Apollo Guidance Computer Part 22: Module B11 fails again!

[@JamieWhitehorn]

Curious Marc's AGC Repair Shop - Solid Work, Reasonable Rates, All Work Guaranteed. The only AGC repair shop in the Solar System 💫 😀

[@Eo_Tunun]

Let's bet the Roswell with happened with time traveling Aliens who wanted to stop by for a quick fix at CM Repair Shop. ^^)

[@JohnDavidDunlap]

* Best AGC repair shop in the Solar System

[@zachz96]

What are the chances Louis Rossmann could repair an AGC?

[@oswaldjh]

@@zachz96 Louis Rossmann Would likely jump at the chance to participate in this venture. It would certainly give him a view into how things are properly engineered and built to last decades. Unlike the status symbols he normally works on.

[@PWingert1966]

I hear there is a some ship named the millenium something or other (Some kind of large flying native predator) due in for computer repair. All its circuits are potted. It apparently got damaged during a race to deliver cargo.....

[@agenericaccount3935]

Thanks for leaving Mike's explanation unabridged. That was honey sweet. Overall this is better than reality TV. Cause its capital R real.

[@g00st62]

this is 500% better than reality tv. Reality tv is the lowest form of entertainment you can find and its almost an insult to compare these videos to that shit...

[@Elios0000]

@@g00st62 i could watch these guys repair old hardware like this for DAYS

[@anonymousarmadillo6589]

@@Elios0000 Not just old hardware. Louis Rossmann

[@francisernens1795]

So what you're saying is that it has a terrible pain in all the diodes down its left side. The Sirius Cybernetics Corporation would be so proud. Amazing to watch this restoration in near real-time. Well done and thank you!

[@Oldbmwr100rs]

Those B11 modules were always so troublesome!

[@tomgeorge3726]

Hi, you guys need your own t-shirts... I'd buy one. "APOLLO AGC REPAIR SERVICE" Awesome video series Marc...

[@dogmannz]

I think quite a few of us would like Mike to not have one at all actually.

[@johnfrancisdoe1563]

Tom George Apollo GC repair.

[@craigs5212]

Nice work guys, brings back memories of working on Minuteman II electronics back in the 60's Had a similar diode problem back in the early 70's with some GE diodes used for a large sampling scanner. Had to change thousands of diodes. Some how the mfg process left some carbon particles inside the glass envelope. Depending on how you handled the diode the carbon particle could dislodge and semi short across the junction. IIRC there was a bizarre test you could do which involved testing while tapping the glass with the cathode up (down?) on a sample of your suspect lot, very bizarre.

[@ShainAndrews]

Mike, once again what a great display of your diligence. Very evident how many hours you logged learning and understanding these details. Undoubtably as fluent in this system as the original engineers were. That was an interesting detail that the vendor was removed from the approved list. I'm assuming they were originally approved for more than just diodes?

[@andromedaturnbull3512]

Is it just me or does anyone else think that Mike is super cute? 🔥 He manages to make debugging a diode failure sound enchanting.

[@MarcelHuguenin]

I'm so impressed by this whole series of videos, it's absolutely amazing to see so much brain power at work. It has been very addictive to watch, same as the series about the Alto restoration. Amazing, thank you Marc for creating these wonderful videos with this great touch of humor and team for the great work you all do.

[@QuintinMassey]

I certainly appreciate the time and effort it takes Mike to repair the diodes, but man is it so satisfying to watch for some reason!

[@skfalpink123]

When you think what could have happened, had those diodes managed to survive "just" long enough to leave Earth's atmosphere. Pretty scary really...

[@SweetBearCub]

Thankfully, the real LMs had an independently designed and built Abort Guidance System that would get them back to the CSM, it could substitute for the Apollo Guidance Computer during any phase (technically), except for the Lunar Landing phase. It was somewhat less accurate, but it was good enough. It also used less electrical power, which was a benefit to Apollo 13.

