102:15:02 Duke: Columbia, Houston! We're standing by. Over. (Long Pause) Columbia, Houston. Over.

102:15:41 Collins: Houston, Columbia. Reading you loud and clear. How me?

102:15:43 Duke: Rog. Five-by, Mike. How did it (that is, the DOI burn) go? Over.

102:15:49 Collins: Listen, babe. Everything's going just swimmingly. Beautiful.

102:15:52 Duke: Great. We're standing by for Eagle.

102:15:57 Collins: Okay. He's coming along.

102:16:00 Duke: We copy. Out. (Pause) And, Columbia, Houston. We expect to lose your high gain sometime during the powered descent. Over.

102:16:19 Collins: Columbia. Roger. You don't much care do you?

102:16:22 Duke: No, sir.

[Both spacecraft are equipped with high-gain antennas which must be pointed accurately at Earth to maintain top-quality communications. Because the Command Module attitude relative to Earth changes as Mike maneuvers so that he can track the LM, the antenna must be moved to maintain pointing. However, the system of gimbals used to point the antenna has physical limits and Houston is telling Mike that, at some point during the descent, one or more of those limits will be reached. In order to re-acquire the high-gain, Mike would have to change the spacecraft attitude and, because he can maintain voice communications with both Earth and Eagle on his omni-directional antennas and because Houston will be pre-occupied with the landing, no attitude adjustment will be made.]
[Comm Break. Buzz Aldrin calls about 45 seconds after Eagle AOS.]

102:17:27 Aldrin: Houston, Eagle. How do you read?
102:17:29 Duke: Five-by, Eagle. We're standing by for your burn report. Over.

[Armstrong - "They are referring to the strength and clarity of the signal on a scale of five. 'Five by five' meant 'loud and clear."]
102:17:36 Aldrin: Roger. The burn was on time. The residuals before nulling: minus 0.1, minus 0.4, minus 0.1, X and Z nulled to zero (Static) (Garbled) nulling (garbled). (Long Pause)
[David Woods writes, "In this case, there has obviously been a large burn. They would like the burn to achieve a particular result in terms of velocity change (delta-v) in the three axes. In most cases. a single engine burn will fail to precisely achieve the required change of velocity - there will usually be a small under or over performance of the engine. The differences between the desired and achieved delta-v across the three axes are called the 'residuals'. For some (not all) burns, the crew are to use the RCS thrusters to bring the three residual values of delta-v to zero or null, essentially making up the deficit or cancelling the overperformance of the main engine to achieve a perfect result. This is called 'nulling the residuals'. However, it is often desirable for the engineers to know what the residuals were before they were taken out - 'before nulling'. This probably gives insight into the performance of the system."]
[Aldrin - "If the burn had been perfect, the computer would have displayed residuals in the three body-axes of 0, 0, 0. If there was some deviation, it was displayed to a tenth of a foot (per second). Apparently, we were supposed to null X and Z (the vertical and fore/aft axes, respectively). They didn't care about Y (left to right). The way they were nulled was (by firing the RCS) with the handcontroller. Neil was looking at them, and I was looking at them and I recorded them."]

[The spacecraft's X axis coincides with the thrust axis, with the positive direction being up - away from the engine toward the rendezvous hatch. The Z axis runs fore and aft, with the positive direction being out the windows. The Y axis runs left to right, with the positive direction being on Aldrin's (right) side of the spacecraft. By nulling the X and Z residuals, they are trying to avoid landing long or short of the target. They are ignoring the small north/south error.]

102:18:25 Duke: Columbia, Houston. We've lost all data with Eagle. Please have him re-acquire on the high gain. Over.
102:18:37 Collins: Eagle, this is Columbia. Houston would like you to re-acquire on the high gain. They've lost data with you. Over. (Pause)

102:18:50 Collins: Eagle, did you copy Columbia?

102:18:54 Duke: Eagle, Houston. Did you call? (Pause; static clears)

102:19:05 Aldrin: Eagle, Houston...(correcting himself) or Houston, Eagle. How do you read now?

102:19:08 Duke: Rog. (Making a mis-identification) Five-by, Neil. We copied up to the AGS residuals. Would you please repeat the AGS residuals and the trim - correction - the Sun check? Over.

