Build Your Own Apollo Guidance Computer

PingXao writes "Well, if you can't exactly give the Moon you can give the gift of a computer to get you there. Almost a year ago this Slashdot story about the Apollo 11 Guidance Computer referenced a pretty cool Dr. Dobbs Journal article from their History of Computing series. Now there's this guy who built one in his basement! It took him 4 years, $2,980 in cash, 2,500 hours of labor and 15,000 hand-wrapped wire connections with 3,500 feet of wire to build. It might be next Christmas before you could build one of your own to give as a gift, but he promises you can build your own for less and it will be better than his. The perfect gift for the space geek who has everything. This guy is my hero."

Comments

Kinda makes you wonder,

[DoraLives]

with those old boxes, how in hell did they ever make it to the moon and back alive.

Re:Kinda makes you wonder,

[Detritus]

They had IBM 360s and other big iron on the ground to do the heavy-duty calculations. If you have a choice between doing something on the ground and doing it on board the spacecraft, it's almost always better to do it on the ground.

Re:Kinda makes you wonder,

[Average_Joe_Sixpack]

Easy, they weren't bogged down with a GUI.

Re:Kinda makes you wonder,

[UniverseIsADoughnut]

well, you could do it without computers, would just be even harder, plus astronauts would have one sucky time flying the craft by hand to the moon. But it could be done, there is always a way without a computer.

The other thing is they took a very simple approach to thing, to do it today would be even harder because we would over complex thing thing with uber redundancy and sensors for everything and so forth. Thus why we could get to the moon, or russians get space stations, but the space shuttle and space station suck.

Not that getting more computers involved is bad, it just makes it easier for things to crap out and not know why. Mechanical stuff is easy to figure out why it's not working, electrical not so much, and code and semiconductors very hard.

I too look at how we did it, am are amazed it all worked. But then, look at a Model T or a Steam Locomotive, today it seams amazing people would trust those thing cross country or that they would be very durable, but they did it just fine.

I'm pretty sure my powermac has failed on my more then my atari 800 ever did.

Re:Kinda makes you wonder,

[Glonoinha]

Heh.
Most of us wear watches with more horsepower than a single System 360 of the time.
I would imagine that every computation performed at all of NASA from T minus 10 until splashdown could grind through my desktop in less time than it took me to reply to this message.

The difference wasn't in the hardware.
It was in the people, their abilities, and in the working relationship those people had with each other.
It was in the management of those people, putting success and excellence above all else.
It was in the work - putting men on the moon wasn't just a job, it was an adventure and it was a dream.

Re:Kinda makes you wonder,

[node 3]

with those old boxes, how in hell did they ever make it to the moon and back alive.

It takes more computational power to provide a retarded paperclip assistant than it does to go to the Moon.

Re:Kinda makes you wonder,

[bigberk]

The difference wasn't in the hardware. It was in the people... it was an adventure and it was a dream.

I agree with you. Look at the incredible computation abilities we now have, it really boggles the mind. We have made leaps forward in speed, miniaturization, and power usage. Materials science has also brought us an entirely new set of possibilities since then. Now, if we had a real goal -- like to start human exploration of space in earnest (longer missions, more frequent), I think we could really do some amazing things.

Personally I think we're being really stupid by not funding more space exploration. Yes, I know people on earth are starving. But both you and I know that it's not the starving Ethiopians competing with NASA or ESA for funding...

Re:Kinda makes you wonder,

[fmaxwell]

well, you could do it without computers, would just be even harder, plus astronauts would have one sucky time flying the craft by hand to the moon. But it could be done, there is always a way without a computer.

Actually, there is not always a way without a computer. Some modern fighter jets are inherently unstable (in order to provide faster response) and no human being in the world could react quick enough to keep those planes from wadding themselves into little silver balls. Their computers make multiple control surface adjustments per second.

Re:Kinda makes you wonder,

[GuruBuckaroo]

There's a Heinlein quote that I'm trying to think of, that goes something along these lines:

There are three stages to the development of any technical project. In stage one, the device is simple, does only what it needs to, and works most of the time. In stage two, the device is vastly overcomplicated, overpowered, does far more than it needs, and works occasionally. In stage three, the "improvements" are thrown out, the device is again simple, does only what it needs to, and works all the time.

I've been waiting for Stage Three for a long time now. My money is on Burt Rutan.

Re:Kinda makes you wonder,

[UniverseIsADoughnut]

well, they are unstable for the purpose of getting better performance, not to make the concept of planes to fly work. You can build a plane without a computer, and could achieve the same basic concepts, but it wouldn't do it as well.

