NASA held a press conference about Hubble Space Telescope’s malfunction. They put out a press release about it. Here are the high points:
1) Hubble is currently in "safe mode", basically sleeping. The Science Instrument Command and Data Handling System has failed (arrowed in the picture at left). It worked for 18+ years, which is pretty good. The specific part that failed is the Science Data Formatter, which takes the observational data, formats it into packets, adds headers, then sends it down to Earth at a rate of 1Mb per second. Without it, Hubble can take data but cannot send it to us! It’s not known exactly what piece of the SDF failed, but the whole schmeer is basically on the fritz.
There is hope, though. There’s a spare SICDHS on board Hubble; NASA built in redundancy. Called Side B (that’s a generic term; the redundant systems on any piece of hardware — even the cameras themselves — are called Side B), it has not been turned on since before Hubble was launched in 1990. NASA will try to switch it on, but it’s not like just hitting a button; a lot of work has to be done first. They’ll have to put the ‘scope in hardware safe mode, basically shutting it down to prevent any problems. They’ll also have to turn on the last remaining gyro to make sure it will stay pointed correctly. Not just the SICDHS but also the entire Data Management System has to be switched to Side B as well, which is a "major event". Then, we have to hope that the Side B electronics, which have been sitting in space for nearly two decades, will turn on without a hitch.
All in all, this is quite an undertaking, and NASA understandably wants to take its time figuring this out.
2) NASA is also working up a plan to bring up a spare SICHDS and put it on board Hubble, which is very cool. I was concerned about what would happen if Side B fails; that’s a single point of failure now that Side A is down. Having a spare launched is a Good Idea. They’ll have to get the astronauts trained to install that, so the servicing mission has to be delayed. That’s a problem given that the spacewalk schedule is packed. But the changeout should only take a couple of hours; if they can do one of the camera repairs quickly enough, they can pack the new SICHDS into the existing schedule.
They have to test the spare SICDHS first, and that takes time as well; it may be early January before it can be delivered to Kennedy for installation on the Shuttle.
3) Endeavour will launch (most likely on time) to go to the ISS in November. They’ll have to figure out when they can launch the delayed servicing mission, but it’ll be a while before they will know. It’ll be months from now for sure. Just so’s you know, it costs about $10-11 million/month to keep the Shuttle on the ground, so they hope to launch sooner rather than later. One number thrown out for the delay was three to four months, but they just don’t know for sure.
Making things worse, Atlantis is already on the pad, so they’ll have to roll it back to the Vehicle Assembly Building if they delay the launch for months.
So the upshot is that this is pretty bad, but possibly not a disaster. If Side B comes up, then Hubble will work and we’re fine until the next servicing mission, when they can install the new cameras, gyros, and other hardware — including the spare SICDHS. If Side B does not come up, well then, yeah, things are bad. Hubble will have to stay down until the new servicing mission can be launched, which means three months at least without a working observatory.
However, once the servicing mission goes up, one way or another NASA currently thinks they’ll be able to get Hubble back up, shined up, and running smoothly.
It will take some time, days at least, before we get more news. But I’ll add that things could have been much, much worse; had this happened after the servicing mission — and it only happened two weeks before the scheduled launch! — and Side B didn’t go up, that would be it. Hubble would be dead. So it’s actually rather good it happened now. It’s almost funny, but that’s the bright side to this.
September 29th, 2008 at 4:40 pm
Phew! Hubble will be ok! Now it better live for a few more years so it can take amazing pictures!
September 29th, 2008 at 4:51 pm
[...] BA Blog: Hubble Update __________________ 0 1 1 0 1 0 0 1 1 0 0 1 0 1 1 0 1 0 0 1 0 1 1 0 0 1 1 0 1 0 0 1 1 0 0 1 0 1 1 0 0 1 1 0 1 0 0 1 0 1 1 0 1 0 0 1 1 0 0 1 0 1 1 0…. [...]
September 29th, 2008 at 4:53 pm
Thanks for the in-depth Phil… I’ve been following this all day but haven’t had time to do more than updates on my site, so I’ll point everybody to you…
cc
September 29th, 2008 at 5:14 pm
Maybe the taikonauts can help with the repairs!
September 29th, 2008 at 5:21 pm
Phil,
Thanks for the updates. As I’m sure is being asked at Goddard right now, are there any other mission critical systems whose *Side B* has been sitting for a long time? Were the SIC&DH units ever even considered to be replaced in any mission so far?
18 years is a long time, but the really smart people at Goddard build great hardware. Besides, no Bill Gates involvement here (inside joke for Dragon*Con attendees), just good ol’ assembly and some C!
