Malfunctions aboard Scott Carpenter’s orbital flight

Scott Carpenter was born in Boulder, Colorado, on 1 May 1925. He joined the US Navy in 1943 but was discharged at the end of the Second World War. He rejoined the US Navy and was commissioned in 1949. He was given flight training at Pensacola, Florida and Corpus Christi, Texas and designated a naval aviator in April 1951. During the Korean War he flew anti‑submarine, ship surveillance, and aerial mining missions in the Yellow Sea, South China Sea and the Formosa Straits. In 1954 he attended the Navy Test Pilot School at Patuxent River, Maryland, and subsequently was assigned to the Electronics Test Division of the Naval Air Test Center, where he flew tests in every type of naval aircraft including multi‑and single‑engine jet and propeller‑driven fighters, attack planes, patrol bombers, transports, and seaplanes.

From 1957 to 1959 Carpenter attended the Navy General Line School and the Navy Air Intelligence School and was then assigned as Air Intelligence Officer to the aircraft carrier, USS Hornet. In April 1959 he was selected as one of the original seven Mercury Astronauts.

Robert B. Voas was a US Navy psychologist posted to NASA, whose first job was to select suitable men for space duty. Warren North was chief of manned space flight. At an early briefing North confirmed that the potential spacemen would be chosen from pilots:

They would monitor and adjust the cabin environment. They would operate the communications system. They would make physiological, astronomical, and meteorological observations that could not be made by instruments. Most important, they would be able to operate the reaction controls in space and be capable of initiating descent from orbit. This was the key part, that the astronaut could take over control of the spacecraft itself.

Carpenter endured 30 hours of tests at the Lovelace & Wright‑Patterson Centre in 1959.

The chosen seven had shown emotional maturity, engineering, flight experience and motivation. Eighteen others were unreservedly recommended.

On 27 April 1959 they began work at Langley Air Force Base. NASA was considering 10 flights carrying chimpanzees but this changed on 12 April 1961. The Soviets made the first manned flight. Shepard followed on 5 May 1961. On 25 May President Kennedy presented his vision to Congress.

A couple of weeks later, Gilruth and Webb were aboard one of NASA’s R4Ds when over the radio the president was addressing Congress, pledging NASA to a lunar expedition. Gilruth was “aghast.” He looked at Webb, who knew all about it. In his special message to Congress, delivered on May 25, 1961, President Kennedy set out his vision on a number of “urgent national needs,” one of them the conquest of space. In a resonant call to arms, the president asked the nation to “commit itself to achieving the goal, before this decade is out, of landing a man on the moon and returning him safely to the earth.” No other space project, Kennedy declared, would be “more impressive to mankind, or more important for the long‑range exploration of space.”

Glenn’s flight had been launched by the Mercury Atlas 6 (MA‑6) which carried 122 tons of kerosene and liquid oxygen, more than four times the fuel load of the Redstone rockets which had powered Shepard and Grissom’s sub‑orbital flights. Scott Carpenter had been the back‑up to Glenn, who had left the face plate of his helmet open on re‑entry. At the formal inquiry he was cleared of the charge of panicking. Deke Slayton was scheduled to make the next orbital flight but during a G force test in August 1959 it had been noticed that his heartbeat was erratic. In January 1962 an Airforce cardiologist recommended that he should be grounded and Carpenter was given the next flight, designated MA‑7. Carpenter named his capsule Aurora 7.

Carpenter’s daughter, Kris Stoever, was six years old at the time and later helped her father write his account. Carpenter’s wife’s name was Rene. Kris Stoever:

Finally, at a little after seven forty‑five, the great Atlas engine were fired, sending out billows of steam, flames, dust, smoke, fumes and heat. At this signal, all four Carpenter children abandoned their posts in front of the television, where all three networks were covering the launch live, and dashed out to the beach. Already the pale morning sky was streaked with contrails, and in the distance they could see the Atlas lifting off. Against the low slant of the sun, Rene saw the Atlas streak into the sky and then disappear.

Just before liftoff, Scott had been thinking about his grandfather, Vic Noxon. “At last I’ll know the great secret,” the old man had told Dr Gilbert on a golden Sunday morning on the Front Range. He was dying. He knew he was dying. He wasn’t afraid. Scott was confident that May morning, like his Grandpa Noxon, that everything was going to be all right – that this experience so long anticipated had finally arrived. As the rocket engines began to rumble and vibrate beneath him, he became preternaturally alert to the many sounds and sensations of liftoff.

There was surprisingly little vibration, although the engines made a big racket and he felt the rocket swaying as it rose. The ride was gentler than he expected. He looked out his window, placed directly overhead, to see the escape tower streaking away like a scalded cat. One especially odd thing, for one accustomed to level flight after the required climb, was to see the altimeter reach seventy, eighty, then ninety thousand feet and yet know that he was still going straight up.

No one noticed at the time – there was no dial to measure its functioning – but the capsule’s pitch horizon scanner (PHS) had already started malfunctioning. The Mercury capsule was chockfull of automatic navigational instruments, among them the PHS, which does just what the name implies: it scans the horizon for the purposes of maintaining, automatically, the pitch attitude of the capsule. For MA‑7, however, the PHS immediately began feeding erroneous data into the Automatic Stabilization and Control System (ASCS), or autopilot. When this erroneous data was fed into the ASCS, the autopilot responded, as designed, to fire the pitch thruster to correct the perceived error. This in turn caused the spacecraft to spew precious fuel from the automatic tanks. Fuel was a finite commodity.

Forty seconds after tower separation, the pitch horizon scanner was already 18 degrees in error. It was indicating a nose‑up attitude, or angle, of plus‑17 degrees while the gyro on the Atlas showed pitch to have been minus 0.5. By the time of spacecraft separation, the pitch gyro aboard the capsule had “slaved” to the malfunctioning pitch‑scanner output and was in error by about 20 degrees. NASA later found that the error would persist, intermittently, to greater and lesser degrees, throughout the three‑orbit flight, with near‑calamitous effect as MA‑7 readied for reentry less than five hours later.

At the moment, Scott was focused on the gravitational forces, which peaked at a relatively gentle 8 Gs. He marvelled at the intense silence, but then experienced an even greater sensation of weightlessness. At five minutes, nine seconds into the flight, he reported to Gus listening as Capcom at Cape Canaveral: “I am weightless! – and starting the fly‑by‑wire turnaround.”

The sensation was so exhilarating, his report to the ground was more of a spontaneous and joyful exclamation than the routine report he had expected to make. The fly‑by‑wire manual controls were exquisitely responsive and quickly placed the Mercury capsule into a backward‑flying position for the beginning of Scott’s first circumnavigation of the earth. John had accomplished this maneuver on autopilot, as specified by his flight plan, causing the system to expend more than four pounds of fuel in the process. In the fly‑by‑wire control mode, it could be done using only 1.6 pounds.

