We continue with noted Pan American Airlines radio officer Almon Gray’s analysis of the radio problems that Amelia Earhart encountered during her final flight.  Before we proceed, a word from the late Art Kennedy, an aircraft technician for the Pacific Airmotive Company in Burbank, Calif., during the 1930s, who directed the repairs of the Electra when it was shipped back to the Lockheed facility following the “ground-loop” at Luke Field, might be instructive.  In Kennedy’s 1992 autobiography, High Times, Keeping ’em Flying, he was quite frank in his appraisal of Amelia’s radio skills, or lack of same.

Kennedy believed that Earhart’s cavalier attitude toward radios led to her undoing.  “In her unique fashion Earhart was quite a lady, although it is well known that she punctuated her airport conversation with a spectacular lexicon of aviation vulgarities,” Kennedy wrote.  “This was especially the case when she had trouble contacting the tower, because she was notoriously lazy about learning how to use the radio properly.  She would get so frustrated that her language became unprintable and Burbank tower operators often found it necessary to reprimand her.  That failure to learn radio procedures may be significant in light of the apparently frantic transmissions before she disappeared.  I remember Paul Mantz telling her that she must be up to speed on frequencies for daylight and night transmissions, but she flippantly replied that if she couldn’t get what she wanted she’d just keep trying until she got a response.” 

Art Kennedy, circa 1937, who directed the repairs for Amelia Earhart’s Electra at the Lockheed facility at Burbank, Calif.  Kennedy said that Earhart was “notoriously lazy about learning to use the radio properly.”

“AMELIA EARHART AND RADIO,” Part II of III

By Almon A. Gray

ANATOMY OF A GOOF

While we shall never have a positive and complete answer to the above questions, it is possible to deduce a great deal.  Therefore there follows a hypothetical scenario which, it is believed, reflects quite accurately what actually transpired.  It is emphasized that some parts are conjecture.

1. Earhart was at Bandung having maintenance done on the plane when the query came in from Itasca as to what radio frequencies she wished Itasca, Ontario and Swan to use in supporting her flight from Lae to Howland.  Time was running out and she had to provide the answers right away.  It had been pounded into her head time and time again that-she needed low frequency radio beacons for homing purposes.  She knew that was what she wanted from the ships but she did not know what particular frequencies to specify.  She therefore sought advise from the best local source available and arranged for herself and Noonan to meet with the top KLM airline communications man.

2. The KLM man did not speak English very well and was accustomed to talking in terms of wavelength and meters rather than frequency and kilocycles.  From his service in the British Navy, Noonan was familiar with the wavelength/meters system so he and the KLM man did most of the talking.  Earhart scribbled notes.  Among them they developed the following plan:

(a) Ontario and Itasca would both use the same frequency but transmit at different times. This would allow Earhart to receive signals from both ships without the necessity of re-tuning her receiver.  To avoid any uncertainty as to which ship’s signals were being received, Ontario would transmit the Morse code character for the letter “A” rather than the customary Morse “M O” as its homing signal. Itasca would transmit the Morse character for the letter “N” as its homing signal. These same characters (A and N) were used it identify the quadrants of the four-course radio ranges in the United States and Earhart could readily recognize them.

Apparently it was envisaged that there would be an overlap of signal coverage over a good part of the leg, and that Earhart would be able to take bearings alternately on the two stations and thus keep on course.  The frequency chosen for Ontario and Itasca was 400 kilocycles, which is equivalent to a wavelength of 750 meters.  It was a frequency assigned worldwide for aeronautical radio-navigation and was an excellent choice.  It probably was chosen over equally good frequencies in the same band because it was easy to remember and easy to find on the receiver tuning dial.

(b) Swan used the frequency of 333 kilocycles which is equivalent to a wavelength of 900 meters.  Use it for voice communication with the plane if possible, but in any event be prepared to send homing signals on it.  333 kc was in the band allocated worldwide for aeronautical radio navigation and air-ground communications.  It was widely used in Europe, the Commonwealth nations and other countries having close ties with Europe, as a calling frequency and for ground-air communications.  Earhart had probably received on it during earlier legs of her flight but called it “nine hundred meters.”  It was an excellent direction-finding frequency.

In this rarely seen photo taken from Last Flight, Amelia is shown shortly after her arrival at Lae, New Guinea on June 30, 1937,

3. Noonan left the meeting satisfied that the radio navigational plans were adequate, or at least as good as could be developed.

4. Earhart went back to the hotel and drafted and dispatched her message of June 27 to Itasca (Black).  She did not show the message to Noonan.

5. It had been difficult for Earhart to understand the adviser’s English, and she had experienced great difficulty in following the discussion as it shifted rapidly back and forth among “frequency,” “wavelength,” “megacycle,” “meter,” “kilocycle,” etc.  Perhaps too she was suffering from dysentery and was actually ill.  Whatever the reason, the message she drafted suggested frequencies for the Swan and Itasca vastly different from those settled on in the meeting.  Specifically:

(a) The frequency for Swan was changed from an intended 333 kilocycles (900 meters) to 900 kilocycles.  One can readily deduce that the wavelength in meters was used but was labeled as frequency in kilocycles.

