Moon landing Hoax?

[Lobreiter]

im not denying or supporting the moon landings, but there are some things that are questionable

like for example, the mass amounts of cosmic radiation experienced would have surely killed the astronauts, unless they in the year 1969 they had somthing that could generate a magnetic feild around them, gamma radiation is the stongest on earth and it takes 4 feet of solid lead to protect you,

and what about the onboard computers that they used, dont you ppl tell me they could land somone on the moon with a computer that has not even half the compteing power of my graphing calculator, because if im correct, i is possible to control a Saturn V rocket, with a TI-83+.

but on the other side, in 1969 there were some prtty smart ppl and u cant really beleave anything on the internet these days, who know the entire hoax thing could be the work of some ****ed off dude in is basement, hell i could creatively edit most of the footage and pictures to make it look fake, its called photoshop.

[Bob B.]

like for example, the mass amounts of cosmic radiation experienced would have surely killed the astronauts...

What mass amount of cosmic radiation?

and what about the onboard computers that they used...

How much computing power is needed to run an autopilot?

[JayUtah]

...that are questionable

"Are questionable," or you just don't know the answers?

...the mass amounts of cosmic radiation experienced...

Exactly how much radiation does it take to kill someone, and exactly how much would someone get by traveling to the Moon, and how do you know?

...gamma radiation is the stongest on earth and it takes 4 feet of solid lead to protect you.

How much gamma radiation is there between the Earth and Moon?

...dont you ppl tell me they could land somone on the moon with a computer that has not even half the compteing power of my graphing calculator...

How much computing power, exactly, is required to land on the Moon? And how do you know?

...i is possible to control a Saturn V rocket, with a TI-83+.

I've seen rockets than can be controlled with a few well-placed fins. Where were those computers? How much computing power does a baseball require to get from the pitcher's mound to home plate? How much computing power does a comet have? How much does it need?

[Swift]

Hi Lobreiter. Welcome to BAUT. I suggest you read the rules and the FAQs (for example, this is a family friendly board and language must be G rated).

There are some great resources for specific moon landing hoax questions.
clavius.org is a great website that answers these specific questions as does the Bad Astronomy website. You can also search around the conspiracy threads right here, we have discussed both of these quite often.

I will give you two quick answers - others can give you a lot more detail. Cosmic rays and charged particles from the sun are a very different form of radiation than gamma radiation, and so the means of stopping them are very different - you don't need all that solid lead.

I don't know how old you are, but a lot of pretty advanced technology can be run without computers. And the computers on Apollo weren't all that bad; remember, they were designed for some very specific tasks, which they were well qualified to perform. A lot of the power of modern computers is used for things like graphical user interfaces and the ability to run a lot of different types of programs, neither of which the Apollo computers needed to do.

On the flip side, the evidence in support of the landings is absolutely overwhelming, and it is not just still and movie films - rocks, core samples, radio transmissions that can be tracked to and from the moon, and tens of thousands of people who were involved in making it all work, none of who have ever come forth to "spill the beans" (there are no beans to spill).

If you have other questions, or need more details, just say the word.

[Bob B.]

- you don't need all that solid lead.

Heck, you don't need all that lead to stop even gamma rays. Outrageous numbers like four feet of lead (or six feet, or whatever the number of the day is) are erroneous.

[JayUtah]

Even if there were gamma rays that needed stopping.

[Van Rijn]

and what about the onboard computers that they used, dont you ppl tell me they could land somone on the moon with a computer that has not even half the compteing power of my graphing calculator, because if im correct, i is possible to control a Saturn V rocket, with a TI-83+.

From http://en.wikipedia.org/wiki/TI-83

The TI-83 Plus is a graphing calculator made by Texas Instruments, designed in 1999 as an upgrade to the TI-83. The TI-83 Plus is one of TI's most popular calculators. It uses a ZiLOG Z80 microprocessor running at 6 MHz, a 96×64 monochrome LCD screen, and 4 AAA batteries as well as backup CR1616 or CR1620 battery. A link port is also built into the calculator. The main improvement over the TI-83, however, is the addition of 512 KiB of Flash ROM, which allows for OS upgrades and applications to be installed. Most of the Flash ROM is used by the OS, with 160 KiB available for user files and applications.
[snip]

Programming may also be done in TI Assembly, made up of Z80 assembly and a collection of TI provided system calls. Assembly programs run much faster, but are more difficult to write.

