applied

sciences
Communication
Computational Design of a Broadband In-Line
Coaxial-to-Rectangular Waveguide Transition
Qiongyue Zhang 1 , Songyuan Xu 1 , Jiwon Heo 1 , Erdenesukh Altanzaya 1 , Galsan-Yondon Ariunbold 1 ,
Delger Otgonbat 1 , Chan-Soo Lee 1 , Bierng-Chearl Ahn 1 , Shu Li 2, * and Seong-Gon Choi 1, *

1 School of Electric and Computer Engineering, Chungbuk National University,

Cheongju 28644, Republic of Korea; [email protected] (Q.Z.); [email protected] (S.X.);
[email protected] (J.H.); [email protected] (E.A.); [email protected] (G.-Y.A.);
[email protected] (D.O.); [email protected] (C.-S.L.); [email protected] (B.-C.A.)
2 Jiangsu Hengxin Technology Co., Ltd., 138 Taodu Road, Yixing 214222, China
* Correspondence: [email protected] (S.L.); [email protected] (S.-G.C.); Tel.: +86-186-6129-9939 (S.L.)

Abstract: This Communication presents a computational design approach for a broadband in-line
coaxial-to-rectangular waveguide transition having a 2.83:1 bandwidth. Existing designs have
a bandwidth ranging from 1.36:1 to 2:1. The proposed transition consists of a four-step ridge
transformer and three sections of the rectangular waveguide. The proposed design approach extends
the bandwidth of the transition by a theory-based design of the width and height of the three-section
waveguides and by a computer-aided optimization of the stepped-ridge transformer. The dimensions
of the three waveguide sections are determined such that the transition operates from very close to
the TE10 -mode cutoff to slightly less than the TE30 -mode cutoff. A four-step ridge transformer was
computer-optimized for a low reflection coefficient over a broad bandwidth. The proposed design
approach was applied to the design of a coaxial-to-WR75 transition whose reflection coefficient is less
than −22.6 dB at 8.14–23.00 GHz (2.83:1 bandwidth). The dimensions of the proposed transition were
obtained using an electromagnetic simulation tool. The increased bandwidth of the transition can be
very useful, especially in measurement applications.
Citation: Zhang, Q.; Xu, S.; Heo, J.;
Altanzaya, E.; Ariunbold, G.-Y.;
Keywords: transition; coaxial; rectangular waveguide; broadband
Otgonbat, D.; Lee, C.-S.; Ahn, B.-C.; Li,
S.; Choi, S.-G. Computational Design
of a Broadband In-Line
Coaxial-to-Rectangular Waveguide
Transition. Appl. Sci. 2024, 14, 74.
1. Introduction
https://doi.org/10.3390/ Rectangular waveguide-to-coaxial transitions are commonly used in communication
app14010074 and measurement systems to convert the coaxial TEM-mode signal into the rectangular
waveguide TE10 -mode wave and vice versa. In the fields of advanced multi-system radars,
Academic Editors: Giovanni Maria
Sardi and Walter Fuscaldo
radio telescopes, near-field measurements, and material dielectric constant measurements,
broadband operation is often required to cover a standard waveguide band and its adjacent
Received: 20 November 2023 ones. For example, using the WR75 waveguide (10.0–15.0 GHz), one can cover WR90
Revised: 12 December 2023 (8.2–12.4 GHz), WR62 (12.4–18.0 GHz), and WR51 (15.00–22.00 GHz) bands if higher-order
Accepted: 18 December 2023 modes are suppressed. A coaxial-to-rectangular waveguide transition can be either of right-
Published: 21 December 2023
angle type or in-line type. For some applications, in-line transitions are more convenient.
Commercial products are available in either type.
Research on the coaxial-to-waveguide transition is believed to have started during
Copyright: © 2023 by the authors.
World War II. Volume 9 (1948) of the MIT Radiation Laboratory Series describes the design
Licensee MDPI, Basel, Switzerland. of the coaxial-to-rectangular waveguide transition in a significant number of pages [1].
This article is an open access article This volume deals with transitions employing a right-angle probe of circular cylindrical or
distributed under the terms and wine glass or hemispherical shape along with various matching structures such as a coaxial
conditions of the Creative Commons stub, a symmetrical probe gap, a dielectric cylinder around the probe, and a cross bar.
Attribution (CC BY) license (https:// In [1], wideband right-angle or in-line transitions are also presented where a single or dual
creativecommons.org/licenses/by/ continuously curved-ridge transformer is employed. The book also describes narrow-band
4.0/). in-line transition designs with E- or H-plane loop coupling or resonant-slot coupling.

Appl. Sci. 2024, 14, 74. https://doi.org/10.3390/app14010074 https://www.mdpi.com/journal/applsci

Appl. Sci. 2024, 14, 74 2 of 15

A journal paper on the coaxial-to-waveguide transition dates back to a 1947 work by S.

B. Cohn [2], where various designs of the right-angle transition were presented along with
a design theory that includes the transmission line and the characteristic impedance. As
one of the fundamental high-frequency devices, the enduring importance of the coaxial
transition can be seen in its continued research interest nowadays [3–19].
The in-line coaxial transition has been investigated from the early days of microwave
technology developments [1–9], when the engineering designs were carried out by ex-
periments guided by an elementary theory. Theoretical analysis has been restricted to
transitions constructed with a simple rectangular loop mode converter [6,8]. A right-angle
transition employing a single or double continuously tapered ridge was already known in
the 1940s [1,2].
G. J. Wheeler appears to be the first engineer to employ a stepped TEM-TE10 mode
transformer in the design of an in-line coaxial-to-rectangular waveguide transition [3]. In
1957, he achieved a VSWR of less than 1.10 at 2.6–3.8 GHz and a VSWR of less than 1.19
at 4.8–6.5 GHz with his transition. In 1963, Dix used a three-step ridge transformer plus
a coaxial quarter-wave transformer to design and test an in-line transition operating at
2.5–4.1 GHz (1.64:1 bandwidth) with a VSWR < 1.22 or a reflection coefficient < −20.1 dB [4].
Das and co-workers presented the theoretical analysis and the design of an in-line transi-
tion employing an inverted-L shaped loop probe [5]. Their transition showed a narrow
bandwidth. It appears that Deshpande, Das, and Sanyal are the ones who presented a
rigorous theoretical analysis of the inverted-L shape end launcher assuming a sinusoidal
current distribution on the inverted-L probe wire [6]. Later, Saad improved the theory
of Deshpande, Das, and Sanyal to include more general cases [7]. Gangadharan used a
stepped wire loop to experimentally design an in-line transition operating at 8.2–12.0 GHz
(1.46:1 bandwidth) with a VSWR < 1.25 or a reflection coefficient < −19.1 dB [8].
Chan and co-workers employed a dual stepped-ridge transformer along with a rectan-
gular coaxial matching section for the theoretical design of an in-line transition operating
at 0.85–1.16 GHz (1.36:1 bandwidth) with a reflection coefficient < −32.3 dB [9]. They
analyzed the transition using the mode-matching method. Levy and Hendrick used an
equivalent-circuit method and an in-house computer program to theoretically design an in-
line transition with a three-step ridge transformer [10]. Their transition showed a reflection
coefficient < −30.0 dB at 3.05–4.60 GHz (1.51:1 bandwidth). Durga and co-workers [11] and
Carmel and co-workers [12] used a stepped-ridge transformer in their design of an in-line
transition. Their transition showed performance similar to that in the preceding works.
Cano and Mediavilla achieved a 2:1 bandwidth in an in-line rectangular waveguide-to-
coaxial transition [13] using oversized mode conversion and a stepped-ridge transformer.
Subsequent works [14–19] on the in-line transition were not targeted for extending the
bandwidth beyond 2:1. In [14], a two-step ridge transformer was employed in a transition
working at 9.5–10.8 GHz with a reflection coefficient of less than −20 dB. Ye and co-workers
used a single waveguide and a three-step ridge transformer to design an in-line transition
operating at 11.5–19.0 GHz (1.65:1 bandwidth) with a reflection coefficient < −20.0 dB [15].
Bilik employed a two-step ridge transformer in a high-power in-line transition design
working at 2.4–2.5 GHz with a reflection coefficient of less than −25 dB [16]. In [17], a
triple-stepped probe extended out of a coaxial cable was used to design an in-line transition
working at 7.6–12.2 GHz (1.61:1 bandwidth) with a VSWR of less than 1.5 (reflection
coefficient < −14.0 dB).
Simone and co-workers realized a five-step ridge transition operating at 33–50 GHz
(1.52:1 bandwidth) with a reflection coefficient < −25.0 dB [18]. Shu and co-workers
used a four-step ridge transformer to realize WR28, WR22, and WR19 transitions op-
erating at 26.5–40.0 GHz (1.51:1 bandwidth), at 33–50 GHz (1.52:1 bandwidth), and at
40–60 GHz (1.50:1 bandwidth) with a reflection coefficient of less than −23.0 dB, −28.0 dB,
and −25.0 dB, respectively [19]. A patent for an in-line coaxial-to-rectangular waveguide
transition was granted as early as in 1973 [20]. In [20], a four-step ridge transformer was
Appl. Sci. 2024, 14, 74 3 of 15