(b) The frequency for Itasca was changed from an intended 400 kilocycles (750 meters) to 7.50 megacycles.  Again it appears that the figures for the wavelength in meters were used but labeled as a frequency.

Had normal air-ground communications existed between Itasca and the plane, the homing problem could almost certainly have been solved quickly.  All that was needed was for Itasca to tell Earhart to home on 500 kHz, which frequency was already being transmitted (in addition to 7.50 MHz) by Itasca.  She should have been able to get bearings on that frequency that would have taken her right in to the ship.  Unfortunately she was unable to hear signals from Itasca on 3105 kHz, although the ship was hearing her well.  It thus was impossible for Itasca and Earhart to coordinate their actions.

THE AIR/GROUND COMMUNICATION PROBLEM

Why could Earhart not hear Itasca‘s transmissions on 3105 kHz?  Here again we probably shall never know for sure, but from the information which is available it is possible to hypothesize an answer which is reasonable and probably reflects quite accurately the actual situation.  Following are some of the things that are known which are germane to the question:

1. There was but one radio receiver aboard the plane and it was used for both communication and radio direction finding purposes.  There were two antennas aboard, a conventional fixed antenna and a rotatable shielded loop.  Either of these, but not both simultaneously, could be connected to the input of the receiver by means of an antenna selector switch on the receiver.  Radio signals could be received on either antenna but usually were stronger when using the fixed antenna, therefore it was the one generally used for communications.  Direction finding could be done only when using the loop antenna.

2. The fixed antenna was used for both receiving and transmitting purposes.  There was a so-called “send-receive” relay in the transmitter which switched the antenna back and forth between the units.  Normally the antenna was connected to the receiver, but when the relay was energized by pushing the “push to talk” button on the microphone, the antenna was switched over to the to the transmitter and remained that way until the microphone button was released.

3.  Energy from the loop antenna went directly to the antenna selector switch of the receiver. Energy from the fixed antenna passed through the “send-receive” relay mentioned above before reaching the antenna selector switch of the receiver.

4. The receiver had six frequency bands; however, the vacuum tubes, voltage determining resistors, bypass capacitors etc., were for the most part, common to all bands, and it was rare that a single band would fail.  It usually was none or all.

5. The radio equipment aboard the plane was checked at Lae by Harry Balfour, the Guinea Airways wireless operator, and was found satisfactory.  The only unusual thing noted was a roughness of the transmitted signal on 6216 kHz, which made Earhart’s speech difficult to understand.  Two-way communication was maintained during a 30-minute test hop at Lae.

Harry Balfour, circa 1937, the radio operator at Lae, New Guinea and the last person to carry on a two-way radio conversation with Amelia Earhart.

6. After takeoff from Lae to Howland it appears that two-way communication with Lae was maintained until about 0720 Greenwich Mean Time (GMT) (6 p.m. Lae time) July 2, at which time Earhart shifted to her “night” frequency (3105 kHz).  Several times after that, throughout the night, she was heard by Nauru and Itasca broadcasting at the pre-arranged times, but little of what she said was intelligible.  Nauru, and later Itasca, called her numerous times but there is no indication she heard any of the calls.  At 1744 GMT (seventeen hours, 44 minutes into the flight), she asked Itasca for a bearing on 3105 kHz and made a signal upon which the bearing was to be taken. Itasca made a response but Earhart did not acknowledge receiving it.  The same thing happened at 1815 GMT.  At 1912 GMT (0742 Howland Island Time), Earhart said the following to Itasca:

“WE MUST BE ON YOU NOW BUT CANNOT SEE YOU. RUNNING OUT OF GAS. ONLY ONE-HALF HOUR LEFT (there is controversy about that phrase).  BEEN UNABLE TO REACH YOU BY RADIO. WE ARE FLYING AT ONE THOUSAND FEET.”

At this time the signals from the plane were very strong.  It is known that the Itasca was putting out strong signals and was on the correct frequency.  (They were heard in San Francisco.)  Therefore the statement “BEEN UNABLE TO REACH YOU BY RADIO” clearly indicated that a failure had occurred in her radio receiving system, and that it probably had occurred early in the flight. Inasmuch as she could still transmit it was obvious that the fixed antenna was intact; beyond that there was no clue as to the nature of the failure.  That clue was given very shortly however.  AT 1925 GCT Earhart asked Itasca to transmit signals “on 7500,” meaning 7.50 MHz. This indicated that she intended to take radio bearings on Itasca with the plane’s direction finder.

Itasca complied immediately and sent the desired homing signals.  The transmitter had no radiotelephone capability so it was impossible to also talk with the plane by voice on that frequency. Earhart responded immediately saying, “WE RECEIVED YOUR SIGNALS ON SEVENTY FIVE HUNDRED BUT UNABLE TO GET A MINIMUM.  PLEASE TAKE BEARING ON US AND ANSWER THREE FIVE NAUGHT FIVE (3105 intended) WITH VOICE.”  This was followed by a series of long dashes on 3105 kHz on which bearings were expected to be taken by Itasca/Howland.  This was the first (and only) time Earhart acknowledged hearing signals from Itasca.  From the fact that Earhart asked for the homing signals it is clear that she intended to take a bearing, which could be done only with the loop antenna.  From her report of hearing the homing signal but being unable to get a minimum on it, it is obvious that she, in fact, shifted the receiver to the loop antenna, and that the homing signals were received on the loop antenna.