September 29th, 2008 at 5:31 pm
Phil, what are the costs involved in keeping the shuttles grounded for it to cost that huge amount of money?
September 29th, 2008 at 5:36 pm
Phil, I thought the same thing – that if there had to be a major malfunction, it’s a blessing it happened now rather than after the final service call!
September 29th, 2008 at 5:59 pm
Typical. An 18 year old space telescope has faster downloads than our brand-spanking-new ADSL2+ broadband.
I blame Thomas Moore.
September 29th, 2008 at 6:06 pm
Some of the best “songs” have been recorded on “Side B”. Thanks for the insights.
September 29th, 2008 at 7:56 pm
Yer right, Phil! It’s very fortunate that it failed now, rather than next month.
Looks like it’ll be an easy enough repair, very accessible.
Bummer that the shuttle has to go back to the VAB, though, but that’s the breaks. We did get to see the cool pix of 2 shuttles on the pad at the same time.
Rich
September 29th, 2008 at 7:57 pm
Thanks, Phil. NASA deserved a big piece of good luck, and the timing of this problem certainly qualifies.
Are Hubble’s procedure manuals on paper, microfiche, etc, or are they on computer? 18 years plus the pre-launch period is a long time. Fiche seems a likely answer.
September 29th, 2008 at 8:30 pm
While I agree that it’s great that this just happened before the mission, I’m also kind of bummed!
Let me explain. I’ve lived in Florida for about the last 6 years. I’ve never gone to a launch, even though I’ve been very interested in the space program. Last December I moved to California, and started reading about how STS is going out in 2010. Unfortunately, things didn’t work out in CA, and I had to move back here. As soon as I found out I was coming back, I decided to go see a launch. I chose STS-125 as it was the next one. I’ve been organizing for time off of work, my brother, and a few friends from out-of-state so we could go see it. I’ve already had to change the whole plan twice, and now this!
I suppose we’ll go watch STS-126, but the wait is killing me!
Thanks for the in-depth, Phil. Always a pleasure
Daily BABlogee
-Fauxnetikz
September 29th, 2008 at 9:00 pm
For anyone interested there are 5 high-resolution spherical panoramas of the Hubble flight hardware in the clean room at KSC here: http://i-ota.net. More pics and pans of the STS-125 mission here: http://nasatech.net
September 29th, 2008 at 10:02 pm
You have to admire any piece of technology that survives 18 years of hard radiation.
September 29th, 2008 at 10:46 pm
What, you mean that Hubble is not constant?!!
September 29th, 2008 at 11:04 pm
As you say, at least this happened BEFORE the servicing mission. Everything still crossed. Hang on Hubble, we’re on our way.
September 30th, 2008 at 1:49 am
As the PI on 2 cycle 17 proposals, I was rather bummed when I got off the plane this afternoon to the news. Thanks for pointing out the silver lining, especially the point about it failing before SM4 as opposed to after.
September 30th, 2008 at 2:15 am
Is the hardware to test and validate, and people with knowledge of the SICHDS systems, still available? I can imagine that they aged a bit too… so yet another big task ahead of the “youngsters” at NASA.
Hope all works out fine and we can still enjoy the great sience of the HST
September 30th, 2008 at 3:24 am
Say Phil – What’s the status of the NICMOS equipment. I heard a few weeks ago that it went down too due to suspected cooling line problems. Has NASA updated taht information?
(Thanks for monitoring this)
September 30th, 2008 at 5:48 am
If a back-up system is an exact copy of the primary, why don’t they periodically switch between systems? I suppose the switch-over itself could be a point of failure, but it still seems to make sense to exercise as much of the system as possible so your backup doesn’t silently die undetected.
September 30th, 2008 at 6:23 am
John, I hear ya. I was still in grad school and working on an instrument team when HST went up and we found out about spherical aberration… the mood on our team was somber for weeks…
September 30th, 2008 at 6:53 am
Does NASA keep spare Hubble “Science Instrument Command and Data Handling System” units laying around? If they weren’t planning on replacing it, they I doubt they have one tested and ready to put in the cargo hold of the shuttle. Did they go digging through the closet and find one in a dusty cardboard box?
BTW: I hope this isn’t the beginning of a larger string of failures for the aging Hubble. If another important system fails while they are preparing the SICDHS, they may decide to simply give up on the thing and abandon the repair.
