Hubble set for critical switch to backup electronics
BY WILLIAM HARWOOD
STORY WRITTEN FOR CBS NEWS "SPACE PLACE" & USED WITH PERMISSION
Posted: October 14, 2008
After two weeks of engineering reviews and discussions, NASA managers have decided to press ahead with plans to put the Hubble Space Telescope into electronic hibernation Wednesday to carry out what amounts to long-distance neurosurgery. To work around the failure of a science data formatter that shut down science operations late last month, the engineering team will reconfigure Hubble to work with redundant data management subsystem components that have not been powered up since launch some 18 years ago.
"It is obviously a possibility that things will not come up," said Art Whipple, manager of the Hubble Space Telescope Systems Management office at the Goddard Spaceflight Center in Greenbelt, Md.
That's the bad news. The good news is "there is very little aging that goes on with an unpowered component in space."
"It's actually a very benign storage environment," he said. "We have very good confidence this will work. In addition, we have contingency plans built in at each step of the transition where if something does not go the way we expect it to, we'll be able to back out and go down an alternate path."
NASA was in the process of preparing the shuttle Atlantis for launch - the target date was today - when channel A of the telescope's control unit science data formatter, or CU/SDF-A, began acting erratically. The telescope's flight computer, following pre-programmed instructions, then acted to "safe" the payload computer and science instruments. An attempt by ground controllers to reset the formatter was not successful and engineers quickly determined the box had suffered a "hard" failure.
While a backup system was available, NASA managers decided to postpone Atlantis' launch on mission STS-125, also known as Hubble Servicing Mission 4, until mid February at the earliest to give engineers time to test and certify a spare unit, used for ground testing, that will be added to the shuttle manifest.
Whipple said ongoing paperwork reviews have not turned up any show stoppers. Vibration, thermal-vacuum and electromagnetic testing will begin next week and engineers should have a good idea by early November whether February is a real possibility for launch. In the meantime, program managers decided to go ahead and attempt a switchover to side B of the telescope's data management subsystem to restore normal science operations.
"The Hubble team has developed and tested the process for switching the Hubble Space Telescope observatory over to its side B and received the necessary approvals for proceeding," Jon Morse, director of NASA's astrophysics division, told reporters today. "The process will commence tomorrow morning and it's expected to take a couple of days to bring the observatory back to science operations.
"Switching to side B accomplishes two main things. One is to recover Hubble's science productivity using its main science instruments, especially the Wide Field and Planetary Camera 2 and the solar blind channel of the Advanced Camera for Surveys. It will also give the team the opportunity to restart the NICMOS cryo cooler, to bring that NICMOS infrared camera back to operational status. Another benefit is to test the side B functionality of the observatory in order to verify HST's redundancy. Restoring that redundancy was the main reason for delaying STS-125."
The transition to side B will involve 40 to 50 engineers at Goddard. Commanding will begin around 6 a.m. Wednesday, with the most critical phase between 8:30 a.m. and 11 a.m. After that, the team will work to bring Hubble out of an induced coma, or "safe mode." If all goes well, Hubble should be back in normal science mode early Friday.
Here's how Whipple described the initial failure and the transition procedure:
"On Sept. 27, just a little over two weeks before the planned October 14th launch of SM-4, the main flight computer on Hubble detected an error signal from the science instrument command and data handling subsystem and correctly responded by putting the SIC&DH and the four science instruments into protective safe mode. The failure was quickly isolated to a hard failure in one of the science data formatters in the SIC&DH. There are two formatters in that unit, one that failed and a second that has been kept as an unpowered backup since Hubble was launched in April 1990.
"There are no indications that this failure affected any other components in the SIC&DH, the science instruments or, in fact, anywhere else on the spacecraft. In its current configuration, Hubble can perform all of its normal health, safety and housekeeping functions but it can only perform astrometry science with the fine guidance sensors, since they do not communicate through the SIC&DH.
"Starting on Wednesday, we will reconfigure Hubble to use the redundant science data formatter in the SIC&DH and six redundant associated components in the spacecraft data management system to restore science operations. Five of the six redundant components in this data management system that will be brought on line have also not been powered since 1990. The command procedures to accomplish this transition have been thoroughly tested. ... So beginning early tomorrow morning, engineers at Goddard will start commanding the reconfiguration and we expect to see the first science data by midnight on Thursday with Hubble back in its science mode on Friday morning."
"Over the last few weeks, the HST operations team has worked hard to be sure the procedures are in place to accomplish an efficient and safe transition. We are confident that all preparations are complete and the team is ready to go."
Because of the way Hubble was designed in the 1970s, it isn't possible to simply power up the B side data formatter and add it to the A side electronics. Instead, ground controllers must power down the telescope and switch over a half-dozen other components as well.
"When we talk about bringing up the B side of the science data formatter, it's actually in a component with the control unit," Whipple said. "You sometimes see people refer to the CU/SDF, that's the box that will be on the B side. That involves the redundant power bus into the SIC&DH and with that comes the use of the redundant computer in that SIC&DH. You can actually run either computer from either bus, but the most straight forward configuration is running it straight through the B side. So that's how the SIC&DH will be configured.
"Because of the way the science payload is wired to the spacecraft bus and the whole data management system on the spacecraft, there are six other boxes that have nothing to do with the science payload - it's a telemetry format control module, a timing interface module, a communications module, a command data interface module, a data interface unit interface and then data interface unit No. 5 - all of those boxes, just because of the way the architecture was designed in the 1970s, are not fully cross-strapped. ... so those six boxes need to come over with the CU/SDF."
Assuming the switchover works, the telescope will again be able to downlink photos from the Wide Field Planetary Camera 2 and the infrared NICMOS camera. The observatory's other two major instruments, the partially operational Advanced Camera for Surveys and the Space Telescope Infrared Spectrograph, are awaiting repairs during the upcoming servicing mission.
During the five-spacewalk mission, the Atlantis astronauts also plan to install two new science instruments, six new batteries, six stabilizing gyroscopes, a fine guidance sensor and replacement insulation. The upgrades are expected to extend Hubble's life at least five years. It's not yet clear where the computer unit swap out needed to restore redundancy in the science data management subsystem will be inserted in the timeline.
Shuttle engineers are currently planning for a possible launch around Feb. 17. But space station operations, the readiness of the new computer unit and a variety of other factors could force NASA to delay Hubble Servicing Mission 4 to early May.
"We think in the first week or two in November we will have a much better handle on the actual state of the hardware," Whipple said. "The paperwork says February should be supportable, but we should have much higher confidence (in November)."