[@paavotorkkola523]

I have the most respect for your wonderful humility fun engineering across the divide of age in guys and technology. pure fun and honest skills. and off course the anticlimax when it worked... as if it's over. exploration in all directions dimension is never over. it is evolution

[@yuppiehi]

What are the value of the diodes? I never picked up on that. Just coincidence... Last week, I visited my neighborhood electronics shop. In one of the aisles, there was a cardboard box filled with old components. I bought the entire box after bargaining with the owner - $10 for the contents (he actually wanted to keep the box, so we threw the contents into another box- weird!) Anyway, I actually bought the batch because there were a bunch of TIL306 modules, and they alone are worth $$$$. When I got home and started to sort the box out, I came across 48 unopened envelopes of diodes. After some research and the help of EEVBlog, I discovered that the diodes were U.S. government stock, and were produced in 1962 by Raytheon. The diodes are MIL-SPEC. The AGC was being designed/built around that period of time. It would be really cool if the batch of diodes I have were actually used (or rejected?) during the build of the AGC.

[@johnfrancisdoe1563]

yuppiehi Well there's no chance Raytheon would trust that brand when building the AGC.

[@mohammadbuhamad7744]

@@johnfrancisdoe1563 any designs of the originals avaliable ?

[@RufftaMan]

Mike is crazy (in a good way). That was some nice debugging right there.

[@cristiancruz5079]

Big cheers for all the team Marc....its been a long way since the repair of the AGC started on that hotel room, quite fascinating to learn how a computer works in a very rudimentary version from the 60s. I shared the landing video( Part 21)yesterday on my facebook wall to commemorate the 50 anniversary of Apollo 11. Looks like the subscribers has been up and up since i started watching your channel. Congratulations again guys. ;-)

[@malcolmbacchus421]

This has been an absolutely stunning achievement. I knew a lot about the programming of the AGC and the very basics of the electronics and, knowing that much, I am totally in awe with the sheer technical knowhow and expertise that you have all put into this. I know you have day jobs to go back too, but I'm sad in some ways that thus project is coming to an end, it has been compulsive viewing.

[@matthewenderby]

This is so exciting, love the progress updates. A bit of a set back but repairable which is all that matters.

[@FredClausen]

Marc you’re on a roll. I love it!

[@hqqns]

@curiousmarc Have you thought about using the curve tracer on the remaining diodes to see if any are already on the way out? I.e Some preventative maintenance?

[@CuriousMarc]

Of course, we took a trace of all the remaining diodes. They all look good at the moment.

[@hqqns]

@@CuriousMarc That's promising I guess. I hope they stay like that!

[@DEW409]

I'm going to miss this series. So interesting and exciting to watch genius at work on this ground breaking artifact. And, yes I watched Armstrong land on the moon fifty years ago.

[@radiumdude]

This is extremely fascinating... I have not the slightest clue what on earth you’re doing there... but I can’t stop watching you guys 👍

[@randalltufts3321]

Question Marc??? Knowing the amazing expertise of this group, wouldn't it be possible to build your own 21st century block for demonstration purposes at a later date or just bite the bullet later on and do a full repair and unpotting of this module? 50 years on crappy diodes is a good run of luck but more failures are expected I'm positive of that at this point. As this AGC computer goes on tour and is transported, handled, temp. Changes etc. It Will undoubtedly plague this antique electronic module. This is a fantastic example of what American technology can produce when we all are of one mind and one purpose. This AGC was literally 50 years ahead of it's time. Congratulations to you gentlemen for a wonderful job. You have given tears of pure joy and childish excitement to this old man's heart. Seeing young men like you re-engineering and bringing back to life a piece of history this important to mankind shows me we are in good hands into the next century and beyond. Much appreciated 👍 PS: Any chance the luminescent dsky displays will be available for sale? I would commit to purchase immediately! Thx Marc

[@ssnoc]

This series was so interesting and enjoyable to watch ... I plan on visiting the Cradle of Aviation museum to see the DSKY .... Excellent work 👍

[@HarryStar56]

I just saw that this computer will be at the museum near me in Bethpage, NY where the LEM was built. I've been glued to this series and now have found out that it's going to be up here in the next two days.

