Electromagnetic Interference Control Techniques - I-Scover

EMC’14/Tokyo
16P2-S1
Electromagnetic Interference Control Techniques for
Spacecraft Harness
A. Junge, J. Wolf
P. Pelissou
Electromagnetic Compatibility Section
European Space Agency – ESTEC
Noordwijk, The Netherlands
[email protected]
Electromagnetic Compatibility Section
Astrium SAS
Toulouse, France
N. Mora, F. Rachidi
Electromagnetic Compatibility Laboratory
École Polytechnique Fédérale de Lausanne (EPFL)
Lausanne, Switzerland
Abstract—The electrical harness is an essential part of all the
spacecraft. The prime contractors of spacecraft are especially
active in efforts to optimize harness size and mass. The shielding
of the harness for Electromagnetic Interference (EMI) control is
an important element in this optimization, because cable shields
and especially connector backshells have a significant mass
impact. We present results of a recently finished study of the
European Space Agency (ESA) conducted by a consortium of
Astrium SAS, École Polytechnique Fédérale de Lausanne (EPFL)
and axon’ cable. First some basic principles of harness shielding
are recalled, then important test and simulation results are
presented, followed by a brief summary of the derived guidelines
on EMI control techniques for spacecraft harness. These
guidelines also consider a trade-off with mass increase and
implementation constraints.
and Earth observation satellites tend to be less constrained by
the mass for the launch, but their performance demands are
constantly increasing while their dimensions are being reduced
in order to be launched by small launchers. Consequently the
relative harness mass and volume compared to the other system
elements can exceed 10% and can also create problems of
layout. For this reason research and development efforts
continue at spacecraft manufacturer level and also at harness
manufacturer level to optimize the harness size and mass.
Keywords—electromagnetic shielding; cable shielding; satellite;
spacecraft
In a recently finished study of the European Space Agency
(ESA) conducted by a consortium of Astrium SAS, École
Polytechnique Fédérale de Lausanne (EPFL) and axon’ cable
[1], we first reviewed briefly basic principles of shielding and
analyzed existing shielding and backshell implementations;
then some promising new techniques have been identified. In
an extensive test campaign, various implementations and
techniques have been validated and compared with numerical
simulations with the aim of defining selection criteria and
guidelines [2]. These guidelines include a trade-off with other
engineering requirements like cost, mass, volume, assembly,
and integration.
I.
INTRODUCTION
The harness is an essential part of all spacecraft. As a nonrecurring element that requires adaptation for each space
mission, it is also a critical element in the development
schedule. The reliability of the harness must be superior to
guarantee that not one of the tens of thousands of connections
fails. To minimize the complexity of the harness is thus a
natural approach, which is just as important as the
minimization of mass and volume.
The prime contractors of telecommunication spacecraft are
specially active in efforts to reduce harness mass. This is not
only because each gram gained allows to increase fuel and thus
lifetime, but also because of the tight schedule, which is on
average just twice as short as that of Earth observation
spacecraft for example. State-of-the-art harnesses account for
only approximately 3% of the mass of telecommunication
spacecraft, but the activities to improve the harness continue
with the focus on integration and tests. Scientific spacecraft
The shielding of the overall harness for Electromagnetic
Interference (EMI) control is an important factor in this
optimization, because the individual shielding elements like
cable shields and especially connector backshells have a
significant mass impact.
In this paper we briefly recall some basic principles and
performance parameters, present important test and simulation
results, and then provide a brief summary of the derived
guidelines.
II.
In general the efficiency of electromagnetic shielding
methods significantly depends on the type of field.
ESA Study on “Advanced Shielding Techniques for Spacecraft Harness”,
Contract No. 4 000 102 505
Copyright 2014 IEICE
BASIC PRINCIPLES
840
EMC’14/Tokyo
16P2-S1
Implementations of a Faraday cage using an electrically
conductive material can be used to confine (quasi-)static
electric fields in a volume or excluded them from a region in
space. For (quasi-)static magnetic fields materials with high
magnetic permeability become necessary to concentrate the
magnetic field inside such materials. For time-dependent
electromagnetic fields, the electric and magnetic field are
coupled and will be attenuated again by electrically conductive
materials. The efficiency will not only depend on the material
thickness in relation to the wavelength of the electromagnetic
field, but also on the continuity of the shield. This aspect is
especially important for a harness, where this continuity needs
to be provided not only by the cable shield but also the
connector including its backshell and the contact to the socket
at equipment side.
