Failure to communicate

The short answer to the question "What went wrong?" is "What we have here is ... failure to communicate." "There was always more mix-up than there was mystery."

Noonan's techniques--a combination of celestial navigation and dead reckoning brought the Electra close to Howland Island. The Itasca radio logs show that the aircraft was steadily approaching Howland. The men aboard the Itasca could hear Earhart loud and clear (S5). But Earhart and Noonan heard only one transmission from the Itasca on 7500 kHz. Earhart was unable to obtain a null on that frequency. The Itasca, for its part, could not obtain a null on Earhart's transmissions; although the crew could not have transmitted the bearing to the Electra because of the radio problems aboard the aircraft, getting a bearing on Earhart's transmissions could have aided the search for Earhart and Noonan.

Inadequate coordination

"Much careful planning by William Miller, on 'loan' from the Bureau of Air Commerce, well documented and distributed to the US Navy and US Coast Guard, was carried over to the second attempt. Unfortunately, Bill Miller was assigned to investigate the potential for air mail routes to New Zealand shortly after Earhart’'s abortive crash take-off in Honolulu, and was unavailable to provide the detailed planning necessary for a successful second attempt for the world flight. George Putnam took over the logistics and planning, but appeared not to be as well versed in this art as Miller" (Jacobson). Evidence of Earhart, Noonan, and Putnam's failure to develop and communicate a coherent protocol for radio communications and direction finding is evident from the flurry of telegrams in the week before the final flight. Earhart and others seem to be making things up as they go along; "the best laid plans of mice and men gang aft aglay"--and these plans were far from "the best laid."

Bob Brandenburg:

On June 26th, Earhart sent the following to Richard Black

SUGGEST ONTARIO STANDBY ON FOUR HUNDRED KCS TO TRANSMIT LETTER N FIVE MINUTES ON REQUEST WITH STATION CALL SIGN LETTER REPEATED TWELVE END EVERY MINUTE.

Note that here she is asking Ontario to transmit on request. But since Ontario could only communicate on frequencies below 600 kHz, and only on CW, there was no way Earhart could have made such a request.

On July 1st, Earhart sent the following to Black, modifying the previous plan

ASK ONTARIO BROADCAST LETTER N FOR FIVE MINUTES TEN MINUTES AFTER HOUR GMT FOUR HUNDRED KCS WITH OWN CALL LETTERS REPEATED TWICE END EVERY MINUTE.

This message could indicate that Earhart realized that she had no way to request Ontario to transmit the desired signal, so she changed the procedure from "on request" to "broadcast". This would have fixed the problem had the message been relayed to the Ontario. But there's no record of Black having relayed the message to the Ontario. The message was relayed from Earhart in Lae to Black on the Itasca via the Navy radio station at Tutuila, so it is possible that Tutuila took the initiative and relayed the message directly to the Ontario - - although there is no record of that having happened.

N.B. In this next paragraph, Brandenburg talks about a suspicion that Earhart's direction finding system may have been incapacitated on takeoff at Lae. His line of reasoning is not given here, but it may involve assumptions about the use of the dorsal antenna in conjunction with the loop antenna. The sole point of quoting this material is to call attention to the inadequate communications between Earhart and the Ontario.

In any case, it would be reasonable for Earhart to assume that the second message was relayed to the Ontario, and that the Ontario would be broadcasting the requested signal on the requested schedule. If Ontario did get the message, the broadcast would have occurred and if, as we suspect, Earhart's DF (direction finding) gear was rendered inoperable on takeoff at Lae, she would not have been able to get a bearing on the Ontario--and she could have turned back to Lae. If Ontario did not get the message, the signal expected by Earhart in her second message would not have been sent by Ontario, and Earhart would have heard nothing. In this case, as well, Earhart would have had a warning that her DF gear wasn't working, and could have turned back to Lae.

Ric Gillespie, 25 March 1999 Forum.

Earhart, by commercial radio messages sent prior to her departure from Lae

• told Itasca what frequencies she would be using,
• told them to send only voice messages to her,
• told them what times she would be transmitting and what times she would be listening,
• told them that she would be using Greenwich time,
• asked that they send signals on 7500 kHz for her to home on and
  • specifically asked if that frequency would be okay.

