Amelia Didn’t Know Radio
6 views
Skip to first unread message
D.J.J. Ring, Jr.
Jun 30, 2024, 4:44:02 PM (7 days ago) Jun 30
to Radio Officers Google Group
Amelia waited while Fuzz Furman fixed a gauge. Above the cockpit—the critical loop and stub-mast antennas.By Captain Almon A. Gray, U.S. Naval Reserve (Retired)
ALSO Log Jam
The Radio Logs of the USCG Itasca
Bandoeng, Java:
Amelia waited while Fuzz Furman fixed a gauge. Above the cockpit—the critical loop and stub-mast antennas.
By Captain Almon A. Gray, U.S. Naval Reserve (Retired)
December 1993 Naval History Volume 7 Number 4
Almost certainly, Amelia Earhart could not get a bearing on the radio beacon on the U.S. Coast Guard Cutter Itasca (WPG-321), lying off the beach at Howland Island, because the frequency that she had designated—7.50 Mcs*—was so high that her direction-finder (DF) was inherently incapable of taking bearings on it.
That Earhart and Fred Noonan failed to reach Howland Island on their 1937 around-the-world flight because of radio problems has been said before—but little has been written about the specifics.
A failure in the plane’s antenna system, which made it impossible to receive signals on the fixed antenna, also was a factor. Had she or Noonan known enough about the system to work around the failure, they could have established voice communications, with the Itasca, where someone surely would have suggested they try taking bearings on the vessel’s 500-kilocycle beacon. It could have made all the difference.
BACKGROUND: In early 1937, several weeks before departing Oakland, California, for Honolulu—the first leg of an intended west-about flight around the world— Earhart met at Alameda, California, with George Angus, the Superintendent of Communications for the Pacific Division of Pan American Airways (PAA). Angus directed the radio communication and DF networks that supported the PAA clippers on their Pacific crossings, and she was looking for help to augment Noonan’s celestial navigation.
The airline then had specially designed versions of the Adcock radio DF system in service at Alameda, Mokapu Point on Oahu in the Territory of Hawaii, Midway Island, Wake Island, Guam, and Manila in the Philippines. They could take bearings on frequencies much higher than could conventional loop-type direction finders—like Earhart’s— and were effective over much greater distances. These high-frequency DFs were the only ones of their type in the United States and its territories. Angus agreed to help and went to work on the details.
This was complicated, inasmuch as PAA could receive but not transmit on either of Earhart’s communication frequencies—3105 and 6210 KHz.—and could not transmit voice on any frequency. Earhart and Angus decided that the aircraft would request a bearing by voice on the frequency in use—usually 3105 KHz. at night and 6210 KHz. during the day—and follow the request with a series of long dashes lasting in the aggregate a couple of minutes. The PAA DF station would take a bearing on the transmission and transmit it to the plane on another previously agreed upon PAA frequency, using continuous wave (CW) telegraphy sent at such a slow speed that the individual dots and dashes could be copied on paper and later translated into numbers.
This arrangement was tested on the flight from Oakland to Honolulu; PAA took the bearings on 3105 KHz. and tansmitted the bearings in Morse code on 2986 KHz. The flight was handled much the same as a routine Clipper flight. Captain Harry Manning, former captain of the SS Roosevelt—the ship that brought her home from Europe after her 1928 trans-Atlantic flight—and a long-time friend, was an experienced radio operator and handled the Elec- tra’s radio and DF gear while regular PAA professional radio operators manned the ground stations. Radio bearings furnished the plane at frequent intervals, first from Alameda and later from Mokapu Point, checked well with the positions Noonan determined by celestial navigation. Nearing Oahu, Manning set up the plane’s DF to home on the 290 Khz. marine radio beacon at Makapuu Point, near Diamond Head, and Earhart homed in on it to a successful landfall.
While attempting take-off for Howland Island from Luke Field, near Honolulu, on 20 March 1937, Earhart ground-looped the Electra, damaging it to the extent that it was shipped back to the Lockheed plant in California for repairs. The radio gear sustained no major damage, but the Western Electric Model 20B radio receiver and its remote-control apparatus were replaced by a Bendix aircraft radio receiver and accessories. The stub mast supporting the V-shaped fixed antenna also was moved a bit forward, and the antenna feed line was rerouted. The late Joseph Gurr, then a moonlighting United Airlines technician, did the work.
THE NEW RECEIVER: The receiver installed at Lockheed was an experimental model incorporating the latest improvements. Only three experimental units were built, although Bendix later marketed an almost identical design as the Type RA-1 Aircraft Radio Receiver.
The experimental model was a continuous-tuning superheterodyne that covered the spectrum from 150 to 10,000 Kcs. in five bands. It could receive voice, CW, or modulated CW (MCW) signals and could be controlled remotely from the cockpit. A switch permitted the operator to connect the receiver to either the conventional wire antenna or the loop antenna. When the loop was used, the combination became an effective radio DF system capable of accurate bearings on frequencies between 150 and approximately 1800 Kcs. Signals on frequencies higher than 1800 Kcs. could be heard, but very seldom could accurate bearings be obtained. Earhart was apparently unaware of this. The receiver was powered by a dynamotor operated by storage batteries charged by the main engines.
The Radio System: When the plane left the Lockheed plant, the radio system consisted of the following elements:
• The experimental Bendix aircraft radio receiver
• A Western Electric Model 13-C 50-watt aircraft transmitter with three crystal-controlled channels—500, 3105, and 6210 KHz.—capable of voice or CW transmissions. It was mounted in the cabin, but there were remote controls in the cockpit.
• A prototype of a Bendix Type MN-20 rotatable shielded- loop antenna. It was mounted on the fuselage above the cockpit; the knob that rotated it was on the cockpit overhead between the pilots. It was used primarily for taking radio bearings but was useful as a receiving antenna in static caused by heavy precipitation.
• Fittings at each side of the cockpit for connecting a microphone, headphones, and telegraph key
• A telegraph key and a jack for connecting headphones at the navigator’s table
• A 250-foot flexible-wire trailing antenna on an electrically operated, remote-controlled reel at the rear of the plane. The wire exited the lower fuselage through an insulated bushing and had a lead weight, or “fish,” at the end to keep it from whipping when deployed. A variable loading coil used in conjunction with this antenna permitted its use on 500 KHz., and the antenna was long enough to give excellent radiation efficiency on all three transmitting frequencies.
• A fixed, Vee-configured wire antenna with its apex at a stub mast mounted on the top of the fuselage, over the center section of the wing, and its two legs extending back to the two vertical tail fins. The antenna was so short that its radiation efficiency was extremely low; it was adequate for local communications around an airport when it was not feasible to have the trailing antenna deployed, but not for the long-distance communication Earhart required for her transoceanic flight.
Either wire antenna could be selected from the cockpit. The one selected both transmitted and received by means of a send-receive relay that switched the antenna from the receiver to the transmitter when the microphone button was depressed, and switched it back to the receiver when the button was released.
MISTAKES AT MIAMI: After deciding to change her route to east-about, in late May 1937 Earhart flew the plane to Miami, where she had the trailing antenna and associated gear removed completely. John Ray, an Eastern Airlines technician who had his own radio shop as a sideline, did the work. Once again, Amelia obviously did not comprehend the devastating impact this would have on her ability to communicate and to use radio navigation. With only the very short fixed antenna remaining, virtually no energy could be radiated on 500 KHz. This not only foreclosed any possibility of contacting ships and marine shore stations but precluded ships—most important, the Itasca—and marine shore-based DF stations from taking radio bearings on the plane, inasmuch as 500 KHz. was the only one of her frequencies that fell within the range of the marine direction finders. Any radio aid in locating Howland Island would have to be in the form of radio bearings taken by the plane on radio signals from the Itasca. Earhart had cut her options severely.
The shortness of the remaining antenna also drastically reduced the power radiated on the two high frequencies. Paul Rafford Jr., a NASA expert in this field involved in forecasting long-range communication requirements to support astronaut recoveries, estimated that the radiated power on 3105 KHz. was about one-half watt. This obviously was a tremendous handicap in the high static level of the tropics.
The fixed antenna also may have been at least partly responsible for the distortion in Earhart’s transmitted signals reported by the operators at Lae, New Guinea, and Howland as affecting the intelligibility of her voice transmissions. A mismatch between the antenna and the final amplifier of a WE-13C transmitter could cause the transmitter to over-modulate and thus introduce distortion.
After a few days in the Pan American Airways shops during which all systems, including the antennas, were tuned and peaked, the plane departed Miami on 1 June 1937 to resume the flight around the world.
Despite these shortcomings, Earhart got as far as the Dutch East Indies without major incident. There, however, because of her unfamiliarity with radio matters, she unwittingly made the mistake that ultimately led to her failure to reach Howland Island.
THE FAULTY PLAN: The legs from New Guinea to Howland Island and from Howland to Hawaii were the most difficult navigational portions of the flight, and three small vessels were stationed along the way to assist. Each planned to use the ship’s transmitter as a radio beacon for Earhart and Noonan to supplement Noonan’s celestial navigation:
• The USS Ontario (AT-13) was on station midway between Lae and Howland.
• The USS Swan (AVP-34) was positioned midway between Howland and Hawaii.
• The USCGC Itasca was at Howland. Her beacon was particularly important; should Noonan’s celestial navigation not put them within visual range of the small, low-lying island, homing in on the Itasca’s signal would be their only chance.
By 23 June these vessels were on or approaching their respective stations but had not been issued their radio beacon frequency or procedures. That day, in a message addressed to Earhart at Darwin or Bandoeng, Richard Black— Earhart’s representative on board the Itasca—advised her of the radio frequencies available on the three ships and asked her to designate the frequency she wished each ship to use when transmitting beacon signals. This message caught up with Earhart at Bandoeng, Java.
Precisely what happened next remains unknown, but it appears that Earhart conferred with a local aeronautical communication specialist to get information she could use to base a reply to Black. A plan was developed that fulfilled her requirements with a minimum of receiver tuning on her part. The Ontario and the Itasca were to transmit on the same frequency but at different times, and each would transmit a distinctive Morse identification signal. The Ontario’s identifier was “N”, and the Itasca’s was “A.” (These were the characters used to identify the quadrants of the four-course radio ranges, then the principal navigation aid in the United States. Hence, Earhart was familiar with them.)
In 1937 it was still common to describe radio emissions in wavelengths expressed in meters, rather than in frequencies expressed in cycles per second. It seems apparent that the specialist did this during his discussions with Earhart and that he suggested that the ships transmit beacon signals on wavelengths as follows:
• Ontario 750 meters (400 Kcs.)
• Swan 900 meters (333 Kcs.)
• Itasca 750 meters (400 Kcs.)
These were excellent choices. All were allocated internationally for aeronautical radio navigation and were ideal for use with the direction finder in the Earhart plane.
Unfortunately, Earhart did not understand the relationship between wavelength and frequency nor how to convert from one to the other. Consequently, when she replied to Black on 27 June, she confused the figures and unwittingly specified incorrect frequencies for the Swan and the Itasca; she was correct with the Ontario.
In the case of the Swan, she apparently confused the wavelength and frequency figures, and specified that the Swan transmit on 900 Kcs. (rather than 333 Kcs.). This was a bad error in that 900 Kcs. was in the broadcast band and not available for aeronautical use. It also was inferior to the intended frequency of 333 Kcs. for DF purposes. It was not necessarily devastating, however, and fair bearings probably could have been taken on it with the aircraft DF.
In the Itasca’s case, however, it was to have grave consequences when she again apparently reversed the numbers and told Black to use 7.50 Mcs. (rather than 400 Kcs.) on the Itasca. The 7.50 Mcs. frequency was so high that there was practically no possibility of obtaining usable radio bearings on it with the aircraft DF.
