Note: Descriptions are shown in the official language in which they were submitted.
CA 02211406 1997-07-24 PCT~S 9 6/ 01~80
IPEWS O S SEP '96
DESCRIPTION
Method For Increasinq Power Handlinq
Capabilities of Hiqh Temperature Superconducting Devices
Field of the Invention
This invention is directed to devices formed from high
temperature superconductors. More particularly, it is
directed to electronic devices formed from such super-
conductors into devlces having substantial power handlingcapability.
' Background of the Invention
The discovery of "high-temperature superconductivity"
(HTS) in 1987 opened up exciting possibilities for future
technology, most of which are yet to be realized. Prior
to that time superconductivity could be utilized only at
near-absolute-zero temperatures that were achieved by the
use of expensive and hard-to-handle liquid helium. Now,
with the discovery of HTS, superconductivity can at
present be obtained at temperatures of up to 130 K. These
temperatures are easily provided by the use of liquid
nitrogen which boils at 77 K, and is much cheaper and
easier to handle than is liquid helium. In addition, the
needed temperatures can also be provided for many applica-
tions by electrical cryogenic coolers of relatively smallsize. Some of these are presently available and, no
doubt, more powerful and less expensive models will be
available in the future. These developments provide
microwave engineers with means for achieving microwave
circuits with small size yet with extremely low loss.
Microwave systems have been greatly advanced by the
application of planar circuit technology which makes
possible the photoetching Gc -omplex, compact circuits
with many components on a single substrate. These are
often referred to as "hybrid microwave integrated cir-
cu-ts" if the solid-state devices are produced in the
AMENDED SHEE~
CA 02211406 1997-07-24 ~r~s 9 6 / O 1 7 8 0
- IPEWS O 5 SEP '96
circuit processing. These techniques have made feasible
many sophisticated systems that would not have been
practical if it had been necessary to use older non-planar
circuitry such as circuits using waveguide and coaxial
lines which are relatively large and heavy. Examples of
non-superconducting, lumped element planar arrangements
are shown, for example, in Swanson U.S. Patent No.
4,881,050, issued November 14, 1989. Further, work has
been done towards the goal of reducing propagation losses
in normal metal coplanar waveguides. For example, in F.
Williams, et al, "Reduction of Propagation Losses in
Coplanar Waveguide", 1984 IEEE MTT-S Digest, pp 453-4,
coplanar waveguide propagation losses are reduced by
forming an air-gap between a center conductor and coplanar
ground plane portions. Two embodiments are proposed, one
in which the coplanar waveguide is formed on a thin
dielectric layer which in turn is formed on a thicker,
higher dielectric constant layer, and a second in which
trenches are formed into the dielectric substrate which
supports the conductive portions of the coplanar
waveguide. A local minima in propagation loss is achieved
in both embodiments where the thickness h of the first
dielectric or trench depth is between approximately 10 to
30~ of the spacing between the center conductor and the
ground plane.
However, conventional planar microwave circuitry has
a serious limitation in that it has high losses compared
to, say, waveguide and most coaxial circuitry. This
limits the use of conventional microwave integrated
circuits to applications where considerable loss in the
circuitry can be tolerated. For example, it is not
feasible to realize narrow-band microwave filters in
conventional integrated circuit form because high-Q (i.e.,
very l~w loss with Q's in excess of 400) resonators are
required. Also, in many conventional microwave circuits
the circuit components are made larger than they would
otherwise need to be, in order to reduce the losses. If
AMENOED SHEEI
CA 02211406 1997-07-24 ~ ~ S 9 6 1 0 1 7 8 0
IPEWS O 5 SEP '96
loss were not a consideration many microwave integrated
circuits could be made to be even smaller and lighter.
With the advent of HTS microwave planar circuits it
has become feasible to fabricate microwave integrated
circuits with very low losses equivalent to and sometimes
even lower than are realized by large, heavy waveguide
circuits. Very compact "semi-lumped" or "lumped" elements
can be used and still obtain phenomenal resonator Q's of
20,000 or more in many situations. This technology opens
the way to high performance, compact microwave integrated
circuits of types that simply were not possible before.
