Slashdot Mirror
Antique Voyager Technology
sea_stuart writes with a story from the Tidbinbilla space tracking station, outside Canberra, Australia. It is still communicating with the two Voyager spacecraft 30 years after they were launched and 18 years after Voyager 2 passed close by Neptune. Here's a little background on Voyager 1 and Voyager 2. "The bank of computers that would look at home in black-and-white episodes of Doctor Who cannot be junked... [T]he 1970s hardware is now our world's only means of chatting with two robot pioneers exploring the solar system's outer limits. Today Voyager 1 is humanity's most remote object, 15.5 billion kilometers from the sun. Voyager 2 is 12.5 billion kilometers from it. Both continue beaming home reports, but now they are space-age antiques. 'The Voyager technology is so outmoded,' said Tidbinbilla's spokesman, Glen Nagle, 'we have had to maintain heritage equipment to talk to them.'"
Even after being flung across the solar system, I am sure Captain Janeway will find a way to repolarize the deflector dish to emit a warp bubble that combined with future Borg technology and that from Species 4971, some old fashioned ingenuity, a transwarp generator, a friendly if dull-witted Talaxian, a half-human half-Klingon baby, a group of Maquis rebels, a hot-shot pilot who doesn't give a damn for regulations, and a hot Borg in a skin-tight leotard will be able to make it back, and the ship will probably be in better condition then when it left!
I'm sorry... I'm bitter...
(32 bits) x (60 seconds) x (60 minutes) x (24 hours) x (365 days) x (30 years) = (30,274,560,000 bits)
(30,274,560,000 bits) / (8 bits) / (1024 bytes) / (1024 KiB) / (1024 MiB) = (about 3.5 GiB over 30 years)
I don't think a modern computer would help, because it's clear that Comcast is seriously throttling their torrent connection.
Not really. As long as you have people who understand the hardware and a supply of old machines for spare parts you should be able to keep things ticking along for decades.
In my last job we ran the entire Melbourne traffic signal system off PDP 11/84's and 83's. Its a good way to keep your wire wrap skills up to scratch.
http://michaelsmith.id.au
Not that TFA can be trusted (honestly how would something be "too slow" for a computer? Does my processor get impatient?) but it kinda implies that these guy's primary responsibility is this computer. For the price of two senior engineers it really seems like they could cook up a modern replacement.
Seems odd that they don't just salvage the analog components and connect it to a modern computer... I guess I'd understand not touching it if it's deemed fragile...
Anyone know if the Voyagers rely on a heartbeat or something? If it's just a receiver I can't see why building a modern backup isn't worthwhile.
I studied Space Tech for a while so while this is still a guess I like to think it a fairly educated one:
In order for something to be acceptable to NASA for use in the space program it has to be very thoroughly tested. This means you could write a software emulator that did everything required, but then it would have to run flawlessly for 10 years in order to be approved for use. You have to remember that these computers can also send commands to the satellites, so if they crash and send an erroneous command out, then that command will be actioned by the satellite.
I know this is highly unlikely, but it is not impossible so why risk it when the result of that one command could be that we lose both satellites for ever.
There is a mantra when it comes to dealing with any computer system that is running a mission critical app:
"If it ain't broke, don't fix it."
I would suggest that anyone wanting to be sysadmin, learn this. There are times when it doesn't apply but that is usually when the benefit of change out way the risks. In this case what is the benefit of upgrading the system at our end?
I dont read
...when NASA inspired me, and the projects in which it was engaged filled me with wonder and curiosity. Nowadays the only thing that amazes me about NASA is the bureaucracy. Well, and the big explosions of course.
I'm sure you understand why... I think the conversation would go something like this:
IT: "This new system will cost $1bn, and will save $3bn/year in maintenance on the old system".
Management: "The previous system was supposed to cost $1bn to develop, and ended up costing $10bn. If I sign off on this it will be my ass on the line when the budget blows out, so I'll stick with known quantities thanks."
Whenever I come across news about the Voyagers, I generally dig deep and read a lot. I am utterly in awe -- of the spacecraft themselves, that they are still functioning, that they are so mind-bogglingly far away, and that humans have created them with the tools of their time. Wow. The link you posted shows in what incredible detail the mission was thought through.
