Apollo 11, a Half Century Past

21648738532_5cfa370616_o.jpg
21648738532_5cfa370616_o.jpg
21648739442_7dec5721a8_o.jpg
21648739442_7dec5721a8_o.jpg
21689769522_4ea2e8b4f8_o.jpg
21689769522_4ea2e8b4f8_o.jpg
367292main_GPN-2000-000048_full.jpg
367292main_GPN-2000-000048_full.jpg
48293732417_0c85e1ede5_o.jpg
48293732417_0c85e1ede5_o.jpg
Agc_flatp.jpg
Agc_flatp.jpg
Agc_nor2.jpg
Agc_nor2.jpg
Agc_rope.jpg
Agc_rope.jpg
Agc_view.jpg
Agc_view.jpg
apollo_11_uss_hornet_simex_jul_1969_ha-030.jpg
apollo_11_uss_hornet_simex_jul_1969_ha-030.jpg
Apollo_CSM_lunar_orbit.jpg
Apollo_CSM_lunar_orbit.jpg
Apollo_DSKY_CHM.agr.jpg
Apollo_DSKY_CHM.agr.jpg
Intel_CPU_Core_i7_6700K_Skylake_perspective.jpg
Intel_CPU_Core_i7_6700K_Skylake_perspective.jpg
Apollo 11, a Half Century Past

The Apollo project was a spaceflight program to get people to the moon. The program began in 1961, and followed the Mercury program (six crewed flights from 1961-63) and the Gemini program (ten crewed flights in 1965-66).

The first crewed Apollo flight, Apollo 7, launched in 1968. Apollo 8 and its crew made ten orbits of the moon later the same year. In May, 1969, Apollo 10 flew to within 10 miles of the moon in what was called a "dress rehearsal" of the lunar landing. Two months later, people were walking on the moon.

On July 16, 1969, Apollo 11 was launched.

48293732417_0c85e1ede5_o.jpg

On July 20, 1969, the Lunar Module of Apollo 11 landed on the moon. Six or seven hours later, which ended up being July 21, Neil Armstrong and Buzz Aldrin walked on the moon.

21648739442_7dec5721a8_o.jpg

On its return, the Apollo 11 spacecraft was recovered in the Pacific Ocean by Pryor veteran journalist Michael Wheat, with the assistance of a few hundred others on the USS Hornet. That had to be an incredible experience. Check out Mike's Facebook posts if you have a chance.

apollo_11_uss_hornet_simex_jul_1969_ha-030.jpg

There were six more moon landings in 1969, 1970, 1971, and 1972. There were humans on the moon in those four years, but never before and none since.

At the moon, the Apollo 11 spacecraft consisted of a Command and Service Module, which orbited the moon, and a Lunar Module, which would separate from the Command Module and descend to the moon for landing.

Apollo_CSM_lunar_orbit.jpg
Command and Service Module

21689769522_4ea2e8b4f8_o.jpg
Lunar Module, headed to the moon

The Command and Service Module weighed about 30 tons, and the Lunar Module weighed about 15 tons. Only part of the 15-ton Lunar Module returned from the moon to the Command Module, about 5 ton's worth. (It would be more accurate to call these masses instead of weights, but sometimes that confuses people.)

That is a lot of weight (or mass) to send to the moon. It's no wonder they needed a giant rocket.

367292main_GPN-2000-000048_full.jpg
Saturn V booster

Apollo Guidance Computers (AGC) were used to guide the spacecraft. There were two on the Command Module and one on the Lunar Module.

Agc_view.jpg
Apollo Guidance Computer

These computers were advanced at the time, but simplistic by today's standards. They were equivalent in power to the TRS-80 Model I or the Apple II, early personal computers introduced about 10 years later. Apollo Guidance Computers had about 72 kilobytes of read-only memory, 4 kilobytes of read/write memory, and a clock speed of 2 megahertz.

That's around 1000 times slower with 700,000 times less capacity than my cell phone. The effective speed was even slower than that, because the native instruction set of the AGC was extremely limited.

The memory of the AGC was magnetic core memory, which uses small magnetic "donuts" threaded with wire to make magnetically positive or negative bits.

Agc_rope.jpg
AGC Magnetic Core Memory

Magnetic core memory was used in the Apollo project because semiconductor memory was not yet available. However, magnetic core memory was used into the 1980s on the Space Shuttle because it resists radiation better than early semiconductor memory. In addition, magnetic core memory is non-volatile. In other words, it has the property of retaining its data when the computer is reset or turned off, much like the flash memory used in cameras and cell phones.

