It's the other Hubble problem, overshadowed by the tragic mirror flaw that robbed the $1.5 billion Space Telescope of its full promise of early discovery in the distant and faint universe.
More than six months after launch, the telescope's two solar panels are still "jittering" when they pass between day and night in orbit, a vibration apparently caused by improperly designed metal support rods that heat unevenly in the intense sunlight.
"It's a significant problem that affects in a fundamental way the stability needed for precise pointing," said Jean Olivier, Hubble deputy project manager at NASA's Marshall Space Flight Center in Alabama. "These arrays are always kind of doing something they shouldn't."
Although engineers at the National Aeronautics and Space Administration hope to tame the jitter with modified pointing-control software, the flaw will remain unless new arrays are added in 1993, during the same shuttle mission that will carry a camera fitted with prescription lenses to correct the bad mirror.
"There was an oversight in design, certainly," said Robin Laurance, Hubble project manager for the European Space Agency, which oversaw construction of the panels in Britain and West Germany. "But this was something impossible to test for on the ground, unlike the mirror."
On top of the months-long agony surrounding the focusing flaw, the continuing solar array problem has caused tension between the U.S. space agency and ESA. One NASA source sarcastically referred to "the problem our friends in Europe gave us."
But Dr. Laurance cautioned that, "to some extent, the problem is overstated." He added that blame goes "to all parties involved," including NASA officials who participated in the review process during design and construction of the panels.
The lightweight 40-by-8-foot panels -- collections of thin solar cells that convert sunlight to electricity to power the spacecraft -- are so large and flimsy they would have collapsed under their own weight if unrolled for testing on Earth, he said.
For launch aboard the shuttle April 24, the panels were rolled up tight in "cassettes" tucked against the sides of the spacecraft, then rotated into position perpendicular to Hubble and unrolled like window shades once in the weightlessness of the 281-mile-high orbit.
The jitter showed up right away, each time the telescope crossed the day/night "terminator" -- on each 97-minute orbit, there is one sunrise and one sunset, nearly 30 thermal cycles a day in which temperatures range from minus 150 degrees to more than 200 degrees.
"It's a flapping motion like butterfly wings, every 10 seconds or so," said Edward Weiler, Hubble program scientist at NASA headquarters. The vibrations last six minutes when passing from day to night and up to 20 minutes after entering sunlight from the cold dark.
The result, in combination with the focusing problem, was a further erosion of the telescope's abilities to perform science, particularly during the daytime part of the orbit.
"If you want to take a picture of something now, you have to schedule it twice, because roughly half of the observations get knocked out by the jitter," said David Skillman, senior system engineer for Hubble at Goddard Space Flight Center in Greenbelt.
Dr. Skillman said daytime observing remains "possible," and two spectrographs aboard are relatively unaffected by the jitter. But best viewing comes at night "when the arrays go quiet for 10 to 20 minutes" and the telescope's vaunted pointing precision is almost achieved.
To complicate the situation, there is at least one other vibration that randomly shows up after the day/night jitters have died down, a smaller, faster oscillation of still mysterious origin.
"The panels will go quiet, and all of a sudden, these little 'pings' occur" that vibrate the spacecraft about every two seconds, said Dr. Skillman. "Probably 50 to 60 percent of every orbit has detectable vibrations of some kind. It's sort of an unquiet environment."
An explanation for the day/night problem was in hand within weeks of launch, with suspicion focused on the flexible metal rods, or bistems, which pull the solar cell blankets out of the cassettes during deployment and provide lengthwise support once the panel is fully extended.
When the rods first enter daylight, the front side facing the sun is heated more than the back side, and a temperature difference, or gradient, is established that slightly bows the bistem -- and the solar panel -- away from the source of heat.
This was no surprise to the European designers of the panels, who allowed for a slow bending of the rods over a five-minute period in which the temperature difference between front and back gradually reached its maximum amount, Mr. Olivier said.