Itasca did not advise her that

• it would be almost impossible for them to take bearings on the frequencies she said she would be transmitting on,
• did not tell her 7500 was far too high a frequency for efficient direction finding,
• ignored her request for voice-only messages and
• sent most of their in-flight transmissions to her in code.

They also ignored her stipulation that the radio schedule use Greenwich time, which caused considerable confusion.

There is no doubt that Earhart made mistakes, but there is plenty of blame to go around in the disappearance of NR16020.

Misunderstanding of equipment: Amelia Earhart

Earhart's resistance to developing radio skills

Randy Jacobson, commenting on Earhart's first attempt to fly around the world:

On top of all this, a Bureau of Air Commerce official, Mr. Reining, sent a telegram to Earhart, with a copy to Miller saying that she should contact Mr. Marriott or Mr. Bedinger, the supervisory aeronautical inspector, for nonscheduled instrument rating or flight check ability to fly entirely by instruments, as her flight permits would not be released – Earhart’s pilot’s license was about to expire, and she needed to take her biennial flight review prior to take-off from Oakland. The next day, Bedinger gave Earhart the instrument flight check, but the written and radio tests "... not given account her desire to expedite and save engines."

This is most curious, as the engines were relatively new and would not need servicing for quite some hours, and a simple one hour test to verify her radio skills would not significantly contribute to engine wear. Nor was Earhart in a big hurry, as there were still four more days planned prior to take-off. The desire to avoid the radio test was probably the first contributing factor to Earhart’s failure to reach Howland in July. Regardless, the BAC in Washington DC put her permit in an airmail package to Oakland.

Apparent ignorance of the limitations of her equipment

This is a problem faced by everyone involved: Putnam, Earhart, and the Coast Guard.

June 25th at 2245GMT, the Coast Guard San Francisco office sent a radio message

... Mr. Putnam now at Oakland and advises Ms. Earhart at Bandoeng Java for repairs to motors and departure indefinite. She will cable details communications from Port Darwin direct San Francisco and you will be given all information immediately. All communications from plane will be on 500, 3105, or 6210 kHz by voice, positions being given at 15 and 45 minutes past the hour. Itasca adjust transmitter for possible use 3105 kHz for voice. Direction finder on plane covers range of about 200 to 1400 kHz.

Before the second world flight, Earhart removed the 500 kHz antenna. She could not understand CW messages sent on that frequency. It would have been a good frequency, given the Itasca’s low-frequency DF equipment.

At 0720GMT, June 26th, Earhart responded with

...Itasca transmit letter A, position, own call letters, as above on half hour at 7.5 MHz. Position ships and our leaving will determine broadcast times specifically. If frequencies mentioned unsuitable night work inform me at Lae. I will give long call by voice 3105 kHz quarter after hour, possible quarter to.

This contradicts the June 25th telegram that position reports would be given "given at 15 and 45 minutes past the hour."

Jacobson: "It is unclear why Earhart wants the Itasca to broadcast at 7500 kHz, since she is asking for code and cannot use that frequency for direction finding. The actual wording in the telegram is 7.5 megacycles."

At 1930GMT, June 26th, the San Francisco office of the USCG sent out a priority radio message stating

Following information from Earhart this date quote homing device covers from 200 to 1500 and 2400 to 4800 kHz any frequencies not repeat not near ends of bands suitable unquote. ... assume continuous signals after her direction finder in range. See broadcast on quarter after and quarter before hour on 6210 and 3105 kHz. Am advising Earhart that Itasca will voice radio her on 3105 on hour and half hour as she approaches Howland. ... Advice priority if adjustments Tare ten transmitter satisfactory for use on 3105.

The information on the 25th was that Earhart's DF equipment operated between 200 and 1400 kHz; now two ranges are given: "200 to 1500 and 2400 to 4800 kHz."

Jacobson: "There is no contemporaneous documentation to verify that the information regarding the double frequency band of Earhart's RDF came from Earhart or George Putnam. Based upon the quoted passage, it should be taken at face value and accepted as coming from Earhart."

