Apollo Guidance Computer Part 12: Return of the AGC

The AGC is in my lab for a short while. We will rush to complete the restoration in time for the 50th anniversary celebration. Step one: finish the testing of the Tray B modules. Bad surprises await.
For PCBs, use our sponsor PCBWay: www.pcbway.com
Buy the CuriousMarc DSKY and other retro-computing T-shirts to support the channel here:
teespring.com/...
More information about what's shown in the videos on my web-site:
www.curiousmar...
Some relevant links:
Playlist of the restoration series: • Apollo Guidance Comput...
Block I AGC period documentary: • MIT Science Reporter-"...
Inertial navigation system documentary: • Video
Schematics: www.ibiblio.or...
and: archive.org/de...
Mike's AGC backplane viewer: apolloguidance....
AGC software repo: github.com/vir...
The Ultimate Apollo Guidance Computer Talk: • 34C3 - The Ultimate A...

[@spikeypineapple552]

This is my favourite series on KZbin. Order of magnitude better than normal tv!

[@digitalblasphemy1100]

I have been obsessed with Apollo for 20 years. What has attracted me to the program so much has always been discovering a problem in the development that had to be solved by engineers. In that respect, this series most definitely satisfies me. Keep solving problems gentlemen!

[@timrb]

I'm absolutely loving this series Marc! Keep up the great work.

[@ArthurGoelzer]

10:27 Such an invasive surgery, now i cant wait for the next episode. Thank you Marc!

[@jetraid]

I'm scream "nooooo" when I see that machine over the module

[@ArthurGoelzer]

​@@jetraid Me too, its like when i watch some brain surgery videos on youtube, when the doctor uses a saw to open the patient's head. Just a wrong move and the module dies =\

[@bobl78]

opening the module and take a look inside for the first time in 50 years.. awesome

[@ShainAndrews]

I don't think I would have the moxy to put any of those modules anywhere near a mill. Really pulling for you guys. This is something truly special, and appreciate sharing it with the world.

[@aserta]

Sometimes, in the pursuit of historical repairs, you have to bite the proverbial bullet. This isn't any different than a restorer bringing a scalpel to a 300 yo painting to remove crud and old repairs. :)

[@rasitsimsek9400]

Guys you are incredible. I have learned a lot how the AGC internally is working.

[@MrBombastic1st]

I love everything about this series!!! Thank you for doing this. I've been inspired by the US space program to become an engineer and the fact that there's people as dedicated as you restoring this stuff just brings me a lot of joy.

[@RobLion]

Great to see this video popping up, and see myself on screen. It was an absolute thrill to be able to contribute to this project!

[@CuriousMarc]

Thanks so much for the help Rob. I need to do a proper intro in an upcoming episode.

[@kevinkatz7027]

When they broke out the drill press with a 5/16 bit, I knew they were professionals....

[@13bigerdave]

I may not fully understand all of what you are doing it is way cool and I love watching all your videos can't wait till you get it working and I know you will

[@dr.strangelove5622]

You guys are great! It is so much fun to watch your videos. I don't understand half of the things you guys say, but the passion with which you make these videos is infectious!!

[@yorgle]

aw! these episodes are too short! Love them! :D

[@RobertGallop]

Such a great series! So jealous you got so much hands on time with hardware made at the height of the space race, so cool. Thank you so much for sharing with us!

[@MikeF1189]

I really hope to see this computer work. Go get snoopy and swap out the new old stock parts.

[@3ctjhseolere]

Anyone got Solar System GPS?

[@15743_Hertz]

I just tossed a C64 power supply due to the circuit board being under 1/2 inch of epoxy and can't imagine what you're going through with the module. Godspeed, AGC restoration team!

[@jaimebarr8426]

It's beautiful engineering magic. Wondrous to behold.

[@paulnelson8946]

Thanks for posting all of these videos... Its so fun to be fascinated by something you didn't ever think about before. Also love the videos on the Xerox Alto!

[@72polara]

With all the skilled people you have and maybe a bit of luck, you guys will fix it! By the way, I happened to be in the area of the Computer Museum and stopped in for a couple of hours. I will be back when I have the entire day to spend. Great work you guys do there, too!