The three‑axis control stick (or hand controller) designed for Mercury was a nifty device that allowed the pilot to fly the capsule in either the “manual proportional” or “fly‑by‑wire” (“wire” here meaning electrical) systems. The manual proportional system required minute adjustments of the control stick – of perhaps 2 or 3 degrees – to activate the one‑pound thrusters. Fore and aft movements controlled pitch, which is the up or down angle of the spacecraft, side‑to‑side movements controlled roll. The pilot could control or change the direction left or right, by twisting the control stick – a hand control that replaced the old rudder pedals used in airplanes The MA‑7 flight plan specified only limited use of the ASCS.

Gus Grissom, as Capcom, gave Carpenter the good news: “We have a Go, with a seven‑orbit capability.” Carpenter replied: “Roger. Sweet words.”

Carpenter described:

Sweet words indeed. With the completion of the turnaround maneuver, I pitched the capsule nose down, 34 degrees, to retroattitude, and reported what to me was an astounding sight. From earth‑orbit altitude I had the moon in the center of my window, a spent booster tumbling slowly away, and looming beneath me the African continent. But the flight plan was lurking, so from underneath the instrument panel I pulled out my crib sheets for the flight plan written out on three 3 x 5 index cards, and Velcroed for easy viewing. I could just slap them up on a nearby surface, in this case the hatch, covered with corresponding swaths of Velcro. Each card provided a crucial minute‑by‑minute schedule of in‑flight activities for each orbit. They gave times over ground stations and continents, when and how long to use what type of control systems, when to begin and end spacecraft maneuvers, what observations and reports to make on which experiments. In short, they told me, and the capcoms, who had copies, what I was supposed to be doing every second of the flight – every detail of which had been worked out, timed, and approved before liftoff. A brief investigation of these cards is enough to suggest constant pilot activity. But to get the best appreciation of just how busy we all were during those early flights, read the voice communication reports between the capcoms and the astronaut.

It was time to open the ditty bag. Stowed on my right, it contained the equipment and the space food for the flight. First out was the camera, for I needed to catch the sunlight on the slowly tumbling booster still following the capsule. The camera had a large patch of Velcro on its side. I could slap it on the capsule wall when it wasn’t in use. Velcro was the great zero‑gravity tamer. Without it, the equipment would have been a welter of tether lines – my idea, incidentally, and not a very good one, for John’s flight. He had ended up in a virtual spaghetti bowl full of tether lines and equipment floating through his small cabin.

Also in the ditty bag were the air‑glow filter, for measuring the frequency of light emitted by the air‑glow layer, star navigation cards, the world orbital and weather charts‑adjuncts to the earth path indicator (EPI) globe mounted on the instrument panel. The EPI was mechanically driven at the orbital rate so that it always showed the approximate spacecraft position over the earth. There were also bags of solid food I was to eat (a space first), and the densitometer.

But the most important items at this point in the flight were probably the flight plan cards. I had been tracking the booster since separation, maneuvering the capsule with the very good fly‑by‑wire system: “I have the booster in the center of the window now,” I reported, “tumbling very slowly.” It was still visible ten minutes later, when I acquired voice contact with Canary capcom.

Carpenter: “I have, west of your station, many whirls and vortices of cloud patterns. [Taking] pictures at this time – 2, 3, 4, 5. Control mode is automatic. I have the booster directly beneath me.”

The brilliance of the horizon to the west made the stars too dim to see in the black sky. But I could see the moon and, below me, beautiful weather patterns. But something was wrong. The spacecraft had a scribe line etched on the window, showing where the horizon should be in retro‑attitude. But it was now above the actual horizon I checked my gyros and told Canary capcom my pitch attitude was faulty.

Carpenter reported: “I think my attitude is not in agreement with the instruments.”

Then I added an explanation – it was “probably because of that gyro‑free period” – and dismissed it. There were too many other things to do.

John had also had problems with his gyro reference system. Kraft described it in an MA‑6 postflight paper, where he wrote that the astronaut “had no trouble in maintaining the proper [pitch] attitude” when he so desired “by using the visual reference.” All pilots do this – revert to what their eyes tell them when their on‑board tools fail. But future flights, he said, would be free of such “spurious attitude outputs” because astronauts would be able to “disconnect the horizon scanner slaving system,” called “caging the gyros” in these future flights. Because my flight plan for the follow‑on mission called for so many large deviations from normal orbital attitude (minus‑34‑degree pitch, 0‑degree roll, 0‑degree yaw), I was often caging the gyros when they weren’t needed for attitude control.

The Canary Capcom picked up on my report, and asked me to “confirm orientation.” Were my autopilot (ASCS) and fly‑by‑wire operating normally?

Carpenter reported: “Roger, Canary. The manual and automatic control systems are satisfactory, all axes…”

The procedure for voice reports on the attitude control system did not call for determining agreement in pitch attitude as shown by (a) the instrument and (b) the pilot’s visual reference out the window. The reporting procedure also assumed a properly functioning pitch horizon scanner, in the case of MA‑7 a false assumption. Because of the scanner’s wild variations careening from readings of plus 50 degrees at one place over the horizon and then lurching back to minus‑20 over another, without any discernible pattern – I might have gotten a close‑to‑nominal, or normal, reading at any given moment in the flight.

A thorough ASCS check, early in the flight, could have identified the malfunction. Ground control could have insisted on it, when the first anomalous readings were reported. Such a check would have required anywhere from two to six minutes of intense and continuous attention on the part of the pilot. A simple enough matter but a prodigious block of time in a science flight – and in fact the very reason ASCS checks weren’t included in the flight plan. On the contrary, large spacecraft maneuvers, accomplished off ASCS, were specified, in addition to how many minutes the MA‑7 pilot would spend in each of the three control modes‑fly‑by‑wire, manual proportional, and ASCS. Because of this, I would not report another problem with the ASCS until the second orbit. I had photographs to take and the balky camera to load.

When I spoke with Kano Capcom, over Nigeria, on that first pass, I was able to relay a lot of valuable orbital information as well as data on the control and capsule systems. I also checked out the radios and, as ordered by the flight surgeon, telemetered my blood pressure reading. While preparing to take the M.I.T. pictures of the “flattened sun” halfway through that pass, I saw I was getting behind in the flight plan and reported that I wouldn’t be able to complete the pictures on that pass. Just as I was making that report, I figured out the problem, managed to install the film, and was able to take the pictures after all.

Before I lost voice contact with Kano Capcom, I was able to get horizon pictures with the M.I.T. film. The first picture was at f8 and 1/125 taken to the south directly into the sun. The second picture was taken directly down my flight path, and the third was 15 degrees north of west at “capsule elapsed time” (elapsed time since launch) of 00 30 17.1 was very busy.

