volts, relative to plasma potential. For cells with silver-coated interconnects, a threshold ... ground and. LEO experiments. The arc rate beyond the threshold voltage depends .... ground tests, and x is about. 8 for the fronts and backs of. 5.9x5.9 cm cells ... ground and orbital data. This threshold may be a function of solar cell.
N'Sb-2555s SOLAR
ARRAY
ARCING
IN
PLASMAS
Dale C. Ferguson Space Environment Effects Branch NASA Lewis Research Center 21000 Brookpark Road Cleveland,
Ohio
44135
ABSTRACT Solar cells known to interconnects of a few potential.
in arc
hundred For
interconnects, arcing exists
ther tions,
space plasma conditions are into the plasma when the are at a negative potential volts, cells a at
V,
voltage as found
called
by arcs in cells isolation from a Models for the on the choice
conductor and possibly
material on the
for in
copper
at
a few relative
hundred to the
volts local
of flying with it
rounding (Kennerud
low
cells, voltages,
which are
individually typically
in space
will
contact
the
the insulating electrons from
plasma, 1974)
tion about arcing arc rates, as well electron currents, Because arcing systems of some
large. another
standard
The are
techno-
space
plasma.
the
first and
in later
cover glasses plasma as
if
effect, known understood in generated on over to be
of a few plasma, into the ground in
hunsolar sur-
tests orbital
of
the
voltage thresholds and as about the "snapover" in LEO conditions. obvious
implications
and anomalous current exposed to the space interest to understand
collection plasma, these
of on it is plas-
ma interactions with spacecraft systems. This paper reviews current progress in understanding solar array arcing in plasmas.
generate strung
In
applicaacross
flight tests (PIX-I and PIX-II, Grier and Stevens 1978, and Grier 1983). PIX-I, because of limited plasma diagnostics, essentially only confirmed that arcing was "not an effect caused by the plasma chamber walls. PIX-II, however, yielded informa-
power systems capable of generating large amounts of power. To keep cable masses low, with no loss of efficiency, high voltages must be used; much higher than the 28 V systems typically orbited until now, and even higher than the occasional i00 V used on Skylab.
Solar
quite cell to
At high negative potentials dred volts relative to the cells were observed to arc
negative plasma.
large structures a need for
may become from one
power voltage
they were conductors. This as "snapover", has been terms of secondary electrons the cover glasses "hopping" collected at the conductor.
INTRODUCTION
The prospect space brings
high total
interconnects.
ma, even collected
to the plasma and choice of
adjacent insulator material. Finally, new technology solar cells use copper for the cell interconnects, a material which may have a lower arcing threshold voltage than silver. It is expected, from ground tests of simulated solar cells, that any junction of conductor and insulator exposed to space plasma conditions will arc into the plasma potential,
for the
Early plasma testing of solar cells in simulated low Earth orbit (LEO) plasmas (Cole et al 1968, Stevens 1978) revealed that at high positive array potentials of a few hundred volts relative to the plas-
having insuffisource Of high arcs are highly of interconnect
exposed geometry
series that
logy solar arrays, the interconnects are coated with silver, for ease in soldering, and are exposed to the surrounding environment. Newer technology cells are welded to a copper trace from the back, so that little conductor is exposed in front. If the cell backs are not well insulated,
both ground and LEO experiments. The arc rate beyond the threshold voltage depends nearly linearly on plasma density, but has a strong power-law dependence on voltage, such that for small increments in operating voltage there is a large increment in arc rate. The arcs generate broadband radio interference and visible light. In ground tests, interconnects have been damaged cient current. dependent
so
the array connections
relative to plasma with silver-coated
threshold about -230
in
toge-
509
PRECEDING PAGE BLANK NOT FILMED
ARC
RATE
AND
Ferguson
THRESHOLD
(1986)
MEASUREMENTS
shows
that
the
onset
of
arcing in solar array plasma tests may not accurately reflect the voltage threshold. If, for example, the arc rate at some combination of plasma conditions and bias voltage is move on to is observed. ground tained 1984), depended ing
not depend or exposed For a large nects, one
very low, the experimenter may higher voltages before arcing When arcs were observed in
tests, and arc rates could be ob(Miller 1983, Leung 1985, and Grier it was found that the arc rate on the conditions in the follow-
approximate
array designs. Figure 1 shows the arc rate behavior found for several arrays of 2x2 cm and 2x4 cm cells in ground and orbital tests, normalized by the plasma parameters in equation (i). It is worth mentioning that the arc rate observed does strongly on interconnects array with pinhole in
thus arc effectively interconnects were space plasma.