September 30th, 2008 at 9:26 am
[...] De Science Data Formatter De reparatiemissie van de Atlantis Space Shuttle, gepland voor vrijdag 10 oktober 2008, is uitgesteld tot minstens februari 2009. Oorzaak is dat de Hubble in safe mode is geraakt, een ander woord voor slaaptoestand. Dat is op haar beurt weer veroorzaakt door een slecht functionerend onderdeel van de Hubble, de Control Unit/Science Data Formatter – Side A. Op de foto zie je dat onderdeel en het heeft na 18 jaar van trouwe dienst de geest gegeven. Het draait met name om die Science Data Formatter (SDF), die de waarneemgegevens van de Hubble keurig in pakketjes stopt, er een header aan geeft en ze dan met een snelheid van 1 Mb per seconde naar de Aarde stuurt. Geen SDF, dan ook geen data naar de Aarde, simpel. Nou is er één hoop en dat is Side B. Yep, er is van dat hele instrument ook een Side B, een reserveonderdeel. Vóór de lancering in 1990 in de Hubble geplaatst en daarna nooit meer aangeraakt. Maar activeren van Side B is helaas niet een kwestie van een knop genaamd Side B indrukken, want daar komt veel bij kijken. Vijf andere modules die gegevens van Hubble verwerken moeten naar hùn Side B worden overgezet. Dat schijnt allemaal vanaf Aarde te kunnen worden geregeld, maar het is tijdrovend. Als de Side A > Side B overgang goed gaat zou Hubble weer z’n gegevens naar aarde kunnen sturen. Maar dan heb je de situatie dat één reserveonderdeel essentiëel wordt en da’s ook niet wenselijk. NASA wil daarom sowieso een geheel nieuwe SDF plus control unit naar de Hubble brengen. Dat kost al gauw drie maanden voorbereidingstijd en zodoende denkt men dat de Atlantis pas ergens in februari volgend jaar z’n missie kan maken. Bron: Bad Astronomy. [...]
September 30th, 2008 at 10:03 am
You could always go with the classical definition of a planet. That is if you can see it in the sky and it moves against the “fixed” stars, it’s a planet. Else, come up with some sort of scientific name for it. This of course would make the Sun and the Moon planets and Neptune and possibly Uranus would be downgraded, but at least it’s pretty definite.
September 30th, 2008 at 10:05 am
C’mon, Phil, quit showing off with the technical jargon — you know we’re mostly just laypeople here.
~David D.G.
September 30th, 2008 at 1:42 pm
[...] Dr. Phil Plait has more details on his Bad Astronomy blog. [...]
September 30th, 2008 at 2:28 pm
Underdog, oops! I mean Shoeshine Boy asks:
I mentioned in the last post about Hubble, but for the benefit of those who didn’t see it…
Yes, there is a spare SI C&DH unit on the ground. Actually, there are two, but one of them only has one side working(!) The good one was built after launch, and has been used for ground testing of new instruments as they have been built. It was built to fly, so it has rad-hard electronics and all the appropriate paperwork, but it hasn’t undergone environmental testing. It may have some additional liens, like some parts may need to be conformally coated, or connectors may need to be replaced with EVA-friendly ones (ever tried to install a D-submini connector while wearing astronaut gloves?) Or it may be all ready to be tested and flown. But that’s at least a two month process, even if it passes all the tests. Any modifications will add to that time.
October 1st, 2008 at 12:49 am
Someone asked it is time to give up on Hubble and move on. The JWST (often called Hubble’s replacement, but it isn’t really) won’t be ready for launch until 2013 at the earliest. Building and launching a true replacement for Hubble would cost at least $1.5B, or at least 3 times the cost of the servicing mission, and would take many years. (My guess would be it would probably be quicker than the JWST if it was built to Hubble’s specs and used as much spare equipment, the replacement instruments planned for SM4, etc., but would still take years to build.)
JWST will orbit at the Earth-Sun L2 point, about 1.5M km outside the Earth’s orbit. (By contrast, the Earth is about 100 times as far or about 150M km from the Sun.) It is much too far away for the Shuttle to reach it and unlike Hubble, isn’t designed for in-orbit servicing or repairs. The Orion could get there by using an Earth Departure Stage launched by an Ares 5. like a Moon mission, but I don’t think there are any plans for that, even as a contingency. No grapple fixtures, for example. I think it would make a good rehearsal mission before a moon landing (testing out everything in deep space, much quicker and easier than an asteroid rendezvous, and you don’t get stuck if the LSAM fails.)