Tom Wolfe wrote in The Right Stuff that I was having “a picnic” during my flight and “had a grand time” with the capsule maneuvers and experiments. He kindly noted that my pulse rate before liftoff, during the launch and in orbit, was even lower than Glenn’s admirably calm readings. The second part, about my pulse rate, may be true, perhaps because nature wired me that way (and Wally, too, for that matter, if you look at his telemetered readouts). But Wally and I were also following in John’s historic steps, had been fully briefed, and knew pretty much what to expect. Knowledge and training create confidence.

MA‑7 was no picnic. I had trained a long time, first as John’s backup, and then for my own surprise assignment to the follow‑on flight. To the extent that training creates certain comfort levels with high‑performance duties like spaceflight, then, yes, I was prepared for, and at times may even have enjoyed, some of my duties aboard Aurora 7. But I was deadly earnest about the success of the mission, intent on observing as much as humanly possible, and committed to conducting all the experiments entrusted to me. I made strenuous efforts to adhere to a very crowded flight plan.

The cabin became noticeably hot during the first orbit, when I was over the Mozambique channel, forty‑five minutes into the flight. I wasn’t the first astronaut to be bothered by a hot cabin, and all of us were prepared for varying degrees of discomfort, and even pain, while we trained for and went through actual space flight. During the selection process, we ran the treadmill at 100 percent humidity and 115 degrees Fahrenheit – and gladly – just to be chosen.

So the term “tolerable temperature,” something the NASA medics determined was endurable with little loss in performance, is relative. You need to know how long the discomfort will likely last, how hard you have to work during that time, and how badly you need to withstand it. It also helps to have an idea of when you believe relief will come. So after giving the Indian Ocean capcom all the normal voice reports, I explained for the record what I was doing inside to bring the high cabin temperatures down.

During all this time, I was also getting some readings with O’Keefe’s airglow filter. All of a sudden my periscope went dark. It really surprised me.

Carpenter reported: “What in the world happened to the periscope? Oh. It’s dark. That’s what happened. It’s facing a dark earth.”

A simple and elegant explanation: day had become night. I was still getting accustomed to moving 17,500 miles per hour.

My flight plan at this point consisted mostly of photography. I had crossed the terminator, which is the dividing line between the dark and sunlit sides of the earth, which caused the light levels to change very rapidly. It was exceedingly important that I photograph the changing light levels. To myself, I read off a lot of camera F‑stop and exposure values and was thinking aloud about my next capcom.

Carpenter reported: “It’s getting darker. Let me see. Muchea contact sometime – Oh, look at that sun! F11.”

No one was listening, so I reported to the tape: “It’s quite dark. I didn’t begin to get time to dark adapt… cabin lights are going to red at this time. Oh, man, a beautiful, beautiful red, like in John’s pictures. Going to fly by wire.”

A mysterious red light had cascaded through the window just as I went into a new control mode, as specified in the flight plan. It reminded me of the pictures John had taken through his red filter. But mine was only the reflection of the red cabin lights. “That’s too bad.” I was disappointed.

But then I was visited by Venus.

Carpenter reported: “I have Venus now approaching the horizon. It’s about 30 degrees up. It’s just coming into view. Bright and unblinking. I can see some other stars down below Venus. Going back to ASCS at this time. Bright, bright blue horizon band as the sun gets lower and lower – the horizon band still glows. It looks like five times the diameter of the sun.”

The sun completely disappeared at this point in my flight, and I reported the exact time – 00 4734 elapsed – and my total incredulity.

Carpenter reported: “It’s now nearly dark and I can’t believe where I am.”

My wonder gave way to surprise just a minute later, when I saw how much fuel I had already used.

Carpenter reported: “Oh, dear, I’ve used too much fuel.”

“Oh, dear” – a Noxon expression. Over Australia, I would have voice contact with two different capcoms – the first with Deke Slayton at Muchea, the second at Woomera. Over Muchea, Deke and I talked about our Australian friends, John Whettler in particular, who had been a Spitfire pilot during World War II. Then I said “Break, Break,” which is voice communication procedure meaning “change of subject.” We talked about cloud cover, too heavy for me to see the lights in Perth turned on for my encouragement. Deke consulted the flight plan and saw it was time to send some telemetered blood pressure readings. Then some arcane navigational matters – how to determine attitudes, yaw, pitch, and roll – on the dark side.

Carpenter reported: “You’ll be interested to know that I have no moon, now. The horizon is clearly visible from my present position; that’s at 00 54 44 [capsule] elapsed. I believe the horizon on the dark side with no moon is very good for pitch and roll. The stars are adequate for yaw in, maybe, two minutes of tracking. Over.”

In 1962 we didn’t know what was visible on the horizon, on the dark side without moonlight. So Deke and I were discussing how one might establish attitude control under such unfavorable conditions. I relayed what reliable visual references I had out the window or periscope. In the absence of valid attitude instrument readings during retrofire, the pilot can use such external visual references, manually establishing proper retroattitude control with the control stick. Pitch attitude can be established and controlled easily, with reference to the scribe mark etched in the capsule window. Accomplishing the proper yaw attitude, however, is neither easy nor quick.

Attitude changes are also hard to see in the absence of a good daytime horizon. At night, when geographic features are less visible, you can establish a zero yaw attitude by using the star navigation charts, a simplified form of a slide rule. The charts show exactly what star should be in the center of the window at any point in the orbit – by keeping that star at the very center of your window you know you’re maintaining zero yaw. But there are troubles even here, for the pilot requires good “dark adaption” (or a dark‑adapted eye) to see the stars, and dark adaption was difficult during the early flights because of the many light leaks in the cabin. The backup measures (“backup” here meaning human) were absolutely critical to have in place at retrofire – in the event of attitude instrument failure.

Deke and I discussed suit temperature, which like the cabin was hotter than I liked. He suggested a different setting, which I tried. Then Woomera capcom hailed me, and I replied: “Hello, Woomera capcom, Aurora 7. Do you read?” while still in voice contact with Deke, at Muchea. “Roger, this is Woomera,” came the capcom’s voice. “Reading you loud and clear. How me?” Deke was confused. He couldn’t hear Woomera and thought to correct me.

Between Muchea and Woomera, I was trying to see the ground flares, a check for visibility. Deke gave me the attitudes to view the first flare, which involved a whopping, plus‑80 degrees yaw maneuver and a pitch attitude of minus 80 degrees. But the cloud cover was too dense. “No joy on your flares,” I told Woomera and then went to drifting flight, where I found that just by rocking my arms back and forth, like attempting a full twist on the trampoline, I could get the capsule to respond in all three axes, pitch, roll, and yaw.