[The AGS is the Abort Guidance System, the backup navigation system used primarily for an emergency return to orbit. The Primary Guidance and Navigation System provides more information and will be used during the landing. The two systems are cross-checked prior to descent; and close agreement between them gives confidence in both.]
102:19:19 Aldrin: Roger. AGS residuals: (X) minus 0.1, (Y) minus 0.2, (Z) minus 0.7 (feet per second). And we used the PGNS Noun 86 for Delta-VZ which was 9.5, versus yours which was 9.1, and I believe that may explain the difference (between the minus 0.7 residual in AGS versus minus 0.1 for the PGNS). Apogee 57.2, perilune 9.1; Sun check to three marks; Noun 20 minus Noun 22, plus 0.19, plus 0.16, plus 0.11. Over.
102:19:54 Duke: Rog. Copy. Looks great.

[Comm Break, with static during most of the last minute.]
[The LM computer takes actions designated as Verbs, while Nouns are data. The PGNS and AGS both integrate accelerations to estimate the spacecraft velocities. The PGNS uses data from the inertial platform and the AGS uses less accurate data from body-mounted accelerometers.]

[Aldrin - "Somewhere, we had a state vector (three-axis position and velocity) update because of the tracking data that Houston got once we came around. But, how that happened and whether we were aware of it, I don't remember. I know that a lot of people got credit for developing the tracking filter that allowed them to do that. That neat capability contributed to the accuracy of our touchdown, even though nobody knew (exactly) where we were."]

[Armstrong - "I was less concerned about the state vector than I was about inertial platform drift. It had been a fair amount of time since we had last aligned the platform and, in time, it was going to drift. This sun check, here, was a gross check on platform drift. During the half orbit before DOI we rotated (the LM) to make the sextant look directly at the Sun and looked to see if it was in the crosshairs. I think that, to within some predetermined limits, some fraction of a degree, it was alright. The platform was drifted slightly, I think (0.08 degrees); but not enough to worry about."]

[Journal Contributor Paul Fjeld writes: " The predetermined limit for this gross check on platform drift was 0.25 degrees. There was quite a bit of controversy about the usefulness of this and other 'confidence builders' and they were discontinued for the later Apollos. Floyd Bennett, of NASA's guidance section (for whom Bennett Hill at the Apollo 15 site is named), thought that, in planning the descent, they were far too pessimistic in assessing the performance of the whole guidance and navigation system. The drift check, checks of the altitude by tracking the Command Module with the rendezvous radar, etc., caused more problems, and cost more time and money than was warranted by a slight gain in confidence. Bennett's Apollo Experience Report (NASA TN D-6846) discusses they way in which the landings and ascents were planned, and is one of the best of the series."]

[Armstrong - "We did landmark checking to check our position over the ground, and that was another check on the state vector."]

[Neil's double-paned window has scribe marks on both panes in the form of graduated vertical and horizontal scales, marked in degrees. During the landing, these scales will give Neil the means of locating the point on the ground where the computer thinks they are going to land. However, during this period before Powered Descent Initiation, Neil uses the scribe marks to determine how quickly objects on the surface move along the scale and, consequently, the LM's current altitude. Crudely, the LM altitude (in thousands of feet) is 360 times the radius of the Moon (in thousands of feet) divided by the LM's orbital period (in seconds) and, finally, that result divided by the tracking rate (in degrees per second). As an example, if we use a lunar radius of 5700 thousand feet and a LM period of 7200 seconds, the LM altitude (in thousands of feet) is 285 divided by the tracking rate (in degrees per second). In detail, the tracking rate as a function of position along the orbit depends not only on the spacecraft altitude but, also, on the shape of the orbit. As Buzz indicates in the next paragraph, he and Neil have a chart in the cabin on which Neil can compare tracking rates with expected values at various positions along the orbit; and differences between the observations and the expected values allow him to estimate the altitude of the low-point in his orbit - called perilune - and the time at which they will reach it. See his comment two paragraphs below.]

[Aldrin, from the July 31, 1969 Technical Debrief - "We had two methods of computing altitude: one based on relative motion from the CSM and the other based on angular rate track of objects observed on the ground. We superimposed the two of them on one graph and re-arranged the graph a little bit with some rather last minute (pre-flight) data shuffling to give us something that the two of us could work on at the same time and to give indication of what the altitude and its time history appeared to be. With the communications difficulties that we were experiencing in trying to verify that we had a good lock-on (with Earth) at this point, I had the opportunity to get only about two or three range-rate marks (on the CSM). They appeared to give us a perilune altitude of very close to 50,000 feet, as far as I could interpolate them on the chart."]