The AGC did use integrated circuits for its central processing unit (CPU), although not on the scale of today's chips. The AGC used 2,800 integrated circuit chips for its processor, each of which used three internal transistors to make up two logic gates.

Agc_flatp.jpg
AGC Processor

Agc_nor2.jpg
A Single AGC Logic Chip

By comparison, today's CPUs are typically housed on a single chip containing more than a billion transistors, some as small as 70 silicon atoms in width.

Intel_CPU_Core_i7_6700K_Skylake_perspective.jpg
Intel 6700K CPU

The AGC did not have a keyboard or computer display like we're used to. The entire user interface consisted of a few buttons, a few numbers, and a few lights that would turn on or off.

Apollo_DSKY_CHM.agr.jpg
AGC User Interface

The program for the AGC was written in assembly language, in which each statement corresponded to 1 of 12 rudimentary machine code instructions (compared to an instruction set of over 1,000 in modern CPUs). It was very low-level programming, and required a lot of code to perform simple tasks.

https://en.wikipedia.org/wiki/Apollo_Guidance_Computer

https://www.doneyles.com/LM/listing@304x450.jpg

In spite of the limited memory, speed, and programming language, NASA programmed the AGC so it could perform multiple tasks at once, restart the computer without restarting the program, and handle hardware faults in the memory and data input (up to a point), all with very high reliability. This is impressive. I won't go into detail here, but here's an interesting paper by one of the designers:

https://www.doneyles.com/LM/Tales.html

The AGCs were used to guide the both the Command Module and the Lunar Module. In the case of the Lunar Module, the AGC guided the spacecraft to predefined positions, with a target speed and acceleration (and acceleration rate of change) for each position.

In spite of the "very high reliability" I mentioned above, on the trip down to the moon's surface from the orbiting Command Module, the AGC in the Lunar Module experienced a crash. Five times.

The problem was that the computer was overloaded, and unable to complete some tasks within the allotted time. This made everything unreliable, so the computer would display an alarm code and restart.

Like many problems that are hard to pinpoint and hard to duplicate, where were multiple causes to the computer problem.

1. The radar on the Lunar Module had a separate clock from the computer. The clock were synchronized, but the phases were not. In other words, the frequency of the clocks were precisely the same, but clock signals or waves were not aligned. This caused some errors in the radar position data that was fed to the computer, and made it look like the radar antenna was moving back and forth really fast. While this was not a problem in itself, it used some extra computing time.

2. The radar was left on during the landing so it would be warmed up in case of an abort, when it would be used on the return to the Command Module. This was not a problem in itself, but it required some extra computing time that was not used on previous flights or ground tests.

These two factors "stole" 13% of the available processing cycles from the computer, which would normally be operating at 85% of capacity. When Buzz Aldrin instructed the computer to display delta-H, the difference between computed altitude and radar altitude, updated twice per second, it added an additional 10% to the computer's load.

This put it over the top and caused the computer to reset with an alarm. This is not something you want to see during the first human landing on the moon. Neil Armstrong's heart rate went up from 120 to 150 during this period.

The astronauts asked mission control whether to abort the landing. The guidance control team at Houston had a rough idea of the cause and gave the "go" order (several times), and the landing operation continued.

The situation was resolved when Aldrin changed the autopilot mode from automatic to attitude-hold, reducing the computer load and requiring some manual control of the spacecraft. In hindsight, he could also have changed the twice-per-second display of delta-H to an on-demand display.

Then the Eagle landed.

21648738532_5cfa370616_o.jpg

There had been another potentially serious problem with the throttle sensor on the descent engine. The descent engine performance had been improved for this mission, but some program parameters had not been changed to reflect this. This resulted in widely fluctuating thrust during the Lunar Module's descent.

You can check out the operation of the Apollo Guidance Computer on this online simulator. A web browser doesn't even slow down simulating the 70 lb. Apollo computer.

http://svtsim.com/moonjs/agc.html

Hal Laning, Jr. from MIT is credited with designing the operating system for the Apollo Guidance Computer. Its multitasking and fault tolerance allowed the landing to continue instead of the mission ending in an "abort".

https://en.wikipedia.org/wiki/J._Halcombe_Laning

The Apollo program resulted in a large number of technological spinoffs, such as flame-resistant clothing, freeze-dried foods, and most importantly, the computer game Lunar Lander.

The first version of Lunar Lander was a text-based program, written by a high school student named Jim in 1969. By 1978, hundreds of versions of Lunar Lander had been written, and it was the most popular computer game in existence.

https://www.technologizer.com/2009/07/19/lunar-lander/