Following is the text of Earhart’s reply to Black, sent the day before she left Bandoeng for Koepang and Darwin:
From: Earhart via RCA Manila & NPM Navy Radio Honolulu
To: Itasca (Black) June 27, 1937 [Java Date; it was 26 June on Howland east of the International Date Line]
SUGGEST ONTARIO STANDBY ON 400 KILOCYCLES TO TRANSMIT LETTER N FIVE MINUTES ON REQUEST WITH STATION CALL REPEATED TWICE END OF EVERY MINUTE STOP SWAN TRANSMIT VOICE NINE MEGACYCLES OR IF I UNABLE RECEIVE READY ON 900 KILOCYCLES STOP ITASCA TRANSMIT LETTER A POSITION OWN CALL LETTERS AS ABOVE ON HALF HOUR 7.5 MEGACYCLES STOP POSITION SHIPS AND OUR LEAVING WILL DETERMINE BROADCASTING SPECIFICALLY STOP IF FREQUENCIES MENTIONED UNSUITABLE NIGHT WORK INFORM ME LAE STOP I WILL GIVE LONG CALL BY VOICE THREE ONE NAUGHT FIVE KCS QUARTER AFTER HOUR POSSIBLY QUARTER TO [signed] EARHART
Had normal air-to-surface communications existed with the Itasca as Earhart approached Howland, the homing problem could almost certainly have been solved quickly. The ship could have told her to home on 500 Kcs., the frequency already being transmitted (in addition to 7.50 Mcs.), and she should have been able to get bearings that would have led her to the ship. Unfortunately, she was unable to hear signals from the Itasca on 3105 Kcs., although the ship was hearing her well. It was thus impossible for the Itasca and Earhart to coordinate their actions.
THE AIR-TO-SURFACE COMMUNICATION PROBLEM: A report by Guinea Airways Ltd. shows that Earhart’s radio gear was checked at Lae by one of its wireless operators, H. J. Balfour, and found satisfactory. Good two-way communication was maintained during a 30-minute test hop at Lae, although a roughness in the transmitted voice signal made Earhart difficult to understand. Balfour told her that her speech might be more intelligible if she spoke in a higher pitch while transmitting.
After the flight left Lae for Howland, two-way communication with Lae was maintained until about 0720 2 July Greenwich Mean Time (GMT) [now Universal Coordinated Time], when she shifted to her 3105 Kcs. night frequency. Several times throughout the night she was heard broadcasting at the prearranged times by stations on Nauru Island and the Itasca, but little of her transmissions were intelligible. Nauru, and later the Itasca, called her numerous times, but there is no indication that she heard any of the calls. At 1515 GMT, the Itasca picked up Earhart calling to say she would listen on 3105 Kcs. on the hour and half-hour. At 1744 GMT she asked the Itasca for a bearing, to be taken then and given to her on the hour. She then whistled into the microphone on 3105 kcs. to create a signal on which the bearing could be taken. The DF operator on Howland heard this signal but was unable to get a bearing. He remarked that the signal had very little carrier and seemed over-modulated. The plane made no response to numerous calls from the Itasca at this time.
At 1815 GMT Earhart again asked the Itasca for a bearing. She wanted it taken then and reported to her in a half hour (at 1845 GMT), and she whistled into the microphone to provide a signal; she said they were about 100 miles out. Again the Howland DF heard her signal but was unable to get a bearing, and again Earhart made no response to numerous calls from the Itasca. At 1912 GMT, Earhart transmitted the following to the Itasca on voice radio:
WE MUST BE ON YOU NOW BUT CANNOT SEE YOU. RUNNING OUT OF GAS. ONLY ONE HALF HOUR LEFT. BEEN UNABLE TO REACH YOU BY RADIO. WE ARE FLYING AT ONE THOUSAND FEET.
The Itasca was on the correct frequency and putting out strong signals at the time—even San Francisco picked them up. In turn, the aircraft’s signals were very strong when the Itasca picked up her transmission; it was obvious that the aircraft’s fixed antenna and its feeder to the transmitter were still intact. Thus Earhart’s transmission “BEEN UNABLE TO REACH YOU BY RADIO” clearly indicates that her receiving system had failed, probably early in the flight. Beyond that there was no clue as to the nature of the failure—but the clue was not long in coming.
After twice failing to obtain a bearing from the Howland DF on 3105 Kcs., Earhart tried to home on the Itasca radio beacon using the aircraft’s direction finder. At 1925 GMT she broadcast to the Itasca:
WE ARE CIRCLING BUT CANNOT HEAR YOU. GO AHEAD ON 7500 NOW OR ON THE SCHEDULE TIME ON HALF HOUR.
By “7500” she was referring to 7500 Kcs., the radio beacon frequency she had specified for the Itasca. The ship complied immediately and transmitted the specified beacon signal—Morse “A”— on 7500 Kcs. The transmitter had no voice capability, so it was impossible to talk to the plane on that frequency. Earhart responded at once on 3105 Kcs., saying:
WE RECEIVED YOUR SIGNALS ON SEVENTY FIVE BUT UNABLE TO GET A MINIMUM. PLEASE TAKE BEARINGS ON US AND ANSWER THREE FIVE NAUGHT FIVE (3105 intended) WITH VOICE.
This was followed by a series of long dashes. No bearing was taken and there was no reply to the Itasca’s subsequent transmission.
Earhart obviously picked up the Itasca’s 7500 Kcs. beacon signals on the aircraft’s loop antenna, because she reported being “Unable to get a minimum,” (the indication of a bearing) and she would not have expected to get a miniumum except with a loop antenna. That she heard the signal indicates her receiver was functioning on at least one band. It was uncommon for only a single band to fail; usually, if one failed, they all failed, and so it is quite likely that the receiver was also functioning on the frequency band containing 3105 Kcs. Under existing conditions, Earhart should have been able to hear both signals on the loop and on the fixed antenna. She did hear 7500 Kcs. on the loop, where signals went directly from loop to receiver, but she did not hear 3105 Kcs. on the fixed antenna, where the incoming signals had to pass through the send-receive relay before reaching the receiver. It is probable, therefore, that the relay had been damaged by lightning or static discharge so that the contacts were not closing properly on the receive side, thus leaving the receiver without an antenna.
No more requests for a bearing were heard. At 2013 GMT Earhart came up on 3105 Kcs., gave a line of position, and said she was shifting to 6210 Kcs.; that was the last time the Itasca heard signals from the plane.
Had Earhart been more familiar with her radio gear and manipulated the antenna selector switch on the receiver to transmit on the fixed antenna, but receive on the loop, she probably would have established two-way communication with Itasca. She apparently did not attempt it.
THE HOWLAND ISLAND RADIO DIRECTION FINDER: Earhart obviously had misconceptions concerning the radio direction finder on Howland. She apparently thought it was a functional equivalent of the Pan American Adcock systems that had furnished her bearings from her 3105 Kcs. signals during the Alameda-Honolulu flight, and she expected that the DF station would be monitoring her signals and it would take a bearing when she asked the Itasca for one. The bearing would be passed to the ship, which would send it to her on the next schedule. This explains why she repeatedly asked the Itasca for bearings on 3105 Kcs. She did not expect the ship to take the bearings with its own DF gear—she was counting on the Howland Island DF.
I think there was some basis for her misconception. After changing to an east-about route, and while the long Lae-Howland leg was being studied, Earhart and Noonan suggested to the Coast Guard that a radio direction finder be set up on Howland. According to an unpublished manuscript by the late Captain Laurance F. Safford, U.S. Navy, (Retired), it was Richard Black, scheduled to go to Howland in the Itasca, who arranged for the Howland DF. Apparently reacting to Noonan’s suggestion, he recommended to George P. Putnam, Earhart’s husband and business manager, that they borrow a high-frequency radio direction finder from the Navy. Subsequently Black, assisted by Lieutenant Daniel A. Cooper of the Army Air Corps, also going to Howland in the Itasca, obtained the desired apparatus from a Navy patrol plane at Pearl Harbor and took it to Howland, where it was jury-rigged to provide a temporary DF capability. An Itasca radioman operated it.
According to Captain Safford, the apparatus was “. . . a 24-volt aircraft type of loop-direction-finder similar to the one installed in Miss Earhart’s plane—possibly its twin.” Black later described it to author Fred G. Goerner as an “experimental model of some of the direction finders we used in the war.” It may have been one of the three experimental receivers built by Bendix, and thus a twin to Earhart’s. Cooper, who helped Black obtain the DF gear, wrote in his official report: “It is true that an airplane direction finder capable of working 3105 KC had been borrowed from the Navy just prior to sailing. This was set up on Howland mainly as a standby in case the ship’s direction finder on 500 KC should go out.”
This clearly shows that Cooper, Black, and Putnam believed that inasmuch as the frequency range of the receiver included 3105 Kcs., it would be able to take bearings on that frequency. Putnam communicated frequently with Earhart and certainly would have kept her apprised of developments regarding the Howland DF; when he told her (while she was in Darwin) that the Itasca reported the DF had been installed on Howland, she had good reason to believe that en route to Howland she would be provided with bearings taken on her 3105 Kcs. signals just as they had been provided her by PAA on the Alameda-Honolulu flight. She was wrong. The apparatus undoubtedly was an excellent receiver and was capable of receiving a wide array of frequencies well above Earhart’s 3105 Kcs. For direction finding, however, it used a simple rotatable loop type antenna, which, because of the very nature of radio wave propagation, is incapable of obtaining meaningful bearings over significant distances on frequencies higher than about 1800 Kcs. On higher frequencies, signals can be heard but no steady null or “minimum” (which indicates the bearing) can be obtained. It should have been no surprise then that the Howland DF was unable to get bearings on the plane. The operator complained that Earhart did not transmit signals long enough for him to take a bearing, but this was irrelevant; longer transmissions would not have helped.
Post-flight signals.
• Nauru. On 3 July (GMT date) an operator at Nauru radio station VKT sent the following “wire note” to RCA radio station KPH at San Francisco, with the request that it be passed to the Itasca:
VOICE HEARD FAIRLY STRONG SIGS STRENGTH TO S3 0843 0854 GMT 48.31 METERS [6210 Kcs] SPEECH NOT INTERPRETED OWING BAD MODULATION OR SPEAKER SHOUTING INTO MICROPHONE BUT VOICE SIMILAR TO THAT EMITTED FROM PLANE IN FLIGHT LAST NIGHT WITH EXCEPTION NO HUM ON PLANE IN BACKGROUND.
The Nauru operator was a professional wireless operator, well qualified to judge the quality of radio signals. He had heard some of Earhart’s transmissions the night before and was familiar with the sound of her voice and of the cockpit background noise. That he was able to recognize the voice but was unable to understand what was being said, and his diagnosis of probable overmodulation, jibe with the reports of the wireless operator at Lae and the DF operator at Howland. He had nothing to gain by fabricating information. Given this—and because Earhart probably was the only woman in that part of the world transmitting voice signals on 6210 KHz, there is a strong case for ascribing the signals to Earhart’s plane. Since more than 12 hours had elapsed between the time the Itasca last heard the plane and the time the Nauru operator intercepted the signals, the aircraft certainly was no longer in flight. The absence of the “hum” (engine noise) in the intercept tends to confirm this.
• Pan American Airways. Shortly after Earhart became overdue at Howland, the Coast Guard requested PAA assistance in the search. The stations at Mokapu Point, Midway, and Wake almost immediately began to monitor the plane’s frequencies consistent with available personnel, and were prepared to take bearings on any signals reasonably believed to be coming from the plane. The airline established a special radio circuit linking the three stations. Numerous weak signals were heard but nothing of interest was picked up until 5 July (GMT). The following is extracted from a report made by the Radio Operator-in-Charge at the Wake Island station, R. M. Hansen:
“At 0948 a phone signal of good intensity and well modulated by a voice but wavering badly suddenly came on 3105. While the carrier frequency of this signal did not appear to vary appreciably, its strength did vary in an unusually erratic manner and at 0950, the carrier strength fell off to QSA2 [2 on a scale of 0 to 5] with the wavering more noticeable than ever. At 0952, it went off completely. . . .At 1212 [GMT 5 July] I opened the DF guard on 3105 KC. At 1223 a very unsteady voice-modulated carrier was observed on 3105 KC appx. This transmission lasted until 1236. I was able to get an approximate bearing of 144 degrees. In spite of the extreme eccentricity of this signal during the entire length of the transmission, the splits were definite and pretty fair. . . . After 1 obtained the observed bearing, I advised Midway to listen for the signal (couldn’t raise Honolulu). He apparently did not hear it. This signal started in at a carrier strength of QSA5 and at 1236, when the transmission stopped it had gradually petered out to QSA2 during the intervals when it was audible. The characteristics of this signal were identical with those of the signal heard the previous night (0948 GMT) except that at DF the complete periods of no signal occurred during shorter intervals. . . . While no identification call letters were distinguished in either case, I was positive at that time that this was KHAQQ [Earhart’s aircraft call letters]. At this date I am still of this opinion.”
