This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination.
IEEE TRANSACTIONS ON MICROWAVE THEORY AND TECHNIQUES 1
A New Interpretation of Through-Line Deembedding
Kimberley W. Eccleston, Senior Member, IEEE
Abstract—Deembedding a device mounted between a pair of
identical transitions requires the transition two-port parameters.
These can be obtained using the through-line (TL) deembedding
method requiring only through and uniform line deembedding
standards. Other related deembedding methods introduce factors
to scale a pair of eigenvectors to obtain the transition two-
port parameters. In this paper, it is shown that these scaling
factors are interdependent and provide a corrective reference
plane adjustment. Moreover, the transition wave-transmission
matrix can be factored into two wave-transmission matrices with
one encapsulating the transition discontinuity and the other of
a matched line to provide a corrective reference plane shift.
This factorization permits incorporation of desired and corrective
reference plane shifts, using data from a second matched line
standard. With an emphasis on substrate-integrated waveguide,
the proposed TL deembedding method is verified.
Index Terms—Deembedding, substrate-integrated wave-
guide (SIW), through-line (TL) deembedding, through-reflect-
line (TRL) deembedding.
I. INTRODUCTION
A
MEASUREMENT environment supports precise S-para-
meter measurement to a coaxial reference plane, and
electromagnetic simulation environments often restrict exci-
tation to microstrip or circuit-type ports. This means, for
example, that a transition from coaxial cable to microstrip
is required to measure transistors [1]–[3], or microstrip or
coplanar-waveguide to substrate-integrated waveguide (SIW)
transitions are required when simulating and measuring SIW
devices [4]–[6]. For a well-designed 50- microstrip test
fixture, the main effect of the transition is a shift of reference
planes. At the other extreme, transitions to SIW have a
profound impact on the SIW device S-parameters [7], [8].
By a process called deembedding, the effects of the
transition are removed to obtain the S-parameters of the
device-under-test (DUT), which requires two-port models of
the transitions. Obtaining this model is challenging as one
of the transition ports is inaccessible for excitation in a
simulation [9] or direct measurement. Instead, the model is
extracted from data acquired from simulation or measurement
of various configurations of known, or partially known, devices
in place of the DUT, denoted the deembedding standards.
Through-reflect-line (TRL) [10] is a well-known and com-
monly used method that uses three deembedding standards:
Manuscript received May 14, 2016; revised September 2, 2016; accepted
September 4, 2016.
The author was with the Department of Electrical and Computer Engi-
neering, University of Canterbury, Christchurch 8140, New Zealand. He is
now with Lincoln Agritech Ltd., Lincoln 7640, New Zealand (e-mail:
dr.kimberley.eccleston@ieee.org).
Color versions of one or more of the figures in this paper are available
online at http://ieeexplore.ieee.org.
Digital Object Identifier 10.1109/TMTT.2016.2607744
1) through connection; 2) uniform line of known length; and
3) reflective termination. For wideband deembedding, multiple
line standards can be used [11]–[13]. TRL has been extended
to characterize multimode devices [14] including differential-
mode devices [15], [16]. For identical transitions, the transition
model can be obtained using only the through and line stan-
dards [16]–[19]. Moreover, the information gained from the
TRL reflective standard can be synthesized from the through
and line data [19]. By eliminating the reflective standard,
simulation time is reduced, and measurement repeatability
errors associated with another standard can be eliminated.
Furthermore, the simulation of short or open circuits in FDTD
is difficult [17].
Using matrix methods, an eigenvalue problem can be for-
mulated for the through and line configurations [11], [12],
[14]–[18], [20]. The eigenvalues directly give the propagation
constant of the line, and the transition two-port parameters
are related to the eigenvectors. In many situations, the propa-
gation constant is a byproduct, but is useful, for example, in
propagation investigations of SIW [21]–[28], in which case
only the through and line standards are required. On the
other hand, eigenvectors by their nature have arbitrary mag-
nitude and phase and must be adjusted to uniquely deter-
mine the transition model. This usually requires data from
the reflective standard along with part of the original TRL
technique [11], [12], [20] or other formulation [14]–[16].
For matrix formulations, seemingly independent factors are
introduced to scale the eigenvectors [14]–[18], [20].
In this paper, the through-line (TL) deembedding method
is derived from an alternative pathway that gives rise to a
physical interpretation, which requires interdependence the
eigenvector scaling factors. Moreover, the wave-transmission
matrix of the transition can be factored into two wave-
transmission matrices: one encapsulating the discontinuity and
the other of a matched line to provide a corrective reference
plane shift established using the through standard. With this
formulation, the corrective shift can be consolidated with a
desired shift using a second line standard. It is not necessary
to know the lengths of the line standards.
Section II develops the theory of the proposed deembedding
method. Section III applies it to SIW measurement. This paper
is confined to single mode operation, identical transitions, and
a bandwidth constrained by a single pair of through and line
standards.
II. TL D
EEMBEDDING THEORY
Fig. 1 shows the block diagram to represent the DUT
embedded between a pair of transitions. The DUT and transi-
tions are represented by two-port networks with the port num-
bering as indicated. Measured or simulated two-port parameter
0018-9480 © 2016 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.