Also, planes like the Stealth Bomber are said to not be able to fly without a computer for reasons like you mentioned, especially landing it. But really it's a issue of it makes it practical, not needed. The Northrop flying wing worked in the late 40s and it obviously had no computers. The B2 is based off that planes design (actually has the exact same wing span), the computers just made it more feasible and overall better.

In the case of going to the moon, it could be done without a computer, rockets went up without computers, plus people make a great computer. The computers for apollo did pretty straightforward stuff, and were mainly there so the astronauts didn't have to keep doing stuff non-stop. They could still sight stars and calculate there path and manual fire rockets to adjust (like they did in Apollo 13),

The thing is we have all gotten so used to doing stuff with a calculator that we forget you can do it without. When was the last time you did a square root by hand (or even remember how). I think this is the kind of thing that causes people to wonder how things like the pyramids were made, people just can't think of how to do things without modern tools, cause thats all they know. To the Egyptians building them probably wasn't that hard to figure out.

Re:Kinda makes you wonder,

[__aaclcg7560]

As I recall, they use some rather antiquated computers aboard the shuttles for this very purpose.

As I understand it, it's easier to radiation harden a 386 or 486 processor than a Pentium or higher CPU. (Although I read recently that might be changing with a new CPU design but I forget which one.) I heard that the shuttle still has a few 8086 (or is that an 8088) computers on board.

Trivial note: You often see 8086/8088, 80286 and higher CPUs, but have you ever seen an 80186 CPU? IBM used to make an ISA network card (Token Ring?) that used the 80186 as the processor.

Re:Kinda makes you wonder,

[R2.0]

"There is really precious little to be learned from travel outside our own atmosphere with the technology currently available that cannot be learned more conclusively and safely here on Earth. When one combines this fact with the enormous cost of getting all but the most insignificant payloads into orbit, there is a very persuasive argument for forgetting about our space travel dreams, at least for the present."

Which completely misses the point - it's not the learning, it's the doing that matters.

Re:Kinda makes you wonder,

[westlake]

The Model T compared to a modern car SUCKS

The Model T entered the market when there were no hard surfaced roads or trained auto mechanics outside the cities, no high octane gasolines, no gas stations, no certainty that fuel or lubricants would be as advertised.
Under those conditions, a simple, tough, forgiving, automobile with a 20hp engine that can cruise comfortably at 35-40, and gets 20-30 mpg doesn't look half-bad.

Re:Kinda makes you wonder,

[jfengel]

Or a networking stack, or threads, or security, or a clipboard, or plug&play, or localization, or...

Re:Kinda makes you wonder,

[jericho4.0]

Quite right. The payloads we are currently capable of boosting are useless for any real development of space. If we are to have the future we dream of, we're going to have to find (and fund) another way to do it.

Yes, we are certainly capable of sending small probes out to discover facts about the solar system, but those facts will remain useless until we can get there in person.

I remember reading a statement somewhere regarding rocket tech, along the lines of 'a 5% increase in current booster efficiency would allow double the payload'. Anyone know these numbers?

Re:Kinda makes you wonder,

[GekkePrutser]

Trivial note: You often see 8086/8088, 80286 and higher CPUs, but have you ever seen an 80186 CPU? IBM used to make an ISA network card (Token Ring?) that used the 80186 as the processor.

Philips used to make a computer called the 'Yes' based on this CPU. It was supposed to be an IBM compatible PC, but for some reason (I think the 80186 itself) it wasn't 100% compatible, and therefore it failed miserably... It was a good CPU though, better than the 8088 that most PC's of the time used.

Re:Kinda makes you wonder,

[hughk]

Not quite, the Soviet space program was big into sequencers. That is glorified timers that controlled the mission. Pretty easy and reliable tech and well known from missilies. However, they are pretty hopeless if you wanted to change a sequence mid flight. Otherwise they used dedicated guidance systems (coupled gyros), which again was old tech.

The Saturn V itself had quite a lot of processing power for the day. The LEM and CM/SM were fully controllable. The basic programs were in ROM but they could be used quite flexibly. On Apollo 13, I seem to remember that they had to use the computer for the pre-reentry manouvers.

In your example of the B2, I agree. The flying wing was considered extremely difficult to manouver. Now the B2 along with many modern warplanes use instability to increase manouverability, but the control is eased through the use of processing power.

Fantastic

[Icarus1919]

Now all he has to do is build his own apollo 11, and he's all set to go to the moon! He just has to pay a few hundred million to get the rockets to take it up.

Give it as a gift?