• Midway. At 0638 5 July (GMT), the station heard a signal having the same characteristics, and almost certainly the same station. The operator computed a quick bearing of 201° True, but the signal was not audible long enough to take a really good bearing and the 201° figure was labeled “approximate.”
• Honolulu (Mokapu Point). This station also heard the 3106 Kcs. “peculiar signal” several times. From 1523 to 1530 on 4 July (GMT), the station attempted to get a bearing; the signal was weak and shifting, and only a rough bearing was obtained. It was logged as 213° but was by implication a doubtful bearing. Sometime between 0630 and 1225 GMT another bearing was attempted. The log describes it thus: “Signals so weak that it was impossible to obtain even a fair check. Average seems to be around 215 degrees—very doubtful bearing.” It is obvious that the bearings from Honolulu were much inferior to those taken from Wake and Midway; they are useful mainly as indications that the unknown station continued to function.
Few paid any attention to these intercepts at the time because no one was aware that Earhart’s radio signals had been abnormal. Had it been known that she was having overmodulation problems more attention probably would have been given them because the wavering in the carrier strength is consistent with a varying degree of overmodulation rapidly increasing and decreasing carrier power. The gradual drop of signal strength from QSA5 to QSA2 over a span of 13 minutes is consistent with the further discharge of an already partially discharged storage battery power supply. The peculiar signals on 3105 KHz. heard by Wake, Midway, and Honolulu may very well have come from the Earhart plane, and it is likely that the radio bearing taken on those signals by Wake was accurate within a degree or so. The one from Midway may have had a slightly larger error.
FREDERICK J. NOONAN: From personal observation, the writer knows that as of late 1935 Noonan could send and receive plain language at slow speeds, around eight to ten words per minute. Recent research by Noonan biographer Michael A. Lang has revealed that circa 1931 Noonan held a Second Class Commercial Radio operator license issued by the Radio Division of the U.S. Department of Commerce. The license, which was valid for two years, certified that the holder was capable of: “Transmitting and sound reading at a speed of not less than sixteen words a minute Continental Morse in code groups and twenty words a minute in plain language”.
CONCLUSIONS: Earhart failed to reach Howland, because she was unable to use the electronic aids that had been set up to help her find the Island. Her inability to hear the Itasca on the communication channel precluded any possibility of receiving aid from the Howland DF. Therefore she was completely dependent upon bearings she could take on the Itasca beacon with her own DF.
When it became evident that she would get no help from the Howland DF, Earhart prepared to take bearings on the Itasca’s beacon. She tuned in the beacon on her DF and heard the signals clearly. When she tried to take a bearing, however, she was unsuccessful because she could not get a “minimum.” She had no idea why she could not get a bearing, nor did she know what to do to improve the situation. Lack of two-way communication with the Itasca prevented her from getting advice from the ship. Apparently, after a final unsuccessful attempt to have a bearing taken on her 3105 Kcs frequency, she gave up on radio navigation and left the area.
The direct cause of the flight’s failure was Earhart’s unwitting error in designating 7.50 Mcs. as the beacon frequency for the Itasca.
The probable cause for the antenna system failure was malfunctioning of the “send-receive” relay, located physically in the transmitter unit, which left the receiver without an antenna. The relay probably malfunctioned because of damage by lightning or heavy static discharge.
Were They for Real?
If the “peculiar signals” intercepted by Nauru and the PAA stations were in fact from the Earhart plane, then the following may be deduced from the radio signals:
• A successful landing had been made. The plane did not nose over or break up. However damage increased the mismatch between antenna and transmitter, which affected modulation enough to make voice signals unintelligible.
• The landing was not in the open sea. Had it been, salt water seepage would have disabled the transmitting gear in a relatively short time.
• Earhart survived the landing. She was heard by the Nauru operator long after the plane would have run out of gas.
• Noonan survived. A man’s voice was distinctly heard on the “peculiar signal” by Midway. It was unintelligible.
• Either Earhart or Noonan, or both, were alive and with the plane at least until 0948 5 July 5 1937 (GMT). The peculiar signals were last heard then.
• The peculiar signals probably were coming from the eastern or southeastern part of the Marshall Islands.
* Since 1937 the unit of measurement for radio frequencies has been changed from “cycles” to “Hertz" (Hz), consequently Megacycles (Mcs) and MegaHertz (MHz) will he used interchangeably, as will Kilocycles (Kcs) and KiloHertz (KHz).
Amelia waited while Fuzz Furman fixed a gauge. Above the cockpit—the critical loop and stub-mast antennas.
By Captain Almon A. Gray, U.S. Naval Reserve (Retired)
December 1993 Naval History Volume 7 Number 4
Almost certainly, Amelia Earhart could not get a bearing on the radio beacon on the U.S. Coast Guard Cutter Itasca (WPG-321), lying off the beach at Howland Island, because the frequency that she had designated—7.50 Mcs*—was so high that her direction-finder (DF) was inherently incapable of taking bearings on it.
That Earhart and Fred Noonan failed to reach Howland Island on their 1937 around-the-world flight because of radio problems has been said before—but little has been written about the specifics.
A failure in the plane’s antenna system, which made it impossible to receive signals on the fixed antenna, also was a factor. Had she or Noonan known enough about the system to work around the failure, they could have established voice communications, with the Itasca, where someone surely would have suggested they try taking bearings on the vessel’s 500-kilocycle beacon. It could have made all the difference.
BACKGROUND: In early 1937, several weeks before departing Oakland, California, for Honolulu—the first leg of an intended west-about flight around the world— Earhart met at Alameda, California, with George Angus, the Superintendent of Communications for the Pacific Division of Pan American Airways (PAA). Angus directed the radio communication and DF networks that supported the PAA clippers on their Pacific crossings, and she was looking for help to augment Noonan’s celestial navigation.
The airline then had specially designed versions of the Adcock radio DF system in service at Alameda, Mokapu Point on Oahu in the Territory of Hawaii, Midway Island, Wake Island, Guam, and Manila in the Philippines. They could take bearings on frequencies much higher than could conventional loop-type direction finders—like Earhart’s— and were effective over much greater distances. These high-frequency DFs were the only ones of their type in the United States and its territories. Angus agreed to help and went to work on the details.
This was complicated, inasmuch as PAA could receive but not transmit on either of Earhart’s communication frequencies—3105 and 6210 KHz.—and could not transmit voice on any frequency. Earhart and Angus decided that the aircraft would request a bearing by voice on the frequency in use—usually 3105 KHz. at night and 6210 KHz. during the day—and follow the request with a series of long dashes lasting in the aggregate a couple of minutes. The PAA DF station would take a bearing on the transmission and transmit it to the plane on another previously agreed upon PAA frequency, using continuous wave (CW) telegraphy sent at such a slow speed that the individual dots and dashes could be copied on paper and later translated into numbers.
This arrangement was tested on the flight from Oakland to Honolulu; PAA took the bearings on 3105 KHz. and tansmitted the bearings in Morse code on 2986 KHz. The flight was handled much the same as a routine Clipper flight. Captain Harry Manning, former captain of the SS Roosevelt—the ship that brought her home from Europe after her 1928 trans-Atlantic flight—and a long-time friend, was an experienced radio operator and handled the Elec- tra’s radio and DF gear while regular PAA professional radio operators manned the ground stations. Radio bearings furnished the plane at frequent intervals, first from Alameda and later from Mokapu Point, checked well with the positions Noonan determined by celestial navigation. Nearing Oahu, Manning set up the plane’s DF to home on the 290 Khz. marine radio beacon at Makapuu Point, near Diamond Head, and Earhart homed in on it to a successful landfall.
While attempting take-off for Howland Island from Luke Field, near Honolulu, on 20 March 1937, Earhart ground-looped the Electra, damaging it to the extent that it was shipped back to the Lockheed plant in California for repairs. The radio gear sustained no major damage, but the Western Electric Model 20B radio receiver and its remote-control apparatus were replaced by a Bendix aircraft radio receiver and accessories. The stub mast supporting the V-shaped fixed antenna also was moved a bit forward, and the antenna feed line was rerouted. The late Joseph Gurr, then a moonlighting United Airlines technician, did the work.
THE NEW RECEIVER: The receiver installed at Lockheed was an experimental model incorporating the latest improvements. Only three experimental units were built, although Bendix later marketed an almost identical design as the Type RA-1 Aircraft Radio Receiver.
The experimental model was a continuous-tuning superheterodyne that covered the spectrum from 150 to 10,000 Kcs. in five bands. It could receive voice, CW, or modulated CW (MCW) signals and could be controlled remotely from the cockpit. A switch permitted the operator to connect the receiver to either the conventional wire antenna or the loop antenna. When the loop was used, the combination became an effective radio DF system capable of accurate bearings on frequencies between 150 and approximately 1800 Kcs. Signals on frequencies higher than 1800 Kcs. could be heard, but very seldom could accurate bearings be obtained. Earhart was apparently unaware of this. The receiver was powered by a dynamotor operated by storage batteries charged by the main engines.
The Radio System: When the plane left the Lockheed plant, the radio system consisted of the following elements:
• The experimental Bendix aircraft radio receiver
• A Western Electric Model 13-C 50-watt aircraft transmitter with three crystal-controlled channels—500, 3105, and 6210 KHz.—capable of voice or CW transmissions. It was mounted in the cabin, but there were remote controls in the cockpit.
• A prototype of a Bendix Type MN-20 rotatable shielded- loop antenna. It was mounted on the fuselage above the cockpit; the knob that rotated it was on the cockpit overhead between the pilots. It was used primarily for taking radio bearings but was useful as a receiving antenna in static caused by heavy precipitation.
• Fittings at each side of the cockpit for connecting a microphone, headphones, and telegraph key
• A telegraph key and a jack for connecting headphones at the navigator’s table
• A 250-foot flexible-wire trailing antenna on an electrically operated, remote-controlled reel at the rear of the plane. The wire exited the lower fuselage through an insulated bushing and had a lead weight, or “fish,” at the end to keep it from whipping when deployed. A variable loading coil used in conjunction with this antenna permitted its use on 500 KHz., and the antenna was long enough to give excellent radiation efficiency on all three transmitting frequencies.
• A fixed, Vee-configured wire antenna with its apex at a stub mast mounted on the top of the fuselage, over the center section of the wing, and its two legs extending back to the two vertical tail fins. The antenna was so short that its radiation efficiency was extremely low; it was adequate for local communications around an airport when it was not feasible to have the trailing antenna deployed, but not for the long-distance communication Earhart required for her transoceanic flight.
Either wire antenna could be selected from the cockpit. The one selected both transmitted and received by means of a send-receive relay that switched the antenna from the receiver to the transmitter when the microphone button was depressed, and switched it back to the receiver when the button was released.
MISTAKES AT MIAMI: After deciding to change her route to east-about, in late May 1937 Earhart flew the plane to Miami, where she had the trailing antenna and associated gear removed completely. John Ray, an Eastern Airlines technician who had his own radio shop as a sideline, did the work. Once again, Amelia obviously did not comprehend the devastating impact this would have on her ability to communicate and to use radio navigation. With only the very short fixed antenna remaining, virtually no energy could be radiated on 500 KHz. This not only foreclosed any possibility of contacting ships and marine shore stations but precluded ships—most important, the Itasca—and marine shore-based DF stations from taking radio bearings on the plane, inasmuch as 500 KHz. was the only one of her frequencies that fell within the range of the marine direction finders. Any radio aid in locating Howland Island would have to be in the form of radio bearings taken by the plane on radio signals from the Itasca. Earhart had cut her options severely.
The shortness of the remaining antenna also drastically reduced the power radiated on the two high frequencies. Paul Rafford Jr., a NASA expert in this field involved in forecasting long-range communication requirements to support astronaut recoveries, estimated that the radiated power on 3105 KHz. was about one-half watt. This obviously was a tremendous handicap in the high static level of the tropics.
The fixed antenna also may have been at least partly responsible for the distortion in Earhart’s transmitted signals reported by the operators at Lae, New Guinea, and Howland as affecting the intelligibility of her voice transmissions. A mismatch between the antenna and the final amplifier of a WE-13C transmitter could cause the transmitter to over-modulate and thus introduce distortion.
After a few days in the Pan American Airways shops during which all systems, including the antennas, were tuned and peaked, the plane departed Miami on 1 June 1937 to resume the flight around the world.
