𝑛 𝑥′𝑚𝑛
𝛾 𝑇𝑀 2
= 𝑖𝛽 = ± 𝑖 𝜇𝜖𝜔 − 𝛾 𝑇𝐸 2
= 𝑖𝛽 = ± 𝑖 𝜇𝜖𝜔 −
𝑅 𝑅

- Only if β is real, the wave can propagate without decay by means of 𝑒 −𝑖𝛽𝑧
- Leads to so-called cutoff frequencies, beyond which wave may propagate for given m, n and R

2 2
2
𝑥𝑚𝑛 2
𝑥′𝑚𝑛
𝜇𝜖𝜔 ≥ 𝜇𝜖𝜔 ≥
𝑅 𝑅

- Cutoff-frequencies for TM (use xmn) and TE modes (use x’mn):

𝘷 1 𝑥𝑚𝑛 𝑜𝑟 𝑥′𝑚𝑛
𝜔𝑐𝑇𝑀,𝑇𝐸 = 2𝜋𝑓𝑐𝑇𝑀,𝑇𝐸 = 2𝜋 = ∙
𝜆𝑇𝑀,𝑇𝐸
𝑐
𝜇𝜖 𝑅

- The spectrum of possible modes above cutoff frequencies is continuous

- Inserting last in first equations on this page yields: 𝛽 2 = 𝜇𝜖 𝜔2 − 𝜔𝑐2 ; TE or TM modes

 Round Waveguide - Field Components
TM-Modes TE-Modes
𝑥𝑚𝑛 𝑚 𝜔𝑇𝐸 𝜇𝑅 2 𝑥′𝑚𝑛
𝜋𝑅 2 𝜕𝐽𝑚 𝑟 𝑅 𝐸𝑟 𝜔 = 𝑖𝐻0 ∙ ∙ ∙ 𝐽𝑚 𝑟 ∙ 𝑠𝑖𝑛(𝑚𝜑) ∙ 𝑒 −𝑖𝛽𝑧 ∙ 𝑒 𝑖𝜔𝑡
𝐸𝑟 𝜔 = −𝐸0 2 𝑐𝑜𝑠(𝑚𝜑) ∙ 𝑒 −𝑖𝛽𝑧 ∙ 𝑒 𝑖𝜔𝑡 𝑟 𝑥′𝑚𝑛 2 𝑅
𝑥𝑚𝑛 𝜕𝑟

𝑥′𝑚𝑛
𝑚 𝜋𝑅 2 𝑥𝑚𝑛 𝜔𝑇𝐸 𝜇𝑅 2 𝜕𝐽𝑚 𝑟 𝑅
𝐸𝜑 𝜔 = 𝐸0 ∙ 2 ∙ 𝐽𝑚 𝑟 ∙ 𝑠𝑖𝑛(𝑚𝜑) 𝑒 −𝑖𝛽𝑧 ∙ 𝑒 𝑖𝜔𝑡 𝐸𝜑 𝜔 = 𝑖𝐻0 ∙ ∙ ∙ 𝑐𝑜𝑠(𝑚𝜑) ∙ 𝑒 −𝑖𝛽𝑧 ∙ 𝑒 𝑖𝜔𝑡
𝑟 𝐿 𝑥𝑚𝑛 𝑅 𝑥′𝑚𝑛 2 𝜕𝑟

𝑥𝑚𝑛
𝐸𝑧 𝜔 = 𝐸0 ∙ 𝐽𝑚 𝑟 ∙ 𝑐𝑜𝑠(𝑚𝜑) ∙ 𝑒 −𝑖𝛽𝑧 ∙ 𝑒 𝑖𝜔𝑡 𝐸𝑧 𝜔 = 0
𝑅

𝑚 𝜔𝑇𝑀 𝜀𝑅 2 𝑥𝑚𝑛 𝑥′𝑚𝑛

𝜋𝑅 2 𝜕𝐽𝑚 𝑟 𝑅
𝐻𝑟 𝜔 = −𝑖𝐸0 ∙ ∙ ∙ 𝐽𝑚 𝑟 ∙ 𝑠𝑖𝑛(𝑚𝜑) ∙ 𝑒 −𝑖𝛽𝑧 ∙ 𝑒 𝑖𝜔𝑡
𝑟 2
𝑥𝑚𝑛 𝑅 𝐻𝑟 𝜔 = 𝐻0 ∙ 2
∙ ∙ 𝑐𝑜𝑠(𝑚𝜑) ∙ 𝑒 −𝑖𝛽𝑧 ∙ 𝑒 𝑖𝜔𝑡
𝑥′𝑚𝑛 𝜕𝑟