The Cape advised me to keep the suit setting where it was, because the temperature was coming down. I continued in drifting flight, and at capsule elapsed 01 02 41.5, over Canton, we checked attitude readings with telemetry. The Canton Capcom told me my body temperature was registering 102 degrees Fahrenheit, clearly a false reading.

Carpenter reported: “No, I don’t believe that’s correct. My visor was open; it is now closed. I can’t imagine I’m that hot. I’m quite comfortable, but sweating some.”

A food experiment had left crumbs floating in the cabin. I remarked on them, and reported the dutiful downing of “four swallows” of water. At his prompting, however, I could not confirm that the flight plan was on schedule. But I reported what I could: “At sunset I was unable to see a separate haze layer – the same height above the horizon that John reported. I’ll watch closely at sunrise and see if I can pick it up.”

Canton Capcom wished me “good luck,” and then LOS – loss of signal.

Everyone on the ground had had an eye on the fuel levels since the end of the first orbit. Gordo Cooper, capcom at Guaymas, had told me to conserve fuel, which was then at 69 percent capacity for the manual supplies, and 69 percent for the automatic. By the time I returned for my second pass over Kano, they had dropped to 51 and 69, respectively.

Carpenter reported: “The only thing to report is that fuel levels are lower than expected. My control mode now is ASCS.”

I explained to the Kano Capcom: “I expended my extra fuel in trying to orient after the night side. I think this is due to conflicting requirements of the flight plan.”

Live and learn. I spoke to the flight recorder, although Kano Capcom still had voice contact.

I should have taken time to orient and then work with other items. I think that by remaining in automatic I can keep – stop this excessive fuel consumption.

When I went to fly‑by‑wire aboard Aurora 7, very slight movements of the control stick in any axis activated one‑pound thrusters and changed the attitude very slowly. Larger stick movements would activate the twenty‑four‑pound thrusters, which would change the attitude much more quickly but use twenty‑four times as much fuel. If the manual proportional control mode were chosen, the change capsule attitude would be proportional to stick movement, just as an airplane. (Move the stick a little, get a little bit of thrust; move it halfway, get half thrust; move it all the way, get full thrust.) Each increment of movement had attendant increases in fuel expenditure. If, however, both control modes were chosen concurrently – and this happened twice during MA‑7 as a result of pilot error – then control authority is excessive and fuel expenditure exorbitant.

For my flight the twenty‑four‑pound thrusters came on with just a wrist flick, that I then corrected with a wrist flick in the other direction. This countermovement often activated the twenty‑four‑pound thrusters yet again, all for maneuvering power not required during orbital flight. The high thrusters weren’t needed, really, until retrofire, when the powerful retrorockets might jockey the capsule out of alignment. The design problem with the three‑axis control stick as of May 1962 meant the pilot had no way of disabling, or locking out, these high‑power thrusters. Because of my difficulties and consequent postflight recommendations, follow‑on‑Mercury flights had an on‑off switch that would do just that, allowing Wally Schirra and Gordo Cooper to disable the twenty‑four‑pound thrusters. Gemini astronauts had a totally different reaction control system.

But I understood the problem and resolved to limit my use of fuel. Consulting my index cards, I saw that I still had voice reports to make on several experiments – the behavior of the balloon, still tethered to the spacecraft; a night‑adaption experiment; and the ingestion of some more solid food. Holding the bag, however, I could feel the crumbled food. If I opened it, food bits would be floating through my work space. I made a mental note: “Future flights will have transparent food bags.” See‑through bags would make crumb strategy easier during these zero‑G food deployments. I was beginning to regret my lack of training time.

Before loss of signal, Kano Capcom asked me to repeat my fuel‑consumption critique.

Capcom asked: “Would you repeat in a few words why you thought the fuel usage was great? Over.”

Carpenter replied: “I expended it on – by manual and flyby – wire thruster operation on the dark side, and just approaching sunrise. I think that I can cut down on fuel consumption considerably during the second and third orbits. Over.”

The Zanzibar Capcom took over ground communication. Consulting the same flight plan I had, he reminded me I was supposed to be on fly‑by‑wire. I thought better of it and said so:

“That is negative. I think that the fact that I’m low on fuel dictates that I stay on auto as long as the fuel consumption on automatic is not excessive. Over.”

The irony is that even the ASCS control mode, ostensibly thrifty with fuel, was now guzzling fuel because of the malfunctioning pitch horizon scanner. “Roger, Aurora 7,” replied Zanzibar Capcom and then congratulated me on my trip so far. “I’m glad everything has gone‑” but the rest of this message dropped out. “Thank you very much,” I said, hoping he could still hear me.

After Zanzibar was the Indian Ocean Capcom, stationed aboard a United States picket ship called Coastal Sentry , permanently anchored at the mouth of the Mozambique channel. After the usual preliminaries (“How do you read?” ‘Loud and clear. How me?’), he reminded me to conserve fuel and then inquired: ‘Do you have any comments for the Indian Ocean?’ I replied, but not with a greeting. I was having that old ASCS difficulty:

“That is Roger. I believe we may have some automatic mode difficulty. Let me check fly‑by‑wire a minute.”

Going to fly‑by‑wire is the best way to diagnose any problem with the thrusters, the small hydrogen peroxide‑spewing jets that control spacecraft attitudes. I checked them again. The thrusters were fine. We didn’t know it at the time, but the thrusters were receiving faulty information, through the autopilot, from the pitch horizon scanner. Worse, the error from the automated navigational tool was intermittent and thus hard to identify. I reported that the gyros, my onboard navigational tools, were not “indicating properly.” This sort of problem requires patient investigation. I told the Indian Ocean Capcom to wait.

Carpenter reported: “The gyros are… okay, but on ASCS standby [the off position]. It may be an orientation problem. I’ll orient visually and see if that will help out the ASCS problem.”

I went off autopilot to fly‑by‑wire, oriented the capsule visually, and then returned to ASCS autopilot, to see what would happen. My hope was to catch the autopilot misbehaving. It was an angel. Imagine that you own a high‑performance car that develops a quirky habit, when on autopilot, of veering off the interstate as you’re speeding along at 80 miles per hour. You take it to the dealer, describe the trouble, and the mechanics can’t duplicate the malfunction when they take it out to the freeway the next day. Imagine this happening in space, with your space car, and you have only two circumnavigations left on the orbital hightway. Imagine further that your precisely timed exit off the orbital highway will be performed using this intermittently malfunctioning autopilot. This is what I was facing, but didn’t know it. No one did.

