Space Shuttle teleprinter reverse engineering
twitter.comI am fascinated they used 60 pounds and a large amount of space for this teleprinter. At 30K per pound of payload that is $1.8 million per flight. Really shows how important reliable printed updates were considered.
And all the flaws of the printer that were managed around - turning it off to save power and prevent it overheating with specific tones.
The thing that immediately jumped into my head is coded military use. Especially since they mention the printer being based on a military design.
It's still rather obscure, if not still secret, just to what extent we were actually using the shuttles in their intended military capacity.
But when designed, the military use was expected to be much much higher than what panned out.
This printer would be high on the list of the weight budget, to the point that I wonder if it wasn't critical protocol to some still secret military use, similar to the teletype nuke codes / orders on a sub.
Running over the audio system also makes me curious if it was strictly unencrypted comms, or if it could plug in to a decrypted steam. Were they clear broadcasting coded messages, or encrypted-broadcast clear messages (or both or neither).
IIRC, NASA used to have at least some 'private' comms with astronauts that were in the clear, but they basically just didn't rebroadcast to the public or publicize those currently used frequencies, and just sort of trusted those in the know not to listen in.
As far as cryptography, the military AN-UGC/74 teleprinter worked with cryptographic equipment such as TSEC/KG-30, KG-84, KW-7, and KY-57 [1]. You'd send the data stream into the crypto equipment and then to the teleprinter. However, the modifications for the Space Shuttle would have prevented this. Specifically, the FSK demodulation boards for the Shuttle were wired directly to the communication UART board, so there was no place to plug in the crypto box.
[1] https://radionerds.com/images/e/e0/TM_11-5815-602-24.pdf#pag...
STS definitely encrypted all comms- voice, data, telemetry, and printer for the classified portions of all missions (that would be STS-51C, STS-51J, STS-27, STS-28, STS-33, STS-36, and STS-38 along with parts of STS-39 and STS-53).
My assumption is that the decryption happened upstream of the teleprinter- the teleprinter was definitely chosen because they thought it would be useful to print things for all missions, not because it provided special encryption for military purposes. When they did classified stuff everything was encrypted.
Basically, printers are really useful. They are heavy and they can break, but are tremendously useful.
> The thing that immediately jumped into my head is coded military use. Especially since they mention the printer being based on a military design.
a significant portion of the funding for the shuttle came from the USAF, and one of its capabilities (to capture and return a large satellite) was a USAF requirement. They even went as far as building an entire launch pad for it at Vandenberg, which came very close to being used.
Here's my guess. Government bureaucracy is probably very high for anything space related. So "based on the military's AN/UGC-74" could have helped get it through approvals. It would have ticked a lot of boxes for ruggedized parts, use in harsh environments, and so on.
I don't doubt that for a second- but I'm more curious about the requirement for some kind of printout in the first place.
Especially since it ultimately ended up weighing what it did.
Military/ Ruggedized is a smart choice for a printer that needs to go through that environment, no matter the why.
I'm more interested in what it tells us about the underlying purpose (or what it might rule out, like the part about wired directly into unencrypted comms).
I don't think the requirement for printed output requires any appeal to military applications. They would have been exchanging navigation information with mission control over the radio. That's all mission critical and needs to be read back. Anything they got via voice they would have to write down. Getting it by text in the first place would increase reliability and reduce cockpit workload.
You also get in-flight information like weather that you'll want to write down for later reference. Teleprinter writes it for you so you can review it at your leisure.
Airliners and ships at sea have both historically used teleprinters for the same reasons, although the modern navigation computers have mostly eliminated the need.
Oh, I didn't consider that, since I was a young person then. Sticking notes, directives, etc, everywhere was just really common. So if they get a note about extra daily checks on equipment X, they print/affix the note to the equipment. Or other similar needs. There would probably be limited screens, so anything you wanted to be known to everyone, becomes a note.
> I am fascinated they used 60 pounds and a large amount of space for this teleprinter.
IIRC the Zion space habitat in Neuromancer had a state-of-the-art line printer which at one point spews continuous-fold paper into the weightless environment.
They had however got rid of the slide-rules that were essential in early-era Arthur C Clarke spaceships
That was Haniwa, not Zion. The printer was spewing paper because someone had put a laser through its faceplate while Corto was taking the ship.
Also, Haniwa is described explicitly and implicitly as an exoatmospheric luxury yacht, and thus likely wasn't designed for high acceleration. Also, I'm pretty sure it was a thermal printer, not a line printer as here. Also, Neuromancer is period-piece literary sf. So I'm not necessarily sure how relevant it is here.