On 2040GMT, June 28, Itasca sent this message to Earhart

Itasca transmitters calibrated 7500, 6210, 3105, 500 and 425 kHz CW and last three either CW or MCW. Itasca direction finder range 550 to 270 kHz. Request we be advised as to time of departure and zone time to be used on radio schedules.

A transmission on 500 kHz would be ideal for the Itasca's declared DF range. But Earhart had removed the trailing antenna intended for use on 500 kHz. When she asked Itasca to take a bearing on her, she was transmitting on 3105 kHz--and the Itasca attempted but failed to get a bearing from her transmission. The Itasca’s direction finder was designed to work only with lower frequencies (270 kHz to 550 kHz).

At 2015GMT, June 29th, Earhart sent this message to Itasca

...Report in English, not code, especially while flying. Will broadcast hourly quarter past hour GCT. ...

Now she plans just one broadcast per hour.

A possible explanation

Elgen M. and Marie K. Long, Amelia Earhart: The Mystery Solved, p. 116.

By pure happenstance the first three bands of the radio receiver and the direction-finder loop were closely matched, and the direction-finder loop could obtain a bearing on bands 1, 2, and 3. However, the frequencies of the Bendix receiver's bands 4 and 5 were completely out of the design range of the direction-finder, and the loop could not process the signals to obtain bearings. But because both the loop coupler band selector and the Bendix receiver were marked with bands 1 to 5, Earhart thought she could take a radio bearing with her direction-finder loop on any signal between 200 and 10,000 kilocycles. She could not. Without Manning, the information provided earlier by the Bendix representative, Cyril Remmlein, could not be conveyed to Earhart. None of the radiomen was aware of Earhart's mistaken assumption.

Earhart unclear on the concept of what Itasca needed

Earhart apparently thought that she had to make noise in the microphone in order to give them audio information to use to determine the null; but the Itasca could have found the null based on the strength of the carrier wave of her transmitter. All she needed to do was to key the microphone on for a minute or so; sound transmission was not necessary. Earhart apparently turned the microphone off when she ran out of breath, frustrating the efforts of the Itasca to comply with her request.

1744 GMT

Earhart wants a bearing on 3105 kHz on the hour, will whistle in microphone, about 200 miles out approximately, now whistling.

1811 GMT

Earhart requests: "Please take bearing on us and report in half hour. I will make noise in mic -- about 100 miles out."

Failure to anticipate difficulties of finding a minimum

On July 1, Earhart tried to use her loop antenna to take a bearing on the airfield at Lae.

Ric Gillespie, Finding Amelia, p. 74

During the test flight, Earhart found she could receive Lae’s signal, but the intensity of the sound did not change when she rotated the loop. In the terminology of the time, she could not 'get a minimum,' and so could not get a bearing on the sending station. The problem was that although the radio receiver could pick up the signal and she could hear the tone in her headphones, the direction-finding aspect of the system could not respond to such a high frequency. Amelia, however, decided that the test had failed because the airplane was too close to the station and the signal was too strong.

The onboard direction-finding system failed again two days later on the last flight.

1928 GMT: "KHAQQ calling Itasca we are circling [?] but cannot hear you go ahead on 7500 with a long count either now or on the schedule time on 1/2 hour"

1930 GMT: "KHAQQ calling Itasca we received your signals but unable to get a minimum. Please take bearing on us and answer 3105 with voice." Another radioman reports this message as: "Amelia on again at 0800 [local time] says hears us on 7.5 megs go ahead on 7500 again."

Even if Earhart's equipment could do direction finding on 7500 kHz, recognizing where the Itasca’s signal was weakest might have beyond Earhart's capacity after twenty hours in a noisy and exhausting environment. The signal-to-noise ratio of CW/Morse Code transmissions is excellent; extremely weak signals can still be intelligible. Earhart may not have been able to discriminate between the strongest and weakest signals because she had been deafened by exposure to the noise of the engines and propellers for so long. The Hooven Radio Compass would have avoided this difficulty altogether, since, like the equipment on the Itasca, it did not rely on the human ear for detection of where the signal was weakest.