However, superconductive circuits are not without
"~ their problems. For example, use of superconductors as
transmitters or receivers requiring any significant power
handling capability are problematic. While an ideal
superconductor has a perfect crystalline structure which
results in total linearity (up to the critical current
density Jc) as a function of power, real superconducting
crystals have slight irregularities which lead to non-
linearity of surface resistance as a function of current,even for currents below the critical current density Jc-
Intermodulation effects can arise, resulting in sum and
difference frequencies. Thus, for a receiver or transmit-
~ ter, interference can result. While this problem may be
somewhat handled through the use of filters, the optimal
solution in such applications is true linearity and the
avoidance of additional components such as filters. Even
with the use of filters, the problem is not eliminated.
For example, under current technology, if very narrowband
filters are used, input power levels on the order of amilliwatt may lead to an increase in transmission loss as
well as intermodulation effects.
Despite the high desirability of manufacturing sub-
stantially linear devices from high temperature super-
conductors, problems remain which preclude truly lineardevices.
AAIEI~IGED SHECT
CA 02211406 1997-07-24 ~ ~ S 9 6 / O 1 7 8 0
IPEA'~ ' '96
SummarY of the Invention
This invention relates to methods and apparatuses for
increasing the power handling capabilities of high temper-
ature superconductor devices. In a stripline structure,
a long, thin center conductor is surrounded by a dielec-
tric except at the center conductor's thin edges. The
thin edges are preferably exposed to a material having a
lower dielectric constant than a surrounding dielectric,
most preferably air or vacuum. In one embodiment of the
stripline structure, the removed dielectric portions
provide tunnels or gaps of air adjacent the center conduc-
tor edges. Yet another stripline embodiment has air gaps
extending substantially completely between the ground
planes in regions laterally external to the center conduc-
tor edges. Optionally, the dielectric between the centerconductor and one of the ground planes may be removed.
The ground planes may be optionally formed on yet other
support substrates.
In an alternative embodiment, a microstrip arrangement
consists of a center conductor having a generally long
thin shape disposed on a dielectric substrate. Trenches
are formed adjacent and laterally exterior to the exterior
edges of the center conductor. A ground plane, optionally
superconducting, is formed on the side of the dielectric
opposite to that of the center conductor.
The trenches or gaps may be formed preferably by
milling, such as ion, mechanical or laser milling, or may
be etched via isotropic or anisotropic etches. Optional-
ly, the trench or gap may be undercut beneath the super-
conductor, such as through the use of an etch.
Accordingly, it is a principal object of this inven-
tion to form structures including superconductors which
have improved power handling capabilities.
It is yet a further object of this invention to
provide superconductive devices having improved linearity
of characteristics, especially surface resistance, as a
function of current.
AMENDED SHEET
. CA 02211406 1997-07-24 PCT~S 9 ~ / O 1 7 8 0
- ; IPEA/IJS O 5 SEP '96
It is a further object of this invention to minimize
the maximum current density being carried through a
superconductive device for a given power level.
Brief Description of the Drawinqs
Fig. 1 shows a perspective view of a stripline config-
uration.
Fig. 2 shows a perspective view of a modified
stripline configuration.
Fig. 3 shows a perspective view of a microstrip
configuration.
Fig. 4 shows in perspective a detail of dielectric
undercut from a center conductor.
Fig. 5 is a cross-section of a prior art stripline
transmission structure without use of the instant
invention.
Fig. 6 shows the current distribution on a center
conductor of the prior art structure of Fig. 5 as a
function of lateral displacement from the center of the
conductor.
Fig. 7 shows a cross-section of electric field lines
in a stripline configuration of the prior art structure of
Fig. 5.
Fig. 8a shows current distribution as a function of
lateral position from the center of conductors in the
structures of Fig. 2 and Fig. 5, respectively.
Fig. 8b shows the current distribution in the vicinity
of the right edge of the center conductor in the struc-
tures of Figs. 2 and 5, respectively.