employed to design a transition working at 7.0–11.0 GHz (1.57:1 bandwidth) with a VSWR
of less than 1.28 or a reflection coefficient < −18.2 dB.
A comprehensive literature survey shows that the concept of the stepped-ridge trans-
former and the stepped waveguide height transformer has been applied to the design of
the broadband right-angle coaxial-to-rectangular waveguide transitions [21–27]. In [26],
a double stepped-ridge transformer was used. Study of the literature shows that the
stepped-ridge transformer has been a preferred choice in the design of broadband coaxial-
to-rectangular waveguide transitions.
In this Communication, we present a design approach for the broadband in-line
coaxial-to-rectangular waveguide transition and two design examples having a bandwidth
exceeding 2.8:1. The proposed transition employs a stepped-ridge transformer for the
conversion of the coaxial TEM mode into the rectangular waveguide TE10 mode. The
transformer was placed in a converter waveguide whose broad-wall width is slightly larger
than that of the output waveguide. The narrow-wall height of the converter waveguide
was reduced from a standard height (one-half of the broad-wall width) so that the TE11 -
mode cutoff coincided with the TE30 -mode cutoff in the converter waveguide. A matching
waveguide was placed between the converter waveguide and the output waveguide to
reduce reflections from a step discontinuity in the waveguide broad wall.
Using the powerful capability of a modern electromagnetic simulation tool—CST Stu-
dio SuiteTM V2023, a widely used simulation tool in the antenna and RF community—the
dimensions of the proposed transition were optimized for a reflection coefficient < −20 dB
over a bandwidth approaching the theoretical maximum of 3:1.
In many papers, authors present a design with experimental validation. In others,
a design method is validated with computer simulation [9,11,18,28–33]. We believe that
the accuracy of the simulation tool used in this Communication is sufficient to prove the
proposed design approach. All the dimensions of the design examples are given, and
anyone can verify the results presented in this paper.
The construction of this Communication is as follows. In Section 2, we review the
performance of a representative commercial in-line coaxial-to-rectangular transition to
show the state of the art in practice. It is followed by the concept and the equations of
the proposed design approach. In Section 3, two design examples are presented: one
for a WR75 waveguide with a reduced narrow-wall height and the other for a standard
WR75 waveguide. Finally, in Section 4, we draw conclusions and suggest possible future
research topics.

2. Design Approach
There are many commercial products of the coaxial-to-rectangular waveguide tran-
sitions either of right-angle type or of in-line type. Figure 1 shows a commercial in-line
coaxial transition in a WR34 waveguide [28]. By approximately extrapolating the man-
ufacturer’s reflection coefficient data, it can be seen that the transition operates with a
reflection coefficient < −22.7 dB from about 21 GHz to about 34 GHz. The start and end
frequencies (21 and 34 GHz) are 1.21 and 1.96 times the fundamental TE10 -mode cutoff
frequency, respectively (1.62:1 bandwidth). If one can extend the bandwidth close to 3:1, it
will be very useful in many applications. Especially in measurement applications, it will
reduce the number of measurement accessories and the measurement time.
Figure 2 shows the waveguide wall structure in the proposed design approach. The
symbols W, W 1 , and W 2 denote the output waveguide, the impedance matching waveguide,
and the TEM-TE10 mode converter waveguide, respectively; a, a1 , and a2 their broad-wall
width, respectively; b, b1 , and b2 their narrow-wall height, respectively. A stepped-ridge
transformer is placed in the mode-converter waveguide W 2 .
Appl.
Appl.Sci. 2024,14,
Sci.2024, 14,74
x FOR PEER REVIEW 44 of 15
15

Figure 1. A commercial in-line coaxial-to-rectangular waveguide transition by Exceed Microwave

and its reflection coefficient [28].

Figure 2 shows the waveguide wall structure in the proposed design approach. The
symbols W, W1, and W2 denote the output waveguide, the impedance matching wave-
guide, and the TEM-TE10 mode converter waveguide, respectively; a, a1, and a2 their broad-
Figure 1.1. AAcommercial
commercialin-line
in-line coaxial-to-rectangularwaveguide
waveguide transition
transition by
byExceed
ExceedMicrowave
Microwave
Figure
wall width, respectively; b, bcoaxial-to-rectangular
1, and b2 their narrow-wall height, respectively. A stepped-
and its reflection coefficient [28].
and its transformer
ridge reflection coefficient
is placed[28].in the mode-converter waveguide W2.

Figure 2 shows the waveguide wall structure in the proposed design approach. The
symbols W, W1, and W2 denote the output waveguide, the impedance matching wave-
guide, and the TEM-TE10 mode converter waveguide, respectively; a, a1, and a2 their broad-
wall width, respectively; b, b1, and b2 their narrow-wall height, respectively. A stepped-
ridge transformer is placed in the mode-converter waveguide W2.

Figure2.2.Waveguide
Figure Waveguidestructure
structureininthe
theproposed
proposed design
design approach
approach (W,(W, output
output waveguide;
waveguide; W 1W 1, match-
, matching
ing waveguide;
waveguide; W2 , W 2, converter waveguide).
converter waveguide).