[@LordPhobos6502]

Every time I watch one of your videos, trying to reactivate ancient electronics, it inspires me to get in the study to do the same 😁 Keep up the great work, it's very entertaining as well as educational and inspiring ❤

[@rjcroy]

I love following this. I am really looking forward to, hopefully, see the AGC running. Thank you for sharing this.

[@jakeman50]

great seeing young engineers involved in this project, jest because you have the prints and program you still need the human brain to solve problems like this

[@JensAndree]

Talk about cliffhanger ending...!!! I hope you upload the next episode soon because I'm dying to know how it pans out?! Many thanks for saving history, making these videos and for making them available to us!

[@kamanchikamanchi3574]

This is the series to suggest to any moon landing skeptic.

[@andrewrixon2347]

Thanks Marc. Porridge for breakfast and you & the team battling the evils within the AGC !! Great way to end the week. Way better than any role playing game. Lol

[@nicwilson89]

This has become my most anticipated series on KZbin by far. Absolutely fascinating! Can't wait for the next one, guys. Great work!

[@jgifford3561]

what a great cliffhanger!

[@stevec00ps]

Argh those last few seconds! My heart's in my mouth!

[@crashtech66]

Oh, man the suspense!

[@williambello4089]

I am wondering if it is possible to contact the manufacturer of the potting compound, to see if they know of a specific chemical or way to dissolve that potting compound without destroying the circuitry.

[@Shadwhawk]

I talk with Mike in IRC, and we spent a while researching the likely potting compounds and their removal agents. All we could find were agents that require heating to a couple hundred degrees Celsius and soaking it in the agent, which felt like too big a risk to take. Could damage or destroy the whole module.

[@williambello4089]

@@Shadwhawk it sounds like there is no option. That module must be replaced. Fortunately, you have the schematic, and a good engineer can use modern-day components to perform the same functions. Oh, did I mention... I'm a printed circuit board designer. If someone could machine a new casing component, I'm sure a replacement printed circuit board can be made for relatively low cost. Provide me with the re-engineered schematic, and the required dimensions of the printed circuit board. I offer my services gratis for the privelege to contribute to this group. The complicated plug pinout dimentions can be done relatively easily using CAD printed circuit board and a jig.

[@lagia5]

Cant wait to see this beast running

[@littlejason99]

Argh! Just when it was getting good! Can't wait for the next video!

[@Ice_Karma]

Oh... oh my god... I _really_ hope good things come from that drill press shot at the end!! >.>

[@Strothy2]

Can't wait for the next episode it must be nevewrecking to mill this module

[@williambello4089]

You guys are Uber geek! (I mean that in the best possible way) I enjoy this series more than any other KZbin series I've seen. I am wondering if it would be possible to visit you guys during one of your sorries. I am still very much hoping that you can get a reaction control system, (or at least a rcs simulator) and star field sextant. That may be difficult to find, but I have great faith that somebody out there has one sitting in their basement. I am still stunned that you found an actual gyro!!!

[@mb5058]

Can’t wait for the next video. Does this mean I’m really an incurable geek?

[@hingeslevers]

Yes ;)

[@asgerms]

Yes! We all are!

[@glenwoofit]

Top job, I can't wait till the next video....

[@soulrobotics]

2:08. I got an idea to support. I'd loved to have those blue prints, with the sign of the team. I know it has no logic, but for sure there is going to be someone who want to buy it like me.

[@kevinreardon2558]

Cutting into the potting. That's brave.

[@1MinhDuc]

Keep going guys, never give up!

[@MikeBramm]

We've had luck using Methyl Ethyl Ketone, Paint Stripper (Methylene Chloride) and hot air rework stations getting most potting compounds to break down so they are rubbery and can be remove by picking them away. It's a little slow and you must have good ventilation, but it works without damaging the underlying circuitry. That was 30 years ago so these tricks may apply to that old potting compound. These chemicals used to be available at paint stores, but they might not be so available in California any more. You might also try Epoxy Paint Stripper. Be patient and good luck.

[@CuriousMarc]

Unfortunately these miracle strippers are known to the State of Cancer to cause California. Not available here! Supposedly acetone was supposed to do it, but it did absolutely nothing on that compound.

[@timothybarney7257]

@@CuriousMarc No more AvE until you get the AGC working!