It also has 32KB of RAM.

It certainly looks to me that, with the appropriate programming and with the proper hardware interface (which might or might not require major surgery depending on the capabilities of the "link port") that it probably could do AGC like functions. Just because you're used to fancy user interfaces that require enormous resources doesn't mean they're required for this application.

[Obviousman]

My old Tandy MC-10 could run a lunar landing simulation programme with a whole 4K of RAM. I can't remember what processor it used.

[Van Rijn]

My old Tandy MC-10 could run a lunar landing simulation programme with a whole 4K of RAM. I can't remember what processor it used.

It had a 6803. I wasn't familiar with the 6800 series, except it was apparently pretty similar to the 6502. The CPU was okay, though I wouldn't want to have an HCF (Halt and Catch Fire) instruction in the code.

It looks like the memory specs on the MC-10 were below that of the AGC, so it would have had problems.

[Peter B]

but on the other side, in 1969 there were some prtty smart ppl

G'day Lobreiter, and welcome to the BAUT Forum.

Yes, the Apollo program employed essentially the best people available in the USA in a range of fields, including engineering and computing.

and u cant really beleave anything on the internet these days

No you can't. But you can cross-check the claims people make. And you can also cross-check claims in books - not everything is on the Internet, you know.

who know the entire hoax thing could be the work of some ****ed off dude in is basement

Well, we're fairly familiar with the backgrounds of most of the leading Hoax Believers, and what motivates them. It seems to be mostly a desire to be noticed. They're not so much interested in resolving issues as keeping them unresolved.

hell i could creatively edit most of the footage and pictures to make it look fake, its called photoshop.

But Photoshop wasn't available in 1969. Photos and video of astronauts on the Moon have been available since then.

In any case, Photoshop won't create 350 kilograms of lunar samples. The only way to collect that amount of samples in 1969 was to send astronauts to the Moon.

[Laguna]

and what about the onboard computers that they used, dont you ppl tell me they could land somone on the moon with a computer that has not even half the compteing power of my graphing calculator, because if im correct, i is possible to control a Saturn V rocket, with a TI-83+.

Guess what...
The AGC was not run on Windows Vista...

[AGN Fuel]

I don't know how old you are, but a lot of pretty advanced technology can be run without computers. And the computers on Apollo weren't all that bad; remember, they were designed for some very specific tasks, which they were well qualified to perform. A lot of the power of modern computers is used for things like graphical user interfaces and the ability to run a lot of different types of programs, neither of which the Apollo computers needed to do.

Added to which is the frequently overlooked fact that a lot of computing grunt work was actually done on the ground and fed up to the spacecraft (leading to one of my favourite Apollo transmission phrases, "POO and accept").

[Serenitude]

It would amaze most modern Microsoft GUI kids that a fully functional web server could run on a hard drive less than 5 megabytes in size, providing full DNS, mail, Samba, etc... services TODAY, on P3 400ish hardware, in an era when familiar bloated OS's won't even think of installing the "barebones" essentials on anything less than 3 Gigs of hard disk space and dual-core processors, but such LAMP systems are entirely feasible. Because one is unfamiliar that the technology is possible, does not therefore make it so.

[Serenitude]

Oops - forget to welcome you to the Forums! And meanwhile, have you checked out the Bad Astronomer's debunking of some of the more enduring Apollo Hoax myths? They're written in a very accessible, layman's style

[Nicolas]

From http://en.wikipedia.org/wiki/TI-83

The TI-83 Plus is a graphing calculator made by Texas Instruments, designed in 1999 as an upgrade to the TI-83. The TI-83 Plus is one of TI's most popular calculators. It uses a ZiLOG Z80 microprocessor running at 6 MHz, a 96×64 monochrome LCD screen, and 4 AAA batteries as well as backup CR1616 or CR1620 battery. A link port is also built into the calculator. The main improvement over the TI-83, however, is the addition of 512 KiB of Flash ROM, which allows for OS upgrades and applications to be installed. Most of the Flash ROM is used by the OS, with 160 KiB available for user files and applications.
[snip]

Programming may also be done in TI Assembly, made up of Z80 assembly and a collection of TI provided system calls. Assembly programs run much faster, but are more difficult to write.