[Prophetic_Truth]

[me] HI AUNT EDNA! Look what I built for you! Its an exact replica of the Apollo guidance computer!
[Aunt Edna] uh, thanks?

I don't understand..

[IGTeRR0r]

I understand that it took him a long time and it's quite an incredible feat, but how is it usable/testable? Apollogize for my stupidity.
--
http://www.gamercentric.com/ - Now with a clan and tournament system!

Or...

[binderhead126]

You could just hack a Gameboy Advance, and have even more horsepower! To the MOOOOOOOOOOOOON!!!!!!!!

yea but...

[mindwar]

will it run Linux? ... or at least NetBSD?

Re:yea but...

[jd]

NetBSD will just about run on those CPUs they use in washing machines. As for Linux, well, if the ELKS project is still alive, then you can just about run Linux on anything 8 bits or larger.

People can (and have) ported just about anything to anything. There's a version of X11 that runs entirely under Java. There are patches to make Linux run on a VAX. Hell, ply Linus, Alan Cox and Richard Stallman with enough beers, I'd be willing to bet you could talk them into developing the necessary hooks in Linux, GCC and binutils to run Linux on a Z80, a 6502, or (in this case) a lunar module.

You'd have to assume sufficient physical storage of one sort or another, in order to run the kernel, but provided that was done, I can't see any objection.

(I actually wish someone WOULD port Linux to one of the ancient architectures, as it would provide an actual demonstration of Alan Turing's computability theory. At present, it's just words. People don't work too well with words, they work much better with examples they can see, experiment with, and learn from.)

Re:yea but...

[new500]

will it run Linux? ... or at least NetBSD?

Yeah, but based on what i read about the original, i bet BillG wishes he could port NT to it :

"Shortly after liftoff of Apollo 12, two lightening bolts struck the aircraft. The current passed through the command module and induced temporary power failure in the fuel cells supplying power to the AGC. During the incident the voltage fail circuits in the computer detected a series of power trenches and triggere several restarts. The computer withstood these without interruption of the mission programs or loss of data.

quote from http://klabs.org/history/history_docs/mit_docs/170 7.pdf

Which kinda redefines "Uptime" for me :)

Every operating system sucks...

[powerlinekid]

you see, I come from a time in the nineteen hundred and seventies
when computers where used for two things
too either go to the moon or play pong
and nothing inbetween, you see
and You didn't need a fancy operating system to play pong
and the men who went to the moon, god bless them
did it with no mouse
and a plain text only black and white screen
and thiry-two kilobytes of ram

Beyond that, this guy is lucky its christmas because with multiple 4-9 meg pdf files it would be a silent night for his server.

Re:Every operating system sucks...

[neuromortis]

For those who don't know, the above excerpt comes from the comedy genius of Three Dead Trolls in a Baggie, specifically the track "Every OS Sucks".

Re:Every operating system sucks...

[powerlinekid]

No no, you got it all wrong. The thing that looked like pong really was the guidance program. The first paddle was earth, the second was the moon and the little dot was the capsule.

And now it's been linked to by Slashdot.

[Nanoda]

10$ says it's flashing "1202" right about now...

Re:And now it's been linked to by Slashdot.

[agtorange]

Well then all he would have to do it flip it off and take the thing in on manual.

Re:Emulators?

[enosys]

An emulator already exists. It has been released as free software under the GPL. It supports Linux and Windows.

FPGA

[saned]

Not to undermine his job, which I think is a major accomplishment, not only by building it but by reimplementing the whole logic from diagrams. But looking at the logic, it seems it could fit easily in a Spartan 3 FPGA. So yes, it could be done cheaper and faster, but not with the degree of detail this guy put on.

Kudos to him

Re:FPGA

[murderlegendre]

But looking at the logic, it seems it could fit easily in a Spartan 3 FPGA.

Pretty amazing, isn't it; how far we have come in so little time.. And the fact that this guy took the time to properly execute the project, and document it as he went along, really allows one to gain a sense of scale when it comes to computing devices. This thing has about as much computing power as an Atari-2600 and it takes a truck to move it. And just about ten years later, we were playing pong in the living room.

And it took people to space, and back again safely. The AGC I mean, not the Atari.

Sucker!

[bbh]

With that kinda money you could rebuild the sound stage they faked the first trip to the moon on!

-bbh

For the space geek who has everything

[dspisak]

Now he has his own server slashdotted just before [Insert Religous Denomination Holiday Here]. Yup, he sure is the space geek who has EVERYTHING now!

Take that those doing with less!

Your Palm Pilot is not radiation hardened.