[Armstrong, from the 1969 Technical Debrief - "The measurements against the ground course were indicative of altitude directly above the ground. The ground measurements were very consistent. If they made a horizontal line, it would indicate that you were going to hit a particular perilune (say) 50,000 feet. They (the actual measurements) didn't say that. They were very consistent (that is, they didn't jump around), but they came down a slope, which finally said that our perilune was going to be 51,000 feet. It started out at about 54,000 feet...and our last point was 51,000 feet. This indicated that either the ground was sloping (which it wasn't)...or that the line of apsides (the line connecting the high and low points in the orbit) was shifted a little bit (from its planned position). So, actually, perilune was coming a little bit before PDI...This was all very encouraging - that we were, in fact, going to hit the guidance box (an imaginary window in the sky) so far as (perilune) altitude was concerned from both (the radar and ground tracking) measurements. But I was quite encouraged that these (ground) measurements, made with the stopwatch, were consistent (that is, followed a smooth trend), in fact."]

[Aldrin, from the 1969 Technical Debrief - "When you're able to smooth the numbers and plot a reasonable number of them, your accuracy increases considerably. I think the pre-flight estimates were something on the order of a 6000-foot capability; and I think we demonstrated a much better capability than that."]

102:22:37 Duke: Columbia, Houston. We've lost Eagle again. Have him try the high gain. Over.
102:22:46 Collins: Eagle, this is Columbia. Houston lost you again. They're requesting another try at the high gain.

[Comm Break, with the static clearing after a half minute or so.]
[The LM computer tracks the signal it receives via the high-gain antenna and corrects the pointing to maintain maximum signal strength. The computer program which maintains proper pointing also contains a 'map' of the LM so that it can use information about the spacecraft orientation and, thereby, avoid trying to 'see' Earth through the spacecraft. Unbeknownst to anyone at this point in the mission, the computer has an incorrect LM map.]

[Armstrong - "I think that, later, they put in a five degree yaw (a rotation around the thrust axis) so that the high gain wasn't working so close to the spacecraft, so near its limit."]

[Houston will recommend a 10 degree right (clockwise) yaw maneuver at 102:27:22.]

102:23:57 Duke: Eagle, Houston. We have you now. Do you read? Over.
102:24:02 Aldrin: Loud and clear.

102:24:04 Duke: Roger. We see your Verb 47.

[Aldrin - "When we were looking at something on the computer, they could see it, too. So the verbal report was just added confirmation."]
[Armstrong - "Of course, they couldn't get anything when we were out of the line-of-sight, and had to rely on our (post AOS) reports. I don't think they had any ability to store that (post-burn) information in the computer."]

[The LM computer has only a very limited memory, so data that is no longer of use is not kept.]

[Journal Contributor Frank O'Brien notes, "Verb 47 is the command used to initialize the Abort Guidance System (AGS), using PGNS data".]

102:24:12 Aldrin: Yeah. I don't know what the problem was there. It (the steerable high-gain antenna) just started oscillating around in yaw. According to the needle...We're picking up a little oscillation right now, as a matter of fact.
102:24:23 Duke: Roger. We'll work on it. (Long Pause)

[The crew is in the Push-to Talk communications mode and we only hear them when they chose to broadcast to Houston.]
[Aldrin - "There were switches on the handcontroller and on the electrical connector that went to the suit. I was using that one because I would have gotten my hand slapped if I had touched the controller."]

102:24:38 Armstrong: Horizon check was right on time.
102:24:41 Duke: Roger.

102:24:45 Aldrin: Did you copy the star...I mean the Sun check, Charlie?

102:24:48 Duke: That's affirmative. We did, Buzz. Out. (Long Pause; with intermittent static)

102:25:35 Duke: Eagle, Houston. The AGS initialization looked good to us. Over.

102:25:43 Aldrin: Roger. (Long Pause)

[Armstrong - "It was a fairly simple procedure to send a state vector from one computer (the PGNS in this case) to the other (the AGS)."]
[Aldrin - "The AGS initialization was done by a Verb, an instruction with maybe a two- or three-digit code (Verb 47, as noted above)."]

