Euthanizing an old and crippled satellite isn't as simple as flipping a switch or pulling a plug.

In fact, scientists and engineers at the Johns Hopkins University worked unexpectely late into the afternoon yesterday, trying to drain the stubborn batteries of NASA's orbiting FUSE observatory and putting to rest an eight-year mission that tested their ingenuity and patience to the very end.

For one astronomer, it was a particularly melancholy moment. "It's sad - I'm a big loser here," said Steve McCandliss, who has been has been working on the FUSE project since his days as a grad student in 1981.

For the news cameras, the FUSE team had planned a little mission-ending theater, turning the satellite's solar panels away from the sun one-by-one, and letting the control room monitors on Hopkins' Homewood campus go dark.

A group gathered in a concrete basement hallway to toast the space telescope's demise with champagne. "It's a sad thing to turn a satellite off, but we had a lot of great years," said Warren Moos, the Hopkins astronomer and physicist who was principal investigator for FUSE.

But a few feet away, in the FUSE control center, engineers were still laboring before computer screens, watching as battery voltages stubbornly refused to drain.

By design, satellites such as FUSE are hard to kill, so engineers were willing to settle for a deep coma. For example, to prevent overcharged solar batteries that might explode, controllers had to trick the spacecraft's electronic brain into accepting drained batteries as fully juiced.

At 5:27 p.m., the telescope responded to its last command. "It put up a good fight, but we finally got it turned off," said William P. Blair, chief of FUSE operations at Hopkins. That should keep it in a deep sleep for 30 years or so, till it falls back through the atmosphere and mostly burns up.

The telescope's $225 million scientific mission ended July 12, when the last of the telescope's four troublesome "reaction wheels" ground to a stop, making it impossible to aim at stars and galaxies.

It was the last in a series of extraordinary technical problems that forced controllers to devise innovative fixes, all without laying an astronaut's glove on the telescope itself. FUSE flew too high for space shuttle repair missions.

"We had been plagued with this in a way no other science mission ever had," said Blair. But, "you can't complain about getting eight years out of a satellite designed for three."

In all, FUSE (for Far Ultraviolet Spectroscopic Explorer) spent 130 million seconds - about four years - making observations of 2,800 celestial objects.

"That's twice as good as we thought we were going to do ... a good, healthy efficiency," Moos said.

Astronomers who used FUSE have published more than 1,200 scientific papers, Moos said, "and they keep on coming." The processed data are being stored for future researchers at the Multi-Mission Archive at the Space Telescope Science Institute in Baltimore.

Unlike the world-famous Hubble, this one produced no glorious photographs. FUSE was a spectroscopic telescope. It broke starlight down into its constituent spectra, or wavelengths, so scientists could study the chemistry, velocity and temperature of objects and regions invisible to telescopes on the ground or in orbit.

Over the years, for example, scientists confirmed the existence of a halo of hot gas around our Milky Way galaxy - the exhalations of exploding stars.

They also looked for molecular hydrogen on Mars, a remnant of the planet's vanished ocean, and estimated that the ancient global sea was about 100 feet deep.

They probed "empty" space looking for deuterium, which is an exotic form of hydrogen - a "fossil" of the big bang and a key to understanding the subsequent evolution of the universe. And FUSE found some of the universe's "missing" matter - in the form of million-degree gas floating between the galaxies.

"I'm very pleased and very happy about the success of the mission ... It's been a good run," Moos said. But it was anything but smooth. Problems began almost immediately after FUSE was launched in 1999.

The 25 controllers, engineers and staff in Hopkins' basement control center at the Bloomberg Center for Physics and Astronomy noticed within weeks that solar radiation was knocking out data in the craft's memory chips.

So they had to reprogram the satellite to automatically reload lost data from backup memory. No big deal, Blair said. "That was basically just growing pains."

By October 1999, the FUSE team realized that two of the telescope's four mirrors kept drifting out of alignment, apparently because of to temperature swings. The error was less than the width of a human hair, but repeated realignments were eating up precious time. So they rescheduled observations to minimize the problem.

Then, in November and December 2001, disaster struck. Two of four reaction wheels quit working, unusual for typically reliable hardware. Reaction wheels turn, aim and hold the telescope on target. Three are needed and FUSE now had just one and a backup. "NASA thought the mission was over," Blair recalled.

But scientist Jeffrey Kruk had a possible fix." I did some initial back-of-the-envelope calculations and it was physically possible," he said.

Kruk proposed using FUSE's "magnetic torquer bars," designed to be a weak brake against Earth's magnetic field, as a third "wheel" to help turn and aim the telescope. There were skeptics, Blair said, but "we were struggling for our lives."

It was an enormously complex problem. FUSE was 500 miles up, flying through a magnetic field that changes with latitude, longitude, altitude, the time of day and the season. But for some, that was the challenge. "Many of the people involved were jazzed," said Blair. "The rest of us were going, 'I need a beer.'"

But it worked. In eight weeks they were able to hold FUSE on target, and soon the scientific data were rolling in again. That lasted until December 2004, when the third reaction wheel broke down, leaving just one working wheel and the torquer bars, a much bigger problem. It took 11 months for the team to concoct a new pointing system for the telescope.

But, "over time we got the thing back to a very respectable performance," Blair said. From November 2005 until last July, with few interruptions, FUSE continued to explore the universe. It couldn't reach every part of the sky, but its extra longevity was a good trade-off.

Looking back, Blair said, the troubles "probably worked to our advantage. ... People who operate these missions are extremely capable people. So if you had a mission that launched and went exactly as planned, those people would leave. ... It never got boring."

frank.roylance@baltsun.com