First Light

David Pacchioli
September 01, 1997
glass windows

"What this is," John Booth said, in a measured half-drawl, "is a really good idea that got totally out of control." The voice was perfect deadpan, but the lanky, boyish Booth—flannel-shirted, thumbs in his toolbelt—couldn't resist a flicker of a grin. Just behind us in the chilly, cavernous metal dome towered the almost-finished Hobby-Eberly telescope, the HET.

This was no ordinary telescope. Its primary mirror, 11 meters (over 35 feet) in diameter, is one of the largest in the world. Its height, roughly four stories, inspires a gaping awe. Yet its transparent design makes the 65-ton giant seem airy, even graceful. The main structure is an open steel frame, an elongated basket of tubular silver struts. Spanning the opening at the top is the black steel rig that holds the tracker, a white, Volkswagen-sized cylinder containing the brains of the telescope, its moving parts. Nested at the bottom of the frame is a turquoise dish of thinner struts welded together in triangular pattern: the bed, or truss, for the mirror. The huge instrument rests on eight flat feet, chest-high atop a circular concrete pier.

For Booth and the rest of the HET team on a cool morning in early December, it was crunch time. They had sweated and struggled for almost four years to deliver this telescope on time and on budget, and one big hurdle remained: It was time to see whether the thing would actually work. First light, astronomers call it.

First light had been delayed a couple of times already, most recently when a component on the tracker broke right before Thanksgiving. The team—11 employees of the university of Texas at Austin—had been planted here at McDonald Observatory in the middle of the Davis Mountains for some three months, shuttling the seven and a half hours home to Austin only once every other weekend.

The walls echoed with the rowdy Texas blues of Delbert McClinton. Booth, a mechanical engineer, was preparing a special high-resolution camera for installation in the "cheese," otherwise known as the tracker. Nearby, electrical engineer Larry Long was mounting temperature sensors on the dome wall. In the adjacent control room, project manager Tom Sebring, a "No Fear" sticker pasted to the front of his hard-hat, was engaged with systems analyst Joan Sage and consultant Frank Ray. A whiteboard on the opposite wall was crammed with numbered lists and runaway equations in several colors; taped next to it, an intricate flow chart captioned in ink:

Plan the work . . . Work the Plan.

golf ball-like dome on building

From the windy summit of Mount Fowlkes, where the HET stands all by itself, you can see a sizable chunk of West Texas. A mile southwest is Mount Locke, elevation 6,793 feet, where the white domes of the McDonald Observatory's two other major telescopes glow like shrines in the morning sun. Beyond, in all directions, there is only the harsh beauty of scrub brush and barren rock stretching under crystal blue sky all the way to the edge of the mountains, 30 miles distant. Lonely rangeland, by day. At night, this site, owned by the University of Texas since the 1930s, is one of the darkest places in the continental United States.

Penn State astronomers Larry Ramsey and Daniel Weedman must have been dreaming of such a place when they came up with the concept for the HET back in 1983.

They wanted a telescope in the 8-meter class, an instrument that would carry Penn State's astronomy department into the forefront of observational research. A conventional design, they knew, would be prohibitively expensive: the Keck telescope in Hawaii, of comparable size, had come with a price tag of $80 million. So they made a number of concessions.

First, they would limit their instrument to spectroscopy. Instead of pictures of stars, galaxies, and other astronomical objects, it would produce spectra, records of the light from those objects separated into its constituent wavelengths, to be studied for clues about the objects' chemical and physical makeup.

Next, instead of allowing a complete range of motion, they would fix the main body of the telescope at a permanent 35-degree tilt from the vertical. In this way they could greatly simplify the technical requirements for mirror support and high-precision motors. Instead of moving the primary mirror—and thus the entire telescope—to track stars across the sky, they would simply rotate the telescope toward the desired region, then maneuver the tracker, suspended high above the primary mirror and outfitted with secondary focusing mirrors, to follow the reflected light. This so-called Arecibo-type telescope would not be able to observe the entire sky, but could take in a sizable ring (about 70 degrees wide) around it. Stars could be observed as they passed through this ring.

Finally, to avoid the great expense of a single giant curve of glass, Ramsey and Weedman designed their primary mirror as a composite of 91 identical one-meter hexagons, set together in a honeycomb pattern to make one large surface.

By these compromises, Ramsey and Weedman found, in exchange for a moderate sacrifice in viewing capability, they could have a world-class telescope—for roughly one-fifth the cost of a Keck.

In 1985, Penn State formed a partnership with the University of Texas at Austin, which had also been looking for a major telescope project. Three more institutions, Stanford University and the German universities George-August-Universität in Göttingen and Ludwig-Maximilians-Universität in Munich, became associate partners in the early 1990s, and two individual benefactors, Lt. Gov. William P. Hobby of Texas and Robert E. Eberly of Pennsylvania, gave the telescope its name. After a 10-year design phase, groundbreaking on Mount Fowlkes took place in March 1994. Now, 14 years after its conception, the HET had become a reality.