Abandonment of CW (Morse Code)

Preference for voice communication (telephony)

Earhart and Noonan were ahead of their time in asking that the Itasca communicate with them solely via voice transmissions. That is now the worldwide norm for aircraft. In 1937, the worldwide norm was Morse code transmitted on continuous wave (CW) or modified continuous wave (MCW). The Itasca could not transmit voice on all frequencies (e.g., 7500 kHz).

"As the result of a talk with Mr. E. Chater and Mr. Balfour the Lae radio operator it is very apparent that the weak link in the combination was the crew’s lack of expert knowledge of radio. Their morse was very slow and they preferred to use telephony as much as possible."

"On enquiry Miss Earhart and Captain Noonan advised that they entirely depended on radio telephone reception as neither of them were able to read morse at any speed but could recognise an individual letter sent several times. This point was again mentioned by both of them later when two different sets at Lae were used for listening in for time signals."

"Miss Earhart and Captain Noonan spent a considerable time in the radio office and as previously mentioned it was learned that neither of them could read morse at any speed but could only distinguish letters made individually slowly and repeated often; in that case their direction finding apparatus would be useless or misleading unless they were taking a bearing on a station using radiophone which could give the station position on voice. We understand the Itasca was to do this but if the plane was unable to pick up the Itasca it is doubtful if the drection finder would be any use to her."

Removal of equipment

500 kHz was a universal frequency used by ships at sea for two purposes: emergency calls and for making contact with other stations (once contact was made, the subsequent transmissions would take place on a different frequency). As a matter of convention, all of the traffic on 500 kHz was conducted in Morse code using continuous wave or modulated continuous wave transmission.

During the first around-the-world attempt, Manning was equipped and trained to handle communication on 500 kHz. When he was unwilling or unable to participate in the second attempt, Earhart and Noonan decided to rely solely on voice transmission (radiotelephony) on 3105 kHz and 6210 kHz. The 500 kHz equipment was then just dead weight and some (or all?) of it was removed to save weight (and thereby gain range).

Trailing antenna

The removal of trailing antenna and concomitant changes in the NR16020 antennas meant that Earhart had to depend on high-frequency direction finding (HFDF). Although Joe Gurr had attempted to adjust the radios and antennas to allow some 500 kHz transmission capacity, there is no evidence in any of the logs that Earhart tried to transmit on this frequency--which was ideally suited to the Itasca's on-board direction finding equipment which, as noted above, was designed to take bearings on signals in the "550 to 270 kHz" range.

Morse Code keys

The Electra was originally equipped with two Morse code keys. The evidence suggests that both were left behind. It seems that W.C. Tinus had one and Joe Gurr had the other.

Ric Gillespie, 25 March 1999 Forum.

500 kHz would only carry code and Earhart had [virtually] no knowledge of code and no key with which to send code.

Modification of radio equipment: Joe Gurr

Joseph H. Gurr worked on the radio and antenna setup in Burbank both before and after the first round-the-world attempt. Some of the modifications he made to the dorsal antenna--lengthening it by 50%--may have caused some problems with power output and the quality of the signal. It is not at all clear that any changes he made were responsible in any way for the problems Earhart had with hearing transmissions from the Itasca or with her failure to find a minimum on 7500 kHz.

The lost antenna

TIGHAR thinks that the ventral (belly) antenna may well have been lost on takeoff. No firm contemporaneous evidence tells us about the exact radio equipment on board during the final flight nor describes the functions of the various antennas on the aircraft.

Some other electrical failure affecting the receiver

The fuse may have blown again

I think a fuse was replaced in Darwin or Lae. Why did it blow? Did it blow again? Did AE and FN have a spare fuse with them? Could they have replaced it even if they did? [refs needed] MXM

The antenna may have been disconnected again

The first service Joe Gurr performed for Earhart was to locate a disconnected antenna lead for the receiver under the co-pilot's seat. It is not wholly inconceivable that this same lead became disconnected during the final flight.

Mike Everette, 6 Aug 1998 Forum.

The type of antenna connection used on aircraft radios in the 30s and 40s was, almost universally, a type of "push-post" or spring-loaded binding post, which also had a ceramic feed through insulator passing through the radio's case or panel. Screw-on fittings were not in general use until years later.