To quantify the efficiency, suitable parameters need to be
defined, which will differ depending on the considered
frequency range.
A. Shielding Effectiveness
The shielding effectiveness (SE) is originally defined as the
ratio of the electric or magnetic field strengths at both sides of
a shield, i.e. the ratio of field strength of incident (outer) and
the transmitted (inner) fields, Ei and Et. It is often expressed in
decibels [4], e.g. for the electric field the shielding
effectiveness can be expressed as [4]
SE
20 log
dB.
(1)
For a cable, this can also be defined as the voltage or
current induced into the inner conductors of the cable with and
without the shield in place. A number of disadvantages exist in
the measurement and characterization of the shielding
effectiveness, e.g. the influence by the test set-up. In modestirred reverberation chambers, the measurement of shielding
effectiveness has limited sensitivity to the geometry of the test
set-up due to the statistical nature of the method. Since the
lowest usable frequency however is related to the geometric
dimensions of a reverberation chamber, this method can be
applied only for high frequencies. Thus shielding effectiveness
is in best case usable above some 100 MHz, typically above
1 GHz. This measurement method is schematically depicted in
Fig. 1.
B. Transfer Impedance
The transfer impedance Z’t is in contrast a parameter that is
independent from the test results towards at low frequencies,
such that it should be comparable with data obtained on the
same cable in different test set-ups. The transfer impedance
relates the voltage U’i induced per unit length on the inner
conductor of the shield to the current Is injected into the shield:
i
=
t s.
(2)
Note that to correctly characterize a shielded cable, we need
to consider also the so-called transfer admittance which relates
the inner current to the voltage applied to the shield. At low
frequencies, the transfer admittance is usually small compared
to the transfer admittance and can be disregarded. However, as
the frequency increases, especially beyond 1 GHz, the transfer
admittance should be taken into account [5]. At such
frequencies the shielding effectiveness described in section
II.A becomes a good parameter of shielding efficiency.
Therefore the transfer impedance is typically used at low and
medium frequencies up to some 100 MHz.
With different transfer impedance measurements
considering (i) only cable separately, (ii) only connectors, and
(iii) the full harness assembly including terminations, the
contribution of each element can be discriminated.
For the measurement of the transfer impedance different
test set-ups can be utilized, which will be briefly described in
the following sub-sections.
1) Tri-axial Method
The tri-axial method is especially suited for shielded cables.
With an outer cylinder around the shielded cable an additional
outer coaxial transmission line is formed. The general method
is described in [6] and we have adopted the use of current
probes for current injection and measurement as described in
[7].
This general principle was adopted to perform
measurements also on single connectors plus backshells in the
set-up shown in Fig. 2.
Current transformer
Coaxial Connector
(current injection)
Flange
Coaxial Connector
(voltage measurement)
Internal wire
DUT
Short circuit point
Fastener
Fig. 2. Triaxial-like fixture for measuring transfer impedance of connectors
plus backshells.
Fig. 1. Schematic diagram of a mode-stirred chamber with with the cable or
harness as device under test (DUT).
Copyright 2014 IEICE
2) Microstrip Method
Moreover, in analogy to the triaxial method a set-up with a
microstrip line on the exterior of the harness was used to
perform measurements on the full harness assembly with the
set-up shown in Fig. 3.
841
EMC’14/Tokyo
16P2-S1
5) Double-layer aluminium foil shield and standard
backshell (AluPolamco)
The overshielding is made with two layers of aluminium
foil and grounded on the chimney of a standard backshell.
Additional mass about 50 g.
III.
MEASUREMENTS
In the frame of the study numerous measurements have
been performed and compared with analytical formulas and
numerical simulations. At low frequencies, formulas based on
equivalent circuits and transmission line theory in CST Cable
Studio were used to validate test results; at higher frequencies
full-wave simulations using CST Microwave Studio have been
performed.