In aa rectangular
In rectangular waveguide,
waveguide, the
the TE
TEmn
mn/TM -mode
/TMmnmn -modecutoff
cutoffoccurs
occursatatthe
thefollowing:
following:
fmn = (c/2) [(m/a)2 + (n/b)2]1/2 (1)
fmn = (c/2) [(m/a)2 + (n/b)2 ]1/2 (1)
Figure 2. Waveguide structure in the proposed design approach (W, output waveguide; W1, match-
where c is the speed of light, a is the waveguide broad-wall width, and b is the narrow-
ing waveguide; W2, converter
c is theFrom waveguide).
a iswill
where
wall height. speed of light,
now on, we the use
waveguide broad-wall
the TE modes to mean width, TM bmodes
the and is the narrow-wall
as well since
height. From now on, we will
their cutoff frequencies are the same. use the TE modes to mean the TM modes as well since their
Infrequencies
cutoffIn a rectangular
are waveguide,
the same. the TEmn/TMmn-mode cutoff occurs at the following:
a standard output waveguide W, where b is equal to a/2, the cutoff frequency of
the TEIn10a, standard
TE20, TE11output
, and TEwaveguide
30 modes =W,
fmn is where
(c/2)
given bybthe
[(m/a) is2 +equal
(n/b)to a/2, the cutoff frequency of the
2]1/2
following: (1)
TE10 , TE20 , TE11 , and TE30 modes is given by the following:
where c is the speed of light, a is the waveguide f10 = c/(2a) broad-wall width, and b is the narrow- (2a)
wall height. From now on, we will use fthe 10 =TE modes to mean the TM modes as well since
c/(2a) (2a)
their cutoff frequencies are the same. f20 = 2c/(2a) = 2f10 (2b)
f 20 = 2c/(2a) = 2f 10 (2b)
In a standard output waveguide W, where b is equal to a/2, the cutoff frequency of
√
the TE10, TE20, TE11, and TE30 modes is 5c/(2a)==2.24f
given by
f 11 =f11 =5c/(2a) the following:
2.24f1010 (2c)
f10 = c/(2a)
f 30 = 3c/(2a) = 3f (2a)
(2d)
f30 = 3c/(2a) = 3f1010 (2d)
Since the stepped-ridge transformer is symmetric in the H plane of the mode-converter
Since the stepped-ridge transformer f20 = is
2c/(2a) = 2f10 in the H plane of the mode-con-
symmetric (2b)
waveguide W 2 , the antisymmetric TE20 mode is not excited in W 2 , and the next higher-order
verter waveguide W2, the antisymmetric TE20 mode is not excited in W2, and the next
mode will be the TE11 or TE30 mode. In a standard rectangular waveguide, the TE11 -mode
higher-order mode will be the TE11 for 11 =TE5c/(2a)
30 mode.= 2.24f
In a 10standard rectangular waveguide, (2c)
cutoff frequency is 2.24 times the fundamental TE10 -mode cutoff frequency. Thus, the
the TE11-mode cutoff frequency is 2.24 times the fundamental TE10-mode cutoff frequency.
bandwidth limit of a transition constructed using a standard rectangular waveguide is
f30 = 3c/(2a) = 3f10 (2d)
2.24:1. The achieved bandwidth, however, is about 90% of the theoretical maximum or
2:1 since
Sinceit the
is difficult to operate
stepped-ridge a transition
transformer at very close
is symmetric to the
in the cutoffofofthe
H plane themode-con-
TE10 and
TE 11 modes.
verter waveguide W2, the antisymmetric TE20 mode is not excited in W2, and the next
higher-order mode will be the TE11 or TE30 mode. In a standard rectangular waveguide,
the TE11-mode cutoff frequency is 2.24 times the fundamental TE10-mode cutoff frequency.
Appl. Sci. 2024, 14, 74 5 of 15

At frequencies above the cutoff, the TE30 mode can be generated in W 2 along with the
fundamental TE10 mode and subsequently transmitted to the output waveguide W. One
can make a transition operate approximately from the TE10 cutoff to the TE30 cutoff in the
output waveguide (about 3:1 bandwidth) if the TE11 mode is suppressed in the waveguides
W, W 1 , and W 2 . If the discontinuity effect is small between W 2 and W, the generation of
the TE11 mode will be minimized.
This can be achieved by reducing the narrow-wall height of the mode-converter
waveguide W 2 so that the TE11 cutoff of W 2 coincides with the TE30 cutoff of W 2 :

f 11 = f 30 (3)

Then, the narrow-wall height b2 of W 2 can be obtained from the following:

(c/2) [(1/a2 )2 + (1/b2 )2 ]1/2 = 3c/(2a2 ) (4)

or
√
b2 = a2 / 8 = 0.354a2 (5)
With a given waveguide broad-wall width a2 , the TE30 -mode cutoff frequency f 30 is
fixed, and the TE11 -mode cutoff frequency f 11 can be increased by reducing the waveguide
narrow-wall height b2 . Making f 11 larger than f 30 offers no gain since the frequency
upper limit will still be f 30 , and the height difference between the output and converter
waveguides will be increased, resulting in an increased mismatch between W 2 and W.
The broad-wall width of the converter waveguide is chosen to satisfy the follow-
ing condition:
a2 = (1 + ∆)a (6)
where ∆ is a design parameter in the range of 0.04–0.08, and a is the broad-wall width of the
output waveguide W. Equation (6) ensures a transition to operate close to the TE10 -mode
cutoff of the output waveguide. The larger ∆, the closer to the TE10 cutoff the start frequency
of operation. Large values of ∆ will increase the reflection at the waveguide width step
discontinuity, and more than one section of the matching waveguide may be required.
From the foregoing discussion, we find that the proposed transition will work in the
following frequency range:
f1 < f < f2 (7)
where
f 1 = max(f 10_ o, f 10_ m, f 10_ c) (8)
The 10 symbol in the subscript denotes the cutoff frequency of the fundamental TE10
mode. The symbols o, m, and c refer to the output, matching, and converter waveguides, re-
spectively. The frequency upper limit f 2 is determined by the level of the higher-order TE11
and TE30 modes generated in the transition. The frequency f 2 is more or less determined by
the cutoff frequency of TE11 and TE30 modes in the waveguides W, W 1 , and W 2 . Of course,
the bandwidth of the transition is also influenced by the performance of the TEM-to-TE10
mode transformer, which is a stepped-ridge transformer in our case.
The impedance matching waveguide W 1 is employed to match the impedance of W
with that of W 2 . The broad-wall width a1 of W 1 is between a and a2 . The width a1 , height
b1 , and length L1 of W 1 are determined using an automatic optimization function of the
simulation tool. The difference in the height of W and W 2 can be matched in the form of a
single- or multi-step transformer or in the form of a continuously tapered transformer.
It is important to note that the TE11 mode should not be generated in the region
between the output waveguide W and the converter waveguide W 2 in the operating fre-
quency range of the transition. In fact, during the parametric analysis and the optimization,
we observed that the TE11 mode can be generated in the converter waveguide W 2 due
to the presence of the stepped-ridge transformer, indicating that a careful design is re-
 b2 to b. One may use a stepped waveguide transformer, in which case there will be more
parameters to be optimized. If necessary, the linear taper may have a zero slope or a neg-
ative slope to make the height of the output waveguide W1 less than or equal to that of the
converter waveguide W2.
Appl. Sci. 2024, 14, 74 A stepped-ridge transformer S converts the TEM mode of the coaxial connector X to
6 of 15
the TE10 mode of the converter waveguide W2. The coaxial probe P is connected to the
stepped-ridge transformer S across a cavity C through a probe hole H in the back short B.
The
quired. four-step ridge
In the next transformer
section, S plays
we apply an important
the foregoing designrole, as it is crucial
approach to the in determining
design of two
the level of the reflection
coaxial-to-WR75 transitionscoefficient
having a and the degree
bandwidth of the higher-order
exceeding 2.8:1. mode generation.
A stepped-ridge transformer seems to be better in achieving a low reflection coefficient
3.over
Designa broad bandwidth than a continuously tapered transformer since it has more control
Examples
parameters. The
Figure 3 shows number of stepsofinan
the structure the ridgecoaxial-to-rectangular
in-line transformer is important for broadband
waveguide op-
transition
toeration. It has been
be designed found the
according thatapproach
four stepsproposed
are sufficient for our
in this transition design.
Communication. TheThe edges
output
of the first ridge—the one connected to the coaxial probe P—are
waveguide W is a standard waveguide with broad-wall width a and narrow-wall height b rounded in the circular
(bshape
= a/2 of in radius
most ofRtheforstandard
better impedance
waveguides). matching. In addition
The converter to S, the
waveguide W 2cavity C and the
has broad-wall
probe hole H play a role in achieving a low reflection coefficient over a
width a2 slightly larger than a (as given by Equation (6)) and narrow-wall height b2 smaller broad frequency
range.
than b (as given by Equation (5)).

Figure3.3.Structure
Figure Structure of
of the
the proposed
proposed transition. W,output
transition. W, waveguide;WW
outputwaveguide; 1 , matching
1, matching
waveguide;
waveguide; W2,
Wmode-converter
2 , mode-converter waveguide; S, stepped-ridge transformer; C, matching cavity; B, back
waveguide; S, stepped-ridge transformer; C, matching cavity; B, back short; short; P,
P, co-
coaxial
axial probe; X,coaxial
probe;X, coaxialconnector; H,probe
connector;H, probehole; T,linear
hole;T, lineartaper.
taper.