Coaxial cable was not used to interconnect most airborne radio gear until the advent of VHF communications, when designers settled on making everything work into a fixed impedance, standardized at 50 ohms. HF gear fell under the same design philosophy at the same time, and any antenna tuners or loading networks were consequently designed to match a complex impedance like an aircraft antenna to that value. This came about after World War II.

Up until that time, aircraft radios were designed to work into a wide range of antennas, over a broad impedance range. Use of coaxial cable would have caused many problems in such a design environment. High radio frequency voltages developed when working a transmitter into a high impedance load (several hundred or several thousand ohms) which might not be completely matched at the frequency, would cause the cable to heat up or arc over. Therefore, open-wire leads insulated with ceramic beads were in general use. The receiver connection might be as simple as a length of cloth insulated heavy wire between the receiver and the terminal on the transmit-receive (antenna) relay.

Coaxial cables were, however, used to interconnect components where shielding of the leads was important, or where high radio frequency voltages were not involved. One specific instance is the lead from the sense antenna to the antenna input of a DF receiver, and Bendix gear used coaxial input with screw-on fittings. A reason to use shielded cable at the lower frequencies used for DF (200-1500 KHz) is to keep out electrical noises generated inside the aircraft. This is a real problem at low frequencies, somewhat less so at HF.

I would be willing to bet that the Western Electric receiver used a push post.

Gurr stated that when he first looked at the Electra to check the non functioning HF receiver, he found it under the copilot's seat... the antenna lead was disconnected or pulled loose.

These push-posts were not safety-wired. I can't ever recall seeing one with provisions to safety-wire it, and this is a result of experience with Collins, Aircraft Radio Corporation, Philco, Wells-Gardner, Western Electric and other manufacturers' products.

Failure to follow protocol: Itasca

Use of CW (Morse Code) instead of telephony

At 2015GMT [0615 local], June 29th, Earhart sent this message to Itasca:

Plan midday takeoff here [June 30 local time]. Please have meteorologist send forecast Lae – Howland soon as possible. If reaches me in time will try leave today otherwise July 1st. Report in English, not code, especially while flying. Will broadcast hourly quarter past hour GCT. Further information later.

"By 1800GMT [on July 2] ... the radiomen are using more and more Morse code on 3105 kHz, in the (correct) belief that reception should improve at the airplane using code vs. voice. Apparently, they believe Earhart can interpret Morse Code at the nominal 10 words per minute rate previously provided to the Shoshone during the March flight. We do know that Capt. Manning was quite proficient in Morse Code, and that Earhart and Noonan were not proficient.":

"The conventional wisdom was that CW was '10 times as effective' as voice transmission" (Hue Miller, December 5, 2000, Forum.).

2135 GMT. Itasca to Howland Island

Itasca called Earhart on 7500, key, and said "we heard you on 3105, keep using 3105."

Lost track of time?

Earhart seems to have approached the use of radio differently from the way it is used in aviation today. Rather than carrying on a dialog in real time with another station, her plan was that she would talk twice and hour and listen twice an hour. "She intended to receive on the hour and half, and transmit on the quarter and three-quarter hour."

Either the radio operators of the Itasca failed to grasp her plan or they forgot about it as the stress mounted on the morning of July 2.

Randy Jacobson:

To summarize, Earhart is getting closer to Howland, eventually arriving at what she thinks is the approximate location, but due to a lack of two-way communication, cannot get a bearing on Itasca nor can she get the Itasca to get a bearing on her and transmit that information back. The Itasca radiomen have abandoned all resemblance of a radio schedule, as does Earhart, to a lesser degree. Earhart’s message content does not indicate she is frantic yet, but indicates a growing concern that she cannot hear the Itasca.