Despite these shortcomings, Earhart got as far as the Dutch East Indies without major incident. There, however, because of her unfamiliarity with radio matters, she unwittingly made the mistake that ultimately led to her failure to reach Howland Island.
THE FAULTY PLAN: The legs from New Guinea to Howland Island and from Howland to Hawaii were the most difficult navigational portions of the flight, and three small vessels were stationed along the way to assist. Each planned to use the ship’s transmitter as a radio beacon for Earhart and Noonan to supplement Noonan’s celestial navigation:
• The USS Ontario (AT-13) was on station midway between Lae and Howland.
• The USS Swan (AVP-34) was positioned midway between Howland and Hawaii.
• The USCGC Itasca was at Howland. Her beacon was particularly important; should Noonan’s celestial navigation not put them within visual range of the small, low-lying island, homing in on the Itasca’s signal would be their only chance.
By 23 June these vessels were on or approaching their respective stations but had not been issued their radio beacon frequency or procedures. That day, in a message addressed to Earhart at Darwin or Bandoeng, Richard Black— Earhart’s representative on board the Itasca—advised her of the radio frequencies available on the three ships and asked her to designate the frequency she wished each ship to use when transmitting beacon signals. This message caught up with Earhart at Bandoeng, Java.
Precisely what happened next remains unknown, but it appears that Earhart conferred with a local aeronautical communication specialist to get information she could use to base a reply to Black. A plan was developed that fulfilled her requirements with a minimum of receiver tuning on her part. The Ontario and the Itasca were to transmit on the same frequency but at different times, and each would transmit a distinctive Morse identification signal. The Ontario’s identifier was “N”, and the Itasca’s was “A.” (These were the characters used to identify the quadrants of the four-course radio ranges, then the principal navigation aid in the United States. Hence, Earhart was familiar with them.)
In 1937 it was still common to describe radio emissions in wavelengths expressed in meters, rather than in frequencies expressed in cycles per second. It seems apparent that the specialist did this during his discussions with Earhart and that he suggested that the ships transmit beacon signals on wavelengths as follows:
• Ontario 750 meters (400 Kcs.)
• Swan 900 meters (333 Kcs.)
• Itasca 750 meters (400 Kcs.)
These were excellent choices. All were allocated internationally for aeronautical radio navigation and were ideal for use with the direction finder in the Earhart plane.
Unfortunately, Earhart did not understand the relationship between wavelength and frequency nor how to convert from one to the other. Consequently, when she replied to Black on 27 June, she confused the figures and unwittingly specified incorrect frequencies for the Swan and the Itasca; she was correct with the Ontario.
In the case of the Swan, she apparently confused the wavelength and frequency figures, and specified that the Swan transmit on 900 Kcs. (rather than 333 Kcs.). This was a bad error in that 900 Kcs. was in the broadcast band and not available for aeronautical use. It also was inferior to the intended frequency of 333 Kcs. for DF purposes. It was not necessarily devastating, however, and fair bearings probably could have been taken on it with the aircraft DF.
In the Itasca’s case, however, it was to have grave consequences when she again apparently reversed the numbers and told Black to use 7.50 Mcs. (rather than 400 Kcs.) on the Itasca. The 7.50 Mcs. frequency was so high that there was practically no possibility of obtaining usable radio bearings on it with the aircraft DF.
Following is the text of Earhart’s reply to Black, sent the day before she left Bandoeng for Koepang and Darwin:
From: Earhart via RCA Manila & NPM Navy Radio Honolulu
To: Itasca (Black) June 27, 1937 [Java Date; it was 26 June on Howland east of the International Date Line]
SUGGEST ONTARIO STANDBY ON 400 KILOCYCLES TO TRANSMIT LETTER N FIVE MINUTES ON REQUEST WITH STATION CALL REPEATED TWICE END OF EVERY MINUTE STOP SWAN TRANSMIT VOICE NINE MEGACYCLES OR IF I UNABLE RECEIVE READY ON 900 KILOCYCLES STOP ITASCA TRANSMIT LETTER A POSITION OWN CALL LETTERS AS ABOVE ON HALF HOUR 7.5 MEGACYCLES STOP POSITION SHIPS AND OUR LEAVING WILL DETERMINE BROADCASTING SPECIFICALLY STOP IF FREQUENCIES MENTIONED UNSUITABLE NIGHT WORK INFORM ME LAE STOP I WILL GIVE LONG CALL BY VOICE THREE ONE NAUGHT FIVE KCS QUARTER AFTER HOUR POSSIBLY QUARTER TO [signed] EARHART
Had normal air-to-surface communications existed with the Itasca as Earhart approached Howland, the homing problem could almost certainly have been solved quickly. The ship could have told her to home on 500 Kcs., the frequency already being transmitted (in addition to 7.50 Mcs.), and she should have been able to get bearings that would have led her to the ship. Unfortunately, she was unable to hear signals from the Itasca on 3105 Kcs., although the ship was hearing her well. It was thus impossible for the Itasca and Earhart to coordinate their actions.
THE AIR-TO-SURFACE COMMUNICATION PROBLEM: A report by Guinea Airways Ltd. shows that Earhart’s radio gear was checked at Lae by one of its wireless operators, H. J. Balfour, and found satisfactory. Good two-way communication was maintained during a 30-minute test hop at Lae, although a roughness in the transmitted voice signal made Earhart difficult to understand. Balfour told her that her speech might be more intelligible if she spoke in a higher pitch while transmitting.
After the flight left Lae for Howland, two-way communication with Lae was maintained until about 0720 2 July Greenwich Mean Time (GMT) [now Universal Coordinated Time], when she shifted to her 3105 Kcs. night frequency. Several times throughout the night she was heard broadcasting at the prearranged times by stations on Nauru Island and the Itasca, but little of her transmissions were intelligible. Nauru, and later the Itasca, called her numerous times, but there is no indication that she heard any of the calls. At 1515 GMT, the Itasca picked up Earhart calling to say she would listen on 3105 Kcs. on the hour and half-hour. At 1744 GMT she asked the Itasca for a bearing, to be taken then and given to her on the hour. She then whistled into the microphone on 3105 kcs. to create a signal on which the bearing could be taken. The DF operator on Howland heard this signal but was unable to get a bearing. He remarked that the signal had very little carrier and seemed over-modulated. The plane made no response to numerous calls from the Itasca at this time.
At 1815 GMT Earhart again asked the Itasca for a bearing. She wanted it taken then and reported to her in a half hour (at 1845 GMT), and she whistled into the microphone to provide a signal; she said they were about 100 miles out. Again the Howland DF heard her signal but was unable to get a bearing, and again Earhart made no response to numerous calls from the Itasca. At 1912 GMT, Earhart transmitted the following to the Itasca on voice radio:
WE MUST BE ON YOU NOW BUT CANNOT SEE YOU. RUNNING OUT OF GAS. ONLY ONE HALF HOUR LEFT. BEEN UNABLE TO REACH YOU BY RADIO. WE ARE FLYING AT ONE THOUSAND FEET.
The Itasca was on the correct frequency and putting out strong signals at the time—even San Francisco picked them up. In turn, the aircraft’s signals were very strong when the Itasca picked up her transmission; it was obvious that the aircraft’s fixed antenna and its feeder to the transmitter were still intact. Thus Earhart’s transmission “BEEN UNABLE TO REACH YOU BY RADIO” clearly indicates that her receiving system had failed, probably early in the flight. Beyond that there was no clue as to the nature of the failure—but the clue was not long in coming.
After twice failing to obtain a bearing from the Howland DF on 3105 Kcs., Earhart tried to home on the Itasca radio beacon using the aircraft’s direction finder. At 1925 GMT she broadcast to the Itasca:
WE ARE CIRCLING BUT CANNOT HEAR YOU. GO AHEAD ON 7500 NOW OR ON THE SCHEDULE TIME ON HALF HOUR.
By “7500” she was referring to 7500 Kcs., the radio beacon frequency she had specified for the Itasca. The ship complied immediately and transmitted the specified beacon signal—Morse “A”— on 7500 Kcs. The transmitter had no voice capability, so it was impossible to talk to the plane on that frequency. Earhart responded at once on 3105 Kcs., saying:
WE RECEIVED YOUR SIGNALS ON SEVENTY FIVE BUT UNABLE TO GET A MINIMUM. PLEASE TAKE BEARINGS ON US AND ANSWER THREE FIVE NAUGHT FIVE (3105 intended) WITH VOICE.
This was followed by a series of long dashes. No bearing was taken and there was no reply to the Itasca’s subsequent transmission.
Earhart obviously picked up the Itasca’s 7500 Kcs. beacon signals on the aircraft’s loop antenna, because she reported being “Unable to get a minimum,” (the indication of a bearing) and she would not have expected to get a miniumum except with a loop antenna. That she heard the signal indicates her receiver was functioning on at least one band. It was uncommon for only a single band to fail; usually, if one failed, they all failed, and so it is quite likely that the receiver was also functioning on the frequency band containing 3105 Kcs. Under existing conditions, Earhart should have been able to hear both signals on the loop and on the fixed antenna. She did hear 7500 Kcs. on the loop, where signals went directly from loop to receiver, but she did not hear 3105 Kcs. on the fixed antenna, where the incoming signals had to pass through the send-receive relay before reaching the receiver. It is probable, therefore, that the relay had been damaged by lightning or static discharge so that the contacts were not closing properly on the receive side, thus leaving the receiver without an antenna.
No more requests for a bearing were heard. At 2013 GMT Earhart came up on 3105 Kcs., gave a line of position, and said she was shifting to 6210 Kcs.; that was the last time the Itasca heard signals from the plane.
Had Earhart been more familiar with her radio gear and manipulated the antenna selector switch on the receiver to transmit on the fixed antenna, but receive on the loop, she probably would have established two-way communication with Itasca. She apparently did not attempt it.
THE HOWLAND ISLAND RADIO DIRECTION FINDER: Earhart obviously had misconceptions concerning the radio direction finder on Howland. She apparently thought it was a functional equivalent of the Pan American Adcock systems that had furnished her bearings from her 3105 Kcs. signals during the Alameda-Honolulu flight, and she expected that the DF station would be monitoring her signals and it would take a bearing when she asked the Itasca for one. The bearing would be passed to the ship, which would send it to her on the next schedule. This explains why she repeatedly asked the Itasca for bearings on 3105 Kcs. She did not expect the ship to take the bearings with its own DF gear—she was counting on the Howland Island DF.
I think there was some basis for her misconception. After changing to an east-about route, and while the long Lae-Howland leg was being studied, Earhart and Noonan suggested to the Coast Guard that a radio direction finder be set up on Howland. According to an unpublished manuscript by the late Captain Laurance F. Safford, U.S. Navy, (Retired), it was Richard Black, scheduled to go to Howland in the Itasca, who arranged for the Howland DF. Apparently reacting to Noonan’s suggestion, he recommended to George P. Putnam, Earhart’s husband and business manager, that they borrow a high-frequency radio direction finder from the Navy. Subsequently Black, assisted by Lieutenant Daniel A. Cooper of the Army Air Corps, also going to Howland in the Itasca, obtained the desired apparatus from a Navy patrol plane at Pearl Harbor and took it to Howland, where it was jury-rigged to provide a temporary DF capability. An Itasca radioman operated it.
According to Captain Safford, the apparatus was “. . . a 24-volt aircraft type of loop-direction-finder similar to the one installed in Miss Earhart’s plane—possibly its twin.” Black later described it to author Fred G. Goerner as an “experimental model of some of the direction finders we used in the war.” It may have been one of the three experimental receivers built by Bendix, and thus a twin to Earhart’s. Cooper, who helped Black obtain the DF gear, wrote in his official report: “It is true that an airplane direction finder capable of working 3105 KC had been borrowed from the Navy just prior to sailing. This was set up on Howland mainly as a standby in case the ship’s direction finder on 500 KC should go out.”