The matching waveguide W 1 is used to ensure a broadband impedance matching

between the converter waveguide W 2 and output waveguides W, which have different
broad-wall widths. A linear taper T in the narrow-wall height is used in the waveguides
W 1 and W for design simplicity and for a smooth change in the narrow-wall height from
b2 to b. One may use a stepped waveguide transformer, in which case there will be more
parameters to be optimized. If necessary, the linear taper may have a zero slope or a
negative slope to make the height of the output waveguide W 1 less than or equal to that of
the converter waveguide W 2 .
A stepped-ridge transformer S converts the TEM mode of the coaxial connector X to
the TE10 mode of the converter waveguide W 2 . The coaxial probe P is connected to the
stepped-ridge transformer S across a cavity C through a probe hole H in the back short B.
The four-step ridge transformer S plays an important role, as it is crucial in determining
the level of the reflection coefficient and the degree of the higher-order mode generation.
A stepped-ridge transformer seems to be better in achieving a low reflection coefficient
over a broad bandwidth than a continuously tapered transformer since it has more control
parameters. The number of steps in the ridge transformer is important for broadband
operation. It has been found that four steps are sufficient for our transition design. The
edges of the first ridge—the one connected to the coaxial probe P—are rounded in the
circular shape of radius R for better impedance matching. In addition to S, the cavity C
and the probe hole H play a role in achieving a low reflection coefficient over a broad
frequency range.
Figure 4 shows the dimensional parameters of the transition. With the given dimen-
sions a and b of the output waveguide W, the design of the transition starts with determining
a2 and b2 of the converter waveguide W 2 , which was explained in the previous section.
The height b2 of the converter waveguide W 2 is held constant throughout its length to
suppress the generation of the TE11 mode. The length (L2 + L3 in Figure 4) of the converter
waveguide W 2 is one of the optimization parameters.
 H can be optimized by using a powerful modern simulation tool such as CST Studio
SuiteTM V2023. The ridge-edge-rounding radius R is held at a constant value. It is virtually
impossible to obtain a low reflection coefficient over a broad frequency range by an ana-
lytical method since the calculation of the reflection coefficient is extremely complicated
Appl. Sci. 2024, 14, 74 if not impossible. An experimental approach is not suitable either because there are7 too
of 15
many parameters to be optimized.

Figure
Figure4.4.Dimensional
Dimensionalparameters parametersof
of the
the proposed
proposed transition.
transition. b,
b, bb11,,and
and bb22, ,waveguide
waveguidenarrow-wall
narrow-wall
height; L,LL11, ,LL2,2and
height;L, , andLL
3, waveguide length; F, D, probe length and position; a, a1, a2, waveguide broad-
, waveguide length; F, D, probe length and position; a, a , a , waveguide broad-
3 1 2
wall width; R, radius of ridge-edge rounding; G, ridge-to-back short gap; t, back short thickness.
wall width; R, radius of ridge-edge rounding; G, ridge-to-back short gap; t, back short thickness.

Prior to thea1design
The width and theoflength
a WR75L1 oftransition,
the matchinga transition
waveguide with
are aamong
linearthetaper T of zero
parameters to
slope (b = b1 = The
be optimized. b2) was designed.
length of a linearThis aided
taper in estimating
T is the the increased
sum of the output waveguidelevellength
of higher-
L and
order modes due
the matching to in thelength
waveguide linear Ltaper T of nonzero slope. In the final design of the WR75
1 , which is optimized for a reflection coefficient < −20 dB,
transition, a linear taper with a slope
especially at the start frequency in the operatingwas added, and dimensions were re-adjusted. Figure
bandwidth.
5 shows the structure of a transition with a linear
The dimensions of the four-step ridge transformer S, taper T the
of zero slope
cavity whose
C, and dimensions
the probe hole H
are
canlisted in Table 1.
be optimized byAn SMAa connector
using powerful modern was used for the coaxial
simulation inputasport.
tool such CSTFigure
Studio6Suite
showsTM

the reflection
V2023. and transmission coefficients
The ridge-edge-rounding radius R is of theatdesigned
held a constant transition
value. It in both a decibel
is virtually impos-
scale
sible and a natural
to obtain a lowscale. The reflection
reflection coefficient coefficient is lessfrequency
over a broad than −24 range
dB at 8.00–22.54 GHz
by an analytical
(2.82:1
method bandwidth). A resonance
since the calculation of spike was generated
the reflection at 21.24
coefficient GHz, causing
is extremely the reflection
complicated if not
coefficient
impossible. to An
increase from −30approach
experimental dB to −24 is dB.
not suitable either because there are too many
Figure 7toshows
parameters the transmission coefficient of the higher-order modes from the coax-
be optimized.
ial portPrior
to theto output
the design of a WR75
waveguide port.transition, a transition
In the operating frequencywithrange
a linear taper T ofGHz,
of 8.00–22.54 zero
slope
the TE11 = b1TE
(band b2modes
= 30 ) was designed. This aided
are predominant in estimating
the highest ranging thefrom
increased
−70.0 dBlevel of higher-
to −40.2 dB.
order
The TM modes
11 mode due to innext-highest
is the the linear taper T of nonzero
one with a maximum slope. In the
level final
of −72 dBdesign
at 22.54 ofGHz.
the WR75
The
transition,
TE 31 and TM a 31
linear
modestaper
arewith a slope
in the level was added,
of −130 dB, and
−150dimensions were re-adjusted. Figure 5
dB, respectively.
shows the structure of a transition with a linear taper T of zero slope whose dimensions are
listed in Table 1. An SMA connector was used for the coaxial input port. Figure 6 shows
the reflection and transmission coefficients of the designed transition in both a decibel
scale and a natural scale. The reflection coefficient is less than −24 dB at 8.00–22.54 GHz
Appl. Sci. 2024, 14, x FOR PEER REVIEW 8 of 15
(2.82:1 bandwidth). A resonance spike was generated at 21.24 GHz, causing the reflection
coefficient to increase from −30 dB to −24 dB.

(a) (b)
Figure 5.
Figure 5. Structure
Structure of
of aa transition
transition with
with aa linear
linear taper
taper of
of zero
zero slope:
slope: (a)
(a) transparent
transparent perspective
perspective view;
view;
(b) cross-sectional view.
(b) cross-sectional view.

Table 1. Dimensions of the designed transition with a linear taper of zero slope.

Parameter Value (mm) Parameter Value (mm)

a, b, L 19.05, 6.57, 21.27 F, D 2.33, 4.59
a1, b1, L1 19.47, 6.57, 16.92 t, G, R 1.59, 0.79, 0.90
 (a) (b)
Appl. Sci. 2024, 14, 74 8 of 15
Figure 5. Structure of a transition with a linear taper of zero slope: (a) transparent perspective view;
(b) cross-sectional view.

Table
Table1.1.Dimensions
Dimensionsofofthe
thedesigned
designedtransition
transitionwith
withaalinear
lineartaper
taperofofzero
zeroslope.
slope.

Parameter
Parameter Value (mm)
Value (mm) Parameter
Parameter Value
Value (mm)
(mm)
a, b,a, Lb, L 19.05, 6.57, 21.27
19.05, 6.57, 21.27 F, D
F, D 2.33, 4.59
2.33, 4.59
a1, ab1,,bL1, L 19.47, 6.57, 16.92
19.47, 6.57, 16.92
t, G, R
t, G, R
1.59, 0.79, 0.90
1.59, 0.79, 0.90
1 1 1
a2, b2, L2, L3 19.89, 6.57, 2.89, 23.78 Ridge width/height 3.55/0.91, 1.99, 3.75, 5.27
a2 , b2 , L2 , L3 19.89, 6.57, 2.89, 23.78 Ridge width/height 3.55/0.91, 1.99, 3.75, 5.27
Coaxial
Coaxial 1.27/4.11, εr = 2.1 (SMA) Probe hole Dia. = 2.95, Len. = 1.59
connector 1.27/4.11, εr = 2.1 (SMA) Probe hole Dia. = 2.95, Len. = 1.59
connector

(a)

(b)
Figure
Figure6.6.Reflection
Reflectionand transmission
and coefficients
transmission of aoftransition
coefficients withwith
a transition a linear tapertaper
a linear of zero slope:
of zero (a)
slope:
in(a)
dBinscale and (b) in natural scale.
dB scale and (b) in natural scale.