1000
Shielding effectiveness (dB)
Fig. 3. Triaxial fixture for measuring transfer impedance of connectors.
B. Performance summary 30 kHz – 1 GHz
Based on the transfer impedance measurements, the
shielding effectiveness of different harness assemblies
consisting of overshielding, shielding braid and terminations
have been derived from transfer impedance measurements
considering an injection of a common mode emission on the
overshielding. Fig. 4 presents the different shielding
effectiveness curves according to the technologies considered
in this study.
A. Test cases
For this paper we present results from five different test
cases in addition to a reference case. As a standard, Sub-D type
backshell (Polamco BT series) was chosen. The shields were
generally attached to the backshell with 360° contact by a band
termination (axoclamp). The reference case was a braided
cable shield grounded with pigtail on a haloring without
backshell.
100
10
Axo/Glenair
Axo/Polamco
Round-It/Polamco Sliding
C12 tape (one layer)
Double layer alu foil
Reference case (shielding cable)
1
1,00E+04
1,00E+05
1,00E+06
1,00E+07
1,00E+08
1,00E+09
Frequency (Hz)
1) Braided shield and composite backshell (Axo/Glenair)
The overshielding is made with a double-layer braided
shield of silver-plated aluminium braid (axotresse) and
grounded on the chimney of a Sub-D type backshell made from
composite material (Glenair 557-186). Additional mass
compared to reference case for 1 m harness bundle is about
65 g.
2) Braided shield and standard backshell (Axo/Polamco).
The overshielding is made with a double-layer braided
shield of silver-plated aluminium braid (axotresse) and
grounded on the chimney of a standard backshell. Additional
mass about 45 g.
Fig. 4. Shielding effectiveness of different technologies in 30 kHz – 1 GHz.
C. Performance summary in 500 MHz – 18 GHz
The Sub-D connector and associated backshell are not
designed for radio-frequency (RF) applications. Therefore the
shielding effectiveness drops down with the frequency, as can
be seen Fig. 5.
The SE behaviour seems similar for all the tested cases; the
double aluminium foil with backshell presents better results.
The gain on SE with respect to the reference case is around
+35dB at 1 GHz, +20dB at 2 GHz and 0dB at 4 GHz for the
majority of tested cases.
3) Self-wrapping Nickel-Copper cladded polymer fabric
and backshell with sliding lid (RoundIt/Polamco sliding; only
up to 1 GHz)
The overshielding is made of self-wrapping polymer
(polyphenylene sulphide, PPS) sheets (FederalModul RoundIt
EMI FMJ) and grounded on the chimney of a Sub-D type
backshell with sliding lid (Polamco BT series). Additional
mass about 105 g.
100
Shielding effectiveness (dB)
Axo/Glenair
Axo/Polamco
C12/Polamco
Alu/Polamco
Reference case (shielding cable)
Polynomial (Axo/Glenair)
Polynomial (Alu/Polamco)
Polynomial (C12/Polamco)
Polynomial (Axo/Polamco)
Polynomial (Reference case (shielding cable))
4) Nickel-Copper C12 cladded polyester fabric and
standard backshell (C12; C12/Polamco)
The overshielding is wounded with a Nickel-Copper C12
cladded polymer fabric tape (Schlegel C12 EMI) and grounded
on the chimney of a standard backshell. Additional mass about
75 g.
10
100000000
1000000000
10000000000
1E+11
Frequency (Hz)
Fig. 5. Shielding effectiveness of different techniques in 500 MHz–18 GHz.
Copyright 2014 IEICE
842
EMC’14/Tokyo
16P2-S1
In this frequency range, the mating contact between
backshell and connectors and between fixed connectors at
equipment side and mobile connectors of the harness have an
influence on the EMC performances. Fig. 6 shows the SE
measurements in reverberating chamber for two configurations:
(i) with isolating Kapton and (ii) with aluminium tape having
conductive adhesive (Cho-Foil) on connector shell. They
simulate the two extremes of connector shell electrical contact.
The results show a maximum difference of 15dB at 2GHz,
which compensate the SE effect due to the overshield and the
backshell.