[gelfling]

There are two reasons why spaceflight computers are relatively underpowered:

Reliability under conditions your PC would fail, like radiation, shock, vibration, acceleration, heat and cold.

Built to solve unique specialized problems for people who are not entirely computer expert.

Navigation computers have to solve complex solid analytic geometry problems for people who are experts in solid analytic geometry but aren't experts in computers and don't have the luxury to spend lots of time to do that.

Re:Your Palm Pilot is not radiation hardened.

[fmaxwell]

That is complete nonsense, the parent poster was totally correct. And also Palm pilot is about $90 if you use Ebay.

I worked at a satellite company and used to believe that a lead box would work. I was set straight by the guys who know about that kind of thing. Lead is not impenetrable to radiation. In fact, early geiger counters used lead shielding as a means of scaling the count.

Limits of simulation?

[Tablizer]

Can it simulate the part where the sensor loop queue was overloaded because they forgot to turn off the rendevous radar and the warning lights went crazy and Neil or Buzz wet his suit? (I have no official info that they did, but I bet at least one did but never told anyone.)

Finally

[Lord Kano]

A perfect way to add guidance to my Cruise Missile

LK

Time != money

[node 3]

Time is far, far more precious than money. We only trade *some* of our time for money so we can use that money during the remaining time.

Re:parts?

[node 3]

Where on earth are you going to find the vintage IC's for this thing? (Didn't RTFA).

Well, RTFA you lazy sod! Had you done so, you'd have had your answer quicker than it took you to post the question.

No disrespect, but...

[fmaxwell]

I have the utmost respect for the initiative, intelligence, and generosity of the man who built this computer. That said, he didn't build a replica of an Apollo Guidance Computer (AGC). He did not use the same parts, constructing it with higher integration 74LS parts that gave about a 10-to-1 IC package reduction. The original AGC prototype used core memory and his uses static RAM and EPROM. There are countless other differences.

Again, he is deserving of high praise, but he did not replicate the original AGC I prototype. He created a working model which was very true to the original at the block diagram level.

Re:OK, so Apollo can go home, but how about...

[HermanAB]

Nope, even Apollo is long gone. King Constantine retired all the old Gods in 325AD at Nicea near Naples and defined the Trinity to take their place. Only Zeus, Mercurius and Demeter survived - oh, and Isis - she survived too, Contantine didn't want to kill her and her cute little baby, but in return for continued worship, the Gods were morphed. Constantin caused such divine confusion, that the collective memory of the Western World still haven't recovered...

To me,

[StarKruzr]

This begs the question of how we make voluntary sterilization an attractive option.

Re:Oh?

[Detritus]

Your best bet is to take a course in computer architecture or pick up a textbook on the subject, esp. one that has a good survey of older computers that introduced significant architectural advances.

The IBM 360/91 was an important high-performance member of the IBM 360 family. The CDC 6600 was also an innovative system from the same era.

The Space Shuttle uses the IBM AP-101. See Computers in Spaceflight: The NASA Experience.

Re:Obligitory Priceless joke.

[alienw]

Go read the articles and you'll appreciate what a tremendous amount of work this was -- a hell of an achievement of the variety that makes most PhD applications look like a 3rd grade book report.

Unfortunately, it's an achievement akin to digging a large hole in the ground with a spoon. Someone wasted a lot of their time to do something useless in the most inefficient way possible.

Apollo Guidance Computer - Disassembled

[new500]

. . .

This is a link to a a partial tear-down of a Apollo Guidance Computer Logic Unit.

http://klabs.org/mapld04/presentations/session_g/g 1007_hall_s.ppt

on slide three, N.B. the cost : $275,800.00.

now i wonder could the guy in the story have afforded to deal with this as well :

"In the early orbital missions before Apollo, NASA learned that the human animal, confined in a spacecraft for a week or so, was not as clean as might be expected from observations on Earth. This additional constraint had . . far-reaching impact . . All electrical connections and other surfaces had to be corrosive resistant . . . everything had to be hermetically sealed."

eww!

quote from http://klabs.org/history/history_docs/mit_docs/170 7.pdf pages 4-5.

Documentation

[Cryptnotic]

What's the most important thing about what this guy did?

Documentation. He documented every step of the way everything that he did. It's something that's lacking in a lot of geeky projects and it's something that I commend this guy at doing an awesome job at.

Next: Build your own working replica of Little Boy

[Muhammar]

"Early gun-type designs are interesting. Because they're so simple, you can (if you like) actually understand the entire critical assembly process, from the start of fission to the propagation of the produced shockwave"