Why could she receive 7500 kHz signals on the loop but not 3105 kHz on the fixed antenna?  At the distances and time of day involved, propagation would not account for it, so something must have changed in the receiving system.  Actually two changes had been made: (a) The receiver had been shifted from band IV which included 3105 kHz to Band V or VI, both of which included 7500 kHz and (b) The receiver had been shifted from the fixed antenna to the loop antenna.

It is possible that some component peculiar to band IV had failed making reception on that band impossible, whereas reception on other bands would be normal.  However, as mentioned previously, the probability of that happening was small, therefore it is unlikely that shifting bands, per se, made the difference between not receiving and receiving signals.  Shifting antennas however was a horse of a very different color.  With the antenna selector switch in the “DF” position incoming signals picked up by the loop antenna went directly to the input of the receiver.  With the switch in that position Earhart heard signals from Itasca.

With the antenna selector switch in the “FA” (Fixed Antenna) position, signals picked up by the fixed antenna did not go directly to the input of the receiver; instead they passed through contacts on the “send/receive” relay in the transmitter.  With the switch in the “FA” position Earhart did not hear signals from Itasca. This indicates very strongly that signals from the fixed antenna were not reaching the receiver and that the receiver, in effect, had no antenna.

Radio room of USCG Cutter Tahoe, sister ship to Itasca, circa 1937.  Three radio logs were maintained during the flight, at positions 1 and 2 in the Itasca radio room, and one on Howland Island, where the Navy’s high-frequency direction finder had been set up.  Aboard Itasca, Chief Radioman Leo G. Bellarts supervised Gilbert E. Thompson, Thomas J. O’Hare and William L. Galten, all third-class radiomen, (meaning they were qualified and “rated” to perform their jobs). Many years later, Galten told former Pan Am Radio flight officer Paul Rafford Jr., “That woman never intended to land on Howland Island.”

The feed line from the fixed antenna was in two sections.  One was between the antenna and the “send/receive” relay in the transmitter.  This section was used both for receiving and for transmitting. Earhart’s transmissions were being heard, therefore this section, including the “send” part of the relay, was functioning.  The other section was between the receiver input and the “send/receive” relay, including the “receive” part of the relay.  There appears to have been an open circuit or a complete “ground” in this section, either of which would have prevented the receiver from picking up signals.

It is possible that the wire in that section of the feed line broke or came loose from a binding post; however, that possibility is very small.  It is much more likely that the trouble was in the “send/receive” relay.  Those devices were subject to damage from several sources.  Lightening or heavy static discharge sometimes burned the contacts completely off or welded them together. Contacts on the “receive” part of the relay were particularly subject to this type of damage.  Mistuning of the transmitter or antenna sometimes caused arcing and subsequent pitting and sticking of contacts.  And sometimes contacts would stick, or not make good contact, for no apparent reason.

It should not be implied from this that the relays were inherently unreliable; they were not. Most went hundreds of hours between routine replacement with no trouble, but occasionally one would fail.  This appears to have been one of those times.  In this writer’s judgment the odds are about 95 to 5 that Earhart was unable to hear Itasca on 3105 kHz because she was switched to the fixed antenna and the “send/receive” relay was defective on the receive side.

Had she shifted to the loop antenna she no doubt would have heard Itasca very well on 3105 kHz or whatever frequency the ship might be using and she was tuned to.  It probably never occurred to her to do that, however.  Earhart knew very little about the technical aspects of radio and consequently operated the gear by rote.  Obviously she had been taught to turn the antenna selector switch to “FA” if she wanted to talk, and to “DF” if she wanted to take a bearing — and that is precisely what she did.  (End of Part II of Almon Gray’s “Amelia Earhart and Radio.”)

For the pilots and other technically astute readers among you, Almon Gray’s analysis might be easily understood, even if you disagree with some or all of his ideas.  But for the lay person, which includes this writer, it’s not so easy to follow Gray’s narrative with clear comprehension.  Just when I thought Gray was attributing Earhart’s radio failures to a misunderstanding about the meters and wavelengths that the “KLM man” was advising Earhart and Noonan to use during their meeting at Bandung, he launched into completely different set of reasons to explain the communications nightmare that was the final flight.  I must admit that I don’t fully grasp the totality of  Gray’s narrative thus far, and may never.  Still, I think it’s important to present the important and unique work of experts like Almon Gray, regardless of how much I fail to understand.

In the final segment of “Amelia Earhart and Radio,” Gray will examine some of the possible “post-flight signals” that have long been sources of controversy and contention among researchers, take a closer look at Fred Noonan’s role in the proceedings, and present his well-informed conclusions.  Please stay tuned.