Technicians, pilots among them, often make erroneous assumptions when troubleshooting a problem. An erroneous assumption early on can invalidate all subsequent efforts to find a solution. Nobody realized that the problem lay in the pitch attitude indicator. From the pilot’s viewpoint, the problem with the ASCS was an anomaly, and the intermittent failure meant little. When your navigational tools disagree with the view out your window and this persists in any great disparity, the instruments are malfunctioning. When the instruments are malfunctioning, you have no recourse but to navigate visually with reliable reference points – the horizon, the position of a known star, geographical landmarks. This is what I did.

The Indian Ocean Capcom waited patiently. Nearly a minute passed while I tried diagnosing the problem. We were working off a tight flight plan, so he reminded me I was “supposed to, if possible, give a blood pressure.” This was a simple matter of pressing a semi‑automatic device on my suit, which I did, and felt the blood pressure cuff inflate. “Roger,” I said, “I’ve put blood pressure up on the air already. Over.”

Mercury Control had in the meantime picked up on my earlier transmission about the thrusters. During MA‑6, a thruster malfunction had forced John to assume manual control for his final two orbits. Rightly concerned about a repeat of the old problem, Mercury Control pressed the capcom to get me to submit a complete report on the thrusters.

Capcom ordered: “Report to Cape you have checked fly‑by‑wire, and all thrusters are okay. Is there anything else?”

“Negative,” I said. Mercury Control was working on an erroneous assumption about the thrusters malfunctioning and needed to be sure I had checked them thoroughly. Having satisfied my own questions about the thrusters, and done the best I could with the ASCS, I had moved on to grappling with my spacesuit’s coolant and steam‑vent settings and said so: “Except for this problem with steam‑vent temperature.” It wasn’t the heat now, but the humidity, in this case inside my suit: I knew that the cabin temperatures were high, at about 103 degrees. The dry air would at least provide some evaporative relief from the sweat now pouring down my forehead, plowing through my eyebrows, and stinging my eyes with salt.

Carpenter reported: “I’m going – I’ll open the visor a minute, that’ll cool – it seems cooler with the visor open.”

The Capcom persisted. Mercury Control needed me to reconfirm that I had used the fly‑by‑wire control system to check out all the thrusters.

Capcom replied: “Aurora 7, confirm you’ve checked fly‑by‑wire, and all thrusters are okay.”

Carpenter replied: “Roger. Fly‑by‑wire is checked, all thrusters are okay.”

But the information coming from the horizon scanner was faulty. During the orbital phase of spaceflight, a malfunctioning automated navigational system is tolerable – for my flight this was especially so because the ASCS was so rarely used. But during an ASCS‑controlled retrofire – that critical exit off the orbital highway – an accurate horizon scanner is crucial. For retrofire, the spacecraft must be aligned exactly in two axes – pitch and yaw. Pitch attitude, or angle, must be 34 degrees, nose down. Yaw, the left‑right attitude, must be steady at 0 degrees, or pointing directly back along the flight path. The ASCS performs this maneuver automatically, and better than any pilot, when the on‑board navigational instruments are working properly.

If the gyros are broken, all is not lost: a pilot can do two things to bring yaw attitude to zero. The first is to point the nose in a direction he thinks is a zero‑degree yaw angle and then watch the terrain pass beneath the vehicle. This is nearly impossible to do over featureless ocean or terrain. Far better to have a certain geographical feature or cloud pattern to watch. Because the pilot is traveling backward, the geographical features he is trying to track must begin at the bottom of the window and flow in a straight line from there to the top. When this happens, the pilot knows he is in a zero‑degree attitude. This can be done through the periscope too, but it takes a little longer and is less accurate.

My travails with a hot cabin and a humid spacesuit continued over Australia. Deke, the Muchea Capcom, assumed ground communications. It was his unhappy job to tell me that my cabin temperatures had climbed to 107 degrees Fahrenheit (they would peak, during the third orbit, at 108 degrees). Dehydration under such conditions is a worry, and for these and other reasons NASA medics had lobbied for some of the capcom posts, to no avail. By the time I had completed another solid‑food experiment, by eating some Pillsbury‑made morsels, I was within voice range of the next Australian capcom, at Woomera, and still fussing with my suit temperature controls. The capcom there asked me for suit temperature and humidity readings. They were at 74 degrees Fahrenheit, with the “steam exhaust” registering a miserable 71 degrees of humidity inside my suit. Still, the numbers had come down since Australia, so the Woomera Capcom asked rather hopefully:

“Are you feeling more comfortable at this time?”

A noncommittal “I don’t know” was the best I could manage. I was frustrated with the suit controls and realized with exasperation that for all the exhaustive testing of the suits prior to this and other early launches, no one thought to test its cooling capacity with the face‑plate open! And so many in‑flight activities required me to keep my visor up.

Carpenter reported: “I’m still warm and still perspiring. I would like to – I would like to nail this temperature problem down. It – for all practical purposes, it’s uncontrollable as far as I can see.”

Capcom asked: “How about water?”

Carpenter replied: “That would be a no.”

Carpenter reported: “I had taken four swallows at approximately this time last orbit. As soon as I get the suit temperature pegged a little bit, I’ll open the visor and have some more water. Over.”

At this point in the flight, over Canton, I was scheduled to take a xylose pill (which is a biomedically traceable sugar pill for later analysis in my collected urine). I could feel the melted Pillsbury mess in the plastic bag and said, “I hate to do this,” more to myself than to the Canton Capcom. Then, surprise, when I opened it: “It didn’t melt!” I found the xylose pill, but all my cookies had crumbled. Chocolate morsels escaped their confines to float, weightless, around my tiny workspace. The rest of the stuff in the bag was a mess. The Nestle concoction, more fruit and nougat than heat‑sensitive chocolate, held up far better.

I was approaching Hawaii, and my second sunrise in space. Referring to the flight plan, the Canton Capcom prompted me, before LOS, for an update on the balloon experiment: “Which of the five colors was most visible?”

Carpenter reported: “I would say that the day‑glow orange is best.”

Capcom replied: “Roger. For your information, the second sunrise should be expected in approximately 3 to 4 minutes.”

“The Surgeon is after me here,” he added, for another blood pressure check. “Is this convenient?” My in‑flight duties at sunrise called for vigorous physical activity, so I waved him off:

“Negative. I won’t be able to hold still for it now. I’ve got the sunrise to worry about.”

He let me alone.

Sunrises and sunsets were extremely busy time‑blocks during Mercury flights. There were important measurements to make of the airglow and other celestial phenomena and innumerable photographs to take.

John O’Keefe had some solid hypotheses about the “fireflies” John had seen during his flight. But they remained unexplained. Whatever the critters were, they were particularly active, or at least visible, at dawn, adding to the scientist‑pilot’s burden. At 02 49 00 I reported the arrival of a beautiful dawn in space: “I’ll record it,” I told the Canton Capcom, “so you can see it.” As a patrol plane pilot, I had trained to serve as the U.S. Navy’s eyes and ears – a militarily indispensable role. In space, as a Mercury astronaut, I was now the eyes and ears for an entire nation. I felt an obligation to record what few would ever have a chance to see.