Figure 7 shows the transmission coefficient of the higher-order modes from the coaxial
port to the output waveguide port. In the operating frequency range of 8.00–22.54 GHz,
the TE11 and TE30 modes are predominant the highest ranging from −70.0 dB to −40.2 dB.
The TM11 mode is the next-highest one with a maximum level of −72 dB at 22.54 GHz. The
TE31 and TM31 modes are in the level of −130 dB, −150 dB, respectively.
After confirming the performance of a transition with a linear taper of zero slope,
the slope of the linear taper T was adjusted for the WR75 output waveguide, and the
transition was optimized. Starting with initial dimensions, we used the “Trust Region
Framework” optimization in CST Studio SuiteTM V2023 to obtain a final WR75 transition,
whose dimensions are given in Table 2. An SMA connector was used for the coaxial input
port. All the dimensional parameters except the port dimensions of the coaxial connector
Appl. Sci. 2024, 14, 74 9 of 15

and the output WR75 waveguide were optimized. The range of the parameter values to
be set
Appl. Sci. 2024, 14, x FOR PEER REVIEW in the optimization was obtained from the parametric analysis carried out prior
9 of to
15
the optimization.

Figure 7. Higher-order modes in a transition with a linear taper of zero

zero slope.
slope.

TableAfter confirming
2. Dimensions thedesigned
of the performance
WR75 of a transition with a linear taper of zero slope, the
transition.
slope of the linear taper T was adjusted for the WR75 output waveguide, and the transition
wasParameter
optimized. StartingValue with(mm)
initial dimensions,Parameter
we used the “Trust Region ValueFramework”
(mm)
optimization
a, b, L in CST Studio
19.05, Suite 15.01
8.06–9.53,TM V2023 to obtainF,a final
D WR75 transition, whose
2.17, 5.60 dimen-
sionsa1are given
, b1 , L1 in Table 2. An SMA connector
19.47, 6.57–8.06, 15.25 was used
t, G, R for the coaxial input port.
1.81, 1.21, 0.90All the
dimensional parameters except the port dimensions of the coaxial connector and the out-
a2 , b2 , L2 , L3 20.58, 6.57, 1.89, 20.05 Ridge width/height 3.23/0.37, 1.98, 3.86, 5.31
put WR75 waveguide were optimized. The range of the parameter values to be set in the
Coaxial was obtained from the parametric analysis carried out prior to the optimiza-
optimization 1.27/4.11, εr = 2.1 (SMA) Probe hole Dia. = 2.95, Len. = 1.81
connector
tion.

TableFigure
2. Dimensions8 shows thedesigned
of the reflection coefficients
WR75 transition.during an optimization of the designed
WR75 transition, where the curve in red is the reflection coefficient of the final design. In
Parameter Value (mm) Parameter Value (mm)
Figure 8, we observe that the reflection coefficient rapidly approaches −20 dB level and that
a, b, L 19.05, 8.06–9.53, 15.01 F, D
many more iterations are required to lower the reflection coefficient below 2.17,−5.60
20 dB. Over
200a1simulations
Appl. Sci. 2024, 14, x FOR PEER REVIEW , b1, L1 19.47,
of the 6.57–8.06, 15.25
transition’s t, G, R were carried out
reflection coefficient 1.81,before
1.21, 0.9010 of 15
obtaining
a , b , L
the final design.
2 2 2 , L 3 20.58, 6.57, 1.89, 20.05 Ridge width/height 3.23/0.37, 1.98, 3.86, 5.31
Coaxial
1.27/4.11, εr = 2.1 (SMA) Probe hole Dia. = 2.95, Len. = 1.81
connector

Figure 8 shows the reflection coefficients during an optimization of the designed

WR75 transition, where the curve in red is the reflection coefficient of the final design. In
Figure 8, we observe that the reflection coefficient rapidly approaches −20 dB level and
that many more iterations are required to lower the reflection coefficient below −20 dB.
Over 200 simulations of the transition’s reflection coefficient were carried out before ob-
taining the final design.
In the final optimized reflection coefficient, in Figure 8, we observe small ripples at
17–20 GHz, larger ripples in at 20–23 GHz, and a spike at 23.15 GHz. This is caused by
higher-order modes excited in the mode-converter waveguide W2, where the stepped-
ridge transformer acts as an obstacle giving rise to higher-order modes. This can more
clearly be understood with the help of Figure 9, where the transmission coefficients of
higher-order modes are plotted. At 17–20 GHz, the higher-order mode level is about −50
Figure
Figure
dB, 8. Reflection
8.
while Reflection coefficients
at 20–23 coefficients
GHz, duringincreases
during
it steadily anoptimization
an optimization ofthe
of
from −50 the designed
dBdesigned WR75
WR75
to −20 dB. transition.
transition.
At 23.15 GHz,TheThe curve
curve
all three
in red is the final
in red is the final
higher-order reflection
reflection
modes, coefficient.
coefficient.
namely TM11, TE30, and TE11 have a local peak, causing a large spike
in the reflection coefficient of the TE11 mode.
Figure 9 shows the transmission coefficient between the coaxial TEM mode and the
rectangular waveguide high-order modes. Analysis of the higher-order modes is im-
portant in the design of the proposed transition. In Table 3, we show the cutoff frequency
of the TE10, TE30, and TE11 modes in the designed WR75 transition. The lower frequency
Appl. Sci. 2024, 14, 74 Figure 8. Reflection coefficients during an optimization of the designed WR75 transition. The10curve
of 15
in red is the final reflection coefficient.

Figure 9 shows
In the final the transmission
optimized coefficient between
reflection coefficient, in Figurethe coaxial
8, we TEMsmall
observe mode and the
ripples at
rectangular
17–20 GHz, waveguide
larger ripples high-order
in at 20–23 modes.
GHz, Analysis
and a spike of at
the23.15
higher-order
GHz. Thismodes is im-
is caused by
higher-order
portant in themodes
designexcited
of the in the mode-converter
proposed transition. Inwaveguide
Table 3, we Wshow
2 , where
thethe stepped-ridge
cutoff frequency
transformer
of the TE10, TEacts30, as
andanTE
obstacle
11 modesgiving risedesigned
in the to higher-order modes. This
WR75 transition. canlower
The morefrequency
clearly be
understood
limit with the help
of the transition of Figure 9,
is determined bywhere
the TEthe transmission
10 cutoff frequency coefficients
7.87 GHzof(f10higher-order
) of the out-
modes
put are plotted.
waveguide W. In AtFigure
17–20 9, GHz, the higher-order
the spikes at 23.10 GHz modeare level
at theis−13.7
aboutdB−level
50 dB, whileby
caused at
20–23
the TMGHz,
11 mode,it steadily increases
at the −17.1 from
dB level due−to50 the
dB TE −mode,
to 30 20 dB. and
At 23.15
at theGHz,
−20.0all
dBthree
levelhigher-
due to
order
the TEmodes,
11 mode.namely
We noteTM 11 ,the
that TE30 , andof
cutoff TE 11 have
the a local
TE11, TM peak,
11, and TE30causing
modesaoccurs
large spike
at the in the
same
reflection coefficient
frequency of 23.10 GHz, of the
as TE 11 mode.
predicted by the design theory.

Figure 9. Higher-order
Higher-order modes in the designed WR75 transition.