Detailed Description of the Invention
Fig. 1 shows a perspective view of a stripline config-
uration of this invention. A center conductor 12 is
disposed substantially equidistant from a first ground
plane 16 and a second ground plane 18. The center conduc-
tor 12 generally which is wider than the thickness thereof
and extends into the plane of the drawing as shown in
~IIIENDED SHEET
CA ~22114~6 1997-~7-24 ~ 9i
cross-section. The center conductor 12 terminates in
center conductor edges 14. Preferably, the center
conductor 12 is formed from high temperature
superconductor materials. While any of the
superconductive materials may be utilized, the YBCO and
thallium superconductors are preferred for their relative-
ly high critical temperature Tc and power handling capabil-
ities.
The center conductor 12 is supported by dielectric 22.
The dielectric 22 may be of any material compatible with
the center conductor 12, such as lanthanum aluminate,
sapphire, and magnesium oxide. In this embodiment, the
dielectric 22 is disposed between the center conductor 12
and the first ground plane 16, as well as between the
center conductor 12 and the second ground plane 18. Gaps
20 are formed in the dielectric 22 adjacent the center
conductor edges 14. Preferably, the gap 20 is formed of
a material having a dielectric constant which is lower
than the dielectric constant of dielectric 22. Most
preferably, the dielectric comprising the gap 20 is air or
vacuum. The gap 20 runs parallel to the center conductor
12 adjacent the center conductor edges 14. The gap 20 is
formed sufficiently large as to improve the power handling
capability of the stripline structure 10 but not made so
large as to imperil the structural integrity of the
overall device. Generally, the larger the gap 20 is
relative to the overall stripline structure 10, the better
the power handling capabilities. Preferably, the width
and depth of the gap should be of a size comparable to the
width of the center conductor and the substrate thickness,
respectively. In one embodiment, the gap depth and width
is greater than 20 microns. Ideally, the gap depth would
extend to the ground planes as in Fig. 2.
The structure of Fig. 1 is preferably constructed by
hybridizing two units together. A first unit comprising
the upper portion of dielectric 22 has two substantially
parallel faces, the first face bearing ground plane 16 and
AMENDED SHEET
CA 02211406 1997-07-24 P ~ ~ S 9 6 / 01 7 8 0
IPEA/US O 5 SEP '96
the second face bearing all or part of the gap 20. The
lower module consists of substantially planar, parallel
faces whereon the first face bears the center conductor 12
and all or part of the gap 20 and the second ground plane
on the opposite side of the dielectric. The two units are
then hybridized forming an interface 24 of the two mod-
ules. The modules preferably are held together via
pressure clamps.
The gap 20 may be formed through any material removal
process compatible with the other materials in the
stripline structure 10. For example, the gap 20 may be
milled, such as by mechanical milling for relatively large
-
structures (e.g., radio frequency devices) or by ion
milling or laser milling for smaller structures (e.g.,
microwave and millimeter wave devices). Alternatively,
the gap 20 may be etched into the dielectric 22. Etching
provides the opportunity to undercut the dielectric 22
from the center conductor 12. As shown in Fig. 4, the gap
20 (shown in partial) has an undercut 26. The undercut 26
extends a distance u from the vertical line extending
downward from the center conductor edge 14.
Fig. 2 shows a modified stripline structure. A center
conductor 30 is generally planar, having a thickness which
is much less than its width and length. The center
conductor 30 has center conductor edges 32 at the lateral
edges of the center conductor 30. First ground plane 34
and second ground plane 36 are disposed generally parallel
to the center conductor 30 and arranged with the center
conductor 30 being parallel to and equidistant from each
of the ground planes 34 and 36. A first dielectric 38 is
disposed between the center conductor 30 and the first
ground plane 34. Preferably, a second dielectric 40,
optimally having the same dielectric constant is disposed
between the center conductor 30 and the second ground
plane 36. In this embodiment, dielectric, other than air,
is removed from the regions laterally exterior to the
center conductor edges 32, as indicated by arrows 42.
AMENDED SHEET
CA 02211406 1997-07-24 PCT~S 9 6101 780
IPEA/US 0 5 SEP '96
Optionally, the first and second ground planes 34 and 36
may be supported by the substrates disposed on the side
away from the center conductor 30.