[@glitchysoup6322]

8:40 thanks for using GNU/Linux with KDE.

[@rearspeaker6364]

its MUCH safer here, no idiots from somewhere thinking your making something bad.

[@Lazarus7000]

Oh jeez a milling machine?!? Have you tried a little Methyl Ethyl Ketone on the potting? Some plastics just, well, stop being plastic under its influence. I learned this the hard way trying to take the yellowing off of old computer equipment; lo and behold it works great on this, except that there is almost total overlap within the groups "plastic that MEK un-plasticifies" and "plastics that are subject to yellowing". So while you can use it with reasonable confidence on an IBM Model M keyboard, you will almost never have a reason to (unless someone's written on it with marking pen or something) it will almost immediately start making Apple gear all melty. It will also leech the blue out of shop-towels after a while...

[@tommccrystal8917]

MEK may dissolve the insulation on the embedded wiring. I've done that before.

[@Lazarus7000]

@@tommccrystal8917 Might, might not, and, well, to quote AvE, "You can't fuck it more".

[@TheJimbodean67]

Yaaaaayyy another agc episode! Annnnnd another broken module. Tough break, but with the combined mind power in your lab you guys should be able to work the problem. I’m wondering if a chemical solution would work for clearing the pin connectors vs mechanical method. Thanks again for sharing.

[@shifter65]

I need this to be 50 minutes longer!

[@chriswendi]

try using a Phillips ultrasonic adult tooth brush to clean pin socket. may need suction from vacuum to help remove loose debris.

[@alexa.davronov1537]

Cool. Thanks for sharing.

[@PepijndeVos]

Such a cliff-hanger!

[@phonotical]

You can get a pu foam remover, maybe it would have helped

[@swebigmac100]

Omg this is intense

[@maicod]

acknowleged you are now an official filmstar Marc :)

[@MattVileta]

OMG, you're gonna mill away the podding!!!!!

[@informativt]

Cliffhanger!

[@ElectricGears]

I've removed a hard black potting before by just scraping it out with a hot soldering iron tip. It was a modern product so it may have been epoxy. The heat caused it to decompose into a powder (and a bit of smoke) as I plowed out with a blunt tip.

[@tomeucapo]

Great job team! :)

[@rearspeaker6364]

at 1:56- that's one large amphenol connector!!

[@mikewazowski350]

This is geekdom x 100^100

[@AsbestosMuffins]

rocket computer surgery > rocket surgery

[@AaronJackson1]

Aww what a cliff hanger! I can't wait :(

[@doc_sav]

Poor Mike. I don't think I have noticed him looking that discouraged yet in the series.

[@ntrasla]

What a cliffhanger 😅

[@DandyDon1]

These analog ICs manufactured by Fairchild were developed after General Micro-electronics (GMe) attempt at manufacturing MOS chips for the Victor Electronic Calculator?

[@scotty3034]

There’s always a whiz kid in these stories.

[@johnchristopherrobert1839]

You should use windex to clean the ICs. The Greeks swear by it. 😂😜

[@davidkilpatrick1640]

Better call my wife and let her know I'll be late for dinner .. there is a new AGC video to watch!

[@spunkmire2664]

cliffhanger! lol what i cliffhanger, cant wait!

[@maicod]

I'm thinking simple: couldn't the missing diode function between the 2 pins which are open now be fixed by adding an off the shelf diode to those pins externally ?

[@mikestewart8928]

Unfortunately no. The problem isn't the diode; it's the winding around the switch core that's in series with the diode. Without being able to send current through that winding, we won't be able to access large sections of erasable memory. This open circuit could also be called by a wire break in that winding; we're just hoping at this point that it's the diode and not the winding, since the latter would be much more of a problem.

[@maicod]

Hi Mike. thanks for explaining your problem. I'm hoping you guys will find an answer for the de-epoxyfying. I'll keep following your quest to get the AGC in working shape, my eyes are glued to this channel :)

[@lbochtler]

I hate potted assemblys. Have those in one of my microscopes and they are basically imposible to fix

[@justimagine2403]

AGC (Apollo Guidance Computer) - how it works - label added for my benefit in the comments section.

[@kevincozens6837]

um... can they be seriously thinking of using a drill to remove the potting compound? We will just have to stay tuned for the next episode.

