The Block II Apollo Guidance Computer, or AGC, operated on all crewed Apollo flights from Apollo 7 in 1968 through Apollo 17 in 1972.
The AGC was manufactured at Raytheon, a major US defense contractor, and developed at the Massachusetts Institute of Technology (MIT) Instrumentation Laboratory, founded and directed by Charles Stark Draper.
The AGC’s central processor contained about 5,600 NOR gates. A NOR gate, short for ‘NOT OR,’ is a basic logic circuit that produces a binary “1” only when all its inputs are “0.”
Made of transistors and resistors, these Micrologic NOR circuits powered the computing functions of the Apollo command and lunar modules.
For Block II, engineers used dual three-input NOR integrated circuits.
The AGC’s processor was specifically designed for Apollo’s navigation and control needs. Single-chip microprocessors didn’t become available until 1971, when Intel launched the 4004.
The AGC’s main clock operated at about two million cycles per second. In practice, this lets the computer carry out around 85,000 simple instructions each second.
More complex operations, such as addition, ran slower, about 43,000 per second.
The Block II Apollo Guidance Computers stored their programs in two kinds of memory.
They had about 72 kilobytes of fixed “core rope” memory, which was roughly 36,000 slots, each holding a single instruction or number that could not be changed.
The AGC also contained about 4 kilobytes of erasable magnetic-core memory, with roughly 2,000 slots that astronauts or onboard programs could update during the mission.
Producing core rope memory often took six to eight weeks to complete a single module and cost about $15,000 in the late 1960s.
Each AGC used 36 modules, or about $540,000 in 1969 dollars, which equals roughly $4.75 million today.
Skilled technicians carefully hand-wove delicate copper wires through and around compact, doughnut-shaped ferrite cores.
Threading a wire through a core represented a binary “1,” while bypassing the core signified a “0.”
Data was permanently stored by how wires were woven through or around the cores, resulting in a durable and reliable form of non-volatile read-only memory that remains effective even in harsh space environments.
The AGC was accessed through the Display and Keyboard, or DSKY, pronounced “disk-key.”
The DSKY acted as both the control pad and data display, allowing astronauts to communicate with the AGC.
By entering command codes like Verb 16 Noun 68, they could instantly access range, velocity, and time-to-go, simplifying tasks such as system checks and mission alarm responses.
As director of the software engineering division at MIT’s Instrumentation Laboratory, Margaret Hamilton led the team that developed the pioneering onboard AGC flight software for the Apollo command and lunar modules.
She popularized the term “software engineering” to confirm that software was developed with the same precision as was used with the spacecraft’s hardware.
Hamilton and her team built proactive error detection and dynamic safeguards into the software, inspired in part by an incident in which her young daughter, playing with a DSKY prototype, pressed random key combinations and caused unexpected input commands.
Those safeguards were active during Apollo 11, guiding the Lunar Module Eagle as it descended toward the lunar surface July 20, 1969.
Commander Neil Armstrong and Lunar Module pilot Buzz Aldrin had kept the Eagle’s rendezvous radar system in AUTO mode during lunar descent so it could quickly reacquire the command module Columbia in lunar orbit after landing.
Unbeknownst to them, the AUTO mode triggered “cycle steals,” or processing delays that slowed the AGC’s scheduler and caused program alarms 1201 and 1202.
When Armstrong and Aldrin reported the alarms to Mission Control in Houston, guidance officer Steve Bales and computer specialist Jack Garman, working from their consoles, quickly reviewed the AGC’s behavior.
They confirmed that the AGC continued to manage all essential landing and navigation tasks and that the Eagle’s engine and trajectory were unaffected.
Bales relayed his assessment to flight director Gene Kranz, who authorized capsule communicator Charlie Duke to say, “We’re go on that alarm,” giving Armstrong and Aldrin official clearance to land.
According to NASA’s technical transcript, the first 1202 alarm occurred at 102:38:26 ground elapsed time, counted from Apollo 11’s launch at 9:32 a.m. Eastern time (8:32 a.m. Minnesota time) July 16, 1969.
Excerpts from the NASA transcript of the alarm exchanges between Aldrin, Armstrong, and Duke:
102:38:26 Armstrong: “Program alarm.”
102:38:30 Armstrong: “It’s a 1202.”
102:38:32 Aldrin: “1202.”
102:38:48 Armstrong: “Give us a reading on the 1202 program alarm.”
102:38:53 Duke: “Roger. We got you . . . we’re go on that alarm.”
102:42:17 Aldrin: “Roger. Understand. Go for landing. Three thousand feet. Program alarm.”
102:42:22 Aldrin: “1201.”
102:42:24 Armstrong: “1201.”
102:42:25 Duke: “Roger. 1201 alarm. We’re go. Same type. We’re go.”
During the final descent, Armstrong saw that the Eagle was headed into West Crater and its hazardous boulder field, which lay directly in the path of the preprogrammed landing site.
Using the Attitude Controller Assembly (a grip hand control), he manually flew the Eagle away from the crater while still relying on AGC stabilization data.
The hand controllers for both the Apollo Command and Lunar Modules were designed and manufactured by Honeywell Inc., with final assembly completed at its Aerospace Division facilities in the Twin Cities; yes, from our state of Minnesota.
The Eagle carrying Armstrong and Aldrin touched down safely in the Sea of Tranquility.
Buzz Aldrin later recalled in 2016, “We touched down . . . we probably had about 15 seconds of fuel left.”
Margaret Hamilton, whose work helped land humans on the moon, celebrated her 89th birthday Aug. 17 this year.
