Would an astronaut survive burial on the Moon?

[ka9q]

That's an excellent answer! Makes perfect sense.

[Noldi400]

That's an excellent answer! Makes perfect sense.

I have to come up with one occasionally - biology/medicine doesn't come up that often in this group.

Perhaps you can provide me with an excellent answer.  The radar 'ground tracking' that was done during Apollo - I gather it was based on the telemetry signal (well, one of the signals) from the LM and CSM rather than the traditional ping-and-return approach?  How accurate was that in terms of guidance, i.e., measurement of their lunar orbit, velocity and altitude during descent, etc?  If they were getting accurate data from 240,000 miles away, well, that's pretty freakin' amazing.  I've looked at some of the Apollo guidance & navigation documents but they're pretty much gibberish to me.  Can you point me toward a primer-level document or source?

I'll say this for hunchbacked - he's better than most as prompting me to go trotting off to do research (which is more than he does, apparently.)

[Allan F]

It wasn't radar - they tracked the actual telemetry transmissions from the vehicles. The big radio telescopes had enough resolution to follow the actual descent and hover by Apollo 11. I'll search for some info.

Edit: http://depletedcranium.com/fascinating-recording-of-apollo-11-at-jodrell-bank-released/

[grmcdorman]

I think what Noldi400 is asking is how NASA tracked the spacecraft; the Jodrell Bank info, while fascinating (and yet another blow on the smear-on-the-ground corpse that is HB denialism) is not that.

[Allan F]

Yes, Jodrell Bank didn't supply their data to NASA, but did the to satisfy their own curiosity.

Edit: Honeysucklecreek has some information. They apparently had the ability to measure the distance to within 1 meter, but errors from outside the system increased the error margin to 15.2 meters.

 http://www.honeysucklecreek.net/station/technical.html

[Allan F]

I think this passage is the key:

Once the station had locked onto the signal, another special processor called a Tracking Data Processor, (TDP) accepted the ranging data, the speed of the spacecraft relative to the station from the doppler, and the antenna angles relative to the station's geographical location, and coded this information for transmission to Goddard. It was coded in both high-speed data at 2,400 bits per second and in teletype code onto paper tape.

So they had the distance pretty accurate, the speed pretty accurate, the direction to the spacecraft too. All they missed was the attitude, and direction of travel. The former was in the datastream, the latter could be calculated.

[ka9q]

Here's how it worked; I know this stuff pretty well.

The spacecraft had an "S-band transponder", basically a special kind of radio transceiver (transmitter and receiver). What made it special was a linkage between the receiver frequency and the transmitter frequency. When the receiver didn't detect a signal, the transmitter used a local oscillator to transmit on its nominal frequency; this could drift due to manufacturing tolerances, temperature, etc.

But when the receiver did receive an uplink signal from earth, it "phase locked" onto it and produced a transmitted frequency that was exactly 240/221 times the received signal frequency. This became the carrier of the downlink signal, which was designed so that the carrier was always present even with modulation (this isn't always true with some modulation methods).

The ground used a highly stable oscillator (an "atomic clock") to generate its uplink, and by producing a local reference signal 240/221 times its own transmit frequency, could compare that with the incoming downlink signal from the spacecraft. If the spacecraft were at constant distance from the ground, the two signals would be exactly the same frequency and phase. But if the spacecraft moves toward or away, the phase will rotate at some rate. Phase rate is the same as frequency, so there's a Doppler frequency shift.

The phase will rotate 360 degrees for every change in round trip distance of one wavelength, which at S-band is about 13 cm. The round trip distance changed one wavelength for every half wavelength of actual ground-spacecraft distance. And phase changes considerably less than 360 degrees could easily be detected.

So you can see how incredibly precise this method was; it could easily detect the changes in velocity caused by a simple urine dump (which led to an unfortunate misunderstanding during Apollo 13).

Doppler tracking was continuous, and if you knew at time t how far it was, by counting RF cycles you could keep track of its current distance. But how did you find the actual distance to start? With a separate mechanism called PN ranging. The ground could optionally transmit a fast (a little less than 1 Mb/s) pseudorandom data sequence to the spacecraft, which (if enabled by the crew) would repeat that signal on the downlink. The sequence was long enough to not repeat during the several seconds it could take the signal to reach Apollo and return, so the ground could compare it with its transmitted sequence and see to within 300 m (actually much less) what the round trip distance was and remember this figure for updating with Doppler. Then it could turn off the PN ranging signal.

PN ranging is widely used today; it's the basis of GPS, for example, and the Qualcomm CDMA digital cellular system that I worked on. GPS and CDMA all use different PN sequences from Apollo, but the "chip" (random bit) rates are remarkably close: 996 kHz (I think) for Apollo, 1.023 MHz for GPS, and 1.2288 MHz for CDMA. It made E911 position determination relatively straightforward, and if a CDMA phone was involved that's how the police found the Tsarnaev brothers in Watertown last month...