[@johnopalko5223]

That looked like an end mill to me. Not that that's any better...

[@joseph9770]

I don't think they are going to drill _all_ the way down. Maybe they are going to drill a little well and try to dissolve the rest. Not that bad of an approach in my opinion...

[@bobkoller90]

Is this potted with epoxy over the polyurethane foam? If so, it may be possible to remove the epoxy overcoat, and then chip away at the foam. Industrial strength paint remover is about the only way to dissolve either of these, and that carries it's own set of problems. You also might be able to abrasive blast these coatings away with a very fine jet. I think that the welded wire connections would survive most anything.

[@ashman187]

plug it and lets go to the moon. lol What's your threshold?

[@paulhammond7489]

From personal experience that potting compound was extremely tough stuff, let's hope it has lost some of its toughness in the intervening decades since it was potted.

[@panikrev175]

Paul Hammond Unfortunately, that potting compound has the same feel of a hockey puck. This stuff is stubborn and filled with hate. I know this because I used to have to case electronics and pot them....

[@paulhammond7489]

@@panikrev175 the company I worked for used it to pot numerous modules, it was as hard as basalt rock once cured, back in those days such potted assemblies were not classed as repairable once they developed an internal fault. The end mill should get through the upper material, but they'll have to find another way once they reach the pins / components. But I guess we have to wait for the next video to see what they have planned :( Patience was never one of my better traits...

[@panikrev175]

Paul Hammond I’m hoping they fab up a jig to begin building a replacement ferrite-core memory module and make a replacement PCB for the current switch module. It would be awesome if the team could replicate a modernized version of these modules. And yeah, modules got scrapped on the spot if they didn’t pass testing after potting. We made IPC-3 rated products, so it was pretty crucial for our automotive products if they failed. SMT components, modified through-hole, wave soldering... I miss that job.

[@paulhammond7489]

@@panikrev175 When I first worked on computer systems during the late 70's the company still employed a bunch of workers that were skilled at making those ferrite-core stores, they ladies used to challenge new employees like me to see if we could weave a few cores, I can tell you I failed miserably. I can only imagine how long they must have trained / practised until they perfected the skill, I think I could have practised until today and still not been able to do it as well as those skilled ladies.

[@TestTubeBabySpy]

Wow, the drill press clip scares me. Is there no way to CT scan it [or some other type of scan] and plot the precise location of the components underneath the potting material and plug that info into a CNC machine? Or is that just fantasy thinking on my part..

[@carlclaunch793]

NASA archives have exact diagrams of the component layout which we obtained, plus we have pictures of an unpotted version of this module. We do know exactly what is under there and its placement.

[@DavidHarris-qn7em]

Formula 409 and windex mix..spray lightly..brush scrub with soft brush..soak 2 minutes..clean with denatured alcohol..check out BBI's channel

[@juliusheide3643]

hey, invent some bit for cnc milling machine, so when the bit come in touch or very near some wire/component part it move left /right or back the tip, maybe capacitance/or near field something.

[@zbradbell]

1:18 "usually we just..." what was that?

[@CyclingSteve]

Well, you saw the last bit with the drill right? :D

[@b43xoit]

"... start hacking"?

[@ElectronicMechanic50]

if had luck with a heat gun removing potting material.... just a thought

[@alastair-r3676]

Using a hoover is playing with fire as it generates static and then snuff your electronics gets fried, blowing air is safer.

[@alastair-r3676]

Brad. I didn't know, when I heard the hum and saw the hose and brush nozzle I thought yikes no. Hope they were using one then.

[@tpcdude]

great work again -- more work than the teletype but bigger reward .. the guy making the most progress is wearing the white lab coat maybe if everyone wore one ..

[@Yrouel86]

I see at the end you are using the milling machine to I presume remove at least the bulk of the potting. I was wondering if asking a dentist or perhaps an archeologist (one that has experience in cleaning fossils) for assistance might be of any help in doing the fine work in potting removal. Unless of course some of you has such steady hand and proficient use of tools like a dentist drill.

[@CuriousMarc]

You are on the right track...

[@murrij]

I know this goes back a ways, but 6:06 - why would they have filled it with black crud? That potted module reminds me of even modern stuff where the manufacturer is attempting to hide their intellectual property and/or make you have to buy a part from said manufacturer. Was that the case here too?

