Apollo computers

[Kiwi]

An interesting aspect of the Apollo computers is the low-tech way that figures were read up to the crew by the Capcom and written down on a piece of paper. The Apollo 15 Flight Journal records the reasons as follows. I have included dialogue which shows what was done, but have removed some technical information. The explanation follows GET 001:40:29 –- see the journal for the full text:

history.nasa.gov/ap15fj/02earth_orbit_tli.htm

With thanks to AGN Fuel who first posted this information at the Bad Astronomy Bulletin Board in June 2004.



001:31:06 Irwin: Okay, Gordo. I'm ready for the TLI [means TLI plus 90] PAD.

[When all three crewmen are occupying the Command Module, it is usually Jim Irwin who takes the task of being the secretary and writing down the data read up from Earth.]

001:31:09 Fullerton: Okay. This is the TLI plus 90 abort PAD, Jim.

001:31:14 Irwin: Okay.

001:31:16 Fullerton: TLI plus 90, SPS/G&N; Noun 47 is 66938; minus 0.52, plus 1.90; GET for ignition is 004:19:56.99; Noun 81, minus 0425.4, plus 0000.1, plus 4921.7; attitude, 180, 166, 002; HA is N/A [not applicable], HP, plus 0021.0; 4940.1, 6:34, 4920.8. Sextant star is 40, 079.5, 35.9; boresight star, N/A; Noun 61, plus 16.04, minus 030.00; 1099.0, 34492; GET for 05g, 017:43:58; GDC align stars are Deneb and Vega; 112, 128, 356; no ullage. Go ahead.

001:34:15 Fullerton: 15, Houston. [No answer.]

001:34:29 Fullerton: Apollo 15, Houston. Over. [No answer.]

001:34:46 Fullerton: Apollo 15, Houston. We're not reading you. Over.

001:34:50 Scott: Okay, Houston; there is a breakup - breakup in the VHF, and Jim lost some of the first part of the transmission. Could you go through it again?

001:34:58 Fullerton: Okay, just what do you need?

001:35:00 Irwin: We have Noun 47 through roll, pitch and yaw.

001:35:04 Fullerton: Okay, Jim. Noun 47 is 66938; minus 0.52, plus 1.90; Noun 33, 004:19:56.99; Noun 81, minus 0425.4, plus 0000.1, minus 4921.7 - correction on Delta-Vz is a plus 4921.7; roll, pitch, and yaw are 180, 166, 002. Go ahead.

001:36:10 Irwin: Okay; readback: TLI plus 90, SPS/G&N; 66938; minus 0.52, plus 1.90; 004:19:56.99; minus 0425.4, plus 0000.1, plus 4921.7; 180, 166, 002; N/A, plus 0021.0; 4940.1, 6:34, 4920.8; 40, 079.5, 35.9; N/A, plus 16.04, minus 030.00; 1099.0, 34492; 017:43:58, Deneb and Vega; 112, 128, 356, and no ullage.

001:37:13 Fullerton: Okay, Jim, your readback's correct. Lift-off plus 8 PAD is 008:00, 6076, minus 175, 027:06. Go ahead.

001:37:38 Irwin: Roger. 008:00, 6076, minus 175, and 027:06.

[The Lift-off + 8 PAD carries data for P37, a program in the computer to perform a return to Earth. The program requires 4 values, the desired time for ignition of the engine, the desired change in velocity (or Delta-V) and the longitude and GET for splashdown. The other data required for the burn (attitude angles, etc.), have already been read up as part of the "TLI plus 90" PAD. The CSM G&C Checklist provides spaces for the P37 data on pages 4-23 and 4-24.]

001:37:51 Fullerton: Roger. And I will have the TLI PAD when you're ready for that one.

001:38:03 Irwin: All right. Go ahead. I'm ready for the TLI PAD.

001:38:07 Fullerton: Okay. Time base 6 predict 2:40:23; attitude for TLI, 180, 045, 001; burn time 5:55; 10401.1, 35599; SEP attitude, 359, 077, 320; extraction attitude, 301, 257, 040. R2 align, 045.0, 038.0; ORDEAL start, 56:45; Yaw 001; ejection time 4:16:00. Go ahead.

001:39:35 Irwin: Roger. TLI readback: 2:40:23; 180, 045, 001; 5:55; 10401.1, 35599; 359, 077, 320; 301, 257, 040; 045.0 [garble] 56:45; 001; and extraction at 4:16:00.

001:40:15 Fullerton: Okay. A little static on two readbacks. [Could you] read back Yaw for extraction and R2 align, please?

001:40:24 Irwin: Roger. 040 and 045.0.

001:40:29 Fullerton: Okay. Readback's correct, Jim. [Long pause.]

[Within their high-tech environment, it may seem somewhat strange for the crew to be given large quantities of mostly numerical information in such a low-tech fashion. At first glance it may appear easier to simply have the data uplinked to them and stored in the computer. However, the computer was not designed as a repository of data in the sense that we have come to think of computers thirty years after Apollo. It functions more like a real-time controller, albeit a very sophisticated one, and not completely unlike the embedded controller chips found in a VCR or microwave oven. The abort PADs are, in essence, a 'checklist' of items that the crew have to sequence through (Program 30 can be quite long), and although there are minimum keystroke ("minkey") options, there was never a 'scripting capability' that would automatically execute a program using stored responses. Additionally, there are verbal comments included in the PADs which cannot be entered into the computer.]

[In light of later, post-Apollo computer systems, it was an incredible feat to get the programming into the CM computer's 32Kwords of storage; most of this being hardwired into rope core memory. There was only 2Kwords (4K bytes) of erasable storage in the machine, and this was used to the maximum. During the Apollo 11 landings, using the very similar LM computer, the resource that the 1201/1202 alarms were complaining about was the lack of erasable memory.]