[@eddievhfan1984]

OOH, they're breaking out Orbiter! So excite!!!!!

[@mikaras]

That is some serious brainpower gathered in the same room. :-) I really enjoy this series. Thanks!

[@timcurran7841]

Curious Marc and the AGC. The KZbin gift that keeps on giving! Thanks again guys! I’m with the others, I think the Mike’s T-Shirt business would really take off! Love the teaser for the next episode.

[@Pants4096]

I would so love to hang out with any of you guys for a day! The knowledge and love of the technology is infectious. I'm truly in awe at what Draper and NASA accomplished when they did it. They were so far past the state of the art! I don't know why that feels so much more amazing to me than the smartphone I'm holding in my hand, which is literally a supercomputer by previous standards. Their testing for component reliability astounds me, and as far as we know there were no electronic equipment failures in the whole program, correct? What mission was it that had the intermittent switch contact, that Draper was able to hot-fix in-flight? The sheer skill in the whole endeavor is just something to behold.

[@alexandrebustico9691]

That third diode failing is a benediction : it gives us the opportunity to get one more exciting episode :-)

[@mogwopjr]

I thought that calling someone "rocket surgeon" was a play on words... Until we got to see real Rocket Surgeons. I believe all of you have now earned this title.

[@XMarkxyz]

You've done a great work in all this episodes! I'm looking forward to see this module rejoining the rest of the agc

[@beakytwitch7905]

I love that - with his mind our young hero climbs right inside the Apollo Computer ! :-)

[@TheFrozenSoldier]

The cliffhangers are killing me! Way to get me hooked Marc!

[@jamesdenney9653]

Would have been nice if you could have made use of Ben to come up with a chemical method to remove the potting compound.

[@CuriousMarc]

We did ask him. He was not sure if something existed that would remove only the potting without damaging other stuff.

[@ohger1]

@@CuriousMarc It doesn't appear that the potting compound actually encapsulated the board, only covering the top with an air gap unless I'm seeing this wrong. If that's the case, you can thin cut the perimeter and pop the encapsulation off in one shot.

[@compwiz101]

@@ohger1 they mill off the top layer, then Marc spends half a day with dental picks prying the (slightly softer) inner material out.

[@ohger1]

@@compwiz101 Thanks. I've dealt with all kinds of potting compound, but never saw one that had a hard outer shell and soft inner core.

[@jbloodwo]

John-Del it is fully potted. In one of the other videos you see arc scraping at the stuff with a dental pic

[@tycoinreno]

I am continuously astonished how similar that schematic looks to a modern one (stylistically, not necessarily design). Remarkable, for the 1960s.

[@ericyt7589]

so 3 dead diodes, and the rest are iffy enough the original vendor got flat-out banned from selling to NASA. heres hoping no more die on you!

[@simontay4851]

I think more will fail over time. See my comment above.

[@Neofito89]

Didn't know they got banned Edit: and this is what happen when you comment before finishing the video.

[@SidneyCritic]

And that bad diodes were replaced with more bad diodes. lol

[@jishcatg]

I wonder, are these same suspect diodes used in any other modules besides B11?

[@rwdplz1]

Yeah I'd be replacing them all...

[@hinz1]

Now, CuriousMarc is stuck on the moon with a failed AGC ;-)

[@SweetBearCub]

Thankfully, the real LMs had an independently designed and built Abort Guidance System that would get them back to the CSM, it could substitute for the Apollo Guidance Computer during any phase (technically), except for the Lunar Landing phase. It was somewhat less accurate, but it was good enough. It also used less electrical power, which was a benefit to Apollo 13.

[@BigDaddy_MRI]

Great video Marc!! You know un-potting the whole brick would be a good idea. You know that more diodes will fail over time. Seems NASA had the right idea eliminating that vendor. Landing on the moon and having a stuck/restarting/unresponsive AGC would be disaster. Excellent job to your whole team!!