It also has 32KB of RAM.

It certainly looks to me that, with the appropriate programming and with the proper hardware interface (which might or might not require major surgery depending on the capabilities of the "link port") that it probably could do AGC like functions. Just because you're used to fancy user interfaces that require enormous resources doesn't mean they're required for this application.

I didn't know the 83plus (which I own) has 512kb flash. However I did once install the dutch software from CD-rom, so now all menus and functions are in dutch. That must have been an OS upgrade.

I think the TI-83 plus is plenty fast enough for apollo computer functions. But here too, you see that it depends on how well you program it, to name just one thing. Just running something from a few commands will result in slow program execution. I once ran a Mario clone on the Ti-83 plus, which was programmed directly into Assembly (some people must have loads of time). It outperformed the original Gameboy in how fluent it ran.

So if you'd make a very clean program and have the hardware 100% fit for the task, I think 6mhz 32k is just fine for apollo, possibly serious overkill for many applications. Remember that you don't need to calculate the looks of the moon, nor the physics, nor the craft, you just need basic sensor-info and command-servo formulas.

[Laguna]

What did it actually calculate?

[NEOWatcher]

and what about the onboard computers that they used, dont you ppl tell me they could land somone on the moon with a computer that has not even half the compteing power of my graphing calculator, because if im correct, i is possible to control a Saturn V rocket, with a TI-83+.

Controls don't even need computers. Most of the reactions can be hardwired. The computer just allows an easier way to make it adjustable.

Anyway, if a cheap 1K 3.25MHz machine can play chess, then what's wrong with Apollo?

[JayUtah]

This is starting to look rather hit-and-run.

Since we're gravitating toward the computer topic, now would be a good time to mention a couple of things.

First, the AGC is proven to work. That is, modern hobbyists have made emulators and actual working hardware from the original designs and specifications that run the actual flight software. In answer to the question, "How could the AGC have gotten them to the Moon?" several people can point to their work and say, "Just like this." You can't necessarily do the same thing with the F-1 engine (although that would be fun). The AGC is an easily-verified piece of technology, and it has been verified.

Second, NASA in the mid-1960s had no way of knowing that in 20 years computers would take over the planet. That is, they had no way of knowing the degree to which people would come to understand them and rely on them, and each to own several of them. Not many people in the 1960s had a clue how computers worked or what they could and couldn't do. So a conspiracy theory wouldn't have to go into such correct detail in order to be convincing. In fact, usually the more detail you go into in a lie, the easier it is to prove wrong. NASA (well, really MIT) didn't have to design a fully-functional computer just to fool people into thinking they had one. So why did they, if not to actually use it?

[Bob B.]

What did it actually calculate?

JayUtah is the real expert but I’ll give my two bits anyway. I haven’t studied the LM computer very closely therefore I must qualify all my comments with “I think”. I’ll surely be corrected if any of this is incorrect. Hopefully I’m not too far out in left field.

My understanding is that computer performed an autopilot role during landing. Data input came from the inertial measurement unit and the landing radar. The IMU included three accelerometers and three gyroscopes for measuring acceleration and attitude angles in X-Y-Z axes. The landing radar provided altitude, horizontal velocity, and vertical velocity. If I’m correct, that is nine inputs. The computer compared these measurements against a pre-planned trajectory and if any were out of tolerance a command would be sent to the appropriate control to correct it. For instance, if angle X is too high then pulse a designated thruster. If X is too low then pulse a different thruster, etc. If there are nine inputs and two corrective functions for each, then that really isn’t a very demanding a task to perform.

I believe the computer also performed rendezvous computations when returning to the CSM in lunar orbit. The Gemini computer was doing this same thing as early as 1965. Angles and distances to the target would be measured and input into the computer, which would then run a series of computations to determine the transfer orbit required to intercept the target and the necessary engine burns. This again wouldn’t be anything particularly tasking on the computer.

EDIT:
The lunar module also included two additional computers -- the brains of the Commander and Lunar Module Pilot.