Figure
We note9 in
shows
Figure the transmission
9 that coefficient
at 8.14–19.40 GHz, the between
TE12 modethe is
coaxial TEM mode
the highest, ranging and the
from
rectangular
−51.0 waveguide
dB to −41.4 dB excepthigh-order
for a TM modes.
11-modeAnalysis
spike ofof−44.6
the higher-order modes
dB at 17.7 GHz. is important
At 19.40– 23.00
in the the
GHz, design
TM11ofmodethe proposed transition.
is the largest (−51.0 In
dBTable 3, we
to −25.0 show
dB). the cutoff
In Figure 10, frequency of the
we also observe
TE , TE , and TE modes in the designed WR75 transition. The
that the TE11, TM11, and TE30 modes steadily increase at 20–23 GHz from about −50 dB
10 30 11 lower frequency limit
to
of the transition is determined by the TE cutoff frequency 7.87
−25 dB. This is in contrast to the transition10without a linear taper, where the GHz (f ) of the output
10 higher-order
waveguide
mode level is W.less
In Figure 9, the
than −47.4 dBspikes
in theatoperating
23.10 GHz −13.7 dB level caused by the
are at therange.
frequency
TM11Figure
mode,10 the −17.1
at shows dB level due to the
the simulated performance TE mode, and at the −
30 of the designed 20.0 dB
WR75 level due
transition in to
a
the TE11scale
decibel mode.andWe in note that the
a natural cutoff
scale of the
as well. TheTEreflection
11 , TM11 , coefficient
and TE30 modes
of the TEoccurs at the
10 mode is
same frequency
plotted along with of 23.10 GHz, as predicted
the transmission by the
coefficient design theory.
between the coaxial TEM mode and the
output waveguide TE10 mode. The reflection coefficient (|Γ|) is less than −22.6 dB at 8.14–
Table 3. Waveguide modes and their cutoff frequencies in the designed WR75 transition.

TE10 -Mode TE30 -Mode TE11 /TM11 -Mode

Dimension
Waveguide Cutoff Frequency Cutoff Frequency Cutoff Frequency
a × b (mm)
f 10 (GHz) f 30 (GHz) f 11 (GHz)
Output Waveguide W 19.05 × (8.06–9.53) 7.87 23.61 17.59–20.18
Matching Waveguide W 1 19.47 × (6.57–8.06) 7.70 23.10 20.13–24.08
Converter Waveguide W 2 20.58 × 6.57 7.28 21.85 23.95

We note in Figure 9 that at 8.14–19.40 GHz, the TE12 mode is the highest, ranging
from −51.0 dB to −41.4 dB except for a TM11 -mode spike of −44.6 dB at 17.7 GHz. At
19.40– 23.00 GHz, the TM11 mode is the largest (−51.0 dB to −25.0 dB). In Figure 10, we
also observe that the TE11 , TM11 , and TE30 modes steadily increase at 20–23 GHz from
about −50 dB to −25 dB. This is in contrast to the transition without a linear taper, where
the higher-order mode level is less than −47.4 dB in the operating frequency range.
 23.00 GHz (=0.974f30). The theoretical upper frequency limit of the transition is determined
by the level of the higher-order TE11 and TE30 modes generated in the waveguides W, W1,
and W2. A spike in the reflection coefficient is generated at 23.1 GHz, where the TE11, TM11,
and TE30 modes have a transmission peak all at the same time, as shown in Figure 9. The
Appl. Sci. 2024, 14, 74 transition operates from slightly above the TE10 mode cutoff to slightly below the 11TE of3015
cutoff. The resultant bandwidth is 2.83:1 or 94.3% of the theoretical limit of 3:1.

(a)

(b)
Figure
Figure10.
10.Reflection
Reflectionand
andtransmission
transmissioncoefficients
coefficientsofofthe
thedesigned
designedWR75
WR75transition:
transition:(a)(a)inindB
dBscale
scale
and
and(b)
(b)ininnatural
naturalunit.
unit.

InFigure
Table 3,
10we note the
shows thatsimulated
the cutoff performance
frequency of the TE11designed
of the , TM11 andWR75 TE30 modes ranges
transition in a
from 20.13
decibel GHzand
scale to 24.08 GHz in scale
in a natural the matching
as well. waveguide W1 and
The reflection from 17.59
coefficient GHz
of the TEto
10 23.61
mode
is plotted
GHz in the along
outputwith the transmission
waveguide coefficient
W. In Figure 10, therebetween the coaxial
are low-level spikesTEM mode
in the and
higher-
the output
order waveguide TE
mode transmission 10 mode. The
coefficients: reflection
−44.5 coefficient
dB at 17.7 GHz due (|Γ|)
to theis less
TM11than −22.6
mode, dB
−42.4
dBat at
8.14–23.0
20.4 GHz GHz
due(2.83:1
to thebandwidth).
TM11 mode, The
and transmission coefficient
−46.1 dB at 20.5 GHz due is determined via the
to the TE11 mode.
relationship
These 1 − |Γ|2 . closely correlate with the cutoff frequency of the TE11 and TM11
spike frequencies
modesIninFigure 10a, theand
the matching startoutput
frequency for the reflection coefficient < −20 dB is 8.09 GHz
waveguides.
(=1.028f 10 ). The transition maintains a reflection coefficient < −22.6 dB from 8.14 GHz to
23.00 GHz (=0.974f 30 ). The theoretical upper frequency limit of the transition is determined
by the level of the higher-order TE11 and TE30 modes generated in the waveguides W, W 1 ,
and W 2 . A spike in the reflection coefficient is generated at 23.1 GHz, where the TE11 , TM11 ,
and TE30 modes have a transmission peak all at the same time, as shown in Figure 9. The
transition operates from slightly above the TE10 mode cutoff to slightly below the TE30
cutoff. The resultant bandwidth is 2.83:1 or 94.3% of the theoretical limit of 3:1.
In Table 3, we note that the cutoff frequency of the TE11 , TM11 and TE30 modes ranges
from 20.13 GHz to 24.08 GHz in the matching waveguide W 1 and from 17.59 GHz to
23.61 GHz in the output waveguide W. In Figure 10, there are low-level spikes in the
higher-order mode transmission coefficients: −44.5 dB at 17.7 GHz due to the TM11 mode,
−42.4 dB at 20.4 GHz due to the TM11 mode, and −46.1 dB at 20.5 GHz due to the TE11
mode. These spike frequencies closely correlate with the cutoff frequency of the TE11 and
TM11 modes in the matching and output waveguides.
 Dimension
Waveguide Cutoff Frequency Cutoff Frequency Cutoff Frequency
a × b (mm)
f10 (GHz) f30 (GHz) f11 (GHz)
Output Waveguide W 19.05 × (8.06–9.53) 7.87 23.61 17.59–20.18
Matching Waveguide W1 19.47 × (6.57–8.06) 7.70 23.10 20.13–24.08
Appl. Sci. 2024, 14, 74 12 of 15
Converter Waveguide W2 20.58 × 6.57 7.28 21.85 23.95

During the optimization of the WR75 transition, we observed that the resonance
During the optimization of the WR75 transition, we observed that the resonance spikes
spikes of the TE11 and TM11 modes can be generated between 20 GHz and 22 GHz. It was
of the TE11 and TM11 modes can be generated between 20 GHz and 22 GHz. It was found
found that
that this isthis
not is not caused
caused by thebystepsthe steps
in theinwaveguide
the waveguide broadbroad
wall.wall.
ThisThis
cancan be checked
be checked by
by analyzing the structure with the waveguides W, W 1, and W2 only (i.e., with the structure
analyzing the structure with the waveguides W, W 1 , and W 2 only (i.e., with the structure
with
with S, S, B, P, and
B, P, and X X in
in Figure
Figure33removed).
removed).ItItisiscausedcausedby bythe
thestepped-ridge
stepped-ridge transformer
transformer S inS
in the converter waveguide
the converter waveguide W 2 . W 2.