"We only got it running yesterday," Sebring said. "Last night, we pointed it dead west at the Milky Way, and ran the software for the first time. We saw stars in the eyepiece and got some images on a home video camera.

"What we need to do for first light," he went on, "is acquire [i.e., focus on] a random star, track it for a few minutes, tweak the image and see how it holds together. And we need to get some digitized, high-quality images." Tonight, he thought, could be the night.

By 6:48, the whole team was back from dinner on Mount Locke. Larry Ramsey, HET project scientist, had arrived from Penn State to witness the impending big moment, and he and Sebring were in the dome, where the lights had already been dimmed. Sebring was demonstrating the HET's air bearing system, an innovation he had added to the basic design.

Under each of the telescope's feet, he explained, is a donut-shaped bladder, like a large inner tube, only perforated with tiny holes. When the bladders are inflated—to 20 psi, a lower pressure than most bicycle tires—they raise the telescope half an inch above its resting place on the pier. Leaking air creates a "film" between the bladder and the sand-papered, baby-powdered concrete surface, allowing the giant telescope to rotate virtually without friction. Using a hand-held device, Sebring pushed one button, then another, and the huge structure moved clockwise, then counter-clockwise, gliding with the smoothness of a spaceship.

At 7:00, optical engineer Bill Gressler climbed up onto the truss and began lifting the temporary covers off the mirrors. Only seven of the 91 had been installed, but their combined reflective area was already larger than that of the Observatory's 107-inch telescope—more than enough surface to prove the HET concept.

At 7:30, Booth, buckled into a yellow safety harness, stepped into the cage of the diesel-powered cherry-picker, and was hoisted slowly to the top of the dome by its hydraulic arm. Carefully he climbed into the white, bucket-shaped tracker, and the cherry-picker was lowered away in a huff of exhaust. Booth's job would be to spend the evening aloft in the cheese, watching its various systems as they kicked on, checking for malfunctions and reporting via walkie-talkie to the control room.

It took two hours, and a lot of back and forth on the walkie-talkies, to hook up the necessary cameras and align the mirrors. In the control room, the engineers bustled about. Then, at 9:30, with Operation First Light finally ready to move ahead, systems analyst Joan Sage's monitor suddenly flashed an emergency stop, and all 13 motors in the tracker shut down.

There followed two hours of checking software, testing initialization procedures, and scratching heads, interspersed with several phone calls to support staff in Austin, as Sage, Booth, and Frank Ray felt out the possibilities for what had gone wrong.

Building a telescope is a frightfully exacting job. One major worry, for example, is thermal expansion. As air temperatures fall, a steel structure the size of the HET can be expected to swell at least a few millimeters over the course of a viewing evening. Not a problem for an apartment building, maybe, but for an instrument whose focus needs to be precise down to microns, this presents a significant complication. To maintain accurate pointing at astronomical targets, the frame has to be built as light and as stiff as possible.

dome structure with town in background

View from the HET dome, at an early stage in its construction.

Temperature fluctuations inside the dome can also distort incoming light, which means that everything has to be carefully insulated. The HET incorporates a powerful fan system that can flush the entire dome of air 20 times an hour. When the shutter slides open in the evening, the air inside the dome has to precisely match what is outside.

The HET's unusual design presented its builders with additional challenges.

Getting those 91 mirror segments to act as a single curved surface, keeping them precisely aligned toward one another and toward the tracker, requires a system of lasers, precision motors, and an array of 50 temperature sensors spaced across the underside of the mirror truss.

Then, the HET's primary mirror is spherical, instead of the conventional parabolic shape—a rounded bowl instead of an oblong dish. This uniform curve is a much easier shape to make and test, but it also results, inevitably, in distortion: As project scientist Ramsey puts it, "The image formed by a spherical mirror is the size of a golf ball—it's a big fuzzy thing that is of no use to us." To sharpen the picture, the HET requires a set of corrective optics: four secondary mirrors which had to be built into the tracker.

Sebring, an optical engineer, had never worked on a telescope before the HET. But he knows mirrors, having consulted on the Strategic Defense Initiative (Star Wars) project during the late '80s. When the HET collaboration hired him in August 1992, he assembled a team from former Star Wars colleagues, employees of McDonald Observatory, and at least one refugee from the Superconducting Supercollider.

"This is a very engineering-based telescope," Bill Gressler says. "The tracker alone has 13 degrees of freedom, as opposed to three for a traditional telescope. It's one complex piece of machinery."

By 11:15, the lights in the control room were turned off. All the bustle had ceased. The entire project team, five engineers, two computer scientists, and Ramsey, was hunkered wearily around a trio of monitors, motionless, looking for a sign—all but Booth next door in the open dome, perched high above the floor with his fading walkie talkie and his flashlight. The only sounds were keystrokes, and the occasional rhetorical question.

"Well," said Frank Ray at last, "we aren't going to get it tonight."

Ten minutes later, Sebring bolted from his chair. "I know what's wrong!" he blurted, and tired eyes shifted warily toward him. "No music!" he grinned. There were groans. Ignoring them, Sebring went to the stereo and popped in a Miles Davis CD.