[sts60]

You can't necessarily do the same thing with the F-1 engine (although that would be fun).

I've actually fired an F-1 emulator. Faithful in all respects except it was smaller-scale and solid-fueled instead of LH2RP1/LOX. You stick the little nichrome wire into the nozzle, connect the 6V battery, count down from 10, and ...

[Bob B.]

I've actually fired an F-1 emulator. Faithful in all respects except it was smaller-scale and solid-fueled instead of LH2/LOX.

Sounds cool! One nitpick ... that should be RP-1.

[ineluki]

im not denying or supporting the moon landings, but there are some things that are questionable

Leaving the technical details aside, while only NASA performed manned landings, they are not the only ones that went to the moon.

Perhaps you were not aware, that the USSR also sent some probes to moon, Zond to return and Luna to land.

IMHO this pretty much dismisses your "questionable things" without even going into the technical detals. At least unless you want to claim that both sides were faking their programs while ignoring the other side's fake ...

[Eta C]

EDIT:
The lunar module also included two additional computers -- the brains of the Commander and Lunar Module Pilot.

Which, according to Von Braun, "are easily manufactured with unskilled labor"

[Jason Thompson]

im not denying or supporting the moon landings, but there are some things that are questionable

In general, or just to you?

like for example, the mass amounts of cosmic radiation experienced would have surely killed the astronauts,

Radiation in space is not something NASA has an information monopoly on. Where is the data that led you to conclude that it was sufficient to kill the astronauts?

gamma radiation is the stongest on earth and it takes 4 feet of solid lead to protect you,

No it doesn't. But it's academic anyway since gamma radiation is not abundant in space in lethal quantities.

and what about the onboard computers that they used, dont you ppl tell me they could land somone on the moon with a computer that has not even half the compteing power of my graphing calculator,

All right, we won't. What we will tell you is that two highly trained and skilled professionals were able to land on the Moon with the aid of an onboard computer with less computing power than your graphics calculator and a number of skilled professionals and much larger computers on the ground back at Mission Control.

What we will also tell you is that NASA and the USSR had also by that point sent unmanned probes past, around, into orbit of and into the surface of the Moon, some of which made a soft landing. Are you contending that Lunar, Ranger, Lunar Orbiter, Surveyor and Zond probes were also faked? If not, what makes the Apollo landing more difficult to achieve than the Surveyor landing, for instance?

because if im correct, i is possible to control a Saturn V rocket, with a TI-83+.

No, though it might be possible to control it with something with comparable amounts of memory. In these days of desktops and laptops with fancy interfaces, excessive memory and redundant software bundled in even though you never use it, it is apparently hard for many people to realise that the early computers were dedicated to one task. Every bit of memory on the Apollo computer was used for the sole purpose of navigation. No fancy interface, no sounds, no spare memory, no extra programs, no nice graphics of the LM as it approached the surface, not even a user interface that gave messages in English. They taught the astronauts computer code rather than teaching the computer the astronauts' language. Hell, occasionally they even overloaded the computer so it couldn't so everything it was being asked to. Check out Apollo 11 and the 1201 alarm.

but on the other side, in 1969 there were some prtty smart ppl

Indeed there were. By that time we already had the nuclear bomb (fission and fusion), the ICBM, the submarine, the jet fighter, the supersonic aircraft, the SR-71, the passenger aircraft. Quantum physics was fifty years old!

and u cant really beleave anything on the internet these days,

You can believe things on the net, but you shouldn't just believe uncritically. A blanket distrust of one source is as bad and misleading as unquestioning acceptance.

[hplasm]

BTW - magnetic fields won't shield gamma rays, though they will deflect charged particles.

[Jakenorrish]

Well, some people find it hard to comprehend that human beings went to the Moon, but the fact is that we did. Because despite the computers being less powerful than your average 21st century calculator, they were adequate for the task. Despite idiots like Bart Sibrel spouting unscientific mumbo jumbo, the people involved were also more than up to the task.

I've no doubt that in decades to come there will be people claiming that the MIR space station, Cassini Saturn mission and New Horizons missions were also faked. Why? Because these individuals do not have the scientific knowledge to grasp the facts. The information they read or are told sounds too ridiculous to be true to them.