[Armstrong - "At the time, there were inertial guidance systems in aircraft, using inertial platforms. But the computations usually had an altimeter-based smoothing input into the Earth-radius measurement so that the calculation of position was stable. The error would oscillate but wasn't like the error in three dimensions which is unstable and will continue to build. So, although aircraft had similar devices, the calculations were different. Aircraft didn't have anything like the AGS, because I don't think the accuracy would last very long."]

102:26:29 Armstrong: Our radar checks indicate 50,000-foot perilune. Our visual altitude checks are steadying out at about 53,000 (feet).
102:26:37 Duke: Roger. Copy. (Long Pause)

[Armstrong - "The visual check was something that we devised ourselves, barnyard math of v = r w. 'r' would be your altitude (that you wanted to know), omega (w) was your angular rate that you determined by watching a point on the ground, and the velocity (v) was pretty well known. We measured omega by measuring the speed at which an object on the ground passed through a certain number of degrees on the (LPD) grid on the window. We just timed it on a stopwatch and had a little plot to compare it with. As the altitude decreased, we could see it was converging pretty well. It gave us an alternate check of our altitude. The importance of this is that, if we weren't in a reasonably close altitude band to our intended starting altitude, the landing guidance would not necessarily converge - the solution wouldn't converge - so it was important to us at that point that the altitude be about right when we started."]
102:26:55 Aldrin: And, Houston, we got a 500 alarm (code) early in the program. Went to Descent 1, proceeded on it, and we're back at Auto again. Over.
[500 series codes were reserved for radar-related computer functions.]
102:27:06 Duke: Roger. We saw that, Buzz. Thank you much. Out.
102:27:09 Aldrin: Rog. I say again...(Listens) Okay. That wasn't an alarm; that was a code. Okay.

102:27:14 Duke: Rog. We saw that. (Pause)

[Armstrong - "We'd have to have a half day in that simulator, again, to remember some of these things."]
[Aldrin - "We spent a lot of time in the LM trainer - maybe 30 or 40 percent of our time - but there was a lot in the CSM (trainer) and on other things. Your sense of well-being is generated by the degree of familiarity you have with what might go wrong in a challenging way. To me, when you're doing what we were doing for the first time, there's a level of importance that's a good bit different from the second or third time, because it has been done already and because the world has seen it. When you're part of the pioneering effort, there's a focusing of an individual's concentration and level of attention that is at the exclusion of a lot of other things. It's a kind of gun-barrel vision."]

[The 15 July 1969 Apollo 11 Crew Training Summary indicates that out of 959 hours of training, Neil spent 285 hours - 30 percent - in the various LM simulators. Buzz did 1017 hours of training, of which 332 hours - 33 percent - was spent in the LM simulators. These figures do not include Neil's LLTV flights nor the 56 hours each of them spent in briefings about LM systems.]

[Armstrong - "I would add, along that same line, that the time requirements necessitated that we accept the conclusions and recommendations of the crews that went before us. So, we did not spend an extraordinary amount of time on things that had already been done and had worked as expected. We focused a great deal of our attention on those things which had not been done before and which we would be expected to pass on to the crews behind us."]

[Aldrin - "Whether they wanted it or not."]

[NASA photo S69-35504 shows Neil, Mike and Buzz debriefing the Apollo 10 crew on 3 June 1969, about a week after Tom Stafford, John Young, and Gene Cernan returned from the Moon. Clockwise from the near left, the people around the table are Collins, Aldrin, Cernan, Stafford, Armstrong, and Young.]

[Armstrong, from the 1969 Technical Debrief - "We did have one program alarm...prior to ignition, that (indicated) we had the radar out of position...a 500 series alarm...which I don't have any way of accounting for. Certainly the switches were in the right positions. They hadn't been changed since pre-launch. But we did, in fact, go to the Descent position on the antenna and leave it there for a half a minute or so, and then go back to Auto and that cleared the alarm."]

[O'Brien - "The landing radar could be moved between two positions: Before pitchover, the radar was in the Descent position, and as the LM rotated to a more upright attitude during the approach phase, the radar was moved to the Hover position. These positions could be commanded manually, in addition to the Auto setting, where the radar was under the control of the guidance computer. The purpose of moving the radar was to keep the antenna pointed as directly towards the surface as possible."]

102:27:22 Duke: Eagle, Houston. We recommend you yaw 10 (degrees) right. It will help us on the high-gain signal strength. Over. (Long Pause; static)