the
number of cells at high voltage. insulated interconthe insulation will
as all
much exposed
as
if to
the the
way: I0o
R
=
C 1
n
T 0"5
m -0"5
where C 1 is a constant, density, T is the plasma the plasma ion mass, V
V x,
(i)
n is the plasma temperature, m is is the interconnect
voltage relative to the plasma and x is approximately equal to cm cells, 2x4 cm cells, and the 5.9x5.9 about
cm 8
5.9x5.9
cells in for the
cm
cells
tests. The a value for in orbit. fronts
of
x and
contrasted tests.
If the observed voltage between uous shown
tests, and
5.9x5.9
cm
in
x
ground
cells
only
and
the
together may be caused the exposed conducting cell fronts and backs later. The difference
other
10-I
is of
results yielded for 2x4 cm cells between the
for the 2x4 cm x in space may of atomic oxygen with
and backs
together,
PIX-II flight x of about 3 The difference
fronts and backs a difference in materials on the will be discussed between tests presence
ground fronts
potential, 5 for 2x2 fronts of
by
10-2
as
cells in ground be due to the in space, as
gases
used
in
ground
voltage at which arcing is first in a test is interpreted as that at which the average time interval arcs becomes less than the contin-
test that
time at that (Ferguson,
voltage, 1986):
it
may
be O • • t
10-4 Von where C 2 is
=
C 2
n (-I/x),
Von is the a constant,
(2)
apparent and n
fined before. Thus, an dependence of the arcing fact be simply a reflection voltage dependence of the
The 2x4
true voltage cm cells with
nects, which standard polated found plasma, orbital
the arc below
from higher be about -230 from data.
onset voltage, x are as de-
apparent density threshold may in of the steep arc rate.
threshold for silver-coated
defined as the measured deviations to
and
PIX II flight PIX IIground Leung Miller
arcing for intercon-
potential rate is the rate
voltages, V, relative
all available This threshold
function of solar cell geometry ials, and should not be taken the threshold for arbitrary
i0-5
below several extra-
has
ground may
to
1 Threshold 10 .6
been the
i00
and be
' 200
I ,I,I,I 600 -V.V
I I000
__
2000
a Figure standard LEO ram
and materto represent or new solar
510
1
Arc rate interconnect conditions.
versus cells,
voltage normalized
for to
rent EFFECTS Of arc
OF
SOLAR
course, occurrence
craft (1985),
ARRAY
it
is if
operations as part
experiment ofrequency cells. found.
only the of
necessary arcs may
or the
to harm
systems. now-defunct
avoid spaceLeung VOLT
been
on
repeatedly
plasma arrays,
produce with
to
visible optical
spectrum of been measured.
the as
capacitance of also found by EMI generated may be significant. For to generite the high
radiofrequency
light, which experiments. the
If arcs occur insufficiently
arcs
has
in solar isolated
2_
early
experiments
EMI, may The not,
negative self.
In
(diodes,
where
potentials this case,
array in
arrays from
a
I
arcs
which high
large source
his
experi-
experimental resistance array to space of the
be
solar high
may be the array the strength of
on the internal
total array
itthe
array capaconnections
etc.).
Adverse gated
Since standard a large and the
For the
arcs will depend citance and on
arcing
several
effects
may
be
miti-
ways:
Design
the
system
so
that
high
negative potentials relatiye to the plasma will exist nowhere in the system. This will mean, in practice, one of two design solutions. Either the total array voltage, from end to
interfere visible to
arced.
tested. however,
i. addition
in
ence, it has become technique to place between bias sources
potentials necessary for arcing, EMI may be significant if it couples to spacecraft electrical systems. Communications between spacecraft, or between spacecraft and telerobotic systems may be disrupted. In
as
a high voltage power supply was used to bias the arrays, the large arc currents may damage materials at the arcing point. Miller (1983) found partially melted interconnect material in arrays which had
development, measured the radinoise spectrum of arcing solar Figure 2 shows the spectrum he Depending on the strength of the
arcs (which depends the array to space, Snyder in 1985), the negligible or quite arrays large enough
source,
ARCS
date,
end, must be current-collecting
are cur-
at
I
the
I
negative
limited, area end
I
of
or is the
a large provided array,
I
so
I
--.x-- 2000 pF CAPACITOR -'D0
_
W ITHOUT CAPAC ITOR
"--_._. e'--"e"
E .