[@SweetBearCub]

Thankfully, the real LMs had an independently designed and built Abort Guidance System that would get them back to the CSM, it could substitute for the Apollo Guidance Computer during any phase (technically), except for the Lunar Landing phase. It was somewhat less accurate, but it was good enough. It also used less electrical power, which was a benefit to Apollo 13.

[@dr.strangelove5622]

DIODE DEATH RECORDED ON CAMERA!! THE ZOMBIE DIODE!! 😂😂😂 Even though the diode blew out, which is a sad thing, I still love the fact that you guys are uploading the video. You are not screening the things which happen. I loved Mike's explaination. I love datasheets and as I am a newbie to electronic engineering, I fumble and tumble all over the Schmitt trigger and AVR datasheets. But I learn more from them and experimentation rather than any book. What I like about your videos is that even though I am just starting in this field, I can relate to you guys. One thing I learn from you guys is that you have lots of patience!! So like I always say below every video of yours, Excellent work team!!

[@danielson9579]

They removed the vendor.. I predict a few more diode troubles on the horizon 😕

[@alakani]

@@bobroberts2371 Yes, eventually. But also no.

[@littlejason99]

Marc - "Maybe we should just replace all diodes.... How many are in that thing?" Mike - "There's 28 in the current switch module..." Marc - "Ooouch... I don't think I'm up to it..."

[@ScienceANDesign]

I’m waiting for the 50k subscribers special upload..!!😎 great work Mike and Marc..!👏👏👏

[@beansontoast9323]

I love yalls effort and amazkng drive into this wonderful peaice of tech. I LOVE YALL

[@ohger1]

I've NEVER seen potting that didn't soak down and totally encapsulate every bit of surface. It's almost like the module was turned upside down and lowered into a potting tray so the potting mixture wouldn't coat the whole board - but only provide an airtight cover. Seems like you could Dremel a fine cut around the perimeter of this module and pop off the potting in one shot.

[@hamishgrove7722]

Every time you take the milling machine to that module i just want to run out of the room screaming. It looks like you are hammering in a tiny nail with a sledgehammer.😆

[@Pants4096]

Not when you're accurate to a thousandth of an inch, though! A well-run well-calibrated cnc mill in his skilled hands could almost do eye surgery.

[@PapasDino]

I gotta figure out how to give these guys more than one thumbs up! Great job...look forward to the next episode!

[@MostlyPennyCat]

Just build a new B12. Those diodes weren't up to snuff, just build a new B12.

[@alakani]

Great now all I'm going to remember is that diodes are an important part of a balanced breakfast

[@simontay4851]

you should mill out/excavate the potting at ALL the diode locations then replace them all with modern new equivalents (or better) in one go and be done with it. long term it will be less frustrating and take less time than having to keep taking it back to the milling machine every time another diode fails (and they will keep failing over time). Just get on with it and do the whole lot. Strange that they fail open. Ive only seen diodes that failed short.

[@simontay4851]

Who was/is the manufacturer of these faulty diodes that fail open so i know not to use them myself.

[@compwiz101]

@@simontay4851 in fairness, a lot of improvements have been made to semiconductor manufacturing since then - back then it was brand new technology, after all.

[@FrozenHaxor]

@@simontay4851 That comment is completely unfair, these diodes were made over 50 years ago where silicon technology was a totally new thing. It would be naive to think that they would still push on people the early faulty designs knowing what we know now about the technology involved, especially something so trivial as a simple diode.

