[ale] [OT] grounding in space

[Alex Carver]

On 2017-08-28 13:55, Phil Turmel wrote:
> On 08/28/2017 01:50 PM, Jim Kinney wrote:
>> The (bs) EMP bags made me think (usually a good thing!) about circuit
>> design (a field I'm not good in).
>>
>> For a spacecraft, the obvious choice for ground plane is the craft
>> framework itself. Relative to the typical power loads in the various
>> modules on ISS, the frame is a "near infinite supply or sink of
>> electrons".
>>
>> But space weather is nasty. Solar flares are huge problems. How is a
>> system designed to withstand an EM flux greater than the typical power
>> throughput? On the ground, we use fat braided copper wires and deeply
>> buried rods. That's not an option on a satellite. Thick skin can
>> shield. What else?
> 
> You don't shield everything.  You shield the ICs with as small a shell
> as possible, and every other circuit in the whole spacecraft is a
> precision engineered tightly twisted pair w/ carefully balanced current
> flows.  Both power and signal.  Plus optical isolation between any
> devices grounded to different parts of the structure.

No, everything is shielded unless the circuit can handle random induced
currents (validated with EMI/EMC testing) or the effects of such random
currents can be mitigated in other ways (putting circuits to sleep,
using them only in benign environments).  All signal wires are shielded
at all times unless the wire must be unshielded to operate (a Langmuir
probe would fall in this category).  Power wires may or may not be
shielded depending on instrument susceptibility (for example, wires for
pyrotechnics would be shielded to prevent accidental activation).

All electronics are fully enclosed in metal enclosures because this
provides both electrical shielding and charged particle/ionizing
radiation shielding (for low level radiation, high radiation always
makes it through).  There's a limit to the amount of shielding you can
actually put on and not just because of weight.  High energy particles
can create radiation inside enclosures (called secondary
radiation/particles or just "secondaries") which could potentially be
worse than the original radiation.  Thin material reduces secondaries
but leaves you exposed to more of the lower energy primary radiation.
Thick material can quench low energy primary radiation but provides a
significant radiation cross section for generating secondaries.


So back to the first question of "what else?":

Heavy filtration on power leads using reactive components (inductors and
capacitors) can help reduce conducted interference (induced current).
Shielding of the cables (even power) helps reduce radiated interference.
 (Note that this also helps prevent an instruments own internal noise
from leaking into another instrument).  Clamping circuits can help with
some voltage issues caused by charging or induced voltage/current
(separate from EMI/EMC because induction is a low frequency process).
If you can avoid shielding by using good design principles then you can
save a lot of weight.

The spacecraft is considered to be at a floating potential.  Depending
on its size (like the ISS) a contactor may or may not be used.  The ISS,
due to its immense size, uses a plasma contactor because the whole
superstructure actually experiences an induced current as it travels
through Earth's magnetic field.  The contactor prevents voltages from
building up due to this and other processes.  Smaller satellites don't
have this feature and rely on natural leakage from the bus to the
environment.  Ground loops must be avoided if at all possible so shields
are broken at one end of cables.  A star topology for grounding is used.

Instruments must undergo environmental testing to determine if they are
susceptible to any of these issues.  If shielding can not be used on the
cables itself then "natural" shielding (by routing cables such that they
are hidden by other components) or rearranging the components to
minimize cable lengths (also affects mass) or crosstalk.

Parts selection is the next option.  Choosing parts that can handle
large input voltages without stress or adding additional clamping
circuitry to prevent voltage excursions would be done.  These are not as
ideal as shielding because they are accepting a more hostile environment
 but there may not be an alternative.

Much further down the list is concept of operation.  This would include
when and when not to operate equipment, automatic failsafes, live ground
intervention, etc.  All of this requires some level of autonomy or human
interaction and is the least favorable of all methods.  It does get
used.  For example, when one of the space weather satellites detects an
incoming coronal mass ejection (CME) an advisory is issued to all space
customers.  They can then choose to put their hardware into safe mode
(typically powering down many instruments), rotate the satellite into a
more benign attitude or, for the ISS, evacuate the crew to a safer
location inside the structure.