"_,,,"x-----x-,,_
..
COMPOS ITE PAYLOAD
",--..,...._-..._ ''
IN PAYLOAD BAY
....
100
LOCATED ON AFT FL IGHT DECK 50 O.Ol
I O.1
",
I 1
I I0 FREQUENCY
Figure ated cells, with
2
- Spectra of RF a 4 cell array with welded-through and without an added by
511
I I00 (MHz) radiation generof 5.9x5.9 cm interconnects, capacitance.
i i,000
_ I I0,000
there that
will end.
be
a The
low potential first solution
drop
at will
As
mean that large_ cable masses must be used to limit I_R losses. The second solut4on will push the positive end of the array up to high positive potentials relative to that snapover problems important. 2.
Insulate
the
thoroughly
that
the
no For
EMI, etc., in mind,
pendent array unit, so the up linearly
is
so that may be This may
to tolerable though, that
each
insu-
levels. inde-
All the on at
one go arc
4. Use materials for interconnects, coatings, etc., which will not arc a_ the high negative potentials likely to exist on parts of the array. This option will be further examined in the next section.
electric on the
Interesting
exist
in
arc rates between surface materials. tests (Thiemann and have shown that
arcing arrays For Bogus arrays
layer surface
cell or system.
flight arrays oxide atomic testing
threshold. tion techniques
is
Obviously,
of of
part
more
ground tests to investigate dependences
testing. In this case, the orbital were likely covered with a thin coating from interaction with the oxygen in LEO, whereas pre-flight was done in an argon plasma.
materi-
Copper in his the local copvalue less than
indicating a for copper
lower than
be understood if arc rate depend of the materials are two popular
of In
arcing one, a
contaminant the conductor,
of
is
at thin
high di-
built up and suf-
work
Center,
TRW,
some
other
needs
to
spacecraft
be
done
in
and
space flight experiments, the material and geometry of the arc rate and arc
Only be
then can employed.
Shuttle experiment on solar arrays in Solar Array Module Experiment, a joint manifested for late ting continues at
Miller (1983) showed that in ground tests the solar array arc rate started out high, and then decreased to a constant level on
ground tests using plasma. PIX-II in orbit (Ferguson
possibly voltage
for the onset potentials.
Solar arrays for the PIX-II experiment had a much higher voltage for the onset of arcing in pre-flight ground test experiments than they exhibited in orbit (Ferguson, 1986). The voltage dependence of the arc rate of the PIX-II arrays was also very different than found in pre-
effect in background this effect
insulating
and the observations they are meant to explain, arcs may be expected at negative potentials of a few hundred volts negative, relative to the plasma, at conductor-insulator junctions, regardless of whether the junction occurs on a solar
arcing.
a time scale of a few hours, as sites were cleaning themselves testing. Leung (1985) observed
of
tests at and backs
insulator under electron bombardment may lead to an avalanche into the plasma if the electric fields are high enough (Hastings et al 1989). In both of these models material properties play a strong role, as does the presence of high electric fields near conductor-insulator interfaces. On the basis of these models
having the interconnects coated with an "insulating" material sometimes arced at lower voltages than they did before the interconnects were coated. The effect amounted to hundreds of volts in the onset of
fronts
ficiently high electric fields may be produced in the vicinity of the insulator to punch through the layer, triggering an arc (Jongeward e__t al 1985). In the second model, breakdown of gas emitted by the
DEPENDENCES
differences
thresholds and having different instance, ESA 1986 and 1988)
with
of these resu!ts may arcing threshold and the surface properties the arc sites. There
models negative
MATERIAL
covered
about -120 V, arcing threshold for silver.
Segments.
POSSIBLE
the
old found in other tests. tests arced so as to bring per potential in arcs to a
only a dumped limit
segment will arc as total arc rate will as the number of
mentioned,
cells arced in ground rate than the fronts
al, to simulate the conductor-insulator junctions on solar cells. Silver in his tests arced so as to bring the local conductor potential down to about a -230 V level, coincident with the arcing thresh-
so
plasma contact will be made. long-lived arrays, micrometeoroids and/or debris may puncture the lation, nullifying this solution.
the Bear
cm
partially
certain
3. Design the arrays small amount of charge when an arc occurs.
already
together. Here, the significant fact may be that the cell backs had copper exposed to the plasma, rather than silver. Supporting this contention are the ground tests performed by Snyder (1986) on metals
plasma, so may become
interconnects it
was
5.9x5.9 a slower
and
proper One
mitigaapproved
to LEO
investigate arcing is the SAMPIE, or Plasma Interaction NASA/ESA venture now in 1994. Ground tesNASA Lewis Research
elsewhere.
if the arc up during the same CONCLUSIONS
a different also showed 1986).