80
C12/Polamco+Kapton
C12/Polamco+Chofoil
Polynomial (C12/Polamco+Chofoil)
SE in dB
60
Polynomial (C12/Polamco+Kapton)
40
20
0
100
1000
Frequency in MHz
10000
100000
Fig. 6. Shielding effectiveness of different technologies in 500 MHz –
18 GHz, measured in a reverberation chamber.
The results have demonstrated that the random electrical
contact between fix connector, mobile connector, and backshell
leaves slots in between both screw-locks. Moreover, these slot
sizes are close to the half of the wavelength in the L- and Sband and so the shielding effectiveness of this assembly drops
to low values with increasing frequency.
Fig. 7 shows the three potential interfaces in the Sub-D
connector assembly that leaves slots and therefore RF leakages:
-
between fix connector and unit chassis,
between the fix and mobile connector, and
between the backshell and the mobile connector.
mainly performed through the screw-locks. The measurements
show that the transfer impedance could vary with a factor of 4
following the connector mating contact (difference observed
after three different connections).
IV.
The best compromise seems to be the braided shield of
silver-plated aluminum with a standard backshell. It presents
good EMC performances and a low additional mass. This
solution however needs to be implemented at the harness
design stage. When an overshielding solution needs to be
implemented as an EMI mitigation at system level after the
harness delivery, then aluminum foil or Nickel-Copper C12
cladded polyester fabric could be used; however with lower
shielding efficiency. For shielding efficiency almost identical
to braided shield of silver-plated aluminum, the self-wrapping
Nickel-Copper cladded polymer fabric and backshell with
sliding lid appears as an optimum solution but with a
significant mass increase.
To limit the RF leakages at Sub-D connector and backshell
level, a possible solution is to cover completely the socket at
equipment side with the backshell up to the unit chassis. In
order to improve the reliability and repeatability of electrical
contact, EMI gaskets for Sub-D connector and /or backshells
could be used. An alternative option would be design
modifications to the existing Sub-D connectors and backshells
to ensure 360° contact with the socket at the equipment side.
ACKNOWLEDGMENT
The authors would like to thank P. Bisognin, J.-P. Bonzom
and B. Theillaumas from Astrium SAS and L. Bernier,
G. Rouchaud, and E. Streissel from axon’ cable for valuable
discussions, test preparations and execution of tests.
REFERENCES
[1]
[2]
[3]
[4]
[5]
Fig. 7. Potential interfaces for EMI leackages.
[6]
Moreover the tests performed on Sub-D connectors have
demonstrated that the electrical contact between male and
female connector shells is random and the electrical contact is
Copyright 2014 IEICE
GUIDELINES
The reference case showed degradation of the cable
shielding effectiveness by the pigtail wire in high frequency.
Moreover, reliability during the connector mating and demating is poor.
[7]
843
Study final report for “Advanced Shielding Techniques for Spacecraft
Harness”, ESA contract no. 400102505, Technical report
2125.RP.PP.13.1283.ASTR, Astrium SAS, École Polytechnique
Fédérale de Lausanne, and axon’ cable, Jan. 31, 2013
Guidelines and trade-off for “Advanced Shielding Techniques for
Spacecraft Harness”, ESA contract no. 400102505, Technical report
WP410, Astrium SAS, École Polytechnique Fédérale de Lausanne, and
axon’ cable, Sep. 24, 2012
N. Mora, F. Rachidi, A. Junge, P. Pelissou, “Cable crosstalk analysis and
simulation: A comparison between low frequency circuit approach and
transmission line theory”, Proc. ESA Workshop on Aerospace EMC,
May 21-23, 2012
C. R. Paul, Introduction to electromagnetic compatibility. Hoboken, NJ:
Wiley, 2006.
F.M. Tesche, M. Ianoz, T. Karlsson, EMC analysis methods and
computational models, New York,NY: John Wiley and Sons, 1997.
P. Degauque, J. Hamelin, Compatibilité électromagnétique, Paris,
Dunod, 1990.
B. Démoulin, L. Koné, Shielded Cables Transfer Impedance
Measurement. IEEE EMC-S Newsletter Fall 2010, pp. 38−45