Let us review for a moment the intended radio schedules. Earhart states that she will broadcast at quarter past the hour, using GMT time zone. She broadcast at times 1415, 1515, 1623, 1744, 1822, 1912, 1928, and 1930GMT. Only three broadcasts were out of schedule: 1744, 1928 and 1930GMT, with the last two as an attempt by Earhart to get Itasca’s attention prior to their scheduled broadcast. The 1744GMT message requested a bearing on her at approximately 200 miles range. Curiously, what is missing is a 1715GMT transmission. Both Itasca1 and Bellarts indicate both radio stations are listening on 3105 kHz, and that no transmissions from the Itasca were being made. This suggests that at 1715GMT, Earhart may have been at a distance without a bounce path off the ionosphere for radio transmission at 3105 kHz, presuming that she did transmit.

The radio schedule for Itasca was for them to transmit letter A, call sign, position on 7500 kHz on the hour and half hour, and to provide voice on 3105 kHz either upon request or when Earhart is within range. By 1415GMT, Itasca has begun to send weather information on 3105 kHz, in addition to letter As on 7500 kHz. According to the records, the times of transmission varied from 2 minutes before to 6 minutes after the hour, sometime continuously, sometimes not during the early hours. By 1800GMT, Itasca begins transmitting either on 7500 and/or 3105 kHz nearly continuously, sometimes interfering with its own ability to listen on 3105 kHz. ...

The Itasca radiomen are also beginning to become confused as to time zones. Our interpretation is that they have forgotten or ignored Earhart's request to maintain GMT time zones, and are instead using the half-hour time zone, a remnant from their base in Honolulu, which is on +10.5 time zone.

Needed: Ric or Randy's notes showing that the Itasca failed to transmit in voice when Earhart said she would be listening.

The cavalry didn't ride to the rescue

The high frequency direction finding equipment on Howland was out of commission when it was most needed.

Short story about how the stuff got there. Need to find the story of the dead batteries.

Inadequate problem solving: Earhart and Noonan

Safe on dry land, untroubled by more than twenty hours in a noisy, stressful environment, and with the benefit of hindsight, it is easy to second-guess what Earhart and Noonan should or could have done about their radio problems.

The single message received by the aircraft was the repeated transmission of the letter "A" (dot-dash) as Earhart requested. From this transmission, Earhart and Noonan should have realized that the Itasca was hearing them on 3105 Khz. They might have reasoned that there was a problem with their receiver when it was configured to use the normal receiving antenna and tuned the DF system to 3105 Khz. Switching from an antenna that was known to be working to one that had not yet worked seems to have been a bad decision. If the Itasca radio operators had been able to talk to Earhart, they could have told her to hold her microphone key long enough for them to get a null, determine the bearing, and tell Earhart what direction to fly to reach safety.

So near, and yet so far!

Where was Manning when they needed him?

On the first round-the-world attempt, all the pieces were in place for a successful flight from Hawaii to Howland. Manning was comfortable with CW/Morse Code and could have been in contact with the Itasca from far greater range than radiotelephony allowed. If there had been trouble with direction finding on the first attempt, Manning almost certainly could have worked with the Itasca to find a resolution.

The original plan was economical and provided Earhart with maximum support at minimal cost. Mantz was to be left behind in Hawaii, allowing a larger fuel load for the flight to Howland, the trickiest and most dangerous part of the world flight. Manning would be left behind at Howland, again extending the range of the Electra for the longest leg from Howland to Lae.

Flying in the other direction after the disaster at Luke Field made it possible to give the Electra a good shakedown cruise before venturing across the Pacific, but it made it impracticable to rendezvous with Manning (or someone of his caliber) in Lae. It's not hard to see why going with a two-person crew seemed like the most reasonable thing to do (apart from any considerations of whether Manning lost confidence in Earhart as a result of the groundloop at Luke Field).

According to Gurr, "Captain Harry Manning's leave of absence would now run out before Amelia could make her flight, so he bowed out and returned to Washington" (letter to Goerner, 3 May 1982).

W.C. Tinus, Vice President of Bell Telephone Laboratories, 1962:

I was the radio engineer who was responsible for the design and installation of her radio communications equipment [at the Newark Airport, New Jersey in February, 1937] ...

I was less successful in persuading Miss Earhart of the importance of having a qualified radio operator in her crew. I had only a short period one afternoon at Newark Airport to show her and captain Manning (of the United States Lines Sea Rescue fame) how to operate the equipment.

Related

• Air Navigation: State of the Art in 1937