Over a thousand parameters, live, from one aircraft

The context

Sector

State aeronautical development agency, Asia

Platform

High-performance light aircraft, new variant

Scope

Full flight-test instrumentation, live telemetry

In service

First flight 2023 — ETEP on the programme since 2005

The programme is a national aircraft development, and the instrumentation history runs with it. An earlier generation of ETEP recorders equipped the flight-test bench from 2005, and carried the acquisition and recording through some fifteen years of flight testing. When a new variant went into development, ETEP was selected for the complete instrumentation of the aircraft.

That continuity is worth naming, because it is the part a datasheet cannot express: two decades of recordings, in a documented format lineage, still readable by the same tool chain.

The constraint: everything, live

The requirement was a full instrumentation set — structure, loads, thermal, dynamics, buses, video — with more than a thousand parameters, and real-time transmission to the ground for the whole of it. Not a recorded set with a telemetered subset: the flight-test engineers needed to watch the campaign as it flew.

At that scale two things follow immediately. No single unit holds that many channels, so the installation is distributed by arithmetic rather than by preference. And every one of those units has to be aligned to the others, because a structural campaign compares channels for a living — a phase error between two accelerometers on different units is a modal analysis error.

The architecture

About ten acquisition units, each carrying twelve modules, linked through a dedicated network switch. One of them acts as master and holds the network controller: it assembles the whole installation into a single stream, transmits it in IRIG 106 Chapter 4 telemetry, and records onboard. The others are slaves, each with a network controller and a power supply module.

The point of that arrangement is that the aircraft carries ten boxes and the ground sees one instrument. One configuration to manage, one time reference, one stream to decommutate — which is what makes a thousand parameters analysable rather than merely recorded.

The channels

The shape of that table is the story. Strain gauges alone are a third of the installation, which tells you this is a structural campaign, and it is why the conditioning — bridge excitation, four-wire connection, per-channel filtering — mattered more here than any headline throughput figure.

What it changed

The result the customer bought was not a channel count. It was the ability to fly a new variant with the whole instrumentation visible on the ground from the first sortie, using a format their analysis chain had already been reading for fifteen years.

What we take from it

Large installations fail on coherence, not on capacity. Ten units that each work perfectly and disagree about the time produce a dataset nobody can use for a modal analysis. Everything that made this installation work — one master, one clock, one configuration pushed to all of it — is about making many boxes behave as one.