𝑥𝑚𝑛
𝜔𝑇𝑀 𝜀𝑅 2 𝜕𝐽𝑚 𝑟 𝑅 𝑚 𝜋𝑅 2 𝑥′𝑚𝑛
𝐻𝜑 𝜔 = −𝑖𝐸0 ∙ ∙ ∙ 𝑐𝑜𝑠(𝑚𝜑) ∙ 𝑒 −𝑖𝛽𝑧 ∙ 𝑒 𝑖𝜔𝑡 𝐻𝜑 𝜔 = −𝐻0 ∙ ∙ 2
∙ 𝐽𝑚 𝑟 ∙ 𝑠𝑖𝑛(𝑚𝜑) ∙ 𝑒 −𝑖𝛽𝑧 ∙ 𝑒 𝑖𝜔𝑡
2
𝑥𝑚𝑛 𝜕𝑟 𝑟 𝑥′𝑚𝑛 𝑅

′
𝐻𝑧 𝜔 = 0 𝑥𝑚𝑛
𝐻𝑧 𝜔 = 𝐻0 ∙ 𝐽𝑚 𝑟 ∙ 𝑐𝑜𝑠(𝑚𝜑) ∙ 𝑒 −𝑖𝛽𝑧 ∙ 𝑒 𝑖𝜔𝑡
𝑅

1 𝑥𝑚𝑛 1 𝑥′𝑚𝑛
𝜔𝑐𝑇𝑀 = ∙ 𝜔𝑐𝑇𝐸 = ∙
𝜇𝜖 𝑅 𝜇𝜖 𝑅

Note: Cutoff frequency determined by interior medium (permittivity, permeability), tube radius (R), and
roots (xmn, x’mn) of Bessel function of first kind (TM-modes) or its derivative (TE-modes)
44
 TE-Mode Pattern
TE21

45
TM-Mode Pattern
TM01

TM11

46
TM-Mode Pattern
TM21

47
First 10 Beam Tube Modes (E/H-fields)
TE11 TM01 TE21

TM11 TE01 TE31

0
TM21 TE12 TE41

TM02

48
 Cutoff Frequencies of Cavity Beam Tubes
- R = 35-39 mm are typical tube radii for 1.3-1.5 GHz (L-band) SRF cavities, e.g.
EU-XFEL/ILC/LCLS-II TESLA-type cavities or JLab’s CEBAF/FEL cavities
- Larger tube radii are considered for high-current heavily HOM-damped cavities, which lets Higher
Order Modes propagate out of cavity at lower frequencies, e.g. 55 mm for Cornell 1.3 GHz ERL
cavity design

49
 Rectangular Waveguide - Field Components
TM-Modes TE-Modes
𝛽 𝑚𝜋 𝑚∙𝜋∙𝑥 𝑛∙𝜋∙𝑦 𝜔 ∙ 𝜇 𝑛𝜋 𝑚∙𝜋∙𝑥 𝑛∙𝜋∙𝑦
𝐸𝑥 𝜔 = −𝑖𝐸0 ∙ ∙ 𝑒 −𝑖𝛽𝑧 ∙ 𝑒 𝑖𝜔𝑡 𝐸𝑥 𝜔 = 𝑖𝐻0 ∙ ∙ ∙ 𝑐𝑜𝑠 ∙ 𝑠𝑖𝑛 ∙ 𝑒 −𝑖𝛽𝑧 ∙ 𝑒 𝑖𝜔𝑡
2 ∙ 𝑎 ∙ 𝑐𝑜𝑠 𝑎
∙ 𝑠𝑖𝑛
𝑏 𝑘𝑐2 𝑏 𝑎 𝑏
𝑘𝑐

𝛽 𝑛𝜋 𝑚∙𝜋∙𝑥 𝑛∙𝜋∙𝑦 𝜔 ∙ 𝜇 𝑚𝜋 𝑚∙𝜋∙𝑥 𝑛∙𝜋∙𝑦

𝐸𝑦 𝜔 = −𝑖𝐸0 ∙ 2 ∙ 𝑏 ∙ 𝑠𝑖𝑛 ∙ 𝑐𝑜𝑠 ∙ 𝑒 −𝑖𝛽𝑧 ∙ 𝑒 𝑖𝜔𝑡 𝐸𝑦 𝜔 = −𝑖𝐻0 ∙ ∙ 𝑒 −𝑖𝛽𝑧 ∙ 𝑒 𝑖𝜔𝑡
𝑘𝑐 𝑎 𝑏 2 ∙ 𝑎 ∙ 𝑠𝑖𝑛 𝑎
∙ 𝑐𝑜𝑠
𝑏
𝑘𝑐