[@1944GPW]

Marc's curve tracer is a beautiful piece of equipment for diodes, transistors and other components. There is a cheaper device called the Huntron Tracker that does a similar job but still these units cost quite a lot on eBay. Fortunately, if you have a dual trace oscilloscope you can build your own rather-less-fancy version with probably what you have in your junk spares box. It's called 'The Octopus curve tracer circuit', just search it up there plenty of DIY examples on the net eg. www.jammarcade.net/simple-component-tester-a-k-a-octopus-curve-tracer/

[@leisergeist]

That looks brilliant, thanks for sharing. I wonder how well it would work on a digital scope.. I guess the output would just look more messy

[@alakani]

Great tip! You can also use a function generator instead of the transformer for added versatility

[@1944GPW]

@@leisergeist Further to my comment above, the one I built works just fine with my generic Golden Dragon 100MHz digital scope. I fitted it into a plastic electrical junction box I had lying around and put Dymo-labelled banana sockets on the top, just like in some of the examples.

[@leisergeist]

​@@1944GPW Great, thanks for the info

[@72polara]

You guys are getting really good at this! Too bad you have to get this good....

[@AmauryJacquot]

in a couple weeks, all the diodes will have been replaced ;-)

[@gtb81.]

Amaury Jacquot I hope he makes a vid about it too

[@PWingert1966]

This is so much more fun the coloring for relaxation!

[@Gadyuzkha]

You know.. its a really good thing these modules didn't fail while in the LEM during the mission. Can you imagine NASA trying to walk the astronauts through this repair while in space?

[@JanBruunAndersen]

Mike is officially a genius.

[@gavincurtis]

The diode is becomming self aware.and attempting to communicate.

[@CursedLcie]

Even though I'm not worthy I still really want to know where Mike got that Apollo guidance Navigation and Control tshirt haha. Its awesome!

[@1944GPW]

It's been mentioned here that Mike made it himself - presumably supplied the image to a print-your-design t-shirt place?. That logo and text appears on the covers of some of the MIT handbooks which can be found on archive.org eg. Don Eyles AGC manual that Mike himself scanned archive.org/details/agc_handbook_jp2 Well that's my guess on how his great t-shirt came about.

[@CursedLcie]

Ah ok thanks! Ya i can see that now. You're probably right. it would be cool if Mike can and Marc can make some of those Tshirts for his shop but I'm sure there are some possible NASA/MIT trademark/copyright issues with that hence the custom shirt.

[@kooziesan]

Solid work, chaps. Love the videos

[@htroberts]

I was going to ask if you had considered building a new B11. The mechanical construction seems straightforward (if obviously a pain in the ass), and the electronic components are probably available. Then I realized the largest challenge would probably be the transformers. Anyway, really enjoy the series.

[@Megabobster]

How authentic is the blue color of the replica EL DSKY display Ben made? I tried looking it up but I wasn't able to find anything conclusive from my brief investigation.

[@CuriousMarc]

The wavelength of the DSKY display is specified, and it’s more green than it is blue. Unfortunately Ben could not fond any EL paint of that shade.

[@ElvisStansvik]

Hm, 28 diods in that thing, and three down now (?). I guess we have at least 25 more episodes to look forward to! \o/

[@Captain_Char]

I wonder how many diodes failed on NASA during the Apollo program, if it was enough for NASA to remove this particular ones from their vendors and parts list, it must have had a serious failure rate

[@rkan2]

Well if they are failing running within their normal specs...

[@BradyBegeman]

I’m sure many, but then again they weren’t using 50 year old diodes.

[@JonathanStray]

What is that "5451C Fourier Analyzer" in the background at 3:50? Seems like a crazy piece of vintage hardware!

[@analogdesigner-Jay]

Nice work! At least you guys had a better day in the lab than I did...

[@SuzuranMajere]

I SAW MY WORK!

[@ZeroSpawn]

You guys are magical! 🔮

[@Godzilla032]

Great work. Mike your a legend!

[@matteofalduto766]

The Luis Rossmann of Lunar Modules

[@Turboy65]

Due to component ageing and the qualities of late 60s diodes, it might be best to replace all the diodes of that type. I know that's a lot of work but there's a good chance that all of the original ones are going to fail before many more hours of operation.

[@jeffreyplum5259]

it seems there was just to much "Die" in that type of diode. Good Luck guys.

