Restoring the Apollo Guidance Computer with CuriousMarc
ฝัง
- เผยแพร่เมื่อ 30 ก.ย. 2024
- Dr. Marc Verdiell (CuriousMarc) and his equally crazy team mates take on challenging restoration projects of major vintage electronics artifacts, including the electronics from the Apollo spacecraft.
VCF Southwest is made possible by the efforts of the Vintage Computing Collective of North Texas, an 501(C)3 non-profit. Please consider a donation to help us continue to produce shows and content like this. www.paypal.com...
I think I have watched all of the Curiousmarc videos on the Apollo gear but I was not aware of some of the background to the restoration efforts, so this has been a wonderful wallow for me. Thank you for putting this out into the public realm! nb Always keep your receipts!
Marc and the whole team, Carl, Ken and Mike (especially Mike, the lone individual who first designed a gate accurate simulation of the AGC in FPGA) did a fantastic job getting the AGC back up and running. And of course Mike had all the rope core memory dumps one could ever want or wish for, to load into the AGC.22:25 And then there's the proprietary connector, with all the weird preservative that was caked in it. Marc spent a lot of time restoring that.
Marc's video series on this is absolutely stellar, one of the best things on YT in my humble opinion, and it's still going on! They give you an appreciation for all the guys involved in the restoration, the original engineers themselves, and the astronauts who trusted their lives to this stunning example of Human ingenuity. The fact that all this came so close to becoming lost to history should make us all club together and give Marc and his team some kind of award.
The absolute gobsmacked slap of luck that Marc works for Samtec can not be understated: this is a big reason the project was successful, being able to completely re-make connectors that have been long out of production.
Well, connectors aren't rocket science. It may not be as good as this one, but you will be able to make one that fits, even more so when you have the full specs.
There are ways of manually making the contacts using jigs. Just like the watchmakers used to make parts. Jigs and hand files. Now the hand files are special and may well go down to 10,000 grit for some of them - but they’ll do the job. Contacts that have box structure would be assembled from two flat parts. Very tedious process and their time was better spent on other things - hence getting folks at Samtec do it.
"It would take another talk to go through it" - Yes please!
Thank you for the comprehensive story. I watched the original series but hearing it in one go and with the extra background makes it even more amazing.
Hats off to whoever decided to plant the photo bomb of Scott Manley in Marc's presentation lol
This talk was AWESOME
What this team achieved is nothing short of a tour de force in technological archaeology; it is so important to document these milestones in computer design and we are damn lucky to have these guys working on the case.
Fabulous, fabulous stuff. I've been following along on the CuriousMarc channel but have really enjoyed watching this and filling in the gaps.
14:17 I didn't know wire wrapping is more reliable then soldering! So cool :) This was a labour of love! These guys are amazing!
Thanks for this. I watched all the videos when Marc posted them but it a really good presentation by him summarising the whole adventure.
"repair it by shaking it"
That sounds like most of soviet products :D
Shaking could be done by punching the box that contain it - like TVs - picture is not good enough, then just hit the cover on top or sides :D
Absolutely incredible! Amazing effort.
Brilliant, I'm saddened that the AGC went into some obscure place, it should ideally stay with the original restorers.
Hey @curiousmarc, I know you're a guy that documents everything. Do you (or mike/ken) have any article or video about the 1202/1201 scenarios being more bad than harmless? I don't remember seeing them in the restoration videos.
Not yet. But certainly coming. It’s mostly Mike’s work at the moment. We need to bring up the CDU first to put our hypothesis to the test.
@CuriousMarc awesome, can't wait 😀. Love your work BTW
@@CuriousMarc Did you find separate issue with CDU, or it's about determining exact amount of CDU-induced TLOSS and subsequent ill effects of stacking non-reentrant code, potentially submitting stale commands and stomping over global variables?
Thank you @CuriousMarc for this excellent summary and background story of the AGC project. I watched all videos in the series but this presentation filled the gap of information for me. I have tremendous respect for all of the team, the research and what you have accomplished.
🤚Same. This fills in a ton of gaps, and it's cool Mike is still truckin' continuing work on core rope dumps.
I have always wondered where did that restored AGC ended up. I saw all the restoration edited videos. They where the best. Really well produced. Thank you for showing us this jurney.
"Look ma there's a diski in my plane" no one laughed! Well i did. I guess me and Marc have similar humour...
hahahah, I bet those two faulty modules ended up in that unit because the two original modules were 'borrowed' for testing another faulty AGC.
Awesome, having enjoyed the full series this is an excellent summary. Thanks so much. Looking forward to the results of the CDU investigation.
Before saying that "Eagle was lucky not to crash", I would refer you to the old adage "You have to be good to be lucky and lucky to be good." Armstrong was monitoring the vehicle performance closely in the descent phase while the autopilot was on. He said as long as the LM was behaving normally he wouldn't have disengaged the autopilot (and flown manually}or commanded an abort. Depending on where the 1202 loss of control were to hypothetically have occurred, Armstrong might still have had the option to disengage the autopilot, and land manually (as he did eventually!).
As far as I understand that would not have worked. The loss of control would mean you also lose your hand controlled landing, which was basically another program in the AGC that just quit on him. The only option would have been an abort with the AGS. That means dropping the landing stage and go back to orbit and make Apollo 12 ready for the next try two months later.
Unfortunately, in a fly by wire system, if the computer is stuck, then everything is stuck. The interface between the hand controls and the thrusters was software.
If you can find it, I recommend the book "Microelectronic Packaging" by Sideris, the 1968 edition, for historical information about PCBs and ICs of the era.
What would the AGC have been like if IBM had been given the contract to design and build it?
Magnesium case in pure oxygen. What could possibly go wrong?
Amazing and very inspiring stuff! So far ahead of its time! Thanks to Marc and his team!
Love this. And all things Apollo. Great presentation and glad to hear the history of this restoration. For the person that asked about the flammability of magnesium, while it does burn hot, if you have a situation where something is hot enough to start igniting your magnesium components, you've got a much worse problem already that's probably not going to be survivable. I sincerely hope that when we return to the moon that there are people on the team as ingenious and creative as on the original team. I feel like a lot of today's engineers are 'lazy' compared to "the days of yore". I have faith that dedicated space and tech nerds are up to the challenge.
The few attempts to burn a NeXT cube show how hard it is to get magnesium to ignite.
Absolutely fascinating to watch :)
Fantastic presentation
The 1202 warning / error was the AGC complaining it had too much to keep track of and had to dump low priority tasks to keep track of high priority tasks, which at the time the high priority was landing and low priority was updating the DSKY display as one way to explain it.
Just watched this entire video after having followed the entire AGC series, and now also the Apollo S-band series. I am absolutely enthralled, and can't wait for the next videos to be published.
This is truly a great video, I enjoyed it. Manu thanks. Ps I was 18 so remember the Black and White tv coverage :)
Excellent talk Marc. I couldn’t be there in person so I’m glad this was captured.
Can we get Marc to come to VCF east?! Lol. Love his channel.
the way you put Marc's face right above the date in the thumbnail makes it look like he died! 😅 please don't do that!
We have updated the thumbnail at your suggestion.
New subscriber
Mike is such a engineering stallion
This is extraordinary.
Typical Vtech layout.
Bravo Marc!