Check out the scientific facts and the masses of data available, not the (often very incorrect) innuendo put forward by Hoax believers.

[sts60]

Sounds cool! One nitpick ... that should be RP-1.

Doh! Fixed.

[01101001]

I've no doubt that in decades to come there will be people claiming that the MIR space station, Cassini Saturn mission and New Horizons missions were also faked. Why? Because these individuals do not have the scientific knowledge to grasp the facts. The information they read or are told sounds too ridiculous to be true to them.

What a semi-pleasant thought: todays' Apollo Hoax believers having to defend the existence of the ISS 30 years from now.

Or will they do so? With their powers of reasoning, maybe they'll just think back in time to now, and decide they were duped all along and the ISS -- and the Space Shuttle missions and all the cool space science and technology going on now -- was all really done in a secret movie studio with special effects.

Buckle up, hoaxies. Prepare to defend what you are now experiencing, from the claims of the nattering newbies of the future. If you care. You better have some documentation that will convince them. It better conform exactly to their expectations of what it should be. Good luck.

[JayUtah]

JayUtah is the real expert but I’ll give my two bits anyway.

You give yourself far too little credit.

The AGC and LGC (same hardware, different application software) are examples of straightforward closed-loop control. "Closed loop" means that the system is able both to impose control and to sense state directly, so that a difference between what's supposed to be the case and what is the case can be remedied by computing the appropriate control action. In contrast, open-loop control simply applies a control according to some deductive procedure.

My sprinkler system is an open-loop system. The controller's output is the signal to open or close the valves that let water pass to the sprinklers. A timer tells the system when to open and close them. It is up to the user to correlate that to the desire degree of watering. A closed-loop system might measure the amount of water deposited on the ground and close the valve when a predetermined amount of water has been applied. That would more closely correspond to the user's intent and would also be more reliable in the face of faults such as varying water pressure. Any system that uses sensors to adapt the behavior of the system to a measured effect (rather than a deduced one, e.g., "10 minutes of watering is probably enough water") "closes the loop."

The basis for all guidance is the state vector: six numbers representing the position and velocity of the vehicle in a three-dimensional coordinate system anchored at some point. Three numbers give the position and the other three give its velocity as components vectors. You can reckon the state vector according to any "fixed" point that seems suitable to your task. For example, the state vector for an airplane might use the airport radio beacon as its fixed point whereas an interplanetary spacecraft state vector would be hampered by that point moving along with the Earth; the sun might be a better reference point.

While the state vector describes where the vehicle is, it's usually also important to know where the vehicle is pointing because various mission and operational restrictions will apply. Hence you have a guidance problem in both position and orientation.

Closed-loop guidance in the positional sense constantly asks "Where am I?" and compares it against a reference of "Where should I be?" Then if a significant difference is found, it computes and applies a proper corrective action. In the orientation sense, closed-loop guidance asks "Where am I pointing?" and compares it against "Where should I be pointing?" and again takes corrective action. These two problems are completely separable. That means that small programs can be written to solve each problem separately without involving the other -- something computer scientists find comforting.

In turn, both the positional and orientational problems can be broken down even further since the problem is expressed in a 3D vector space. Many 3D problems in such spaces can be reduced to three 1D problems using the same code for each dimension. Routine state-vector maintenance measures the elapsed time since the last update and integrates the new position along some axis by multiplying elapsed time by the velocity along that axis. Repeat for each of the three axes. Because the position and velocity are expressed as orthonormal component vectors, this works mathematically with provable rigor. Orientation is not integrated from rates, but is expressed in the same component-wise fashion that makes working with it largely a matter of the same program code applied in sequence to each of three components.

The questions "Where should I be?" and "Where should I be pointing?" are generally answered by consulting tables uploaded from Mission Control, which were computed by the RTCC for that mission and adapted as the mission proceeded. Different mission phases and the difference in type of flying the CSM and LM do call for different tables and ways of consulting those tables.

