Miles from Earth at the Space Center ...

KENNEDY SPACE CENTER – For some engineers at NASA’s Marshall Spaceflight Center in Huntsville, Alabama, the highlight of this shuttle mission may have already happened.

At about seven o’clock this morning, as super cold liquid hydrogen (-423 degrees F) was flowing into the big orange external fuel tank, some gauges that these engineers have come to know intimately worked precisely as they were designed. Imagine that.

You could almost hear the sighs of relief on the NASA-TV channel.

The devices are called Engine Cutoff Sensors – or ECO for short (pronounced EE-coh). In the past few years, the shuttle team has been stumped by a series of odd, seemingly unpredictable failures of these sensors.

First a word about what they do: ECO sensors are gas gauges – there to sense if the liquid hydrogen is about to run dry – thus allowing the trio of main engines of the orbiter to shut off in an orderly fashion. That is an important function as a space shuttle main engine can break-up catastrophically if it is still running when the fuel stops flowing.

For the past few years, ECO sensors were failing repeatedly in the countdown to launch – not long after NASA started gassing up their rockets.

In December, this nagging issue came to a head and the space agency decided not to fly again until they could figure out why their sensors were failing.

Those Huntsville engineers – led by John Chapman – did some sleuthing at the pad, discovered the problem was somewhere in a 35 pin connector that joins the gauges on the inside of the tank with the wiring on the outside.

They yanked both sides of the connector out of the tank and took it to their lab in Alabama.

They attached a bunch of sensors and bathed the box in LH2. And they noticed the sensors failed every time at about -350 degrees F. That happens to be the temperature that air freezes. Yes air can freeze, folks!

Long story short, somehow air got inside the pin connectors as the shuttle sat on the pad, as the temperature plummeted during the cryogenic fueling, that air became a solid – and interrupted the circuit.

The short-term fix was fairly straightforward: simply solder the pins into those connectors. And this morning, it worked like a charm.

But no one is sure why this only cropped up only in the past few years – and not since day one for the shuttle (April of 1981).

Chapman wonders if maybe the pad rats started wrapping the wires with tape a little differently. That part of the story is still unfolding.

But this is yet another reminder that this is a machine with a million parts, and literally every one of them has to be perfect for it to leave the pad. Even the cheapest connector has to work as intended.

This morning, I asked NASA Administrator Mike Griffin if this is the stuff engineers really live for.

“Yah, in a certain way, yes,” he said with and engineer’s grin. “We’d rather not have problems, but when you get something like this that they can sink their teeth into, they are pretty formidable.”

As the astronaut gallows humor goes – “a million parts – all from the low bidder.”

And just when you think you know every which way every one of those parts can bite you in the behind, something like the mystery of the balky ECO sensors rears is ugly head.

- From Miles O’Brien, CNN's chief technology and environment correspondent

Filed under: NASA