APOLLO GUIDANCE COMPUTER

AGC Restoration Videos and Press Coverage:

Printed Press, Web Articles and Podcast Coverage

Wall Street Journal Article

Livingspace.com article

Samtec Blog Articles

Ken Shirriff Blog Articles

Purdue Space Podcast

An almost 2 hour (!) in depth podcast with the whole AGC restoration team.

https://open.spotify.com/episode/2eMAWWn5f9rpWx56BNUn4B

Detailed Material Related to the Restoration Episodes

The Ron Burkley's and Mike Stewart's Virtual AGC Website is the main reference for all things AGC. Everything you ever wanted to know about the AGC, and way more, can be found on the Virtual AGC Website. That's where Ron and Mike put all the material they've recovered and scanned from many, many sources over many, many years. It's truly an Herculean effort of a site:

https://www.ibiblio.org/apollo/

Below are some selected references that pertain to particular videos.

Episode 1-4: AGC Schematics

Restoring the AGC

We embark on the restoration of the computer that put man on the Moon.

Here are several references to the schematics of our AGC

Download a .zip archive of all the schematics:

agc_schmeatics.zip

Scan of Original Handbook with AGC Block II Schematics

agc_handbook_jp2.pdf

Or browse individual pages here

Top Level Schematics:

Logic Level Schematics

Nicely organized table of the schematics, original from: https://www.ibiblio.org/apollo/ElectroMechanical.html

Episode 7: Erasable Core Memory

Memory Module B12 Video

This video explains what's wrong with our memory module.

Core Memory Explained

This video explains how core memory works, which is helpful to understand the previous Module B12 video.

For those who want to try their luck at spotting the break in our inhibit wire, here are the memory module XRay files

For those who volunteered to re-weave a core memory module, we offer the following documentation as a form of encouragement:

Here is some more information for our intrepid core memory weavers:

The wide temperature lithium cores spec.

The 4 wire weave pattern. Wire gauge is 38.

A partial view of how two core planes address and sense/inhibit wires are connected. The AGC memory is a one giant memory mat, so the wires have to be continuous across the  whole 16 bit plane mat.

The complete view of the 32k core mat.

The folding pattern.

And some of the drive electronics you'll need to replicate for testing. Happy weaving!

Episode 8: Malco Connector Drawings

Restoration Update: 

Samtec remakes the Malco connectors, Mike turns the AGC into a Blinkenlight machine.

The connector drawings, known as SCD (Specification Control Drawings) for the Malco MiniWasp connectors, can be downloaded below.

Episode 9: IRIG Gyro Documents

Unboxing my IRIG gyro: 

We look at a few military gyros and then unbox a real Apollo IRIG gyro

Episode 10: Mike Stewart AGC Emulation Hardware and Software

Mike Flies P63, Lunar Landing

Mike Stewart has developed gate accurate FPGA replicas of the AGC and its test hardware. It represents a massive amount of work. If you are Mike, you can use it to fly the real missions, and simulate every signal in the AGC. Making this work is not for the faint or heart and requires intimate understanding of the AGC hardware and software. We offer no instructions and no support.

Mike Stewart's AGC FPGA replica source code:

AGC FPGA Test Monitor:

Bonus: AGC Verilog Simulation


Many other people have made physical implementations of the AGC, most notably John Pultorak that demonstrated first that this was possible by building a Block I replica in TTL. Their efforts are nicely summarized here:

Episode 15 and 24: Recovering lost code (Core Rope Modules Dumps)

Reading core rope modules

Thanks to our restored AGC, we can now read core rope modules that contain historically significant but previously lost code. So far we have read 9 core rope modules. Two in episode 15 (recovering Retread 50), and 7 during our East Coast tour in episode 24 (recovering part of Sundance and a complete Sundial-E). The links to the recovered software, some of with its disassembled code, are below.

Archived ropes:

Sundial -E (recovered from ropes at the MIT Museum)

Sundial-E is the Command Module equivalent of Aurora 12 which was written for the LM. It contains the full test suite of the AGC, as well as the tests for all the subsystems of the Command Module (Aurora would contain the same AGC test suite, but would have tests for the LM subsystems, which are very different).  The recovered software ran on our successfully AGC after we dumped it from the original ropes.

https://github.com/virtualagc/virtualagc/tree/master/SundialE

Sundance (recovered from ropes loaned by AGC creator Eldon Hall and AGC programmer Don Eyles)

So far we have been able to recover 3 of the 6 rope modules of Sundance, the Apollo 9 LM program. Apollo 9 was the first test of the LM in space, Earth orbit. The modules we have differ as of their version of Sundance. We are in the process of disassembling the code we recovered. We have located 3 rope modules that would complete the software (which fits in 6 rope modules), and will be trying to get access to them to complete the recovery.

https://github.com/virtualagc/virtualagc/tree/master/Sundance302

Retread 50 (recovered from ropes at the Computer History Museum)

Retread 50 is the first functional program written for the Block II AGC. This was used to test the functionality of the newly developed AGC. Parts of the ropes were damaged but we were able to recover the original content using the parity bit. We ran the repaired software succesfully on our AGC, and yes, it passed all the tests.

https://github.com/virtualagc/virtualagc/tree/master/Retread50

Episode 23: Flying Landings in Orbiter/NASSP

Flying landings on the AGC

Mike Stewart modified the Oribiter/NASSP simulation software to run with the real AGC flying the simulated LM. For the rest of us that does not have an AGC around, the non-modified version runs just as well using a software emulation of the AGC, so anyone should be able to fly like we did. Link below for the installation instructions.

Orbiter/NASSP installation instructions link:

https://www.orbiter-forum.com/showthread.php?t=40351

DSKY Relight

In this episode, Marcel lends Carl a flight spare of a DSKY screen, two DSKY relay modules, and a DSKY power supply module. After a lot of rewiring, we finally get to see the glow of the screen, which is far better than one would expect for 1960's technology.

Manoel Da Silva has made a 3D model of all parts of the DSKY following the original NASA drawings:

https://github.com/virtualagc/virtualagc/tree/mechanical/3D-models

Here is some of the documentation we used for this episode, including the spec drawing of the display and the relay module schematics.

3D CAD models

Viewer Gwyllym Suter is making 3D CAD models of some of the AGC modules from the original NASA drawings. He has already posted the current switch module.

https://grabcad.com/gwyllym.suter-4/models