In terms of relevance, I was amused that real and contemporary fictional spacecraft both had paper printers
The question is, would the craft now need help from reaction wheels to offset the inertia from the moving print head / paper feed? Or would there need to be a special version of the printer firmware that kept track of its movement so it could always impart equal but opposite inertia?
Since a thermal printer only needs as much pressure on the platen roller as required to keep the paper registered, the roller itself could be hollow and thus carry negligible inertia. The print head could be net neutral by operating boustrophedonically, which many printers already do - both of my inkjets, for example! - and that might actually count for more, considering the possible need for some thermal mass to dissipate waste heat from the printing elements. (I don't think modern thermal printers need that, but I suppose in the 80s they might have.) Worst case, the firmware might need to ensure the number of head passes is even, I think most simply by counting print passes and if needed performing a final "return" pass with a cold head - again a common behavior, and one also displayed by both my inkjets since that avoids the need for a parking station on either side of the paper path.
The line printer with its heavy drum would probably be more of a concern for the Shuttle, although I doubt a significant one. Reaction wheels function by change in angular momentum, and a line printer drum spins at a fixed rate in operation; as long as spinup and spindown take about the same time, which is easy enough to achieve, the net effect on attitude should be negligible, especially considering the entire printer constitutes only about 1/3000 of the orbiter's dry mass.
You might not want to run the thing during an OMS burn, but even if you did I doubt it'd matter; assuming constant thrust from spinup through spindown, your course would be displaced proportionally to the length of the print job, but given the minimal relative inertia and that line printers typically output in the tens of hundreds of lines per second I think it'd need quite a long job to make a noticeable difference, and you'd almost certainly run out of paper long before.
(Don't take any of this too seriously; I'm the wrong kind of engineer to give authoritative answers here, but the question is fun to think about anyway.)
"boustrophedonically": now there's a word you don't get to use very often (possibly unless you're Indiana Jones). That's got to be worth something in Scrabble!
Tens to hundreds of lines per minute, not second. Good grief, line printers are scary enough as is, without also feeding paper fast enough to start friction fires...
I was pretty stuck by that, too, and it made me wonder if maybe I don't understand or underestimate the value of paper transmission. I suppose if radio communication is good but flaky, or for persistence of instruction, like procedures, then it would be good for them to be able to print. But at such a cost!
Remember that this was also before easily portable electronics.
Today you'd get a PDF and store it on your iPad. Secure it for landing and you're good. All the documents you want.
In the 80s consider the "Mission Control has new landing procedures that you need to follow. Here is a 42 item check list that you'll need to do before initiating task 357 from the mission specification."
How do you get that 42 item check list? Do you write it down? What that transcribed properly? Was that a P or a B that the person heard over the radio? (Yes, I know Bravo Papa).
Here's a secret objective for the mission ( https://ntrs.nasa.gov/api/citations/20110023479/downloads/20... ) to preform while in orbit that has now been approved and was not part of the initial mission profile. The instructions will be printed out and are for the captain and pilots eyes only.
There are a number of reasons that one may need a secure printer to handle new documents while on the Space Shuttle. With 70s and 80s tech, the approach taken is reasonable.
Maybe they remember the situation with Apollo 13 where they had no way to right down the procedures other than by hand and using space pages from the existing printed documents. You then had the issue where the guys were so fatigued and CO2 levels getting to a point of making vision blurry. I could see where that might have factored into the decision of wanting to avoid all of that with the ability of printing new/modified procedures.
too late to edit, but right != write <hangsHeadInShame>
I guess with modern googles one would just pack an "ipad" equivalent tablet. If one worries about accidents and freak space particles disabling it give them 3. If one still worries one can develop a "space rated" tablet. But probably at those radiation levels one should also start worrying about the crew's health.
But of course that is projecting our current capabilities back in time. I looked it up and the "Osborne 1" portable computer[1] was just released 9 days before the shuttle's first flight. It weighed 24.5 lb (11.1 kg) and could display 52 x 24 characters on a small CRT. So yeah, that would not be nice to read manuals with :D
Right it clearly was the best choice for the time the shuttle was developed but just shows how removed from the 1960s and 70s we are today.
My earlier point about payload weight was probably the wrong focus. Mission success and maximizing what the crew can accomplish in orbit are greatly facilitated by one way text from ground control, and doubtlessly paid for the 60 pounds
Astronauts were already used to Telex weather reports as pilots so existing UX. And the crew specialists all had PhDs and thus were experts at reading typed paper. So no training on yet another shuttle subsystem.
The teletype could be effectively shared between crew - just tear off the paper. A portable computer could only be used by one or two people, AND would need to be radiation hardened and aerospace qualified.
iPads are more hazardous than you think:
https://appleinsider.com/articles/23/07/14/fatal-helicopter-...