EMI generated in generate radiofrequency
512
solar
array arcs noise which
may might
disrupt negative surfaces
telerobotic potentials are a
communications. High on other spacecraft possible threat to the
successful operation of spacecraft tems, including automation and electronics and communications. necessary to consider and solar-array-type
solar arcing
Grier,
sysrobotics It is
array in the
arcs may potentials
also of
occur the
during docking
Material thresholds
in system design, and them relies to a great spaceflight experiments, remain to be done.
N.J.
(1978),
Experiment
(PIX)
Spacecraft
1978,
NASA
Charqin_
CP-2071,
pp.
295-
Hastings, D.E., Weyl, G., and Kaufman, D. (1989), "A Simple Model for the Threshold Voltage for Arcing on Negatively Biased High Voltage Solar Arrays", Journal of
docking, vehicles
dependences are important
Stevens,
Results",
Technoloqy 314.
arcing design
differ sufficiently. Thus, control of spacecraft potentials is important in spacecraft design, if reliability and communications are important to spacecraft systems. rates and
and
Interaction
Flight
of sPacecraft power systems and other systems which may be affected by arcing. Systems should be designed to mitigate the incidence and effects of arcing, whenever possible. Although not the topic of this paper, if the
N.T.
"Plasma
of arc factors
Spacecraft
and
Jongeward, Unneutralized
G.A.
submitted.
et al Surface
Potential vol. NS-32,
Arcing", no. 6,
Kennerud,
K.L.
Array
our knowledge of extent on lab and some of which
Rockets,
(1985), Ions
IEEE Dec.,
(1974),
Experiments",
The Role in Negative
Trans. pp. "High
NASA
o_
Nucl. 4087-4091.
Sci.,
Voltage
Solar
CR-121280.
Leung, P. (1985), "Characterization of EMI Generated by the Discharge of a 'Volt' Solar Array", JPL D-2644, Jet Propulsion Lab, California Institute of Technology, Sept.
1985.
Miller, W.L. (1983), "An Investigation of Arc Discharging on Negatively Biased Dielectric-Conductor Samples in a Plasma", a_r_ Environmental Interactions Technoloqy 377. REFERENCES
Snyder, D.B. Arc Currents
Cole, R.K., Ogawa, H.S., Jr. (1968), "Operation Arrays in Dilute Streaming CR-72376. Ferguson, D.C. (1986), Threshold for Arcing for LEO Flight and Ground NASA TM-87259.
and Sellen, of Solar Plasmas",
"The Solar Test
J.M., Cell NASA
pp.
Grier, Coupling Array in Conference,
Cell
Array
Snyder, tion.
Voltage Cells in Results",
Grier, N.T. (1983), "Plasma Interaction Experiment II (PIX If) : Laboratory Flight Results", Spacecraft Interactions Technoloq¥ 1983, NASA 2359,
1983,
in
a
N.J.
1978,
Thiemann, H. lous Current
CP-
Solar-Cell Density Journal,
N.T. (1984), "Dilute Plasma Currents to a High Voltage Solar Weak Magnetic Fields", 19th IECEC San Francisco.
Thiemann,
pp.
CP-2071,
and Bogus, Collection
and
"Interactions
Be-
Charged-Particle Char_inq Technopp. K.
268-294. (1986), "Anomaand Arcing of
K.
(1988),
Voltage Solar Cell Modules in Low-Earth-Orbit Plasma", Journal craft and Rockets, Vol. 25, pp.
513
of Solar
communica-
in a Simulated Orbit Plasma", pp. 43-57. Bogus,
367-
TM-83728.
private
and the Spacecraft
Modules Low-Earth Vol. i0, H.
NASA
(1978),
NASA
and
333-347.
Plasma",
(1986),
tween Spacecraft Environment", loav
CP-2359,
(1985), "Characteristics on a Negatively Biased
D.B.
Stevens,
NASA
HighESA-
"High
simulated of Space278-285.