jwsmail at jwsss.com
Tue Jan 23 09:33:54 CST 2018
On 1/23/2018 7:24 AM, Daniel Seagraves via cctalk wrote:
>> On Jan 23, 2018, at 9:14 AM, jim stephens via cctalk <cctalk at classiccmp.org> wrote:
>> On 1/23/2018 6:30 AM, Daniel Seagraves via cctalk wrote:
>>> The Saturn software, which is what actually flew from Earth to the moon,
>> The navigation and guidance was in the CM and LM processors. The Saturn IBM firmware is lost, but was under command of the LM and CM computers, and is running on simulators, as well as on some hardware replicas.
> Absolutely wrong. The only time the CM computer flew the Saturn was in an abort scenario where the Saturn digital computer failed, and it happened via a data path from the FDAI needles to the Saturn’s analog control computer. At all other points prior to S4 staging the CM was strictly along for the ride. After S4 staging the CM and LM were on their own, but that was after the translunar burn.
As pointed out, the function of the Saturn computers was to do the lunar
injection. The main guidance problem of the mission was the coast and
the landing, which were done by the CM and LM.
They did what the ground uploaded to them for guidance, as full
independent guidance was decided to be too intensive and unnecessary to
have in the onboard computers. The real work was done in the back rooms
@ Mission control, with certain features implemented on the systems
onboard the rocket.
You couldn't carry out a mission w/o the ground supporting either system
with computations to the onboard systems. You didn't punch in the
address of the moon on any system onboard the rocket, you got
pre-computed parameters from ground computations that the flight
computers carried out.
The AGC systems could vary that more than the Saturn stack could, as you
point out the capability was added at astronaut insistence to control
the Saturn part of the rocket, but was never used.
I'm not getting your "absolutely wrong" part.
>> After I reread the thread, I think they were talking about saving the Apollo computer software, not the spacecraft.
> They were.
>> AGC software here, FWIW.
> I know. I am one of the NASSP maintainers. We took the yaAGC core and built a spacecraft around it so we could actually use it instead of just running it to look at the pretty flashing numbers in the idle loop. The project has been in work for more than 10 years now. Right now we have the most complete Apollo simulation ever built, exceeding the capabilities of even the NASA training simulators. See http://nassp.sourceforge.net/wiki/Main_Page <http://nassp.sourceforge.net/wiki/Main_Page>
More information about the cctalk