[@JohnRineyIII]

I watched a video today with Don Eyles walking through some Luminary source listings. I noticed a couple fixed memory modules on his shelf; any thoughts on getting in contact with him about dumping those? I understand he's been generous with providing listings for transcription.

[@patemblen3644]

You guys are amazing.

[@mariodistefano2973]

Oh, it was so pity to see that diode died right under the test equipment ... At This point, I would suggest to remove all that nasty Black resin... Just in case (unlikely!) of another faulty component... A chíll in my spine: Just imagine that diode go faulty while landing on the moon! TERRIFIC!

[@leisergeist]

iirc it took him an entire day of carpal tunnel inducing precision chiseling work just to get to each of the previous squares Marc has some seriously admirable patience, but doing the entire thing would probably be way too much lol My hands ached just watching him do it!

[@electronash]

"Swiss Cheese Module B11" lol

[@MockBrainworthy]

Just showed up to this thread. Have you tried Decap for the potted modules? Used to do destructive physical analysis on potted ICs and modules in the late 80's. Put it in a bath and the potting material dissolves.

[@frac]

That could very well be a recorded first. That clearly was a failure shown as it happened. And that curve tracer paid for itself in an instant.

[@MostlyPennyCat]

To be fair, it's practically 50s tech, 60 years later. I'm surprised any of that gear had survived.

[@pmcgee003]

X ... t ... 3 ... yeah, that guy doesn't know *what* he's talking about. :D :D Excellent work.

[@mymessylab]

It seems that AGC wont to remain with you guy😬(at least a module). Of course.. nobody in the world will take care of it better than you 🤗

[@mechmania8450]

Diode high forward resistance then open, must have worried the designers. What did they do for cosmic ray screening ?

[@jdholbrook33]

Houston???? I hope we get to see it while it's here.

[@FrozenHaxor]

They were damn right banning that diode vendor lol

[@jtveg]

That kid is a genius!

[@MatthijsvanDuin]

Maybe the diode returned from "open" to "somewhat working but iffy" and went back to "open" under test-load due to thermal contraction/expansion?

[@ohger1]

Intermittent semis are not uncommon. Pure speculation but it seems to me it's a bonding issue between the semiconductor and the leads in a diode, and possibly contact issues within layers on transistors.

[@wtmayhew]

One of my speculations a couple of episodes back was the diodes were exposed to mechanical shock when this particular B11 module was assembled. Using diagonal cutters instead of a scissors type cutter to trim the excess lead from the diodes can subject the bond wires and semiconductor to many Gs of shock, leading to a latent failure. My theory seems a bit less plausible knowing the scrupulous conditions under which the modules would have been manufactured - and the banning of the vendor puts the blame squarely on the parts. Other factors also could have led to the problems in this specific module - for example, this module may have been used for temperature cycling tests, mechanical shock tests or electrical stress tests. Whenever I work on equipment with small semiconductors, especially glass encapsulated diodes, I trim leads using a scissors type cutter I found at Electronic Surplus in Cleveland, Ohio in the late 1970s.

[@switchthechannel6317]

Wait, they're going to plug it into a GUI?

[@SuzuranMajere]

That's Project Apollo / NASSP for Orbiter. We implement the yaAGC emulator inside the Orbiter spaceflight engine, and then simulate all the rest of the ship around it. I am the maintainer-pro-tem and wrote the special joystick driver that lets the user have two joysticks for simultaneous translation and rotation controls (the way real spacecraft work).

[@indianajones91]

@Devon Casper You'll probably see it fully explained in the next video. Mike connected his gate exact FPGA AGC (which has been shown a few times in this video series) to NASSP, instead of the Virtual AGC emulator. The simulation already talks to the Virtual AGC like if it was a real AGC, setting the right input signals, processing the output signals from the AGC etc. So that made it feasible to do this and probably also to connect it to the real AGC.

[@jms019]

Worth asking whether something in your environment pops diodes

[@pl2247]

Can you do with the simulator the famous error 1202?