The questions "Where am I?" and "Where am I pointing?" can be answered in various ways. "Where am I pointing?" was most commonly answered in Apollo by consulting the IMU's stable member. That's a big hunk of beryllium in which gyroscopes are embedded. The gyroscopes spin so as to keep the beryllium hunk oriented in the same position in space. The hunk is contained within three orthonormal gimbals that allow the spacecraft to rotate (almost) freely while letting the stable member beryllium hunk to retain its orientation in space. Sensors in the gimbal swivels measure the deflection of each gimbal relative to its outer neighbor, and from this the orientation of the spacecraft relative to the stable member can be derived. In Apollo parlance this is known as the REFSMMAT -- "reference to stable-member matrix".

For those who haven't studied linear algebra, directions expressed as XYZ formulations of a vector can be converted easily from one reckoning to another by a straightforward matrix multiplication, where the matrix is composed of the coordinates of the new coordinate system as expressed in the old one. Not surprisingly, the AGC and LGC software loads contained data types and program libraries for vector and matrix arithmetic.

The stable member is aligned with a fixed-space reference prior to launch. Those gimbal angles can then be correlated to a known orientation in space, and changes in gimbal angles can then be reckoned as changes in orientation in space.

As the stable member drifts (as gyroscopic systems do), it has to be corrected. That's what CMPs and LMPs do in Apollo. One procedure for doing that tells the spacecraft to point at a reference star. A telescope aligned very precisely with the spacecraft axis is used to "shoot" the star. The spacecraft has aligned itself with what it thinks is that star's orientation. It does that by converting the star's absolute orientation through the guidance platform's conversion matrix into relative coordinates, then positioning the spacecraft according to those coordinates. If the telescope shows that the spacecraft's relative orientation is in error, the pilot uses controls on the telescope to bring the reference star into the telescope crosshairs. When the star is "dialed in," the telescope tells the pilot the difference between the spacecraft's axis and the angle the telescope had to adopt relative to the axis in order to dial in the star, in terms of pitch and yaw corrections.

But not roll. The astute linear algebraist will rise up in revolt to tell us that one degree of freedom is still missing. The spacecraft can indeed roll through its entire gamut when properly oriented without changing the telescope error angles. And so corrections to the matrix derived from one star sighting will be incomplete. So at least two -- and in practice three or four -- star sights will be used to correct the matrix. When properly adjusted, the new matrix that describes the orientation of the spacecraft relative to the (drifted) stable member will be useful for guidance. (Unless you're Jim "Shaky" Lovell and you accidentally key in the command to zero out the reference matrix altogether!)

Reference to a stable onboard member is one way to measure your orientation. But it's not the only way. Differential star-, Earth-, and sun-sensors can be used along with the same kind of math to give the guidance computer an idea of the spacecraft's orientation in space. Ballistic missiles can also use horizon sensors. These are optical instruments that "see" various objects in space. They are precisely mounted on the spacecraft structure so that the sensor axis can be directly converted to the spacecraft axis, and the measurement of deflection according to sensor axes can be reckoned easily in terms of the spacecraft axis.

So much for "Which way am I pointing?" For "Where am I?" you have a lot of options, depending on which spacecraft you're flying and what phase of the mission you're in.

Periodically the spacecraft updates the position portion of the state vector by integrating the velocity portion. The spacecraft's electronics generate a very stable 100 Hz reference signal that is coupled through a counter to a register available to the computer. The integrator program "wakes up" every so often (1-2 times per second for Apollo, 20 times per second for the space shuttle), reads the counter and resets it, multiplies its consulted value (which represents the elapsed time since the counter was last reset in hundredths of a second) by the velocity in each of the three cardinal axis and adds the result to the position point (again, for each of the three axes).

The position portion of the state vector could also be updated from Earth. The ground stations can use the radio signal to track the spacecraft in terms of right-ascension, declination, and velocity (via doppler shift). Using their own linear algebra and some orbital mechanics, they can use the RTCC to crunch heavy numbers and compute the spacecraft's position with astonishing accuracy. Then they just beam up the new state vector to the spacecraft.

The velocity portion of "Where am I?" is a little bit more involved. The IMU also has very sensitive accelerometers that measure acceleration along the stable member's axes. They are "pendulous" accelerometers, meaning that a little mass is cantilevered out into space. Acceleration along one axis causes the mass to bend the cantilever a little bit, which can be measured by strain gauges. Another method places a floating mass next to a fixed mass with some pressure-sensitive materials between them. Acceleration pinches the sensor between the floating mass and the fixed mass. Today we use various sensors that don't require moving parts.