I was just beginning to go through my crowded schedule of sunrise‑related work when Hawaii took over from Canton, announcing, “Hawaii Com Tech. How do you read me?” prompting me for a short report. The Navy has a one‑through‑five scale for grading the volume and clarity of voice transmissions. An old Navy quip came to mind, “I read you two by two” – a voice‑report short‑hand for “too loud and too often.” But I reserved the smart answer and said only, “Stand by one. My status is good. My capsule status is good. I want to get some pictures of the sunrise. Over.”

Capcom asked for a fuel consumption report. Carpenter reported that his fuel supplies were 45–62.

The 45–62 figures were the percentages of Aurora 7’s fuel supply. I had less than half my manual fuel supply left; my automatic fuel supplies stood at 62 percent. Not alarmingly low yet, but low enough. Still Kraft, directing the flight from the Cape, later reported that he wasn’t worried: “except for some overexpenditure of hydrogen peroxide fuel,” he wrote in his own postflight analysis of MA‑7, “everything had gone perfectly.” I still had 40 percent of my manual fuel, which, “according to the mission rules,” Kraft figured, “ought to be quite enough hydrogen peroxide… to thrust the capsule into the retrofire attitude, hold it, and then to reenter the atmosphere using either the automatic or the manual control system.”

But I myself was running low on water – hadn’t drunk any even after the prompting over Woomera. This was a mistake. I was in good physical condition and could tolerate dehydration, but I still should have been drinking copious amounts of water to compensate for what I was losing through sweat and respiration. Someone, the flight surgeon, directed the Hawaii Capcom to inquire about my water intake:

“Did you drink over Canton? Did you drink any water over Canton?”

Carpenter replied: “That is negative. I will do, shortly.”

The water would have to wait. But Hawaii Capcom persisted: “Roger. Surgeon feels this is advisable.” More cabin and suit temperature readings were asked for and given. It was at this time that Mercury Control, alert to potential problems, had pondered one of my earlier voice reports (at capsule elapsed 02 08 46) about the difficulty I was having, not with the thrusters, but with the ASCS. It directed Hawaii Capcom to have me conduct an ASCS check:

“Aurora 7. This is Capcom. Would like for you to return gyros to normal and see what kind of indication we have: whether or not your window view agrees with your gyros.”

Sixteen seconds passed. I was feverishly working through the sunrise‑related scientific work, too busy to drink water, too busy to send a telemetered blood pressure reading, and ground control had just asked me to perform an attitude check. “Roger. Wait one,” I replied.

Mercury Control had chosen an awkward moment to troubleshoot the (intermittently) malfunctioning ASCS. They wanted an attitude check, at dawn, over a featureless ocean while I was busily engaged with the dawn‑related work specified in my flight plan. Again, adequate checks for attitude, particularly in yaw, are difficult enough in full daylight over recognizable land terrain, requiring precious minutes of continuous attention to the view of the ground out your periscope and the window. In my postflight report I explained the difficulty.

Manual control of the spacecraft yaw attitude using external references has proven to be more difficult and time‑consuming than pitch and roll alignment, particularly as external lighting diminishes… Ground terrain drift provided the best daylight reference in yaw. However, a terrestrial reference at night was useful in controlling yaw attitudes only when sufficiently illuminated by moonlight. In the absence of moonlight, the pilot reported that the only satisfactory yaw reference was a known star complex nearer the orbital plane.

But Mercury Control had requested an attitude check, and I complied, first reporting that I had to get back within “scanner limits,” that is, to an attitude in which the horizon was visible to the pitch horizon scanner. That required more maneuvering, which required more fuel. I was still trying to cram in more observations.

Capcom asked: “Can we get a blood pressure from you, Scott?”

I sent the blood pressure, reported on the transmission, and continued voice reports on the experiments: the behavior of the “fireflies”; the balloon, still shadowing Aurora 7 like a stray animal, was oscillating some. Just before LOS, I reported I was going to “gyros normal. Gyros normal now.” Hawaii Capcom replied: “Roger, TM [telemetry] indicates your‑zero pitch.” And then “LOS, Scott, we’ve had LOS.”

Loss of signal. I was moving on to voice contact with Al Shepard, California Capcom, and approaching the start of my final, most perilous circumnavigation of the planet.

Kris Stoever continued:

The pilot of Aurora 7 speeded toward California, where Al Shepard was capcom, in charge of ground communications. Scott first gave Al his short report on fuel, cabin‑air temperature, and control mode (“manual, gyros normal, maneuver off”). But then the important issue: the suit steam‑exhaust temperatures. They were “still reading,” he told Al dispiritedly, “70 degrees.”

But Al had good news:

“Understand you’re GO for orbit three.”

While the GO business was nice to hear, it was really hot in the cabin, and Scott still had lots of work to do. As it happened, more than the MA‑7 cabin temperatures were hot. From all reports, Kraft was full‑out fuming as Scott approached the continental United States. The flight director appears to have concluded, erroneously, that the pilot of MA‑7 had deliberately ignored his request for an attitude check over Hawaii. Now, in addition to his anxieties about fuel use, Kraft was nursing a grudge about a snub that never took place.

In his memoir, he writes that as Carpenter approached California, he directed Al Shepard, the California Capcom, to set things right. Al’s new job, Kraft told the famously self‑possessed Navy commander, was “to find what the hell was going on up there,” adding that he left the California Capcom with “no doubt” about his “frustration” with Carpenter. Kraft was in fact bellowing through the earpieces of Al’s headset.

The flight director told Al he needed two things from Scott: an attitude check and a tight curb on fuel use. In an exercise of judgment as California Capcom, Shepard relayed just one of Kraft’s two requests:

“General Kraft is still somewhat concerned about your auto fuel. Use as little auto – use no auto fuel unless you have to prior to retrosequence time.”

Shepard then turned to the matter at hand, which was the heat in the cabin and an apparently malfunctioning heat exchanger in Scott’s suit. He suggested another, more comfortable setting. He omitted Kraft’s request for an attitude check. Al then did unto Scott as he hoped others might one day to unto him, offered the pilot a little time, a little quiet, and some encouragement:

“Roger. You’re sounding good here. Give you a period of quiet while I send Z and R cal.”

The two men carried out these quiet space chores over the next three and a half minutes. Then Al gathered information. Either he knew enough to ask, or he was prompted by the flight surgeon:

“Do you – have you… have you stopped perspiring at the moment?”

No, Scott told him, he was “still perspiring.” A good sign. No impending heat stroke. Catching the drift of the conversation, Scott reported he might open his visor “and take a drink of water.”

Capcom acknowledged: “Roger. Sounds like a good idea.”