Midway through Miles' first buttery solo, Joan Sage let out a small whoop and snatched up the radio. "We're cleared," she told Booth. "We're showing no fault, no emergency."

Half an hour later, nothing had happened. Sage was making another hopeful announcement: "We have achieved initialization." But Ray was back on the phone. "We're still having tracker problems," he told Austin, and the threat of a breakthrough passed.

Outside the dome the wind was up, not cold but restless. The dark, when the control-room door fell shut, was nearly perfect: ink or velvet, a disorienting, fathomless black magnified almost unbearably by the accompanying silence. Not a sign of an artificial light anywhere in the world. Above, though, the vastness of sky came gradually clear, stippled majestically with a billion glimmering pinpoints.

Back inside, at 12:30, Booth's voice came thoughtfully out of a long gloomy silence.

"I hope I get conjugal visits up here."

Soon, an image, a muddy, oblong blotch, appeared on Ramsey's monitor; the room came alive.

"We've got something!" someone shouted.

"But no focus," Ramsey cautioned, "so we don't know what we've got." Nevertheless, he started saving frames, punching keys repeatedly to download as the image moved jerkily across the screen.

At 1:35, Ramsey announced that the Hobby-Eberly Telescope was successfully tracking.

At 2:00, six blurry stars became visible on the monitor, trailing across a murky field. Ramsey peered for a moment, then frowned and said quietly, to nobody in particular, "We're dominated by the angular misalignment."

Sebring countered eagerly: "Is this first light?"

Ramsey replied as if suddenly awakened: "No, but damned close. The tracking is awesome.

"We know what we have to do," he added, after a long pause. "It's not a fundamental problem. Let's call it a night."

By late that morning, Sebring and Booth, cat-napped and coffee-fueled, were back in the dome and immersed in a strategy session, setting priorities. Several things, it was agreed, needed doing. But what needed doing right now?

Through the afternoon, Sage worked on software modifications. ("We need to be able to make mid-course corrections while tracking," she explained.) Larry Long strung ethernet cable. Booth worked on his cameras and his alignment data.

At 4 p.m., the stiff wind that had been blowing all day kicked up to 39 miles-per-hour; the sky was streaky with clouds. Inside, the aluminum dome rumbled like a garden shed.

At 8:30, Sebring was crouched on the truss with a spherometer, measuring angles.

In the control room, somebody punched up the National Weather Service on the World Wide Web. The satellite image showed a thick tongue of cloud-cover closing from the Pacific.

"Tonight," someone said, "looks like it's going to be our last chance for a while."

By 9:30, the mirrors were in good alignment. The group in the control room included a few new faces, staff from other parts of the Observatory who had showed up in anticipation of success.

At 10:00, Booth made his ascent. "We're gonna look west, at Pegasus," Sebring said. "Same as last night."

Soon enough, at 10:40, Ramsey captured a star. A tiny star, and a short exposure, but a good start. At 10:45, the monitor in front of him showed a clump of ten objects, slightly elongated, like a scattering of rice.

For the next three hours, HET's performance swung to and fro, bedeviled by wind conditions and clouds: First there were good, sharp initial images but the tracking wasn't working; then the tracker would kick in but the images were blurred. More than once, the insistent wind blew open the side door to the control room, loosing a loud bang and a rush of air and adrenalin.

Finally, at 2:45 a.m., Sage found and corrected a minor software error.

Shortly there followed a beautiful, trackable image: a pinpoint, single round star, mid-field, surrounded by a slight glimmer of halo. Actual size, 230 microns—about twice the diameter of a human hair. Ramsey smiled. Sebring jumped to his feet, beaming, and pumped every hand in sight.

First light.

The next morning, while Ramsey hung behind to run a few more tests, Sebring, Booth, and the other members of the project team unceremoniously scattered, by various routes, for Austin. There would be lots more to do: mirrors to be installed, essential instruments and fiber-optics to be added to the tracker . . . Many months of optimizing lay ahead before the telescope would be ready to do any science. But the big hurdle, the one they had all been aiming toward, was behind them at last.

The HET was seeing stars.

Lawrence W. Ramsey, Ph.D., and Daniel W. Weedman. Ph.D., are professors of astronomy and astrophysics, at 517 and 525 Davey Laboratory, University Park, PA 16802; 814-865-0333.

Ramsey is project scientist for the Hobby-Eberly Telescope. Thomas J. Sebring is project manager. The remaining members of the HET project team, all employees of the University of Texas at Austin, are John Booth, Frank Ray, Bill Gressler, Joan Sage, Larry Long, Victor Krabbendam, J. Fred Harvey, George Barczak, Robert Poenisch, and Michael Steiner.

The telescope is located at the University of Texas' McDonald Observatory, near Fort Davis, Texas. HET sites on the World Wide Web include those of Penn State's department of astronomy and astrophysics; Larry Ramsey; and the University of Texas McDonald Observatory.

Last Updated September 01, 1997