𝑚∙𝜋∙𝑥 𝑛∙𝜋∙𝑦
𝐸𝑧 𝜔 = 𝐸0 ∙ 𝑠𝑖𝑛 ∙ 𝑠𝑖𝑛 ∙ 𝑒 −𝑖𝛽𝑧 ∙ 𝑒 𝑖𝜔𝑡
𝑎 𝑏 𝐸𝑧 𝜔 = 0

𝜔 ∙ 𝜀 𝑛𝜋 𝑚∙𝜋∙𝑥 𝑛∙𝜋∙𝑦 𝛽 𝑚𝜋 𝑚∙𝜋∙𝑥 𝑛∙𝜋∙𝑦

𝐻𝑥 𝜔 = 𝑖𝐸0 ∙ 2 ∙ 𝑏 ∙ 𝑠𝑖𝑛 ∙ 𝑐𝑜𝑠 ∙ 𝑒 −𝑖𝛽𝑧 ∙ 𝑒 𝑖𝜔𝑡 𝐻𝑥 𝜔 = 𝑖𝐻0 ∙ 2 ∙ 𝑎 ∙ 𝑠𝑖𝑛 ∙ 𝑐𝑜𝑠 ∙ 𝑒 −𝑖𝛽𝑧 ∙ 𝑒 𝑖𝜔𝑡
𝑘𝑐 𝑎 𝑏 𝑘𝑐 𝑎 𝑏

𝜔 ∙ 𝜀 𝑚𝜋 𝑚∙𝜋∙𝑥 𝑛∙𝜋∙𝑦 𝛽 𝑛𝜋 𝑚∙𝜋∙𝑥 𝑛∙𝜋∙𝑦

𝐻𝑦 𝜔 = −𝑖𝐸0 ∙ 2 ∙ 𝑎 ∙ 𝑐𝑜𝑠 ∙ 𝑠𝑖𝑛 ∙ 𝑒 −𝑖𝛽𝑧 ∙ 𝑒 𝑖𝜔𝑡 𝐻𝑦 𝜔 = 𝑖𝐻0 ∙ ∙ 𝑒 −𝑖𝛽𝑧 ∙ 𝑒 𝑖𝜔𝑡
𝑘𝑐 𝑎 𝑏 2 ∙ 𝑏 ∙ 𝑐𝑜𝑠 𝑎
∙ 𝑠𝑖𝑛
𝑏
𝑘𝑐

𝑚∙𝜋∙𝑥 𝑛∙𝜋∙𝑦
𝐻𝑧 𝜔 = 0 𝐻𝑧 𝜔 = 𝐻0 ∙ 𝑐𝑜𝑠 ∙ 𝑐𝑜𝑠 ∙ 𝑒 −𝑖𝛽𝑧 ∙ 𝑒 𝑖𝜔𝑡
𝑎 𝑏

n, m ≥ 1 (m or n = 0  𝐸𝑧 𝜔 = 0) 𝐸𝑥 = 𝑍𝑇𝑀 ⋅ 𝐻𝑦 n, m ≥ 0 (n=m=0 not allowed, 𝐻𝑧 𝜔 static) 𝐸𝑥 = 𝑍𝑇𝐸 ⋅ 𝐻𝑦

TM11 is 1st propagating TM-mode TE10 is first propagating TE-mode
𝐸𝑦 = −𝑍𝑇𝑀 ⋅ 𝐻𝑥 𝐸𝑦 = −𝑍𝑇𝐸 ⋅ 𝐻𝑥

2 2 𝜋 𝑚 2 𝑛 2
𝑚∙𝜋 𝑛∙𝜋 𝑓𝑐 2
𝑘𝑐2 = + 𝜔𝑐 = 2𝜋𝑓𝑐 = ∙ + 𝛽= 𝜇𝜖𝜔 ∙ 1 −
𝑎 𝑏 𝜇𝜖 𝑎 𝑏 𝑓2

Note: Cutoff frequency determined by interior medium (permittivity, permeability)

and waveguide internal dimensions a and b (a = inner width, b = inner height) 50
 Mode Pattern TE- and TM-mode, respectively
TE10

Surface current distribution for the TE10 mode at E-field (solid) lines and H-field (dashed) lines at a fixed time.
a fixed time. The mode moves in z direction.

TM11

E-field (solid) lines and H-field (dashed) lines at a fixed time.

These figures are taken from a textbook from Prof. Z. Popovic,
‘Electromagnetics Around Us: Some Basic Concepts’
 Mode Pattern

TE10 TE11 TE21

TE20 TM11 TM21