This clearly shows that Cooper, Black, and Putnam believed that inasmuch as the frequency range of the receiver included 3105 Kcs., it would be able to take bearings on that frequency. Putnam communicated frequently with Earhart and certainly would have kept her apprised of developments regarding the Howland DF; when he told her (while she was in Darwin) that the Itasca reported the DF had been installed on Howland, she had good reason to believe that en route to Howland she would be provided with bearings taken on her 3105 Kcs. signals just as they had been provided her by PAA on the Alameda-Honolulu flight. She was wrong. The apparatus undoubtedly was an excellent receiver and was capable of receiving a wide array of frequencies well above Earhart’s 3105 Kcs. For direction finding, however, it used a simple rotatable loop type antenna, which, because of the very nature of radio wave propagation, is incapable of obtaining meaningful bearings over significant distances on frequencies higher than about 1800 Kcs. On higher frequencies, signals can be heard but no steady null or “minimum” (which indicates the bearing) can be obtained. It should have been no surprise then that the Howland DF was unable to get bearings on the plane. The operator complained that Earhart did not transmit signals long enough for him to take a bearing, but this was irrelevant; longer transmissions would not have helped.
Post-flight signals.
• Nauru. On 3 July (GMT date) an operator at Nauru radio station VKT sent the following “wire note” to RCA radio station KPH at San Francisco, with the request that it be passed to the Itasca:
VOICE HEARD FAIRLY STRONG SIGS STRENGTH TO S3 0843 0854 GMT 48.31 METERS [6210 Kcs] SPEECH NOT INTERPRETED OWING BAD MODULATION OR SPEAKER SHOUTING INTO MICROPHONE BUT VOICE SIMILAR TO THAT EMITTED FROM PLANE IN FLIGHT LAST NIGHT WITH EXCEPTION NO HUM ON PLANE IN BACKGROUND.
The Nauru operator was a professional wireless operator, well qualified to judge the quality of radio signals. He had heard some of Earhart’s transmissions the night before and was familiar with the sound of her voice and of the cockpit background noise. That he was able to recognize the voice but was unable to understand what was being said, and his diagnosis of probable overmodulation, jibe with the reports of the wireless operator at Lae and the DF operator at Howland. He had nothing to gain by fabricating information. Given this—and because Earhart probably was the only woman in that part of the world transmitting voice signals on 6210 KHz, there is a strong case for ascribing the signals to Earhart’s plane. Since more than 12 hours had elapsed between the time the Itasca last heard the plane and the time the Nauru operator intercepted the signals, the aircraft certainly was no longer in flight. The absence of the “hum” (engine noise) in the intercept tends to confirm this.
• Pan American Airways. Shortly after Earhart became overdue at Howland, the Coast Guard requested PAA assistance in the search. The stations at Mokapu Point, Midway, and Wake almost immediately began to monitor the plane’s frequencies consistent with available personnel, and were prepared to take bearings on any signals reasonably believed to be coming from the plane. The airline established a special radio circuit linking the three stations. Numerous weak signals were heard but nothing of interest was picked up until 5 July (GMT). The following is extracted from a report made by the Radio Operator-in-Charge at the Wake Island station, R. M. Hansen:
“At 0948 a phone signal of good intensity and well modulated by a voice but wavering badly suddenly came on 3105. While the carrier frequency of this signal did not appear to vary appreciably, its strength did vary in an unusually erratic manner and at 0950, the carrier strength fell off to QSA2 [2 on a scale of 0 to 5] with the wavering more noticeable than ever. At 0952, it went off completely. . . .At 1212 [GMT 5 July] I opened the DF guard on 3105 KC. At 1223 a very unsteady voice-modulated carrier was observed on 3105 KC appx. This transmission lasted until 1236. I was able to get an approximate bearing of 144 degrees. In spite of the extreme eccentricity of this signal during the entire length of the transmission, the splits were definite and pretty fair. . . . After 1 obtained the observed bearing, I advised Midway to listen for the signal (couldn’t raise Honolulu). He apparently did not hear it. This signal started in at a carrier strength of QSA5 and at 1236, when the transmission stopped it had gradually petered out to QSA2 during the intervals when it was audible. The characteristics of this signal were identical with those of the signal heard the previous night (0948 GMT) except that at DF the complete periods of no signal occurred during shorter intervals. . . . While no identification call letters were distinguished in either case, I was positive at that time that this was KHAQQ [Earhart’s aircraft call letters]. At this date I am still of this opinion.”
• Midway. At 0638 5 July (GMT), the station heard a signal having the same characteristics, and almost certainly the same station. The operator computed a quick bearing of 201° True, but the signal was not audible long enough to take a really good bearing and the 201° figure was labeled “approximate.”
• Honolulu (Mokapu Point). This station also heard the 3106 Kcs. “peculiar signal” several times. From 1523 to 1530 on 4 July (GMT), the station attempted to get a bearing; the signal was weak and shifting, and only a rough bearing was obtained. It was logged as 213° but was by implication a doubtful bearing. Sometime between 0630 and 1225 GMT another bearing was attempted. The log describes it thus: “Signals so weak that it was impossible to obtain even a fair check. Average seems to be around 215 degrees—very doubtful bearing.” It is obvious that the bearings from Honolulu were much inferior to those taken from Wake and Midway; they are useful mainly as indications that the unknown station continued to function.
Few paid any attention to these intercepts at the time because no one was aware that Earhart’s radio signals had been abnormal. Had it been known that she was having overmodulation problems more attention probably would have been given them because the wavering in the carrier strength is consistent with a varying degree of overmodulation rapidly increasing and decreasing carrier power. The gradual drop of signal strength from QSA5 to QSA2 over a span of 13 minutes is consistent with the further discharge of an already partially discharged storage battery power supply. The peculiar signals on 3105 KHz. heard by Wake, Midway, and Honolulu may very well have come from the Earhart plane, and it is likely that the radio bearing taken on those signals by Wake was accurate within a degree or so. The one from Midway may have had a slightly larger error.
FREDERICK J. NOONAN: From personal observation, the writer knows that as of late 1935 Noonan could send and receive plain language at slow speeds, around eight to ten words per minute. Recent research by Noonan biographer Michael A. Lang has revealed that circa 1931 Noonan held a Second Class Commercial Radio operator license issued by the Radio Division of the U.S. Department of Commerce. The license, which was valid for two years, certified that the holder was capable of: “Transmitting and sound reading at a speed of not less than sixteen words a minute Continental Morse in code groups and twenty words a minute in plain language”.
CONCLUSIONS: Earhart failed to reach Howland, because she was unable to use the electronic aids that had been set up to help her find the Island. Her inability to hear the Itasca on the communication channel precluded any possibility of receiving aid from the Howland DF. Therefore she was completely dependent upon bearings she could take on the Itasca beacon with her own DF.
When it became evident that she would get no help from the Howland DF, Earhart prepared to take bearings on the Itasca’s beacon. She tuned in the beacon on her DF and heard the signals clearly. When she tried to take a bearing, however, she was unsuccessful because she could not get a “minimum.” She had no idea why she could not get a bearing, nor did she know what to do to improve the situation. Lack of two-way communication with the Itasca prevented her from getting advice from the ship. Apparently, after a final unsuccessful attempt to have a bearing taken on her 3105 Kcs frequency, she gave up on radio navigation and left the area.
The direct cause of the flight’s failure was Earhart’s unwitting error in designating 7.50 Mcs. as the beacon frequency for the Itasca.
The probable cause for the antenna system failure was malfunctioning of the “send-receive” relay, located physically in the transmitter unit, which left the receiver without an antenna. The relay probably malfunctioned because of damage by lightning or heavy static discharge.
Were They for Real?
If the “peculiar signals” intercepted by Nauru and the PAA stations were in fact from the Earhart plane, then the following may be deduced from the radio signals:
• A successful landing had been made. The plane did not nose over or break up. However damage increased the mismatch between antenna and transmitter, which affected modulation enough to make voice signals unintelligible.
• The landing was not in the open sea. Had it been, salt water seepage would have disabled the transmitting gear in a relatively short time.
• Earhart survived the landing. She was heard by the Nauru operator long after the plane would have run out of gas.
• Noonan survived. A man’s voice was distinctly heard on the “peculiar signal” by Midway. It was unintelligible.
• Either Earhart or Noonan, or both, were alive and with the plane at least until 0948 5 July 5 1937 (GMT). The peculiar signals were last heard then.
• The peculiar signals probably were coming from the eastern or southeastern part of the Marshall Islands.
* Since 1937 the unit of measurement for radio frequencies has been changed from “cycles” to “Hertz" (Hz), consequently Megacycles (Mcs) and MegaHertz (MHz) will he used interchangeably, as will Kilocycles (Kcs) and KiloHertz (KHz).
Amelia waited while Fuzz Furman fixed a gauge. Above the cockpit—the critical loop and stub-mast antennas.
By Captain Almon A. Gray, U.S. Naval Reserve (Retired)
December 1993 Naval History Volume 7 Number 4
Almost certainly, Amelia Earhart could not get a bearing on the radio beacon on the U.S. Coast Guard Cutter Itasca (WPG-321), lying off the beach at Howland Island, because the frequency that she had designated—7.50 Mcs*—was so high that her direction-finder (DF) was inherently incapable of taking bearings on it.
That Earhart and Fred Noonan failed to reach Howland Island on their 1937 around-the-world flight because of radio problems has been said before—but little has been written about the specifics.
A failure in the plane’s antenna system, which made it impossible to receive signals on the fixed antenna, also was a factor. Had she or Noonan known enough about the system to work around the failure, they could have established voice communications, with the Itasca, where someone surely would have suggested they try taking bearings on the vessel’s 500-kilocycle beacon. It could have made all the difference.
BACKGROUND: In early 1937, several weeks before departing Oakland, California, for Honolulu—the first leg of an intended west-about flight around the world— Earhart met at Alameda, California, with George Angus, the Superintendent of Communications for the Pacific Division of Pan American Airways (PAA). Angus directed the radio communication and DF networks that supported the PAA clippers on their Pacific crossings, and she was looking for help to augment Noonan’s celestial navigation.
The airline then had specially designed versions of the Adcock radio DF system in service at Alameda, Mokapu Point on Oahu in the Territory of Hawaii, Midway Island, Wake Island, Guam, and Manila in the Philippines. They could take bearings on frequencies much higher than could conventional loop-type direction finders—like Earhart’s— and were effective over much greater distances. These high-frequency DFs were the only ones of their type in the United States and its territories. Angus agreed to help and went to work on the details.
This was complicated, inasmuch as PAA could receive but not transmit on either of Earhart’s communication frequencies—3105 and 6210 KHz.—and could not transmit voice on any frequency. Earhart and Angus decided that the aircraft would request a bearing by voice on the frequency in use—usually 3105 KHz. at night and 6210 KHz. during the day—and follow the request with a series of long dashes lasting in the aggregate a couple of minutes. The PAA DF station would take a bearing on the transmission and transmit it to the plane on another previously agreed upon PAA frequency, using continuous wave (CW) telegraphy sent at such a slow speed that the individual dots and dashes could be copied on paper and later translated into numbers.
This arrangement was tested on the flight from Oakland to Honolulu; PAA took the bearings on 3105 KHz. and tansmitted the bearings in Morse code on 2986 KHz. The flight was handled much the same as a routine Clipper flight. Captain Harry Manning, former captain of the SS Roosevelt—the ship that brought her home from Europe after her 1928 trans-Atlantic flight—and a long-time friend, was an experienced radio operator and handled the Elec- tra’s radio and DF gear while regular PAA professional radio operators manned the ground stations. Radio bearings furnished the plane at frequent intervals, first from Alameda and later from Mokapu Point, checked well with the positions Noonan determined by celestial navigation. Nearing Oahu, Manning set up the plane’s DF to home on the 290 Khz. marine radio beacon at Makapuu Point, near Diamond Head, and Earhart homed in on it to a successful landfall.
While attempting take-off for Howland Island from Luke Field, near Honolulu, on 20 March 1937, Earhart ground-looped the Electra, damaging it to the extent that it was shipped back to the Lockheed plant in California for repairs. The radio gear sustained no major damage, but the Western Electric Model 20B radio receiver and its remote-control apparatus were replaced by a Bendix aircraft radio receiver and accessories. The stub mast supporting the V-shaped fixed antenna also was moved a bit forward, and the antenna feed line was rerouted. The late Joseph Gurr, then a moonlighting United Airlines technician, did the work.
THE NEW RECEIVER: The receiver installed at Lockheed was an experimental model incorporating the latest improvements. Only three experimental units were built, although Bendix later marketed an almost identical design as the Type RA-1 Aircraft Radio Receiver.