[@CuriousMarc]

Mike can do anything... Coming up next.

[@pl2247]

@@CuriousMarc Mythical

[@simonmorris3964]

Don't see any servicemen having to change diodes in voyager 1 or 2 ......from one reliability extreme to the other 8 years later.

[@SweetBearCub]

Voyager 1and 2 were also built a bit more than 10 years later, with more refined electronics. Having said that, both V1 and V2 have had to have instruments shut down due to degraded performance, and the RTG power source on both is steadily providing less and less electrical power.

[@Pants4096]

Well, as they discussed, there's a reason this module never flew. They _knew_ it was defective.

[@SomeGuyInSandy]

How hard would it be to recreate an new Current Switch Module from scratch? You have an example and the schematics...

[@gu4xinim]

Could this module be replaced? Or does it have any special component on it? Would be nice to preserve its state.

[@ReneSchickbauer]

There were only a few dozen of these made at the most. So the chance of finding an original replacement part are practically zero. Yes, you could build a new one, Mike has the whole circuit plan - but the goal is to save as much as possible of the original hardware.

[@htroberts]

I know some Apollo electronic component connections were welded instead of being soldered, primarily to save the weight of the solder. Were these internals soldered originally (I think you have a non-flight article, so it may have been soldered even if flight examples were welded)?

[@zachz96]

If a modern CPU dies, you put the whole thing in the trash. If a 1960s CPU dies, you just have to replace some diodes and transistors to make it work again.

[@gordonwelcher9598]

The 1N914 diodes require a strain relief loop when used in cordwood construction. Maybe you should advise NASA quickly before they build the next batch of guidance computers.

[@heredownunder]

I concur!

[@DAVIDGREGORYKERR]

What about dunking the B11 module in a metal container of xylene solvent and be shot of the potting compound for good and encapsulate the Module in Silicon Rubber.

[@grant2053]

9th!! Love this guys!! Hope you got my email Marc! Please keep it up! It gets better every time!!

[@chrisg6597]

If you can measure the diodes (or lack of a working diode) across the pins of the module, then couldn't you add a new diode across the socket on the AGC?

[@ohger1]

Edit: No, according to a comment from Marc on a similar question, the diodes are in series with an inductor, so while the diode can be read and even curve traced externally, adding a diode along the terminal strips would put it out of circuit with the inductor.

[@steve1978ger]

www.ibiblio.org/apollo/klabs/history/history_docs/r713.pdf - E.C. Hall, "Reliability History of the Apollo Guidance Computer", 1972; p.23: "... intermittent short circuits caused by loose conducting particles entrapped within the package" - you're re-living AGC history here :D - it also says they standardized to use only one type of switching diode, but there are 3300 of them in an AGC (p. 42 although that refers to Block 2 I think).

[@SuperAWaC]

you might want to just unpot the whole module at this point- provided it's polyurethane i have had success with N-methyl pyrolidone (NMP) in a vacuum chamber

[@avcomth]

The diodes probably failed due to its aging and impure Germanium elements. Remember, early 1950s-60s era fabrication processes and techniques were quite crude and unsophisticated which resulted in impurities that degrade Germanium and causes the semiconductivity to fail over time.

[@CuriousMarc]

These were silicon. Being unstable in resistance and failing suddenly open like this, I’d think this is a packaging failure or stress fracture. But just like you, this is speculation, which is not a valid way to explain a failure.

[@TarakuT]

Replace them all.

[@boherrmannsen8219]

not sure if it would be possible, but 3d print a frame and build up a complete new module and insert that in to the metal frame and then pot it again make the 3d printed frame so that the potting can be removed again without risking anything

[@johnfrancisdoe1563]

Bo Herrmannsen Or just don't pot it. Most other modules in this AGC are unpotted.

[@bar04z]

Have tour dates been announced?

[@bar04z]

Found the blog post rescue1130.blogspot.com/2019/07/final-preparations-for-public-display.html