Someone in the other thread wondered if it would make more sense to put the Hubble on (or near?) the ISS to make servicing easier. I think it would make sense to put observatories and other large, serviceable satellites in orbits in the same plane but somewhat higher than the ISS. Raising or lowering the orbit of a satellite requires a relatively small amount of fuel. (By contrast, orbital plane changes (such as changing from the Hubble’s 28 degree orbit to the ISS’s 50 degree orbit, or changing between two misaligned 50 degree orbits requires *HUGE* quantities of fuel. ) They could park the observatory in a co-planar orbit a few hundred miles about the ISS, and when they needed to service it, they could send a small space tug-type vehicle (a Soyuz, Progess, Orion, Dragon or ATV could all do the job it it was equipped with a small robot arm with a grappling adapter) that
would rendezvous with it, grab it and pull it back to the ISS. Then ISS-based astronauts could then repair it, replace instruments, etc. using parts sent up on a Progress, ATV or other ISS supply flight. They could take as long as necessary (wouldn’t be restricted to a 14-day shuttle mission) and could check it out before using the tug to send it back to its parking orbit. (By placing it in a higher orbit, it would remove any danger of it colliding with the ISS or lost cameras, tools, etc. or arriving or departing supply or crew flights, it would be exposed to less contamination from rocket exhaust, waste water, etc. than in the immediate vicinity of the ISS, and there would be less air friction to cause its orbit to decay at a higher altitude. (The shuttle servicing missions always boost the Hubble to the highest possible altitude before they leave for this reason.)
(They wouldn’t want to just leave the observatory permanently attached to the ISS because then it would be subject to all the bumps of dockings, reboosts, people moving around inside, pumps moving water and fuel and thus changing the center of mass, the solar wings rotating, etc. For many payloads (e.g. the AMS) this wouldn’t matter but it would wreak havoc with a high-resolution telescope. (Remember the Skylab astronauts running around the inside of it, creating artificial gravity for themselves? I saw what that did to the solar telescope data. Not pretty! Well, actually, it was kind of interesting, but scientifically worthless.)
However, to move the Hubble to such an orbit would require retrieving it and bringing it back to Earth on one shuttle flight, and then re-launching it into the ISS’s orbit on a 2nd flight. For the same cost, they could do 2 servicing missions. And the shuttle may not be powerful enough to launch it to the ISS orbit… Hubble is currently in a 28 degree orbit which provides the maximum payload for a rocket launched from Cape Canaveral. Launching to a higher (or lower) inclination orbit reduces the maximum payload significantly. The ISS is in a 50 degree orbit to maximize the payload from Baikonur. I don’t think Soyuz’s could reach it if it were in a lower inclination orbit, though that would increase the maximum payload for American and European launches to it. That’s one of the things that many people think is wrong with the ISS.
Maybe the next space observatory will take advantage.
October 2nd, 2008 at 10:46 am
[...] Hubble is having its woes right now, and even if it gets fixed its own clock is ticking. It was launched in April 1990 — I still remember it well, as I had just signed up to use it for my PhD project — and nothing lasts forever. But Hubble has taken thousands upon thousands of observations, all of which have been stored away. And while Hubble may eventually be shut off, the images, spectra, and other data it’s taken will live on. Astronomers a century from now will be digging through the archives, looking for an elusive supernova, a feature of a nebula that’s changed in the intervening years, the positions of stars in globular clusters, and the colors of galaxies so far away that when the light Hubble detected left them, the Earth was still a cooling ball of molten rock. [...]
October 14th, 2008 at 5:40 pm
[...] couple of weeks ago, a component on Hubble failed (read that link for the back story here). This critical piece of hardware controls almost all the [...]
October 14th, 2008 at 8:51 pm
[...] to send pictures back down to Earth for a few weeks now, as it’s come down with a bad case of broken electronics. In the meantime, they’ve still been able to do useful science by obtaining ever-more [...]
October 15th, 2008 at 6:12 pm
[...] El componente estropeado, denominado Control Unit/Science Data Formatter, destacado en rojo en esta imagen – Foto NASA vía Bad Astronomy Blog [...]
October 15th, 2008 at 6:33 pm
[...] El componente estropeado, denominado Control Unit/Science Data Formatter, destacado en rojo en esta imagen – Foto NASA vía Bad Astronomy Blog [...]
April 14th, 2009 at 7:00 am
[...] was done before, last September, again for the Hubble servicing mission. But that launch was scrubbed due to a failure of a component on Hubble. The launch was delayed until May of this year so that a spare part could be moved into the Shuttle [...]
January 26th, 2012 at 3:28 am
Touche. Sound arguments. Keep up the amazing effort.