The experimental model was a continuous-tuning superheterodyne that covered the spectrum from 150 to 10,000 Kcs. in five bands. It could receive voice, CW, or modulated CW (MCW) signals and could be controlled remotely from the cockpit. A switch permitted the operator to connect the receiver to either the conventional wire antenna or the loop antenna. When the loop was used, the combination became an effective radio DF system capable of accurate bearings on frequencies between 150 and approximately 1800 Kcs. Signals on frequencies higher than 1800 Kcs. could be heard, but very seldom could accurate bearings be obtained. Earhart was apparently unaware of this. The receiver was powered by a dynamotor operated by storage batteries charged by the main engines.
The Radio System: When the plane left the Lockheed plant, the radio system consisted of the following elements:
• The experimental Bendix aircraft radio receiver
• A Western Electric Model 13-C 50-watt aircraft transmitter with three crystal-controlled channels—500, 3105, and 6210 KHz.—capable of voice or CW transmissions. It was mounted in the cabin, but there were remote controls in the cockpit.
• A prototype of a Bendix Type MN-20 rotatable shielded- loop antenna. It was mounted on the fuselage above the cockpit; the knob that rotated it was on the cockpit overhead between the pilots. It was used primarily for taking radio bearings but was useful as a receiving antenna in static caused by heavy precipitation.
• Fittings at each side of the cockpit for connecting a microphone, headphones, and telegraph key
• A telegraph key and a jack for connecting headphones at the navigator’s table
• A 250-foot flexible-wire trailing antenna on an electrically operated, remote-controlled reel at the rear of the plane. The wire exited the lower fuselage through an insulated bushing and had a lead weight, or “fish,” at the end to keep it from whipping when deployed. A variable loading coil used in conjunction with this antenna permitted its use on 500 KHz., and the antenna was long enough to give excellent radiation efficiency on all three transmitting frequencies.
• A fixed, Vee-configured wire antenna with its apex at a stub mast mounted on the top of the fuselage, over the center section of the wing, and its two legs extending back to the two vertical tail fins. The antenna was so short that its radiation efficiency was extremely low; it was adequate for local communications around an airport when it was not feasible to have the trailing antenna deployed, but not for the long-distance communication Earhart required for her transoceanic flight.
Either wire antenna could be selected from the cockpit. The one selected both transmitted and received by means of a send-receive relay that switched the antenna from the receiver to the transmitter when the microphone button was depressed, and switched it back to the receiver when the button was released.
MISTAKES AT MIAMI: After deciding to change her route to east-about, in late May 1937 Earhart flew the plane to Miami, where she had the trailing antenna and associated gear removed completely. John Ray, an Eastern Airlines technician who had his own radio shop as a sideline, did the work. Once again, Amelia obviously did not comprehend the devastating impact this would have on her ability to communicate and to use radio navigation. With only the very short fixed antenna remaining, virtually no energy could be radiated on 500 KHz. This not only foreclosed any possibility of contacting ships and marine shore stations but precluded ships—most important, the Itasca—and marine shore-based DF stations from taking radio bearings on the plane, inasmuch as 500 KHz. was the only one of her frequencies that fell within the range of the marine direction finders. Any radio aid in locating Howland Island would have to be in the form of radio bearings taken by the plane on radio signals from the Itasca. Earhart had cut her options severely.
The shortness of the remaining antenna also drastically reduced the power radiated on the two high frequencies. Paul Rafford Jr., a NASA expert in this field involved in forecasting long-range communication requirements to support astronaut recoveries, estimated that the radiated power on 3105 KHz. was about one-half watt. This obviously was a tremendous handicap in the high static level of the tropics.
The fixed antenna also may have been at least partly responsible for the distortion in Earhart’s transmitted signals reported by the operators at Lae, New Guinea, and Howland as affecting the intelligibility of her voice transmissions. A mismatch between the antenna and the final amplifier of a WE-13C transmitter could cause the transmitter to over-modulate and thus introduce distortion.
After a few days in the Pan American Airways shops during which all systems, including the antennas, were tuned and peaked, the plane departed Miami on 1 June 1937 to resume the flight around the world.
Despite these shortcomings, Earhart got as far as the Dutch East Indies without major incident. There, however, because of her unfamiliarity with radio matters, she unwittingly made the mistake that ultimately led to her failure to reach Howland Island.
THE FAULTY PLAN: The legs from New Guinea to Howland Island and from Howland to Hawaii were the most difficult navigational portions of the flight, and three small vessels were stationed along the way to assist. Each planned to use the ship’s transmitter as a radio beacon for Earhart and Noonan to supplement Noonan’s celestial navigation:
• The USS Ontario (AT-13) was on station midway between Lae and Howland.
• The USS Swan (AVP-34) was positioned midway between Howland and Hawaii.
• The USCGC Itasca was at Howland. Her beacon was particularly important; should Noonan’s celestial navigation not put them within visual range of the small, low-lying island, homing in on the Itasca’s signal would be their only chance.
By 23 June these vessels were on or approaching their respective stations but had not been issued their radio beacon frequency or procedures. That day, in a message addressed to Earhart at Darwin or Bandoeng, Richard Black— Earhart’s representative on board the Itasca—advised her of the radio frequencies available on the three ships and asked her to designate the frequency she wished each ship to use when transmitting beacon signals. This message caught up with Earhart at Bandoeng, Java.
Precisely what happened next remains unknown, but it appears that Earhart conferred with a local aeronautical communication specialist to get information she could use to base a reply to Black. A plan was developed that fulfilled her requirements with a minimum of receiver tuning on her part. The Ontario and the Itasca were to transmit on the same frequency but at different times, and each would transmit a distinctive Morse identification signal. The Ontario’s identifier was “N”, and the Itasca’s was “A.” (These were the characters used to identify the quadrants of the four-course radio ranges, then the principal navigation aid in the United States. Hence, Earhart was familiar with them.)
In 1937 it was still common to describe radio emissions in wavelengths expressed in meters, rather than in frequencies expressed in cycles per second. It seems apparent that the specialist did this during his discussions with Earhart and that he suggested that the ships transmit beacon signals on wavelengths as follows:
• Ontario 750 meters (400 Kcs.)
• Swan 900 meters (333 Kcs.)
• Itasca 750 meters (400 Kcs.)
These were excellent choices. All were allocated internationally for aeronautical radio navigation and were ideal for use with the direction finder in the Earhart plane.
Unfortunately, Earhart did not understand the relationship between wavelength and frequency nor how to convert from one to the other. Consequently, when she replied to Black on 27 June, she confused the figures and unwittingly specified incorrect frequencies for the Swan and the Itasca; she was correct with the Ontario.
In the case of the Swan, she apparently confused the wavelength and frequency figures, and specified that the Swan transmit on 900 Kcs. (rather than 333 Kcs.). This was a bad error in that 900 Kcs. was in the broadcast band and not available for aeronautical use. It also was inferior to the intended frequency of 333 Kcs. for DF purposes. It was not necessarily devastating, however, and fair bearings probably could have been taken on it with the aircraft DF.
In the Itasca’s case, however, it was to have grave consequences when she again apparently reversed the numbers and told Black to use 7.50 Mcs. (rather than 400 Kcs.) on the Itasca. The 7.50 Mcs. frequency was so high that there was practically no possibility of obtaining usable radio bearings on it with the aircraft DF.
Following is the text of Earhart’s reply to Black, sent the day before she left Bandoeng for Koepang and Darwin:
From: Earhart via RCA Manila & NPM Navy Radio Honolulu
To: Itasca (Black) June 27, 1937 [Java Date; it was 26 June on Howland east of the International Date Line]
SUGGEST ONTARIO STANDBY ON 400 KILOCYCLES TO TRANSMIT LETTER N FIVE MINUTES ON REQUEST WITH STATION CALL REPEATED TWICE END OF EVERY MINUTE STOP SWAN TRANSMIT VOICE NINE MEGACYCLES OR IF I UNABLE RECEIVE READY ON 900 KILOCYCLES STOP ITASCA TRANSMIT LETTER A POSITION OWN CALL LETTERS AS ABOVE ON HALF HOUR 7.5 MEGACYCLES STOP POSITION SHIPS AND OUR LEAVING WILL DETERMINE BROADCASTING SPECIFICALLY STOP IF FREQUENCIES MENTIONED UNSUITABLE NIGHT WORK INFORM ME LAE STOP I WILL GIVE LONG CALL BY VOICE THREE ONE NAUGHT FIVE KCS QUARTER AFTER HOUR POSSIBLY QUARTER TO [signed] EARHART
Had normal air-to-surface communications existed with the Itasca as Earhart approached Howland, the homing problem could almost certainly have been solved quickly. The ship could have told her to home on 500 Kcs., the frequency already being transmitted (in addition to 7.50 Mcs.), and she should have been able to get bearings that would have led her to the ship. Unfortunately, she was unable to hear signals from the Itasca on 3105 Kcs., although the ship was hearing her well. It was thus impossible for the Itasca and Earhart to coordinate their actions.
THE AIR-TO-SURFACE COMMUNICATION PROBLEM: A report by Guinea Airways Ltd. shows that Earhart’s radio gear was checked at Lae by one of its wireless operators, H. J. Balfour, and found satisfactory. Good two-way communication was maintained during a 30-minute test hop at Lae, although a roughness in the transmitted voice signal made Earhart difficult to understand. Balfour told her that her speech might be more intelligible if she spoke in a higher pitch while transmitting.
After the flight left Lae for Howland, two-way communication with Lae was maintained until about 0720 2 July Greenwich Mean Time (GMT) [now Universal Coordinated Time], when she shifted to her 3105 Kcs. night frequency. Several times throughout the night she was heard broadcasting at the prearranged times by stations on Nauru Island and the Itasca, but little of her transmissions were intelligible. Nauru, and later the Itasca, called her numerous times, but there is no indication that she heard any of the calls. At 1515 GMT, the Itasca picked up Earhart calling to say she would listen on 3105 Kcs. on the hour and half-hour. At 1744 GMT she asked the Itasca for a bearing, to be taken then and given to her on the hour. She then whistled into the microphone on 3105 kcs. to create a signal on which the bearing could be taken. The DF operator on Howland heard this signal but was unable to get a bearing. He remarked that the signal had very little carrier and seemed over-modulated. The plane made no response to numerous calls from the Itasca at this time.
At 1815 GMT Earhart again asked the Itasca for a bearing. She wanted it taken then and reported to her in a half hour (at 1845 GMT), and she whistled into the microphone to provide a signal; she said they were about 100 miles out. Again the Howland DF heard her signal but was unable to get a bearing, and again Earhart made no response to numerous calls from the Itasca. At 1912 GMT, Earhart transmitted the following to the Itasca on voice radio:
WE MUST BE ON YOU NOW BUT CANNOT SEE YOU. RUNNING OUT OF GAS. ONLY ONE HALF HOUR LEFT. BEEN UNABLE TO REACH YOU BY RADIO. WE ARE FLYING AT ONE THOUSAND FEET.
The Itasca was on the correct frequency and putting out strong signals at the time—even San Francisco picked them up. In turn, the aircraft’s signals were very strong when the Itasca picked up her transmission; it was obvious that the aircraft’s fixed antenna and its feeder to the transmitter were still intact. Thus Earhart’s transmission “BEEN UNABLE TO REACH YOU BY RADIO” clearly indicates that her receiving system had failed, probably early in the flight. Beyond that there was no clue as to the nature of the failure—but the clue was not long in coming.
After twice failing to obtain a bearing from the Howland DF on 3105 Kcs., Earhart tried to home on the Itasca radio beacon using the aircraft’s direction finder. At 1925 GMT she broadcast to the Itasca:
WE ARE CIRCLING BUT CANNOT HEAR YOU. GO AHEAD ON 7500 NOW OR ON THE SCHEDULE TIME ON HALF HOUR.
By “7500” she was referring to 7500 Kcs., the radio beacon frequency she had specified for the Itasca. The ship complied immediately and transmitted the specified beacon signal—Morse “A”— on 7500 Kcs. The transmitter had no voice capability, so it was impossible to talk to the plane on that frequency. Earhart responded at once on 3105 Kcs., saying:
WE RECEIVED YOUR SIGNALS ON SEVENTY FIVE BUT UNABLE TO GET A MINIMUM. PLEASE TAKE BEARINGS ON US AND ANSWER THREE FIVE NAUGHT FIVE (3105 intended) WITH VOICE.
This was followed by a series of long dashes. No bearing was taken and there was no reply to the Itasca’s subsequent transmission.
