9 ms·
I was just reading the pale blue dot and he mentions that Voyager 2’s scan platform jammed after its encounter with Saturn. Voyager 2 emerged from behind Satur
by superasn 6y ago
I was just reading the pale blue dot and he mentions that Voyager 2’s scan platform jammed after its encounter with Saturn.
Voyager 2 emerged from behind Saturn (as seen from the Earth) in August 1981. The scan platform had been moving feverishly—pointing here and there among the rings, moons, and the planet itself during the all-too-brief moments of close approach. Suddenly, the platform jammed. A stuck scan platform is a maddening predicament: knowing that the spacecraft is flying past wonders that have never been witnessed, that we will not see again for years or decades, and the incurious spacecraft staring fixedly off into space, ignoring everything.
The scan platform is driven by actuators containing gear trains. So first the JPL engineers ran an identical copy of a flight actuator in a simulated mission. This actuator failed after 348 turns; the actuator on the spacecraft had failed after 352 turns. The problem turned out to be a lubrication failure. Good to know, but what to do about it? Plainly, it would be impossible to overtake Voyager with an oilcan.
The engineers wondered whether they could restart the failed actuator by alternate heating and cooling; maybe the resulting thermal stresses would induce the components of the actuator to expand and contract at different rates and unjam the system. They tested this notion with specially manufactured actuators in the laboratory, and then jubilantly found that in this way they could start the scan platform up again in space. Project personnel also devised ways to diagnose any additional trend toward actuator failure early enough to work around the problem. Thereafter, Voyager 2’s scan platform worked perfectly. All the pictures taken in the Uranus and Neptune systems owe their existence to this work. The engineers had saved the day again.
Our hacks are nothing compared to the hacks of these old school nasa programmers :)
- onion2k 6y agoOur hacks are nothing compared to the hacks of these old school nasa programmers :) The amazing thing (to me) is that they developed the Voyager spacecraft in an era when space flight wasn't hugely popular and the budgets at NASA were being cut regularly. NASA went from a peak of $6bn a year in the mid-sixties under Johnson to almost half that under Ford and Nixon. The engineers were well-respected and they clearly felt the work was important, but they weren't particularly well paid. They did this for the love and the science.
- brazzy 6y agoIt's great, but hardly amazing. Most academic research is done by underpaid PhD and postdoc students for the love of science, and most of that is far less glamorous than SPACE.
- abhayhegde 6y agoI believe the aspect of beauty and glamor is a matter of taste and hence very subjective. Maybe some projects may turn out to be less influential, but nevertheless it appeals to a particular community.
- part1of2 6y ago> Most academic research is done by underpaid PhD and postdoc students for the love of science I’d hardly call the graduate papers coming out of academia equivalent to the level of innovation & creativity of what the engineers above described. Perhaps they’re both underpaid, but one is certainly more amazing. Hardly the same.
- matthewdgreen 6y agoYou should read Rhodes’ “The Making of the Atomic Bomb” if you want to understand what incredible engineering feats academic scientists are capable of — when the people are tightly focused on achieving a single goal. You should also read Rhodes’ book if you want to understand why unguided fundamental scientific research itself is so critical, even when it doesn’t immediately seem useful to you: without the discoveries of science, nuclear fission wouldn’t have been understood. (Of course, many will disagree that the development atomic bomb is a good outcome, but that’s a very different conversation.)
- BenjiWiebe 6y agoI love that book. It takes a while to read but it's worth it.
- brazzy 6y ago
- wst_ 6y agoI found out that usually happens when you have some constraints. The more constraints the more clever ways people find to achieve their goal.
- yetihehe 6y agoUp until so much constraints than even the most clever give up.
- Cthulhu_ 6y agoIt makes me think of the demoscene, that is doing things like getting 3D graphics on the Commodore or whatever it was.
- prox 6y agoFuture Crew - Second Reality
- mkl 6y agoThat was a 1993 PC demo, not C64. I had to mess with my config.sys and autoexec.bat to get it to run, but watched it repeatedly.
- jacobush 6y agoYeah but that demo was later re-created on the C64, arguably the C64 version is better. :-D
- Someone 6y agoFor those who don’t know about these demos: - http://www.pouet.net/prod.php?which=63 http://www.pouet.net/prod.php?which=63 (PC; page has link to source code) - http://www.pouet.net/prod.php?which=1216 http://www.pouet.net/prod.php?which=1216 (C64)
- nitrogen 6y ago
- m463 6y agoIt has become commonplace to solve hardware problems with software. But many times it is not for heroic reasons, it's because hardware is cost-reduced.
- ReptileMan 6y agoUnfortunately when preparing something for space missions you have constraints that are not as easily solvable as throwing more money.
- radiorental 6y agoI'm not sure I'd agree with that at all. On the surface of this planet software is used to simulate hardware as it's much quicker to iterate on a design. In space, however, you can only run the hardware in it's destined environment once. Software is critical to the design process regardless of the HW costs.
- patmorgan23 6y agoIntel has patched hardware flaws using microcode updates before.
- ponker 6y agoThey didn't even spin something 350 times before launching it out of the solar system?
- dirktheman 6y agoIf the actuator broke down the 351st time, would you've commented 'they didn't even spin something 400 times...'? What would be a number you'd be comfortable with instead?
- amelius 6y ago640k
- elcomet 6y agoMaybe the expected number of times it would be expected to spin in space (or twice that to make sure it's working)
- amelius 6y agoThe initial Voyager program already ended, and they are now in an "extended" mission.
- curryst 6y agoI would have thought they would run every part until it died so they know roughly what it's lifespan is. It's good if the parts lifespan is twice as long as you need. However, it's even better if you could make it four times the amount you think you need, as long as it isn't horribly expensive. I would tend to assume they've thought of that though, and stuck with this part for reasons. Maybe there werent other space-ready alternatives or something.
- jerf 6y agoI don't think any of us are in a position to be too critical of the engineering done for the Voyager probes. The number of machines running continuously for longer than I've been alive with no maintenance is pretty short; the number of them out past the orbit of Pluto even smaller.
- myth_drannon 6y agoIt looks like they spent a lot of resources to fix it. In our current business environment, the engineers will be moved to another more visible/successful project and this one will be written off.
- TheCondor 6y agoA former coworker/friend was telling me similar stories of his uncle working on missile guidance analog computers from the 1970s. He referred to it as "proper engineering"
- ProAm 6y agoIt's not 'hacking' it's engineering.
- DEADBEEFC0FFEE 6y agoThat the difference between actual engineers and software engineers.