Not really an iPad-specific issue. A clipboard or checklist could have done the same thing.
> At 30K per pound of payload that is $1.8 million per flight.
Yeah, but they could save by using white-label ink cartridges.
TIL a space shuttle printer will fetch $8,659 at auction.
https://www.rrauction.com/auctions/lot-detail/34715670664982...
Wait, do normal printers need gravity to work? If I were to find one and flip it upside-down, would it continue functioning normally?
Not exactly a primary source but according to this thread: https://forum.nasaspaceflight.com/index.php?topic=10617.0
"Toner is held to the paper in a laser printer by electrostatic attraction (opposite charges attract), not gravity. Along those same lines, in an inkjet printer ink droplets are fired at the paper, not just dropped, so once again I doubt gravity is an issue."
But before you even got to the point of printing, the paper in a regular laser or inkjet printer wouldn't even feed properly if upside down!
I'm also willing to bet that whether or not the droplets are "fired" at the paper or not, the inkjet cartridges aren't going to work in an inverted position: they might not even supply ink in that configuration.
An Apple ImageWriter II would work, no matter what. I bet it'd work if they were printing DURING a launch.
The only competition it would have would be a Microline ML184.
Normal printers probably wouldn't handle the shaking of launches and the extra G forces during that which for shuttle were ~3Gs.
And when shuttle was developed, printers barely existed. Both inkjet and laser desktop printers were introduced commercially 1-3 years before the shuttles first flight in 1981, and weren’t very reliable yet. Desktop printers still aren’t as reliable as a teletype or dot matrix printer. There’s a reason airlines use dot matrix for printing flight manifests at the gate.
Ink plotters, teleprinters, and fax machines ruled the world. But plotters are dreadfully slow at writing text. Radio fax machines may have been viable if they were rugged enough. But they probably weighed as much as the teletype and were much slower - only real advantage is printing diagrams and photos.
Inkjets need a gas cartridge.
Imagine a printer in zero gravity just spraying printed pages around the cabin of the shuttle, since pages don't neatly fall into the output tray.
This is cool but it just makes me think the Shuttle was absurdly inefficient across the board. Why waste not just the launch weight but the engineering work on making such a specialized printer? Aren’t we talking about the late 70s/ early 80s? Surely there were commercially available printers that could have been substituted that would weigh less and wouldn’t have the overheating issues. Having unlimited government money is ultimately a curse for efficiency and performance.
It's overkill for normal office use but the space shuttle is going to have some massive vibration and mechanical shock during launch. All systems have to be ruggedized to withstand the environment. A commercial unit might not survive the trip up and building some sort of vibration damping system for it might take up more space. If the manufacturer can't guarantee the uptime for the commercial unit, they might need a spare aboard as well also in its own special dampener case. Sometimes it just makes sense to do a custom design that meets your needs instead of trying to make a square peg fit a round hole. Also the related systems have to be considered. The Shuttle had a power bus of 28VDC, were there any commercial printers that took 28VDC and had proper EMI filtering? Were commercial printers compatible with the types of data signals expected on the Shuttle? You may need to build some sort of adapter to make a commercial printer work with the data and power systems. You may end up with something just as heavy and bulky and still not as reliable.
> It's overkill for normal office use but the space shuttle is going to have some massive vibration and mechanical shock during launch.
Maybe you haven't seen how some delivery drivers treat the packages in their care. I swear I've received boxes that look like the were shaken, not stirred, to the level of a rocket launch.
Printers aren’t usually shipped in operable condition.
You're suggesting that a printer being launched into space on a rocket would also not be secured in a similar fashion? What's your point otherwise?
I have doubts a commercial printer back then would handle the 3G acceleration and vibrations during launch and reentry well. There's also the issue of materials from the safety of them for fire risks to potential off gassing you wouldn't notice in an office but could cause issues in the enclosed recycled atmosphere of the shuttle.
A lot of the cost for space bound items are from R&D being concentrated in a few items but others are because it costs a lot to make sure they're not going fail or kill someone. Using this behemoth saves some of the time by reusing a military line printer that was likely already tested for shock resistance if it was used on ships and for fire safety for similar reasons.
There's also the potential military uses of the Shuttle which informed a lot of it's design. Military printer already has the decoding infrastructure if it needs to be built in with known key management etc you'd have to build seprately for a commercial printer.
The explanation for why they chose this printer is in a conference paper published in the National Telesystem Conference, 1982. Unfortunately, I can't find these conference proceedings anywhere (even in physical form). If anyone happens to have a copy lying around...