Earhart obviously picked up the Itasca’s 7500 Kcs. beacon signals on the aircraft’s loop antenna, because she reported being “Unable to get a minimum,” (the indication of a bearing) and she would not have expected to get a miniumum except with a loop antenna. That she heard the signal indicates her receiver was functioning on at least one band. It was uncommon for only a single band to fail; usually, if one failed, they all failed, and so it is quite likely that the receiver was also functioning on the frequency band containing 3105 Kcs. Under existing conditions, Earhart should have been able to hear both signals on the loop and on the fixed antenna. She did hear 7500 Kcs. on the loop, where signals went directly from loop to receiver, but she did not hear 3105 Kcs. on the fixed antenna, where the incoming signals had to pass through the send-receive relay before reaching the receiver. It is probable, therefore, that the relay had been damaged by lightning or static discharge so that the contacts were not closing properly on the receive side, thus leaving the receiver without an antenna.
No more requests for a bearing were heard. At 2013 GMT Earhart came up on 3105 Kcs., gave a line of position, and said she was shifting to 6210 Kcs.; that was the last time the Itasca heard signals from the plane.
Had Earhart been more familiar with her radio gear and manipulated the antenna selector switch on the receiver to transmit on the fixed antenna, but receive on the loop, she probably would have established two-way communication with Itasca. She apparently did not attempt it.
THE HOWLAND ISLAND RADIO DIRECTION FINDER: Earhart obviously had misconceptions concerning the radio direction finder on Howland. She apparently thought it was a functional equivalent of the Pan American Adcock systems that had furnished her bearings from her 3105 Kcs. signals during the Alameda-Honolulu flight, and she expected that the DF station would be monitoring her signals and it would take a bearing when she asked the Itasca for one. The bearing would be passed to the ship, which would send it to her on the next schedule. This explains why she repeatedly asked the Itasca for bearings on 3105 Kcs. She did not expect the ship to take the bearings with its own DF gear—she was counting on the Howland Island DF.
I think there was some basis for her misconception. After changing to an east-about route, and while the long Lae-Howland leg was being studied, Earhart and Noonan suggested to the Coast Guard that a radio direction finder be set up on Howland. According to an unpublished manuscript by the late Captain Laurance F. Safford, U.S. Navy, (Retired), it was Richard Black, scheduled to go to Howland in the Itasca, who arranged for the Howland DF. Apparently reacting to Noonan’s suggestion, he recommended to George P. Putnam, Earhart’s husband and business manager, that they borrow a high-frequency radio direction finder from the Navy. Subsequently Black, assisted by Lieutenant Daniel A. Cooper of the Army Air Corps, also going to Howland in the Itasca, obtained the desired apparatus from a Navy patrol plane at Pearl Harbor and took it to Howland, where it was jury-rigged to provide a temporary DF capability. An Itasca radioman operated it.
According to Captain Safford, the apparatus was “. . . a 24-volt aircraft type of loop-direction-finder similar to the one installed in Miss Earhart’s plane—possibly its twin.” Black later described it to author Fred G. Goerner as an “experimental model of some of the direction finders we used in the war.” It may have been one of the three experimental receivers built by Bendix, and thus a twin to Earhart’s. Cooper, who helped Black obtain the DF gear, wrote in his official report: “It is true that an airplane direction finder capable of working 3105 KC had been borrowed from the Navy just prior to sailing. This was set up on Howland mainly as a standby in case the ship’s direction finder on 500 KC should go out.”
This clearly shows that Cooper, Black, and Putnam believed that inasmuch as the frequency range of the receiver included 3105 Kcs., it would be able to take bearings on that frequency. Putnam communicated frequently with Earhart and certainly would have kept her apprised of developments regarding the Howland DF; when he told her (while she was in Darwin) that the Itasca reported the DF had been installed on Howland, she had good reason to believe that en route to Howland she would be provided with bearings taken on her 3105 Kcs. signals just as they had been provided her by PAA on the Alameda-Honolulu flight. She was wrong. The apparatus undoubtedly was an excellent receiver and was capable of receiving a wide array of frequencies well above Earhart’s 3105 Kcs. For direction finding, however, it used a simple rotatable loop type antenna, which, because of the very nature of radio wave propagation, is incapable of obtaining meaningful bearings over significant distances on frequencies higher than about 1800 Kcs. On higher frequencies, signals can be heard but no steady null or “minimum” (which indicates the bearing) can be obtained. It should have been no surprise then that the Howland DF was unable to get bearings on the plane. The operator complained that Earhart did not transmit signals long enough for him to take a bearing, but this was irrelevant; longer transmissions would not have helped.
Post-flight signals.
• Nauru. On 3 July (GMT date) an operator at Nauru radio station VKT sent the following “wire note” to RCA radio station KPH at San Francisco, with the request that it be passed to the Itasca:
VOICE HEARD FAIRLY STRONG SIGS STRENGTH TO S3 0843 0854 GMT 48.31 METERS [6210 Kcs] SPEECH NOT INTERPRETED OWING BAD MODULATION OR SPEAKER SHOUTING INTO MICROPHONE BUT VOICE SIMILAR TO THAT EMITTED FROM PLANE IN FLIGHT LAST NIGHT WITH EXCEPTION NO HUM ON PLANE IN BACKGROUND.
The Nauru operator was a professional wireless operator, well qualified to judge the quality of radio signals. He had heard some of Earhart’s transmissions the night before and was familiar with the sound of her voice and of the cockpit background noise. That he was able to recognize the voice but was unable to understand what was being said, and his diagnosis of probable overmodulation, jibe with the reports of the wireless operator at Lae and the DF operator at Howland. He had nothing to gain by fabricating information. Given this—and because Earhart probably was the only woman in that part of the world transmitting voice signals on 6210 KHz, there is a strong case for ascribing the signals to Earhart’s plane. Since more than 12 hours had elapsed between the time the Itasca last heard the plane and the time the Nauru operator intercepted the signals, the aircraft certainly was no longer in flight. The absence of the “hum” (engine noise) in the intercept tends to confirm this.
• Pan American Airways. Shortly after Earhart became overdue at Howland, the Coast Guard requested PAA assistance in the search. The stations at Mokapu Point, Midway, and Wake almost immediately began to monitor the plane’s frequencies consistent with available personnel, and were prepared to take bearings on any signals reasonably believed to be coming from the plane. The airline established a special radio circuit linking the three stations. Numerous weak signals were heard but nothing of interest was picked up until 5 July (GMT). The following is extracted from a report made by the Radio Operator-in-Charge at the Wake Island station, R. M. Hansen:
“At 0948 a phone signal of good intensity and well modulated by a voice but wavering badly suddenly came on 3105. While the carrier frequency of this signal did not appear to vary appreciably, its strength did vary in an unusually erratic manner and at 0950, the carrier strength fell off to QSA2 [2 on a scale of 0 to 5] with the wavering more noticeable than ever. At 0952, it went off completely. . . .At 1212 [GMT 5 July] I opened the DF guard on 3105 KC. At 1223 a very unsteady voice-modulated carrier was observed on 3105 KC appx. This transmission lasted until 1236. I was able to get an approximate bearing of 144 degrees. In spite of the extreme eccentricity of this signal during the entire length of the transmission, the splits were definite and pretty fair. . . . After 1 obtained the observed bearing, I advised Midway to listen for the signal (couldn’t raise Honolulu). He apparently did not hear it. This signal started in at a carrier strength of QSA5 and at 1236, when the transmission stopped it had gradually petered out to QSA2 during the intervals when it was audible. The characteristics of this signal were identical with those of the signal heard the previous night (0948 GMT) except that at DF the complete periods of no signal occurred during shorter intervals. . . . While no identification call letters were distinguished in either case, I was positive at that time that this was KHAQQ [Earhart’s aircraft call letters]. At this date I am still of this opinion.”
• Midway. At 0638 5 July (GMT), the station heard a signal having the same characteristics, and almost certainly the same station. The operator computed a quick bearing of 201° True, but the signal was not audible long enough to take a really good bearing and the 201° figure was labeled “approximate.”
• Honolulu (Mokapu Point). This station also heard the 3106 Kcs. “peculiar signal” several times. From 1523 to 1530 on 4 July (GMT), the station attempted to get a bearing; the signal was weak and shifting, and only a rough bearing was obtained. It was logged as 213° but was by implication a doubtful bearing. Sometime between 0630 and 1225 GMT another bearing was attempted. The log describes it thus: “Signals so weak that it was impossible to obtain even a fair check. Average seems to be around 215 degrees—very doubtful bearing.” It is obvious that the bearings from Honolulu were much inferior to those taken from Wake and Midway; they are useful mainly as indications that the unknown station continued to function.
Few paid any attention to these intercepts at the time because no one was aware that Earhart’s radio signals had been abnormal. Had it been known that she was having overmodulation problems more attention probably would have been given them because the wavering in the carrier strength is consistent with a varying degree of overmodulation rapidly increasing and decreasing carrier power. The gradual drop of signal strength from QSA5 to QSA2 over a span of 13 minutes is consistent with the further discharge of an already partially discharged storage battery power supply. The peculiar signals on 3105 KHz. heard by Wake, Midway, and Honolulu may very well have come from the Earhart plane, and it is likely that the radio bearing taken on those signals by Wake was accurate within a degree or so. The one from Midway may have had a slightly larger error.
FREDERICK J. NOONAN: From personal observation, the writer knows that as of late 1935 Noonan could send and receive plain language at slow speeds, around eight to ten words per minute. Recent research by Noonan biographer Michael A. Lang has revealed that circa 1931 Noonan held a Second Class Commercial Radio operator license issued by the Radio Division of the U.S. Department of Commerce. The license, which was valid for two years, certified that the holder was capable of: “Transmitting and sound reading at a speed of not less than sixteen words a minute Continental Morse in code groups and twenty words a minute in plain language”.
CONCLUSIONS: Earhart failed to reach Howland, because she was unable to use the electronic aids that had been set up to help her find the Island. Her inability to hear the Itasca on the communication channel precluded any possibility of receiving aid from the Howland DF. Therefore she was completely dependent upon bearings she could take on the Itasca beacon with her own DF.
When it became evident that she would get no help from the Howland DF, Earhart prepared to take bearings on the Itasca’s beacon. She tuned in the beacon on her DF and heard the signals clearly. When she tried to take a bearing, however, she was unsuccessful because she could not get a “minimum.” She had no idea why she could not get a bearing, nor did she know what to do to improve the situation. Lack of two-way communication with the Itasca prevented her from getting advice from the ship. Apparently, after a final unsuccessful attempt to have a bearing taken on her 3105 Kcs frequency, she gave up on radio navigation and left the area.
The direct cause of the flight’s failure was Earhart’s unwitting error in designating 7.50 Mcs. as the beacon frequency for the Itasca.
The probable cause for the antenna system failure was malfunctioning of the “send-receive” relay, located physically in the transmitter unit, which left the receiver without an antenna. The relay probably malfunctioned because of damage by lightning or heavy static discharge.
Were They for Real?
If the “peculiar signals” intercepted by Nauru and the PAA stations were in fact from the Earhart plane, then the following may be deduced from the radio signals:
• A successful landing had been made. The plane did not nose over or break up. However damage increased the mismatch between antenna and transmitter, which affected modulation enough to make voice signals unintelligible.
• The landing was not in the open sea. Had it been, salt water seepage would have disabled the transmitting gear in a relatively short time.
• Earhart survived the landing. She was heard by the Nauru operator long after the plane would have run out of gas.
• Noonan survived. A man’s voice was distinctly heard on the “peculiar signal” by Midway. It was unintelligible.
• Either Earhart or Noonan, or both, were alive and with the plane at least until 0948 5 July 5 1937 (GMT). The peculiar signals were last heard then.
• The peculiar signals probably were coming from the eastern or southeastern part of the Marshall Islands.
* Since 1937 the unit of measurement for radio frequencies has been changed from “cycles” to “Hertz" (Hz), consequently Megacycles (Mcs) and MegaHertz (MHz) will he used interchangeably, as will Kilocycles (Kcs) and KiloHertz (KHz).
Amelia waited while Fuzz Furman fixed a gauge. Above the cockpit—the critical loop and stub-mast antennas.