[@CuriousMarc]

Oh, not at all, this is to make it flight worthy: shock, vibration and humidity resistance. It dates back to an early 1950s rocket that failed shortly after liftoff, and the cause was two diodes leads shorting in a power supply due to vibration, although they were pretty far apart. Potting became mandatory from then on on mil rockets. And it’s not crud, it is quite elaborate materials, or stack of materials as in the core memory, with the right softness and hardness. Potting still used today in industrial, mil and aerospace equipment

[@murrij]

​@@CuriousMarc Awesome! Thank you so much for the detailed answer.

[@twilliamson3]

That work space is pretty legit, is it someones home lab?

[@maicod]

Tripp Williamson I believe it is Marc's house's basement

[@dwarftoad]

Did you consider patching in some kind of new replacement circuit in between that module and the computer rather than repairing it?

[@mikestewart8928]

Unfortunately it isn't that simple. The problem isn't the diode; it's the winding around the switch core that's in series with the diode. Without being able to send current through that winding, we won't be able to access large sections of erasable memory. We'd have to replace this whole circuit, and the type of core used in it is nigh impossible to find nowadays.

[@sim4fun]

What does AGC stand for?

[@rearspeaker6364]

Apollo Guidance Computer

[@joseph9770]

What are those machines in the background at: 7:19?

[@CuriousMarc]

HP 9826 and 9836. High end pre-PC era modular scientific computers. I haven’t gotten to them yet.

[@joseph9770]

@@CuriousMarc, Thanks. I'll be looking forward to seeing something on those. :-)

[@mike_van_in]

Marc, what was the joke at 1:23 ?

[@CuriousMarc]

“Usually we just start hacking at it”. The journalist crew kept asking for a detailed plan of what would happen when, so they’d be there for the big moment. Which of course we can’t predict. I was poking fun at them.

[@mike_van_in]

@@CuriousMarc Thx.

[@Saghetti]

hi!

[@freednighthawk]

Acetone will break down PU Foam while leaving the rest of the components intact.

[@Sixta16]

Not really. Ketones can attack some kind of plastics (polystyrenes) that may have been used within that module (holders, wire bushings,..).

[@Dirtlifer]

And another cliff hanger, stay tuned ending :-( See you next time.... same AGC time.... same AGC station...

[@Dirtlifer]

PS Mark.... thanks for what in my opinion is some of the most interesting and well timed restoration! The countdown to 50 is moving in so quickly and this fact is truly making this most awesome content exponentially more entertaining! Thanks to you as well as a great thanks to this most capable team set to this restoration task ladened with pitfalls at nearly every turn.

[@Wobblybob2004]

Looks like a job for a dentist not a milling machine!

[@mcapps1]

A 50 year old hand made computer...wtf did you expect?

[@tomeucapo]

Aarrrgghh! Terrible dirty MALCO female pins!

[@davidk8184]

Acetone

[@cameron1975williams]

What if the pioneers responsible for these masterpieces of engineering, deliberately made it impossible to follow in their footsteps?

[@uberdude2555]

Yikes , it looks like a successful restoration of the AGC is far from certain.

[@physicsguybrian]

With the right team and infinite resources, anything is possible. They have the right team!!!! It will happen!

[@minibikemadman]

Linux ftw!

[@dogmannz]

Mike is still a hottie

[@jeremiefaucher-goulet3365]

Just find the stolen module. It shouldn't be too hard since the module obviously is on Earth. They wouldn't have put an unpotted one in space. It sure narrows how far you need to search ;-) lol

[@CuriousMarc]

Excellent idea. Now where to find the NASA trained module sniffing dog.

[@jeremiefaucher-goulet3365]

@@CuriousMarc Dead by now, considering the lifespan of dogs. Train a new one using your current modules 😂. I would hope they used "Pushinka" back then. Nothing beats the offspring of a Soviet space dog for such a purpose 😉

[@oldschoolelectronics8573]

First.

[@panikrev175]

First Comment on the First Comment.

[@CuriousMarc]

@@panikrev175 OK @The E Lab one gets a vintage bit, but @PannickRev! gets only a vintage parity check bit.

[@oldschoolelectronics8573]

@@panikrev175 Inception much?