He let Scott drink. Sixteen quiet seconds passed. Then Al asked a question. Note the man’s impeccable manners:

“Seven, would you give us a blood pressure, please, in between swallows.”

It was a remarkable moment of earth‑to‑space human solicitude. A minute later, a refreshed Scott reported:

“Twenty swallows of water. Tasted pretty good.”

Capcom replied: “Roger, Seven, we’re sure of that.”

In a final, reassuring exchange before LOS, the California Capcom would send Aurora 7 on her way:

“Seven, this is California. Do you still read?”

Carpenter replied: “Roger, loud and clear.”

Capcom: “Roger, we have no further inquiries. See you next time.”

The “next time” would bring the two men, Shepard and Carpenter, together again in an even more life‑saving conspiracy of astronauts.

After four hours in orbit and a long period of drifting flight, Aurora 7’s cabin temperature had dropped to 101 degrees Fahrenheit; the vexing problems with the suit temperature were being resolved. The balky camera was now a memory. Scott had succeed in shooting all the M.I.T. film for the “flattened sun” photographs. The experiment on the behavior of liquids in zero‑G was a success. Capillary action can pump liquids in space. Over Woomera, Scott described and analyzed various successful valve settings for his suit – in‑flight observations that would assist with a later redesign of the Mercury suit; he also took photometry readings and measurements on Phecda, a star in the Big Dipper, then sinking into the haze layer of the horizon. His short report had good news:

“I’m quite comfortable. Cabin temperature is 101… fuel reads 46 and 40 percent. I am in drifting flight. I have had plenty of water to drink.”

For the next eleven minutes of spaceflight Scott transmitted an uninterrupted flow of observations. The partially inflated balloon, which had failed to jettison as planned over the Canaries, still bumped along behind the capsule. It kept “a constant bearing,” Scott reported, “at all times.” Still transmitting to Woomera:

“I have 22 minutes and 20 seconds left for retrofire. I think I will try to get some of this equipment stowed at this time.”

Coming up on sunrise, rich with “observables,” the pilot of Aurora prepared for one final observation of the airglow phenomenon, reporting for the tape recorder:

“There is the horizon band again; this time from the moonlit side.”

Carpenter complained once more about the light leak:

“Visor coming open now. It’s impossible to get dark‑adapted in here.”

NASA had molded an eye patch for John Glenn so he could keep an eye covered through daytime on one orbit and emerge on the night side with a dark‑adapted eye. But the small cabin was so dusty the sticky tape (designed to keep the patch secured over his eye socket) became covered with dust and would not stick to his skin. NASA did not reattempt this dark‑adaption patch with MA‑7. Managing to get a good view through the filter, Scott continued: “Haze layer is very bright through the air glow filter. Very bright.” He then concentrated on the photometer measurements, reporting with some puzzlement that the width of the airglow layer was exactly equal to the width of the X inscribed on the lens. “I can’t explain it – I’ll have to – to–”

And then the sunrise, at 04 19 22. Scott would remember the sunrises and sunsets as the most beautiful and spectacular events of his flight aboard Aurora 7. “Stretching away for hundreds of miles to the north and the south,” they presented “a glittering, iridescent arc” of colors that, he later wrote, resolved into a “magnificent purplish‑blue” blending, finally, with the total blackness of space. Thinking the camera might help with the air‑glow measurement, he quickly grabbed for it and in doing so inadvertently rapped the spacecraft hatch.

A cloud of tiny, luminous particles swarmed past the window.

“Ahhhhhhhh!” he exclaimed, to the tape recorder. “Beautiful lighted fireflies that time,” explaining, “it was luminous that time.” Banging repeatedly on the hatch, he was rewarded with explosions of cloud after cloud of luminous particles from the spacecraft.

“If anybody reads,” Scott explained excitedly, “I have the fireflies. They are very bright. They are,” he announced with triumph, “capsule emanating!” He quickly explained the cause and effect that proved his finding: “I can rap the hatch and stir off hundreds of them. Rap the side of the capsule: huge streams come out.”

He would yaw around the other way to get a better view, he reported. With his photometer handy, Scott estimated that the fireflies might register at a nine on the device and proposed to find out. “I’ll rap,” he told Woomera, now out of range. “Let’s see.”

The official NASA history of Project Mercury notes that:

Until Aurora 7 reached the communication range of the Hawaiian station on the third pass, Christopher Kraft, directing the flight from the Florida control center, considered this mission the most successful to date; everything had gone perfectly except for some overexpenditure of hydrogen peroxide fuel.

This overexpenditure was traced to a spacecraft system malfunction that went undiagnosed until after the flight.

At 04 22 07, Hawaii Capcom established ground communications.

Carpenter responded: “Hello, Hawaii, loud and clear. How me?”

But the signal from Aurora 7 was weak, so for half a minute pilot and Hawaii Capcom struggled with communication frequencies.

Carpenter asked: “Roger, do you read me or do you not, James?”

Capcom replied: “Gee, you are weak, but I read you. You are readable. Are you on UHF‑Hi?”

Carpenter confirmed: “Roger, UHF‑Hi.”

Reading off the flight plan, the capcom immediately told Scott to reorient the capsule and go to autopilot – the old ASCS. Scott replied six seconds later: “Roger; will do,” and, complying, at 04 22 59, repeated:

“Roger; copied. Going into orbit attitude at this time.” Retrosequence, as both Scott and capcom were aware, was fast approaching. With retro‑rockets to be fired at 04 32 30 – ten minutes away – the flight plan called for equipment stowage and retrosequence checklists to begin at 04 24 00, allotting two minutes for these tasks, and then one more minute until LOS. Hawaii Capcom’s sense of urgency was evident:

“Roger, are you ready to start your pre‑retrosequence checklist?”

Carpenter confirmed: “Roger, one moment.”

The Navy adage, “Aviate, Navigate, Communicate,” always in that order, was never more apt – now for the first time in space. In the grip of this instinct, Scott was properly engrossed with a critical retrosequence maneuver. He finally explained to Hawaii Capcom:

“I am aligning my attitudes. Everything is fine.”

Anticipating the capcom’s request, he said: “I have part of the stowage checklist taken care of at this time.”

Stowage is important. You can’t have equipment flying around the cramped compartment during entry. More important still, however, is aligning the spacecraft. Twice more, at 04 25 11 and 04 25 55, the Hawaii Capcom prompted Scott to begin the pre‑retrosequence list.

Capcom asked: “Aurora 7, can we get on with the checklist? We have approximately three minutes left of contact.”

Carpenter confirmed: “Roger, go ahead with the checklist. I’m coming to retroattitude now, and my control mode is automatic, and my attitudes [are] standby. Wait a minute. I have a problem in–”

Thirty‑three seconds passed. Scott confirmed the “problem.”