[1] https://ui.adsabs.harvard.edu/abs/1982ntc..confR...4S/abstra...
I couldn't resist searching myself. It looks like there is a physical copy of the conference proceedings in a library in Japan!
https://topics.libra.titech.ac.jp/recordID/catalog.bib/BA902...
Let's hope the internet can connect us with someone who lives near by and has a ibrary card!
Its unlikely that it was special-built for the Space Shuttle. Obviously this is speculation, but it most likely was already special-designed for a military application and just re-used in the shuttle. All of those strategic nuclear bombers that used to fly on standby 24/7 needed communications, too.
You don't need to speculate. As I explained in the thread, the Shuttle teleprinter was based on the military AN/UGC-74 teleprinter but had many modifications including new three circuit boards for the FSK decoding.
Ah yes, its right there in the 2nd post! Thanks for the correction.
If you trust that office printer to survive a roll down a steep mountain sure.
Your idea of how electromechanical devices "just work" is based on office environments (and even there they fail with astounding regularity today, half a century later).
As an electrical engineer that has repaired many (also very old) devices I don't think you have a realistic idea of why the thing looks as it does. One point is environmental factors (strong vibration, harsh radiation, potential temperature differences etc) another one is risk managment. If your commercial printer fails, how will it fail? You better know exactly what it's failure modes are, because you are the guy who selected that printer and you are reaponsible for both the failure of the mission and the potential death of astronauts who trust you. Still feel secure about the choice? Then you are probably the wrong person for the job.q
TL;DR: look for a certain (recent) submarine failure to see how well your approach works in practise
There's a much simpler version of a communications 'printer' called Hellschreiber
https://en.wikipedia.org/wiki/Hellschreiber
Some old specimens still work, some new reproductions have been made, and it can also be simulated in software with various popular ham-radio programs.
...also, it's used by at least a few dozen hams every now and then :)
X (twitter) is useless for threads unless you're logged in. How can I read this thread?
This was also posted on mastodon https://oldbytes.space/@kenshirriff/112124283096167861
(I assume it's the same post, couldn't read the one on twitter)
Can confirm it's the same
Three ways you can read the thread: a) Log into Twitter. b) Read the same thread on Mastodon [1]. c) Wait until I turn it into a blog post.
Can I ask as a genuine question: why use twitter at all for a post like this? I’m not trying to be a Luddite or even just anti-twitter. I just don’t understand using it for long form deep topics like this. And yes I fully understand many people do use it this way, and to them I’d ask the same question.
I answered that question elsewhere in this thread: https://news.ycombinator.com/item?id=39768125
Off the top of my head: Clicks/ Hype/Sharing/Reach.
Yeah somebody should mirror it. In general I dont understand why people post things like this on X. Its a horrible platform.
Serious answer: I post things on Twitter because it is easy, lets a huge number of people read my thread, and provides easy interaction. I post the same threads to Mastodon and get orders of magnitude less interest. I also write blog posts, which take a huge amount of time (a week versus an hour) and are hit-or-miss. Sometimes they are very popular and sometimes they disappear without a trace.
I’m always delighted when I visit HN and see one of your blog posts on the front page; they posts are great and they elicit some of the most interesting HN discussions. It took me a little while to realize that all these good posts were on the same blog, but once I did I had a lovely time browsing through the archive and reading more of the “deep cuts.”
The posts about the System/360 consoles come to mind, for example. I’m not sure how “popular” those were(?), but they’ve been very helpful as references for an art project I’m working on!
Why not Threads?
Might it be that it takes a week because you spend time thinking things through in a much more coherent manner rather than just submitting stream of conscious level tweets? Sounding professional instead of trendy?
What are some better alternatives?
Current best answer seem to be Mastodon and Bluesky.
If we are going by the number of users, Facebook has more users than Twitter and Threads has about the same order of magnitude, but we still see a lot more Twitter submissions to HN compared to Facebook and Threads[1].
If users prefer the microblog format, Mastodon and Bluesky appear to be the most compelling alternatives, both have been steadily gaining users[2].
[1] Submissions from various sites:
https://news.ycombinator.com/from?site=twitter.com
https://news.ycombinator.com/from?site=threads.net
https://news.ycombinator.com/from?site=facebook.com
https://news.ycombinator.com/from?site=bsky.app
Not sure how we can find the number of Mastodon links since it has so many domains.
[2] Mastodon and Bluesky user counts:
https://mastodon.social/@mastodonusercount/11213120632905851...
For some; for technically dumb people like me, it's the only platform that's easy to use.
From the guidelines[0]:
Please don't complain about tangential annoyances—e.g. article or website formats, name collisions, or back-button breakage. They're too common to be interesting.