That you can continuously update a one-shot PN ranging measurement with Doppler is one of my disagreements with Hunchbacked. He insists the PN signal has to be on all the time, and the fact that it can be turned off is one of his many discovered "incoherences". I've tried to explain how the Doppler tracking is coherent and continuous so the PN ranging only need be done at first acquisition and after any loss of signal, but to no avail...

[ka9q]

So they had the distance pretty accurate, the speed pretty accurate, the direction to the spacecraft too. All they missed was the attitude, and direction of travel. The former was in the datastream, the latter could be calculated.

Right. The accuracy of the direction measurements depended on the size of the ground station antenna and thus its beamwidth. Apollo used either 85' or 210' (25.9 or 64m) antennas. At S-band the smaller dish had a beamwidth of +/- 0.15 deg at the 3dB points and the larger had a beamwidth of about +/- 0.06 deg at the 3dB points.

I've written a program to determine a satellite's orbit from measurements like these; you design it to accept whatever information you have, with a weighting on each measurement to indicate how confident you are in that number. Then you turn the crank on a big least-squares fit to produce a state vector (or orbital element set) that most closely matches all those measurements.

[Allan F]

So you can see how incredibly precise this method was; it could easily detect the changes in velocity caused by a simple urine dump (which led to an unfortunate misunderstanding during Apollo 13).

Poor Fred....

Fascinating. I bet this info perplexes many people. I understand the random-sequence-bit - they frame-matched to get a range-estimate.

[ka9q]

That's right. And the Doppler measurement was not much different from how the microwave motion sensors on automatic supermarket doors work. They beam a continuous carrier at 10 GHz toward you and detect your reflection. They can even tell which way you're going by which way the phase difference rotates; this is done with a pair of detectors in quadrature to pick up both the sine and cosine components. Police radars work much the same way except I don't think they bother with the quadrature detectors.

The only real difference between the supermarket door detector and the Apollo transponder was that the former relied on passive reflection while the latter required the spacecraft to actively amplify and repeat the signal. It necessarily had to do this on a different frequency to avoid interference, but even this difference was effectively eliminated by having the ground compare the received signal to that of a local reference transponder that doesn't move.

[Allan F]

As I understand it, the deliberate phase-locked frequency shift eliminates the doppler shift, so the return carrier signals's frequency is independent of the spacecraft's speed and direction. Is that correct?

Edit: No, that didn't make so much sense when I read it again.

[Allan F]

So they had the range to within one wavelength, and the up/Down/left/right position to an area approx. 200 x 200 km (edit: at the far end of the journey)? That explains the need for the rendezvous radar and transponder system on the LM/CSM.

[ka9q]

I'd have to look up the actual numbers, but from the radio link design I'd estimate they had the range to probably several meters (a fraction of a PN chip time) and the range-rate (velocity projected along the line of sight) to something like a millimeter per second (a few degrees per second of carrier phase rotation). These are the accuracies you'd likely get in real time; you could do better if you averaged out the noise over a longer tracking run.

There were several reasons for the rendezvous radar and transponder. One, neither spacecraft could be tracked on the lunar far side, and much of the rendezvous took place there. Two, while the ground can measure the range and range-rate along its line of sight very accurately, it can't measure position and velocity at right angles to that line of sight. It has to rely on antenna angles, and while the beamwidths are small they're still quite large compared to the half degree angular size of the moon at the earth. Three, while you can solve for orbital elements by doing a Kalman filter or a big least-squares fit with whatever measurements you have, you also need a model of all the forces affecting the spacecraft which, at the moon, is largely lunar gravity -- and we didn't have a very good model of lunar gravity in those days. (We're just getting a really good one now from the GRAIL mission.) Four, good orbital estimates require a series of measurements over the longest possible period of time when the spacecraft is not being maneuvered, and neither assumption was a good one during the relatively quick rendezvous sequence.

I'm not sure which was the dominant factor, but #1 and #3 were probably the most important limitations of earth-based tracking during rendezvous.

This isn't to say that earth-based tracking wasn't also done and used as a cross-check; after the DOI burn on at least Apollo 11, the ground tracked Eagle as soon as it came around the edge of the moon as a double-check to ensure that they didn't overburn and put themselves on a lunar impact trajectory. Supposedly there would have been enough time to perform an emergency bailout burn to avoid that.

[ka9q]

Back in the late 1970s I did an informal experiment with a TV broadcasting satellite that worked much like the Apollo tracking system. In my college years I worked as a summer intern at the Baltimore PBS station. This was just before domestic communication satellites, and after PBS installed ground stations at their affiliates I paid the engineering team a visit to take a look. It was Friday evening and they were uplinking Wall Street Week to the network. I punched up the downlink of their own signal and put it on a vectorscope using house subcarrier as the reference.