The accelerometers are "integrating," meaning that they have electronics that reads out the deflection in terms of counter-suitable pulses. When the accelerometer mass passes each increment of deflection, it pulses a "count up" wire. On its way back to rest position it pulses a "count down" wire for each returning increment. If you wire those outputs up to the "tick up" and "tick down" pins of a digital counter, you have an instrument that reads the momentary (snapshot) change in velocity (acceleration) as the counter value.

Imagine your bathroom scale was rigged that way. Imagine that you're the mass cantilevered out there, and that the deflection of the accelerometer arm occurred as Earth's gravity tried to accelerate your mass. For each increment of deflection, your scale's accelerometer would pulse its count-up output. And through the magic of Sir Isaac we know that acceleration, deflection, mass, and weight are all correlated quantities. When you stepped on the scale there would be a flurry of count-up signals. As you shifted your weight on the scale, there would be a lot of "noise" in count-up and count-down signals. But the counter would meta-stabilize around the number of deflection-caused pulses that your standing on the scale had (finally) induced. And then when you stepped off, there would be a commensurate flurry of count-down pulses. If the system is well-designed and well-built, that counter should read zero at the end -- i.e., the number of count-up pulses balanced the number of count-down pulses.

As a matter of fact, digital bathroom scales are rigged this way, but they use the pinch-type accelerometers.

But this is all just another job for the all-purpose integrator. Before computing the position portion of the state vector, it updates the velocity portion. After recording elapsed time, it reads the registers corresponding to the accelerometer counters to see if any acceleration is being recorded. If so, it integrates the acceleration in each axis and updates the velocity in each axis prior to updating the position.

The astute embedded-system programmer has just realized that this integration process is the same algorithm in the acceleration-to-velocity case as it is for the velocity-to-position case. On a general purpose RISC or CISC architecture, this algorithm could be implemented in four instructions. On the AGC archicture it can be implemented in three, because of some unique addressing modes. You simply point that algorithm at different data sources for the rates and quantities to be integrated into, and you have a very competent, very simple guidance computer.

The good news is that this method works for any inertially-detectable accelerations, whether they come from engine burns, atmospheric interface, or an astronaut on EVA kicking the spacecraft. Engineers love solutions that handle effects of entire classes of causes, not just those few causes they can think of. The bad news is that not all acceleration is inertially-detectable, such as that deriving from orbital mechanics.

For orbital motion you need the orbital integrator. While it is not possible to measure accelerations directly that occur during orbital motion, you can deduce the changes in velocity with reasonable accuracy if you know the geometry of the orbit. If you know the orbital elements -- the handful of numbers that together uniquely describe an orbit's shape -- you can determine your velocity at each point according to elapsed time by solving simple orbital mechanics equations for that orbit and that elapsed time. So the orbital integrator uses given periapsis and apoapsis values and some simplified math to deduce how the velocity should change from moment to moment according to orbital motion.

Another way to measure velocity reasonably directly is with radar. If you assume the thing the radar waves are bouncing off of is fixed in your coordinate system, then radar returns measure your velocity in that coordinate system, if you take into account the radar pulse's orientation. And radar can sometimes also give you position directly, if you know the position of your radar target according to a fixed point in your coordinates.

Any practical guidance system will use as many of these measurement techniques as are possible and applicable, averaging the diverging results if necessary. So a landing spacecraft may use both radar and inertial means to maintain its state vector. It may compute its velocity according to idealized curves from its descent profile, and according to velocity information obtained from the radar. It may update its position both from radar altimeter measurements and from intertial integration. This happens because there is no one true and accurate sensing mechanism. During Apollo 11's descent you can hear a conversation about which guidance methods are going to be used, which are currently operating correctly, and finally a comment that the methods are "converging," meaning that the error between the guidance methods is growing smaller and smaller.

[Fazor]

I propose that our ancestors never ate fish, because they did not have the graphite rods and reels that I have today to catch them with.

Just because the technology that would be used to do something today wasn't available in the past, it doesn't mean it couldn't be done.