Careful
Careful efforts
efforts are
are required
required for for aa transition
transition design
design having
having aa bandwidth
bandwidth approaching
approaching
the theoretical limit of 3:1. It is because the TE 11 and TM11 modes can be generated by the
the theoretical limit of 3:1. It is because the TE11 and TM11 modes can be generated
stepped-ridge transformer
by the stepped-ridge in the converter
transformer waveguide
in the converter and then transmitted
waveguide to the match-
and then transmitted to
ing and output waveguides. Thus, the design effort should
the matching and output waveguides. Thus, the design effort should be focused be focused on the suppression
on the
of the TE11 and
suppression TMTE
of the 11 modes while achieving a low reflection coefficient over a broad fre-
11 and TM11 modes while achieving a low reflection coefficient over a
quency range. Therange.
broad frequency designedTheWR75designed transition
WR75was obtained
transition wasbyobtained
simultaneously optimizing
by simultaneously
for both the for
optimizing lowboth
reflection
the low coefficient
reflectionand the lowand
coefficient leveltheoflow
higher-order modes.
level of higher-order modes.
To
To achieve a much smaller reflection coefficient in the range of −30 dB withpro-
achieve a much smaller reflection coefficient in the range of −30 dB with the the
posed
proposed transition,
transition,a further
a further investigation
investigation is is
necessary
necessary that
thatinvolves
involvesa are-design
re-designof of the
the
stepped-ridge transformer,including
stepped-ridge transformer, includinganan increase
increase in the
in the number
number of steps.
of steps. On theOnother
the other
hand,
hand,
it mayiteven
mayrequire
even require a different
a different approach approach to the TEM-to-TE
to the TEM-to-TE 10 mode 10 mode transformer
transformer design. de-
A
sign. A literature
literature survey survey
revealed revealed that a with
that a design design with a reflection
a reflection coefficientcoefficient in the
in the level of level
−30 dBof
−30
overdB over
such such abandwidth
a broad broad bandwidth
seems toseems be raretoifbenotrare if not impossible.
impossible. The problem Theof problem
designingof
designing a very-low-reflection
a very-low-reflection broadbandbroadbandtransition is transition is left as
left as a future a futuretopic.
research research topic.
To
To ascertain
ascertain the validity
validity of of the computational
computational design, design, we we compared
compared the the time-
time- and
and
frequency-domain simulationsofofthe
frequency-domain simulations theWR75
WR75 transition
transition in Figure
in Figure 11. The
11. The agreement
agreement be-
between
the two
tween thesimulation
two simulationmethods is excellent.
methods is excellent.

Figure
Figure 11.
11. Reflection
Reflectionand
andtransmission
transmission coefficients
coefficients of
of the
the designed
designed WR75
WR75 transition
transition by
by the
the time-
time- and
and
frequency-domain simulations.
frequency-domain simulations.

In Table
In Table4,4,we
wecompare
compare the
the proposed
proposed transition
transition design
design with
with previous
previous works.
works. We
We per-
performed a thorough literature search for comparison and listed in Table 4 only
formed a thorough literature search for comparison and listed in Table 4 only those pre- those
previous
vious works
works thatthat
are are worthy
worthy of comparison
of comparison in this
in this paper.
paper. ForFor completeness,
completeness, wewe include
include a
a commercial transition by Exceed Microwave [28], which is shown in Figure 1 along with
its reflection coefficient graph. Compared with the bandwidth of 1.36:1 to 2.00:1 in the
previous works, the design approach presented in this Communication achieves a 2.83:1
bandwidth, a great increase from existing results. The proposed transition uses a WR75
waveguide and operates from 8.14 GHz to 23.00 GHz, fully covering the WR90, WR62, and
WR51 waveguide bands.
Appl. Sci. 2024, 14, 74 13 of 15

Table 4. Comparison with previous works.

Frequency (GHz)/ Reflection

Work Year Feature
Ratio Bandwidth Coefficient (dB)
Three-step ridge and
[4] 1957 2.5–4.1/1.64:1 −20.1
coaxial quarter-wave transformer
[8] 1979 Stepped wire loop 8.2–12.0/1.46:1 −19.1
[9] 1990 Dual stepped ridge 0.85–1.16/1.36:1 −32.3
[10] 1998 Three-step ridge 3.05–4.60/1.51:1 −30.0
Oversized mode conversion and
[13] 2017 10.0–20.0/2.00:1 −30.0
four-step ridge
[15] 2018 Three-step ridge 11.5–19.0/1.65:1 −20.0
[18] 2021 Five-step ridge 33–50/1.52:1 −25.0
[19] 2021 Four-step ridge 40–60/1.50:1 −25.0
[28] In production Commercial product 21–34/1.62:1 −22.7
Four-step ridge and
This work 2023 8.14–23.00/2.83:1 −22.6
linear taper

4. Conclusions
The design approach proposed in this paper enables a realization of an in-line rect-
angular waveguide-to-coaxial transition with a bandwidth close to the theoretical limit of
3:1. The maximum bandwidth in previous works was 2:1. The proposed design approach
employs three waveguides—output, matching, and converter waveguides. A standard
waveguide was used at the output, and a mode-converter waveguide was used at the
input. A matching waveguide was placed between the input and the output. Making
the broad-wall width of the converter waveguide slightly larger than that of the output
waveguide ensured an operation close the TE10 -mode cutoff. The narrow-wall height of the
converter waveguide was reduced such that the TE11 -mode cutoff frequency was the same
as the TE30 -mode cutoff frequency. For low reflection over a broad frequency range, a four-
step ridge transformer was used for the coaxial TEM to waveguide TE10 mode conversion.
With the proposed scheme, a transition can operate with a maximum bandwidth of 3:1. It
was found that the discontinuity-induced generation of the unwanted higher-order modes
should be suppressed by a careful optimization of the stepped-ridge transformer. A WR75
transition designed according to the proposed approach showed a 2.83:1 bandwidth, close
to the theoretical limit of 3:1 with a reflection coefficient < −22.6 dB. We expect that the
proposed design approach will pave the way for a significant extension in the bandwidth of
microwave devices and antennas that employ coaxial-to-rectangular waveguide transitions.
Possible future research topic may include (1) reduction of the transition’s reflection coeffi-
cient to −30 dB level, (2) further extension of the transition’s bandwidth by suppressing
higher-order modes, and (3) design integration of the proposed transition in such devices
as near-field antenna measurement probes and materials measurement fixtures.

Author Contributions: Conceptualization, S.L. and S.-G.C. methodology, C.-S.L., Q.Z., S.L. and
S.-G.C.; validation, Q.Z., J.H., E.A., D.O., S.X., G.-Y.A., C.-S.L., S.L. and S.-G.C.; formal analysis, Q.Z.,
S.X., J.H., C.-S.L., D.O., S.-G.C. and B.-C.A.; formal data curation, G.-Y.A., D.O., J.H., E.A. and B.-C.A.;
writing—original draft, E.A., Q.Z., S.L. and G.-Y.A.; writing—review and editing, S.X., S.-G.C. and
B.-C.A. All authors have read and agreed to the published version of the manuscript.
Funding: This research was supported by Basic Science Research Program through the National Re-
search Foundation of Korea (NRF) funded by the Ministry of Education (No. 2020R1A6 A1A12047945)
and by the MSIT (Ministry of Science and ICT), Korea, under the Grand Information Technology
Research Center support program (IITP-2023-2020-0-01462) supervised by the IITP (Institute for
Information & communications Technology Planning & Evaluation).
Appl. Sci. 2024, 14, 74 14 of 15

Institutional Review Board Statement: Not applicable.

Informed Consent Statement: Not applicable.
Data Availability Statement: The data presented in this study are available in this article.
Conflicts of Interest: Author Shu Li was employed by the company Jiangsu Hengxin Technology Co.,
Ltd. The remaining authors declare that the research was conducted in the absence of any commercial
or financial relationships that could be construed as a potential conflict of interest.