A vectorscope was a special-purpose oscilloscope used in analog TV to display the amplitude and phase of the NTSC color subcarrier in a polar plot. White, grey and black all showed up as a point in the center with saturated colors around the outer perimeter. One point was always supposed to say fixed: the color burst between each horizontal scan line that was a reference sample of the 3.579545... MHz NTSC color subcarrier. Since my scope was synchronized to house subcarrier (coming directly from the station's rubidium frequency standard, which was also used to synchronize the uplink to the satellite) any change in color subcarrier phase in the downlink would show up as a rotation of the color burst (and all the colors, actually) on the scope face. I wanted to see if the satellite was moving -- and it was. I saw a slow steady rotation of the entire chroma signal, indicating the satellite was moving along our line of sight.

I was looking at the color subcarrier, not the C-band (6/4 GHz) RF carrier, so each rotation represented one cycle of the color subcarrier frequency rather than the RF. The wavelength of the subcarrier is about 80 meters, so one full rotation would have been a change of about 40 m in station-to-satellite distance. I don't remember the exact numbers (it's been 35 years!) but it was maybe one rotation every few minutes. That would have been a range-rate of a fraction of a meter per second, and it was easily visible.

Because the Apollo system did the same thing directly on the S-band carrier, each rotation of the carrier phase represented a much smaller change in distance (by a factor of over 600!) so as you can see the system was very sensitive.

[Noldi400]

Here's how it worked; I know this stuff pretty well.

The spacecraft had an "S-band transponder", basically a special kind of radio transceiver (transmitter and receiver). What made it special was a linkage between the receiver frequency and the transmitter frequency. When the receiver didn't detect a signal, the transmitter used a local oscillator to transmit on its nominal frequency; this could drift due to manufacturing tolerances, temperature, etc.

But when the receiver did receive an uplink signal from earth, it "phase locked" onto it and produced a transmitted frequency that was exactly 240/221 times the received signal frequency. This became the carrier of the downlink signal, which was designed so that the carrier was always present even with modulation (this isn't always true with some modulation methods).

The ground used a highly stable oscillator (an "atomic clock") to generate its uplink, and by producing a local reference signal 240/221 times its own transmit frequency, could compare that with the incoming downlink signal from the spacecraft. If the spacecraft were at constant distance from the ground, the two signals would be exactly the same frequency and phase. But if the spacecraft moves toward or away, the phase will rotate at some rate. Phase rate is the same as frequency, so there's a Doppler frequency shift.

The phase will rotate 360 degrees for every change in round trip distance of one wavelength, which at S-band is about 13 cm. The round trip distance changed one wavelength for every half wavelength of actual ground-spacecraft distance. And phase changes considerably less than 360 degrees could easily be detected.

So you can see how incredibly precise this method was; it could easily detect the changes in velocity caused by a simple urine dump (which led to an unfortunate misunderstanding during Apollo 13).

Doppler tracking was continuous, and if you knew at time t how far it was, by counting RF cycles you could keep track of its current distance. But how did you find the actual distance to start? With a separate mechanism called PN ranging. The ground could optionally transmit a fast (a little less than 1 Mb/s) pseudorandom data sequence to the spacecraft, which (if enabled by the crew) would repeat that signal on the downlink. The sequence was long enough to not repeat during the several seconds it could take the signal to reach Apollo and return, so the ground could compare it with its transmitted sequence and see to within 300 m (actually much less) what the round trip distance was and remember this figure for updating with Doppler. Then it could turn off the PN ranging signal.

PN ranging is widely used today; it's the basis of GPS, for example, and the Qualcomm CDMA digital cellular system that I worked on. GPS and CDMA all use different PN sequences from Apollo, but the "chip" (random bit) rates are remarkably close: 996 kHz (I think) for Apollo, 1.023 MHz for GPS, and 1.2288 MHz for CDMA. It made E911 position determination relatively straightforward, and if a CDMA phone was involved that's how the police found the Tsarnaev brothers in Watertown last month...

That you can continuously update a one-shot PN ranging measurement with Doppler is one of my disagreements with Hunchbacked. He insists the PN signal has to be on all the time, and the fact that it can be turned off is one of his many discovered "incoherences". I've tried to explain how the Doppler tracking is coherent and continuous so the PN ranging only need be done at first acquisition and after any loss of signal, but to no avail...

PROGRAM ALARM: 1201 . . . . . . . EXECUTIVE OVERFLOW

OK, wow, nicely answered, even if most of it is over my head.

It does match up nicely with the tech reports I've found since asking the question.  I got my answer, at least enough of a broad overview that I understand what they were doing, even if I quickly get lost in the technical details.

Sometimes I get the feeling that most HBs think that NASA claims to have basically slapped something together, fired it off in the direction of the moon, and made the rest up as they went along.

Then too, that may not be confined to HBs; I remember someone who accepted the reality of Apollo asking if anyone knew how the crew on Apollo 11 decided who was going to land and who stayed in the CM.  Did they draw straws or something?  A lot of people, I believe, really don't understand the sheer volume of the planning, designing, testing, training, etc. that went into Apollo, or the extent to which every scrap of data was analyzed and incorporated into subsequent missions.