Fig. 3 shows a perspective view of a microstrip
configuration. The microstrip structure 50 includes a
substrate 52 having generally planar, parallel disposed
faces, a first face 54 and a second face 56. A center
conductor 58 is formed on the first face 54 of substrate
52. Generally the center conductor 58 has a thickness t
which is substantially less than its width s. Trenches 62
are formed laterally adjacent to the center conductor
edges 64. The trench 62 has a depth h and a width w.
Ground plane 60 is disposed on the second face 56 of the
substrate 52.
Fig. 4 shows a gap 20 having an undercut portion 26
under the center conductor 14. The undercut amount of
distance u may be as desired, consistent with maintenance
of structural integrity.
In the preferred mode of practicing the inventions of
this patent, it is preferable that the conductive materi-
als, namely the center conductor and the ground planes, be
formed from superconducting materials. However, replace-
ment of one or more of these structures with normal
metals, such as gold, high purity copper, or other materi-
als compatible with transmission of high frequency elec-
tromagnetic radiation is consistent with this invention.
HTS microwave circuits do have limitations in the
amount of power they can carry. If the power level in a
clrcuit gets too high the current density in some regions
of the circuit will exceed a "critical" level Jc, which
depends on the temperature, frequency and microstructure,
and the HTS in those regions will no longer operate as a
superconductor. This results in markedly increased
circuit loss and in nonlinear effects such an
intermodulation between signals at different frequencies.
While a function of many variables such as temperature,
frequency and local structure, this crit cal current
AMENDED SHEET
CA 02211406 1997-07-24 P ~ ~ S 9 6 / O 1 7 8 0
IPEA/US O 5 SEP '96
density Jc is typically of the order of 3 x 106 amps/cm2 in
high quality thin films at 77K. In some planar microwave
circuit applications where the transmission lines are well
matched (i.e., they do not have reflections at their ends)
the HTS transmission lines may be able to carry as much as
hundreds of watts (or possibly more depending on the line
cross-sectional dimensions) without appreciable effects
due to the critical current density being exceeded.
However, in cases such as narrow-band filters, the cur-
rents within the resonators of the filter may be as muchas 100 times or more as large as the currents in the lines
which connect to the ports of the filter. This is a
result of the resonance conditions that exist in the
resonators plus the loose couplings between the resona-
tors. Since power varies as the square of the current,this means that the power rating of the transmission lines
used in the resonators may need to be 10,000 times or more
greater than is required for the transmission lines
leading into or out of the filter. Other parts of an
integrated circuit which may have relatively high currents
due to high standing wave ratios may also need to have
high power ratings. For these reasons apparatus and
techniques for increasing the power handling ability of
HTS transmission lines is quite important for some micro-
wave integrated circuit applications which must handlerelatively high power.
By way of example, Fig. 5 shows an HTS strip transmis-
sion line which consists of a center conductor 160 sur-
rounded by dielectric 162 with a relative dielectric
constant ~r with ground planes 164 at the top and bottom.
The center conductor extends into and out of the paper.
The current density on the center conductor is distributed
as shown in Fig. 6, where there are high peaks of current
density at the outer ed~-es cf -he center conductor. High
peaks of current density exist near the center-conductor
edges, and it is in these edge regions where current
AMENDED SHEET
: - CA 02211406 1997-07-24 PCT~S 9 6/ 01780
IPEA/IJS O 5 S~P '96
saturation will first occur as the power level is being
raised on a transmission line.
Some idea of why the high current density occurs at
the strip edges in Fig. 5 can be seen from Fig. 7 which
shows the same structure as in Fig. 5 but with electric
flux lines D sketched in. The flux lines shown begin on
positive surface charge on the center conductor 160 and
end on negative surface charge on the ground planes, and
the surface charge density at any point on the surface is
equal to the electric flux density D associated with that
point. As these flux lines and surface-charge distribu-
tions propagate into (or out of) the paper, the propagat-
ing charge distributions comprise surface currents on the
conductors flowing into or out of the paper. Flux lines
from a sharp edge with positive charge will emanate out
radially, as shown in Fig. 7, and this requires a high
concentration of charge and current along the sharp edges
in order to provide the needed electric flux D. Consis-
tent with this invention, in order to reduce the high
current density along the edges of the center conductor
this invention reduces the fringing flux along the edges
of the strip.