Carpenter reported: “I have an ASCS problem here. I think ASCS is not operating properly. Let me – emergency retrosequence is armed and retro manual is armed. I’ve got to evaluate this retro – this ASCS problem, Jim, before we go any further.”

Thirty seconds of silence ensued, for good reason. The automatic pilot was not holding the capsule steady for retrosequence. Again, at retrofire (an event that determines your landing point three thousand miles away) the capsule’s pitch attitude must remain steady 34 degrees, nose down. Yaw angle, too, steady at zero degrees. These two attitudes, in conjunction with a precisely timed retrofire, precisely determine the capsule’s landing point. At retrofire, two‑thirds of the impulse, or thrust, delivered to the capsule at 34 degrees, nose down, tends to slow the capsule down; the remaining third tends to alter the capsule’s flight path downward. If yaw and pitch attitudes, together with the timing of retrofire, are correct, then both events – the reductions in speed and altitude – would send Aurora 7 homeward along the predetermined reentry path, somewhere in the waters southeast of Florida.

Mindful of these contingencies, the Hawaii Capcom, Jim, replied: “Roger,” told Scott he was standing by, and squeezed in two critical retrosequence items – the pilot was to switch off the emergency drogue‑deploy and emergency main fuses. Scott replied:

“Roger, they are. Okay, I’m going to fly‑by‑wire, to Aux Damp, and now – attitudes do not agree. Five minutes to retrograde, light is on. I have a rate of descent, too, of about 10,12 feet per second.”

Hawaii Capcom did not hear and transmitted: “Say again? Say again?” He had a rate of descent, Scott repeated, “of about twelve feet per second.” The capcom asked: “What light is on?” Things were happening quickly. Scott replied only, “Yes, I am back on fly‑by‑wire. Trying to orient.” With only a minute until LOS, the Hawaii Capcom finally proposed a run through the checklist. Carpenter finally said: “Okay. Go through it, Jim,” and then, prompting, once more, “Roger, Jim. Go through the checklist for me.”

Approaching the most critical moment of the flight, Hawaii Capcom and the pilot of Aurora 7 used the remaining minute of voice contact to report back and forth on the arming of various squib switches, the periscope levers, up or down? Manual fuel handles (as backup for the ASCS) – were they in or out? Finally:

“Roll, yaw, and pitch handles are in.”

Capcom transmitted: “Transmitting in the blind… We have LOS. Transmitting in the blind to Aurora 7. Make sure all your tone switches are on, your warning lights are bright… Check your face‑plate is closed.”

With Aurora 7 nearing reentry, Kraft learned with as much dismay as the pilot himself that the spacecraft’s ASCS was not, in the best traditions of astronaut understatement, “operating properly.”

Scott had in fact noticed the symptoms, now and then, of a malfunctioning pitch horizon scanner, and was puzzled, at times, by some instrument readings. He reported them as the anomalies they were. But the intermittent nature of these instrument failures made repeated checking of little value. The view out the window was a very good backup, and it was impervious to failure.

It was clear at Mercury Control that day that Kraft’s indignation, simmering since the second‑orbit incident over Hawaii, was now compounded by the man’s genuine anxiety. Speaking of MA‑7 Kranz explains: “A major component of the ground team’s responsibility is to provide a check on the crew.” And the ground, Kranz says, “waited too long in addressing the fuel status and should have been more forceful in getting on with the checklists.” A thoroughgoing attitude check during the first orbit would probably have helped to diagnose the persistent, intermittent, and constantly varying malfunction of the pitch horizon scanner. By the third orbit it was all too late. MA‑7’s fuel problems dictated drifting flight. A third‑orbit attitude check, particularly in yaw, would have used prodigious amounts of fuel – at reentry, an astronaut’s lifeblood.

Scott, meanwhile, was busy aviating and navigating. The California Capcom, Al Shepard, took over voice communications for the retrofire sequence, one minute away:

“Seven, this is Cap Com. Are you in retroattitude?”

Carpenter replied: “Yes. I don’t have agreement with ASCS in the window, Al. I think I’m going to have to go fly‑by‑wire and use the window and the [peri]scope. ASCS is bad. I’m on fly‑by‑wire and manual.”

Capcom responded: “Roger. We concur.”

But in going to fly‑by‑wire, Scott forgot to shut off the manual system that he’d activated during the pre‑retro checklist over Hawaii as backup for the automatic system. So his efforts to control attitude during retrofire were accomplished on both fly‑by‑wire and manual control modes, spewing out fuel from both tanks. Halfway through his fifteen‑minute flight the year before, Al himself had committed the identical error. Retrosequence was coming up.

Capcom radioed: “About ten seconds on my mark… 6, 5, 4, 3, 2, 1.”

Carpenter reported: “Retrosequence is green.”

Green is good. Green means that everything is set right for automatic retrofire. But not in this case, because the automatic system was locked out and the gyros were caged. Scott would have to fire the retros manually, throwing switches upon Al’s count, coming up fast. Suddenly a critical intervention from Al. If Scott’s gyros were caged, Al reported, he would have to “use attitude bypass.” His gyros were off, Scott answered, and Al repeated his remark:

“But you’ll have to use attitude bypass and manual override.”

Carpenter reported: “Roger.”

Then two seconds. Al counted down, “4, 3, 2, 1, 0.”

Before, during, and after the retrofire firing, Al offered Scott two crucial observations. Because of the instrument failures, the cockpit was in a configuration never before envisioned, and Al perceived the effect it would have on the required cockpit procedure. His contribution to Scott’s safe reentry was a resounding endorsement of the decision made a few years before to place astronauts in the communication loop, as knowledgeable buffers between the ground‑control people and the man doing the flying. Al’s insight at a crucial moment probably kept Scott’s landing from being even farther off target than it was.

Carpenter continued:

The last thirty minutes of my flight, in retrospect, were a dicey time. At the time I didn’t see it that way. First, I was trained to avoid any active intellectual comprehension of disaster – dwelling on a potential danger, or imagining what might happen. I was also too busy with the tasks at hand. Men and women who enter high‑risk professions are trained to suppress, or set aside, their emotions while carrying out their duties. After the job, and after the danger has passed, is the time for emotions.

Without the ability to detach oneself from the peril in a situation, one has no chance of surviving it. What perils did I face? They were the same perils faced by Al, Gus, and John – and later by Wally and Gordo. The retros might not fire. They might explode or not burn properly. The heat shield might not work. The drogue or the main chute might not deploy or reef properly. Thinking about all the things that could go wrong, no one would ever climb into a spacecraft. A pilot counts on all those things going right, not because he needs to believe in a fairy tale, but because he has confidence in the hardware, in the systems, in the men and women on the ground. In himself.

Still, I do remember being surp