EGSE: where the satellite meets the ground before leaving it

A satellite is an object designed to be unreachable. Everything that will be asked of it over ten or fifteen years in orbit has to be demonstrated while it is still standing in a cleanroom, wired to connections that will be disconnected once.

This is the part of a space programme where ground support equipment does its work. It is also the part that rarely appears in the images that reach the public: no launch, no separation, no first signal. Just racks, harnesses and a flight model under test, surrounded by equipment that will never leave the building.

Electrical Ground Support Equipment, EGSE, is that equipment. Its purpose is narrow and demanding at the same time: to stimulate the satellite and its subsystems exactly as the mission will, to observe how they respond, and to record the evidence that they behaved as specified.

The last place where a design issue can still surface

During Assembly, Integration and Testing (AIT), the flight hardware exists in a state that will never occur again. It is complete enough to be exercised as a system, and still accessible enough to be instrumented, adjusted and repaired.

Once that window closes, the options change completely. An anomaly detected after launch is not a technical problem to be solved on a bench, it is a mission decision, taken with partial telemetry, limited time and no possibility of physical access.

This asymmetry explains why the requirements placed on EGSE are, in several respects, stricter than those placed on ordinary industrial test equipment. The equipment has to be trusted enough that its own behaviour is never a candidate explanation for what the satellite is doing. When a measurement looks wrong during a qualification campaign, the first question is whether the unit under test is at fault or the bench is. Good ground support equipment is designed so that this question can be answered quickly, and ideally so that it does not arise.

Two levels of check-out

Check-out equipment operates at two distinct levels, and the distinction matters more than the acronyms that describe it.

At system level sits the Overall Check-Out Equipment, the OCOE. It addresses the satellite as a whole: it establishes the telemetry and telecommand link, runs the operational sequences, monitors the responses and holds the reference against which the entire campaign is evaluated. It is the level at which a test operator asks whether the spacecraft, as an integrated object, does what the mission requires.

Below it sit the subsystem benches, the SCOE, each dedicated to one part of the satellite. Power units, communication modules, data handling, payload electronics, sensors and actuators are each exercised by equipment designed around that subsystem's own interfaces, signal levels and failure modes. A bench built for a power conditioning unit shares very little with one built for a payload data interface, and trying to make one serve both usually produces equipment that serves neither well.

The two levels are complementary. The subsystem benches allow problems to be isolated and characterised where the physics is still tractable. The system level ensures that subsystems which are individually correct also behave correctly together, which is a different question and frequently a harder one.

Simulating what the satellite will not have on the ground

Some parts of a mission environment cannot be brought into a cleanroom. They have to be simulated with enough fidelity that the flight hardware cannot tell the difference.

Electrical power is the clearest case. In orbit a satellite is powered by its solar array, whose behaviour is nothing like a laboratory power supply: the current-voltage characteristic is non-linear, it moves with illumination and temperature, and it changes over the mission as the cells degrade. A power conditioning unit designed for a source of this kind has to be tested against a source that reproduces that behaviour, not against an approximation of it.

Solar Array Simulators reproduce this behaviour electronically, allowing the satellite power chain to be exercised across operating points, eclipse transitions and end-of-life conditions that would otherwise be impossible to reproduce on the ground.

The same logic applies one level down, to the cells themselves. Solar Cell Test Systems characterise photovoltaic devices from single cells up to strings and complete panels, producing the electrical characterisation that qualification requires and that feeds back into the simulation used at system level. IPSES developed SoCRATes for exactly this purpose, as an instrument dedicated to solar cell measurement.

What actually makes ground support equipment good

Beyond the architecture, a small number of properties separate equipment that supports a campaign from equipment that slows it down.

• Repeatability. A test that cannot be reproduced identically has limited value as evidence. The bench has to produce the same stimulus and the same measurement conditions today, next month and after a component has been replaced.
• Traceability. Every measurement has to be attributable: to a configuration, to a calibration, to a moment in time. Qualification is a documentary exercise as much as a technical one, and equipment that makes documentation difficult makes the whole campaign difficult.
• A common time reference. When several instruments observe the same event, their timestamps have to mean the same thing. Otherwise the correlation between a command and a response becomes an interpretation rather than a measurement.
• Longevity. Space programmes outlive electronic components. Equipment delivered at the start of an integration campaign has to remain maintainable, calibratable and reproducible years later, which is a design constraint from the beginning rather than a maintenance question afterwards.
• Operability. The people using the bench during a campaign are working under schedule pressure on hardware that cannot be replaced. The equipment has to be legible to them, not only to the engineers who built it.

Turnkey systems, or single elements

Not every programme needs a complete system, and not every organisation wants one.

Some customers require a full delivery: hardware, custom electronics, test and supervision software, documentation and acceptance. Others already have an established test environment and need one element that does not exist on the market, a specific instrument, a conditioning stage, a software layer that integrates equipment from different suppliers into a single sequence.

Both are legitimate, and they call for different kinds of supplier relationships. Delivering a single element into someone else's architecture requires understanding constraints that were set before the request arrived, which is often more demanding than designing a system from scratch.

Where the competence comes from

IPSES did not arrive at ground support equipment from the space sector alone. The company's work in functional test systems for industrial, defence and broadcast electronics has, for years, addressed requirements that carry different names but ask the same things: measurements that hold up under scrutiny, systems that stay reproducible over long production or qualification cycles, and a time reference that is coherent across an entire installation.

That last point is not a coincidence. IPSES's work on GNSS-disciplined timing and network time distribution comes from a different product line, but the underlying question is identical to the one a test operator faces in a cleanroom: when two instruments report the same event, do they agree on when it happened?

The European framework for space engineering, the ECSS standards, formalises much of what is described above. Working within it is less a matter of compliance than of shared language: it defines what the customer expects to receive, in what form, and with what evidence attached.

What comes next

Ground support equipment is not the visible part of a space programme, and it is not supposed to be. Its success is measured negatively, by what does not happen after launch.

In the coming weeks we will look at individual areas in more detail, starting with how dedicated subsystem architectures reduce risk during qualification, and later at solar array simulation as a discipline in its own right.

If you are building satellite electronics and the bench that will validate them is still an open question, that is exactly the conversation we are interested in having.