I am very glad that there are still people who monitor and maintain the Voyagers. They deserve it.
"Good news, everyone!"
What is little known externally except by those that actually worked on this project is that the radios work at all is amazing. Motorola GEG built the radios in the Voyager spacecraft. Right after launch of both space crafts there was a failure of a critical capacitor that sets the bandwidth of the acquisition loop filter. The net result of that failure was that the signal acquisition of the radios was severely impaired. In order to compensate for this NASA engineers developed an emperical model of the entire spacecraft while it was on it's initial loop around the sun for it's slingshot to Jupiter. Since it was relatively close they could hit the spacecraft with a very large signal thus ensuring acquisition of the transmitted commands. The model consisted of predicting exactly where the front end input LO would be depending upon the temperature of the space craft, the added doppler due to movement, aging of the crystals, etc, etc. Basically anything that could effect the LO was factored in. Once the model was complete, the ground stations would then use and probably still use, this model to predict what the frequency for lockup needs to be. Due to the efforts of the engineers at NASA, they were able to "save" both spacecraft and the mission. And they still work today!!! Pretty amazing.
"TV, a medium as it is neither rare nor well done." Ernie Kovacs
A typical mainframe of 30 years ago would have done a lot of batch processing. But it still multi tasked. Only an embedded system would have had deterministic timing. And that is true of today as well.
I funded a hitch hiking holiday in Tasmania in 1986 by doing small withdrawals in the middle of the night when ATM's couldn't connect to the banking systems because overnight jobs were running.
http://michaelsmith.id.au
"Er, Hi, This is Ranesh from Advanced Emulation Solutions..."
Or more likely:
"Hello sir, my name is Tom, calling from your Houston of your Texas. With the client we are noticing a problem. Please to do the needful."
Slashdot quality declines as the number of hot grits posts decreases. - Provolt's Law, Apr-09-2005
a day? cripes! I dont even need to build hardware just a soundcard and a little bit of software, an hour tops.
This kind of stuff is done daily in ham radio. I build a interface to read the old abandoned weather satellites slow scan TV signal with a soundcard and a connector plug. wrote the app in C in 2 hours and had a picture on screen the next pass.
I would have been faster if I though to record the last pass's audio and replay it for debugging, but no It took me 30 minutes to find a different bird passing over that I could receive.
Do not look at laser with remaining good eye.
Nope.
1) It's not a software receiver, it's hardware
2) There IS a newer receiver, the Block V, that uses a whole board full of custom ICs and demodulate all the old legacy stuff. I would imagine this is what's really being used at Canberra.
3) There IS a newer (software based) receiver in the works, but not done yet
4) The processing load is substantially more than your old Dell can handle (we're talking about finding and demodulating a very, very faint signal that is a few Hz wide in a substantially wider bandwidth)
5) The interfaces to the analog RF hardware are somewhat unique.
6) Doing the processing requires a very stable reference clock (I assume your Dell doesn't have a hydrogen maser driving the on-mobo sound card? DSN receivers do.
7) There are a limited number of people in the world (probably 100) who really understand this stuff well enough to do the emulation, an even more limited number who are not retired, and a good fraction of them are busy working on #3, above.
this might be a good comparison.
I worked at bellsouth in '97; 1 of the midrange systems I was responsible for was a freaking ancient magnetic core drum & reel to reel monster that ran the switches that "a critical emergency system" used.
yes, you could replace the hardware components with something a little more modern, but you would have to be absolutely certain that the replacement component was exactly 100% the same as the legacy hardware, down to once-every-3rd-monday bugs; otherwise you could damage the system.
why not replace the whole thing? the company that made the system was out of business since the late 70's; no source for the software it was running was available, no complete logic diagram of the workings of the system. maybe it would have been possible, maybe not, but you are talking about a system that had been running without flaw for about 25 years. why replace it?
to the best of my knowledge, it's still running.
Why, yes, I AM a Pagan Libertarian.
It doesn't. Therefore /. must have made it into the "Mainstream Media" cabal.:-)
I don't know whether I should celebrate or commiserate. I fear the latter.
Anyway, anybody know what comps. etc are being used at the Tidbinbilla space tracking station?br I'm old enough to be genuinely interested.