References
1. Niemann, F.L. Transitions from coaxial line to waveguide. In Microwave Transmission Circuits; Ragan, G.L., Ed.; McGraw-Hill:
New York, NY, USA, 1948; pp. 314–361.
2. Cohn, S.B. Design of simple broad-band wave-guide-to-coaxial-line junctions. Proc. IRE 1947, 35, 920–926. [CrossRef]
3. Wheeler, G.J. Broadband waveguide-to-coax transitions. In Proceedings of the 1958 IRE International Convention Record, New
York, NY, USA, 21–25 March 1957.
4. Dix, J.C. Design of waveguide/coaxial transition for the band 2.5–4.1 Gc/s. Proc. Inst. Electr. Eng. 1963, 110, 253–255. [CrossRef]
5. Das, B.N.; Sanyal, G.S. Coaxial-to-waveguide transition (end-launcher type). Proc. Inst. Electr. Eng. 1976, 123, 984–986. [CrossRef]
6. Deshpande, M.D.; Das, B.N.; Sanyal, G.S. Analysis of an end launcher for an X-band rectangular waveguide. IEEE Trans. Microw.
Theory Tech. 1979, 27, 731–735. [CrossRef]
7. Saad, S.M. A more accurate analysis and design of coaxial-to-rectangular waveguide launcher. IEEE Trans. Microw. Theory Tech.
1990, 38, 129–134. [CrossRef]
8. Gangadharan, T.S. Broadband end launcher for X band rectangular waveguide. Electron. Lett. 1982, 18, 354–355. [CrossRef]
9. Chan, K.K.; Martin, R.; Chadwick, K. A broadband end launched coaxial-to-waveguide transition for waveguide phased arrays.
In Proceedings of the IEEE Antennas and Propagation Society International Symposium, Atlanta, GA, USA, 21–26 June 1998.
10. Levy, R.; Hendrick, L.-W. Analysis and synthesis of in-line coaxial-to-waveguide adapters. In Proceedings of the 2002 IEEE MTT-S
International Microwave Symposium Digest (Cat. No.02CH37278), Seattle, WA, USA, 2–7 June 2002.
11. Durga, M.; Tomar, S.; Singh, S.; Suthar, L. Millimeter wave in-line coaxial-to-rectangular waveguide transition. In Proceedings of
the 2011 IEEE Applied Electromagnetics Conference (AEMC), Kolkata, India, 18–22 December 2011.
12. Carmel, N.; Elmakayes, D.; Matzner, H. Investigation of a coax-to-waveguide transition element. In Proceedings of the 2011 IEEE
International Conference on Microwaves, Communications, Antennas and Electronic Systems (COMCAS 2011), Tel Aviv, Israel,
7–9 November 2011.
13. Cano, J.L.; Mediavilla, A. Octave bandwidth in-line rectangular waveguide-to-coaxial transition using oversized mode conversion.
Electron. Lett. 2017, 50, 1370–1371. [CrossRef]
14. Gatti, R.V.; Rossi, R.; Dionigi, M. In-line stepped ridge coaxial-to-rectangular waveguide transition with capacitive coupling. Int.
J. RF Microw. Comput.-Aided Eng. 2018, 29, e21626. [CrossRef]
15. Ye, J.; Zhang, H.; Chu, R. Analysis of a new method to design a coaxial-to-rectangular waveguide transition. In Proceedings
of the 2019 International Applied Computational Electromagnetics Society Symposium-China (ACES), Nanjing, China, 8–11
August 2019.
16. Bilik, V. High-power in-line waveguide to coaxial adapter. In Proceedings of the 2020 30th International Conference Radioelek-
tronika (RADIOELEKTRONIKA), Bratislava, Slovakia, 15–16 April 2020.
17. Kim, D.-H.; Yang, D.-Y. Analysis and design of multiple-stepped coaxial coupling launcher for wideband rectangular waveguide
excitation. J. Electr. Eng. Technol. 2021, 16, 2149–2155. [CrossRef]
18. Simone, M.; Fanti, A.; Lodi, M.-B.; Pisanu, T.; Mazzarella, G. An in-line coaxial-to-waveguide transition for Q-band single-feed-
per-beam antenna systems. Appl. Sci. 2021, 11, 2524. [CrossRef]
19. Shu, Y.; Ren, L. Novel waveguide connectors to simplify microwave and millimeter wave component packaging. In Proceedings
of the 2022 IEEE/MTT-S International Microwave Symposium-IMS 2022, Denver, CO, USA, 19–24 June 2022.
20. Gaudio, J.G.; Station, J.; Debski, T.R. Broadband Waveguide to Coaxial Line Transition. U.S. Patent 3,737,812, 5 June 1973.
21. Fábregas, I.; Shamsaifar, K.; Rebollar, J.M. Coaxial to rectangular waveguide transition. In Proceedings of the IEEE Antennas and
Propagation Society International Symposium 1992 Digest, Chicago, IL, USA, 18–25 June 1992.
22. Yi, W.; Li, E.; Guo, G.; Nie, R. An X-band coaxial-to-rectangular waveguide transition. In Proceedings of the 2011 IEEE
International Conference on Microwave Technology & Computational Electromagnetics, Beijing, China, 22–25 May 2011.
23. Zhou, Y. A wideband millimeter-wave coaxial to rectangular waveguide transition structure. Channels 2011, 4, 5G.
24. Wang, C.; Guo, G.; Wang, J.; Li, E. Analysis and design of Ku band coaxial-waveguide transition. In Proceedings of the PIERS,
Guangzhou, China, 25–28 August 2014.
25. Song, D.; Zhao, G. A novel design of 110GHz broadband coaxial to waveguide adapter. In Proceedings of the 2020 International
Conference on Microwave and Millimeter Wave Technology (ICMMT), Shanghai, China, 20–23 September 2020.
26. Zalabsky, T.; Bezousek, P.; Krejci, T. Broadband waveguide to coax adapter for measurement of harmonic distortion power
amplifier. In Proceedings of the 2021 International Symposium ELMAR, Zadar, Croatia, 13–15 September 2021.
Appl. Sci. 2024, 14, 74 15 of 15

27. Mohamadzade, B.; Dunning, A.; Hayman, D.B.; Smart, K. Broadband coaxial-to-rectangular waveguide transition. In Proceedings
of the 2023 International Conference on Electromagnetics in Advanced Applications (ICEAA), Venice, Italy, 9–13 October 2023.
28. Exceed Microwave. WR34 Adapter End Launch 2.92 mm-F. Available online: https://exceedmicrowave.com/products/
combiners-adapters (accessed on 30 September 2023).
29. Kehn, M.N.M.; Kildal, P.S. Miniaturized rectangular hard waveguides for use in multifrequency phased arrays. IEEE Trans.
Antennas Propag. 2005, 53, 100–109. [CrossRef]
30. Bai, Y.; Wang, L.; Zhang, L.; Wang, P.; Peng, B. 16-channel wavelength division multiplexers based on subwavelength grating.
Appl. Sci. 2023, 13, 1833. [CrossRef]
31. Simoncini, G.; Rossi, R.; Alimenti, F.; Vincenti Gatti, R. Single-ridge waveguide compact and wideband hybrid couplers for
X/Ku-band applications. Electronics 2022, 11, 1538. [CrossRef]
32. Bermúdez-Martín, D.; Gillard, R.; Molero, C.; Legay, H.; García-Vigueras, M. Methodology for improving scanning performance
loading an array element with a 3D all-metal WAIM. Electronics 2022, 11, 2848. [CrossRef]
33. Ahmed, F.; Hayat, T.; Afzal, M.U.; Zhang, S.; Esselle, K.P.; Whittow, W. 3D printable synthetic metasurface to realize 2D
beam-steering antenna. IEEE Open J. Antennas Propag. 2023, 4, 506–519. [CrossRef]

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.