It is possible to obtain a more complex estimate of
the benefits of the structure in Fig. 2 with regard to
reducing the current peaks at the edges of the center
conductor. If the center conductor in Fig. 5 is infinite-
ly thin, and if the ground plane spacing h' approaches
infinity, the current distribution on the center conductor
will be exactly of the form
Al
J1(x)= (1)
(l-(2x/w) 2) 0.5
where Al is a constant which depends on the applied voltage
and the structure of the line, x is the distance from the
center of the conductor and ~W" is the width of the
conductor. It is found that this current-density
distribution function is still a useful approximation in
many cases even when the ground-plane spacing h' is finite
AMENDED SHEET
CA 022ll406 l997-07-24 ~ ~ S 9 6 / O 1 7 8 0
IPEA/US O 5 S~P '96
as in Figs. 5 and 7. (See, e.g., G.L. Matthaei, et al.,
"A Simplified Means for Computation of Interconnect
Distribution Capacitances and Inductances", IEEE Trans. or
Computer-Aided Design, Vol. 11, pp 513-524, April, 1992.)
It is easily shown that the function J1(x) goes to
infinity as
(w/2-x)-~5 as x approaches w/2. R. Mittra et al. in
"Analytical Techniques in the Theory of Guided Waves," at
pp 10-11, 1971, analyzes the singularities at the edges of
thin conductors at dielectric interfaces and indicates
that at the edges of the center conductor in Fig. 2 the
current distribution should go to infinity as (w/2-x)-0128
as x approaches w/2 if ~=24 as for LaAl03.- If we modify
(1) to exhibit this behavior at the edges of the strip we
get
A2
J2(x)= (2)
(l-(2X/W2)-0.12~
which is a good approximation for the current distribution
on the center conductor for the case in Fig. 2 where A2 is
a constant which depends on applied voltage and the
structure of the line. The dashed line in Figs. 8(A) and
8(B) shows a plot of the dashed line in Figs. 8(a) and
8(b) show a plot of Eq. (1) while the solid lines show
plots of Eq. (2), where the coefficients A1 and A2 have
been adjusted so that the area under both curves is the
same to equalize total current in the two transmission
lines. Thus the dashed lines would apply to cases like
that in Fig. 5 (the prior art), while the solid lines
apply to the case in Fig. 2 with ~=24 (the invention). As
can be seen, particularly in Fig. 8B, the presence of the
dielectric discontinuity greatly reduces the intensity of
the current singularity at the edges of the center
conductor. Thus the structure in Fig. 2 carries
considerably more power than that in Fig. 5 without having
part of the current density exceed the critical value Jc
and cause intermodulation and excess loss. Of course, due
to the finite thickness of the HTS and the limited current
A~JIEN~E~ SHEE~
- CA 02211406 1997-07-24 P ~ ~ S 9 6 / O 1 7 8 0
IPEAJUS O 5 SEP '96
penetration depth associated with the HTS, the peaks of
the current distribution would be rounded off some even
before current saturation is included. In practice, the
divergence towards extremely high values in the current
density is somewhat reduced or cut-off by an effective
penetration depth which depends on the wavelength and film
thickness t.
The apparatus and methods of this invention are useful
in connection with electronic circuits carrying power
through waveguide type structures having relatively thin,
wide conductors in which fringing affects would otherwise
occur. Of particular utility is the application of these
techniques to superconductive circuits, especially high
temperature superconducting circuits. Because these
techniques serve to improve the linearity of device
operation, particularly the linearity of the surface
resistance as a function of current, the structures are
particularly useful in connection with any signal process-
ing electronics, such as receivers, transmitters and
filters. By reducing the non-linearities, interference is
reduced substantially.
Although the foregoing invention has been described in
some detail by way of illustration and example for purpos-
es of clarity and understanding, it will be readily
apparent to those of ordinary skill in the art in light of
the teachings of this invention that certain changes and
modifications may be made thereto without departing from
the spirit or scope of the appended claims.
A~ENDED SHEE~