[The crucial importance of the data requires that the crew write it down and have 'hard-copy' available to them in case of the very systems failure that might invoke such an abort. Say, for instance, that the guidance computer fails. Having the abort PADs stored electronically would make them inaccessible. Or, say an oxygen tank blows on the way to the Moon, and you have to power down the entire Command Module, computer and all à la Apollo 13. It's tough to beat having a piece of paper with all the vital information for getting home written on it.]

[It's important to realize that although the computer is a critical part of the spacecraft, it isn't an absolute requirement for its operation. Early in the development of the computer, there were even serious doubts that it would remain functional for the entire mission! As a result, Apollo was designed to be flown without an operational computer. All the tasks that it normally manages could be done manually. (Making attitude adjustments, firing the engine, etc.) An essential design philosophy: Always try to have survivable options even when a critical piece of equipment fails.]

[Scott, from 1998 correspondence - "The design philosophy was even more precise than 'survivable' options - survivable' being exactly what? The back-up system was usually of a completely different design, never two - prime and backup - of the same 'kind.' This was one of the major factors in 'What Made Apollo a Success?' Operating, maintaining, and learning two completely different systems for one purpose was far more difficult and costly than having two identical systems for redundancy - but the concept proved its worth, time and again."]

001:41:00 Fullerton: 15, Houston. We have a question regarding the primary and secondary isolation valve. Over.


For more information about the computers, see:
The Apollo On-board Computers, by Phill Parker
history.nasa.gov/afj/compessay.htm


<Edited to add:>
There is more information about the computers in this thread:
apollohoax.proboards21.com/index.cgi?board=theories&action=display&thread=1129996341

[gwiz]

kiwi said:

It's important to realize that although the computer is a critical part of the spacecraft, it isn't an absolute requirement for its operation. Early in the development of the computer, there were even serious doubts that it would remain functional for the entire mission! As a result, Apollo was designed to be flown without an operational computer. All the tasks that it normally manages could be done manually. (Making attitude adjustments, firing the engine, etc.) An essential design philosophy: Always try to have survivable options even when a critical piece of equipment fails.

That paragraph is all you need to completely rebut the "computers of the day weren't adequate" arguments of the HBs.

[JayUtah]

I tend to agree because I speak from experience in system design. Not everyone has that perspective, and when we tell them that part of the miracle of Apollo was our advancement in computing power -- especially the on-board computer -- they believe us. Everything that I design to be controlled by a microcontroller has also a set of manual controls, both electrical and mechanical. If these systems had to be operated manually it would be tedious and ineffective. But the system will not fail if the computer fails.

Dispassionate people generally fall into two categories on points like this. Some people simply want the assurance of an expert. If you tell them -- truthfully -- that the Apollo spacecraft would not fly off into oblivion without its computer, then they're satisfied. The other camp wants to learn about the underlying principles and practices. It's not that they don't trust the experts; they're just owners of a brain that doesn't like to defer all understanding to other people. There's nothing wrong with wanting to know why.

[Retrograde]

Well there's no way the computers of 1903 were capable of flying an airplane...

[Data Cable]

kos said:

Well there's no way the computers of 1903 were capable of flying an airplane...

Of course they were, all they'd need is a little training.

[Sticks]

kos said:

Well there's no way the computers of 1903 were capable of flying an airplane...

I remember when they built a replica of the Wright Brother's plane, they could not get it to fly , so obviously that first flight was faked and all flights have been faked

[Retrograde]

sticks said:

I remember when they built a replica of the Wright Brother's plane, they could not get it to fly , so obviously that first flight was faked and all flights have been faked

Which also disproves this theory about hijacked airplanes on 11.09.2001...

[nomuse]

On the other hand, when they built a replica of DaVinci's glider, it did fly....proving that the historical use of aerial warfare by the Medici's has been covered up by cowardly historians!

[Data Cable]

Wow, I'm surprised nobody jumped on me (facesiously or otherwise) for using the obsolete definition of "computer."

[colinr]

Perhaps that's because in the days before mainframes were capable of modelling the stress on an aircraft's structure aerospace companies had rooms full of "computers" , people who computed..

The idea being that the problem was broken down into small discrete areas, that were worked on indivdually , the final results being intergrated later.

I've got some books at home showing Lockheed's "computer room" in the '40's roughly football field sized hanger full of men and woman woking away on old adding machines .. No adware/Spyware in those days!

Human computers - no investment needed to create a new one , just some unskilled labour, and time! ;D

[Fnord Fred]

colinr said:

Perhaps that's because in the days before mainframes were capable of modelling the stress on an aircraft's structure aerospace companies had rooms full of "computers" , people who computed..

The idea being that the problem was broken down into small discrete areas, that were worked on indivdually , the final results being intergrated later.

I've got some books at home showing Lockheed's "computer room" in the '40's roughly football field sized hanger full of men and woman woking away on old adding machines .. No adware/Spyware in those days!

Human computers - no investment needed to create a new one , just some unskilled labour, and time! ;D

No investment? I think my parents might have some problems with that statement.

[Retrograde]

fnordfred said:

No investment? I think my parents might have some problems with that statement.

Also, the mechanical computers don't form unions and go on strike...

[colinr]

Sorry badly phrased - plus being in the UK, and going to college in the 70's everything was paid for me - I even got a grant to cover living expenses - those were the days!

[Sticks]

colinr said:

Sorry badly phrased - plus being in the UK, and going to college in the 70's everything was paid for me - I even got a grant to cover living expenses - those were the days!

Cue DeVorjak's New World Symphony

[Data Cable]

kos said:

Also, the mechanical computers don't form unions and go on strike...

Not yet...