By Captain Almon A. Gray, U.S. Naval Reserve (Retired)
December 1993 Naval History Volume 7 Number 4
Almost certainly, Amelia Earhart could not get a bearing on the radio beacon on the U.S. Coast Guard Cutter Itasca (WPG-321), lying off the beach at Howland Island, because the frequency that she had designated—7.50 Mcs*—was so high that her direction-finder (DF) was inherently incapable of taking bearings on it.
That Earhart and Fred Noonan failed to reach Howland Island on their 1937 around-the-world flight because of radio problems has been said before—but little has been written about the specifics.
A failure in the plane’s antenna system, which made it impossible to receive signals on the fixed antenna, also was a factor. Had she or Noonan known enough about the system to work around the failure, they could have established voice communications, with the Itasca, where someone surely would have suggested they try taking bearings on the vessel’s 500-kilocycle beacon. It could have made all the difference.
BACKGROUND: In early 1937, several weeks before departing Oakland, California, for Honolulu—the first leg of an intended west-about flight around the world— Earhart met at Alameda, California, with George Angus, the Superintendent of Communications for the Pacific Division of Pan American Airways (PAA). Angus directed the radio communication and DF networks that supported the PAA clippers on their Pacific crossings, and she was looking for help to augment Noonan’s celestial navigation.
The airline then had specially designed versions of the Adcock radio DF system in service at Alameda, Mokapu Point on Oahu in the Territory of Hawaii, Midway Island, Wake Island, Guam, and Manila in the Philippines. They could take bearings on frequencies much higher than could conventional loop-type direction finders—like Earhart’s— and were effective over much greater distances. These high-frequency DFs were the only ones of their type in the United States and its territories. Angus agreed to help and went to work on the details.
This was complicated, inasmuch as PAA could receive but not transmit on either of Earhart’s communication frequencies—3105 and 6210 KHz.—and could not transmit voice on any frequency. Earhart and Angus decided that the aircraft would request a bearing by voice on the frequency in use—usually 3105 KHz. at night and 6210 KHz. during the day—and follow the request with a series of long dashes lasting in the aggregate a couple of minutes. The PAA DF station would take a bearing on the transmission and transmit it to the plane on another previously agreed upon PAA frequency, using continuous wave (CW) telegraphy sent at such a slow speed that the individual dots and dashes could be copied on paper and later translated into numbers.
This arrangement was tested on the flight from Oakland to Honolulu; PAA took the bearings on 3105 KHz. and tansmitted the bearings in Morse code on 2986 KHz. The flight was handled much the same as a routine Clipper flight. Captain Harry Manning, former captain of the SS Roosevelt—the ship that brought her home from Europe after her 1928 trans-Atlantic flight—and a long-time friend, was an experienced radio operator and handled the Elec- tra’s radio and DF gear while regular PAA professional radio operators manned the ground stations. Radio bearings furnished the plane at frequent intervals, first from Alameda and later from Mokapu Point, checked well with the positions Noonan determined by celestial navigation. Nearing Oahu, Manning set up the plane’s DF to home on the 290 Khz. marine radio beacon at Makapuu Point, near Diamond Head, and Earhart homed in on it to a successful landfall.
While attempting take-off for Howland Island from Luke Field, near Honolulu, on 20 March 1937, Earhart ground-looped the Electra, damaging it to the extent that it was shipped back to the Lockheed plant in California for repairs. The radio gear sustained no major damage, but the Western Electric Model 20B radio receiver and its remote-control apparatus were replaced by a Bendix aircraft radio receiver and accessories. The stub mast supporting the V-shaped fixed antenna also was moved a bit forward, and the antenna feed line was rerouted. The late Joseph Gurr, then a moonlighting United Airlines technician, did the work.
THE NEW RECEIVER: The receiver installed at Lockheed was an experimental model incorporating the latest improvements. Only three experimental units were built, although Bendix later marketed an almost identical design as the Type RA-1 Aircraft Radio Receiver.
The experimental model was a continuous-tuning superheterodyne that covered the spectrum from 150 to 10,000 Kcs. in five bands. It could receive voice, CW, or modulated CW (MCW) signals and could be controlled remotely from the cockpit. A switch permitted the operator to connect the receiver to either the conventional wire antenna or the loop antenna. When the loop was used, the combination became an effective radio DF system capable of accurate bearings on frequencies between 150 and approximately 1800 Kcs. Signals on frequencies higher than 1800 Kcs. could be heard, but very seldom could accurate bearings be obtained. Earhart was apparently unaware of this. The receiver was powered by a dynamotor operated by storage batteries charged by the main engines.
The Radio System: When the plane left the Lockheed plant, the radio system consisted of the following elements:
• The experimental Bendix aircraft radio receiver
• A Western Electric Model 13-C 50-watt aircraft transmitter with three crystal-controlled channels—500, 3105, and 6210 KHz.—capable of voice or CW transmissions. It was mounted in the cabin, but there were remote controls in the cockpit.
• A prototype of a Bendix Type MN-20 rotatable shielded- loop antenna. It was mounted on the fuselage above the cockpit; the knob that rotated it was on the cockpit overhead between the pilots. It was used primarily for taking radio bearings but was useful as a receiving antenna in static caused by heavy precipitation.
• Fittings at each side of the cockpit for connecting a microphone, headphones, and telegraph key
• A telegraph key and a jack for connecting headphones at the navigator’s table
• A 250-foot flexible-wire trailing antenna on an electrically operated, remote-controlled reel at the rear of the plane. The wire exited the lower fuselage through an insulated bushing and had a lead weight, or “fish,” at the end to keep it from whipping when deployed. A variable loading coil used in conjunction with this antenna permitted its use on 500 KHz., and the antenna was long enough to give excellent radiation efficiency on all three transmitting frequencies.
• A fixed, Vee-configured wire antenna with its apex at a stub mast mounted on the top of the fuselage, over the center section of the wing, and its two legs extending back to the two vertical tail fins. The antenna was so short that its radiation efficiency was extremely low; it was adequate for local communications around an airport when it was not feasible to have the trailing antenna deployed, but not for the long-distance communication Earhart required for her transoceanic flight.
Either wire antenna could be selected from the cockpit. The one selected both transmitted and received by means of a send-receive relay that switched the antenna from the receiver to the transmitter when the microphone button was depressed, and switched it back to the receiver when the button was released.
MISTAKES AT MIAMI: After deciding to change her route to east-about, in late May 1937 Earhart flew the plane to Miami, where she had the trailing antenna and associated gear removed completely. John Ray, an Eastern Airlines technician who had his own radio shop as a sideline, did the work. Once again, Amelia obviously did not comprehend the devastating impact this would have on her ability to communicate and to use radio navigation. With only the very short fixed antenna remaining, virtually no energy could be radiated on 500 KHz. This not only foreclosed any possibility of contacting ships and marine shore stations but precluded ships—most important, the Itasca—and marine shore-based DF stations from taking radio bearings on the plane, inasmuch as 500 KHz. was the only one of her frequencies that fell within the range of the marine direction finders. Any radio aid in locating Howland Island would have to be in the form of radio bearings taken by the plane on radio signals from the Itasca. Earhart had cut her options severely.
The shortness of the remaining antenna also drastically reduced the power radiated on the two high frequencies. Paul Rafford Jr., a NASA expert in this field involved in forecasting long-range communication requirements to support astronaut recoveries, estimated that the radiated power on 3105 KHz. was about one-half watt. This obviously was a tremendous handicap in the high static level of the tropics.
The fixed antenna also may have been at least partly responsible for the distortion in Earhart’s transmitted signals reported by the operators at Lae, New Guinea, and Howland as affecting the intelligibility of her voice transmissions. A mismatch between the antenna and the final amplifier of a WE-13C transmitter could cause the transmitter to over-modulate and thus introduce distortion.
After a few days in the Pan American Airways shops during which all systems, including the antennas, were tuned and peaked, the plane departed Miami on 1 June 1937 to resume the flight around the world.
Despite these shortcomings, Earhart got as far as the Dutch East Indies without major incident. There, however, because of her unfamiliarity with radio matters, she unwittingly made the mistake that ultimately led to her failure to reach Howland Island.
THE FAULTY PLAN: The legs from New Guinea to Howland Island and from Howland to Hawaii were the most difficult navigational portions of the flight, and three small vessels were stationed along the way to assist. Each planned to use the ship’s transmitter as a radio beacon for Earhart and Noonan to supplement Noonan’s celestial navigation:
• The USS Ontario (AT-13) was on station midway between Lae and Howland.
• The USS Swan (AVP-34) was positioned midway between Howland and Hawaii.
• The USCGC Itasca was at Howland. Her beacon was particularly important; should Noonan’s celestial navigation not put them within visual range of the small, low-lying island, homing in on the Itasca’s signal would be their only chance.
By 23 June these vessels were on or approaching their respective stations but had not been issued their radio beacon frequency or procedures. That day, in a message addressed to Earhart at Darwin or Bandoeng, Richard Black— Earhart’s representative on board the Itasca—advised her of the radio frequencies available on the three ships and asked her to designate the frequency she wished each ship to use when transmitting beacon signals. This message caught up with Earhart at Bandoeng, Java.
Precisely what happened next remains unknown, but it appears that Earhart conferred with a local aeronautical communication specialist to get information she could use to base a reply to Black. A plan was developed that fulfilled her requirements with a minimum of receiver tuning on her part. The Ontario and the Itasca were to transmit on the same frequency but at different times, and each would transmit a distinctive Morse identification signal. The Ontario’s identifier was “N”, and the Itasca’s was “A.” (These were the characters used to identify the quadrants of the four-course radio ranges, then the principal navigation aid in the United States. Hence, Earhart was familiar with them.)
In 1937 it was still common to describe radio emissions in wavelengths expressed in meters, rather than in frequencies expressed in cycles per second. It seems apparent that the specialist did this during his discussions with Earhart and that he suggested that the ships transmit beacon signals on wavelengths as follows:
• Ontario 750 meters (400 Kcs.)
• Swan 900 meters (333 Kcs.)
• Itasca 750 meters (400 Kcs.)
These were excellent choices. All were allocated internationally for aeronautical radio navigation and were ideal for use with the direction finder in the Earhart plane.
Unfortunately, Earhart did not understand the relationship between wavelength and frequency nor how to convert from one to the other. Consequently, when she replied to Black on 27 June, she confused the figures and unwittingly specified incorrect frequencies for the Swan and the Itasca; she was correct with the Ontario.
In the case of the Swan, she apparently confused the wavelength and frequency figures, and specified that the Swan transmit on 900 Kcs. (rather than 333 Kcs.). This was a bad error in that 900 Kcs. was in the broadcast band and not available for aeronautical use. It also was inferior to the intended frequency of 333 Kcs. for DF purposes. It was not necessarily devastating, however, and fair bearings probably could have been taken on it with the aircraft DF.
In the Itasca’s case, however, it was to have grave consequences when she again apparently reversed the numbers and told Black to use 7.50 Mcs. (rather than 400 Kcs.) on the Itasca. The 7.50 Mcs. frequency was so high that there was practically no possibility of obtaining usable radio bearings on it with the aircraft DF.
Following is the text of Earhart’s reply to Black, sent the day before she left Bandoeng for Koepang and Darwin:
From: Earhart via RCA Manila & NPM Navy Radio Honolulu
To: Itasca (Black) June 27, 1937 [Java Date; it was 26 June on Howland east of the International Date Line]
SUGGEST ONTARIO STANDBY ON 400 KILOCYCLES TO TRANSMIT LETTER N FIVE MINUTES ON REQUEST WITH STATION CALL REPEATED TWICE END OF EVERY MINUTE STOP SWAN TRANSMIT VOICE NINE MEGACYCLES OR IF I UNABLE RECEIVE READY ON 900 KILOCYCLES STOP ITASCA TRANSMIT LETTER A POSITION OWN CALL LETTERS AS ABOVE ON HALF HOUR 7.5 MEGACYCLES STOP POSITION SHIPS AND OUR LEAVING WILL DETERMINE BROADCASTING SPECIFICALLY STOP IF FREQUENCIES MENTIONED UNSUITABLE NIGHT WORK INFORM ME LAE STOP I WILL GIVE LONG CALL BY VOICE THREE ONE NAUGHT FIVE KCS QUARTER AFTER HOUR POSSIBLY QUARTER TO [signed] EARHART
Had normal air-to-surface communications existed with the Itasca as Earhart approached Howland, the homing problem could almost certainly have been solved quickly. The ship could have told her to home on 500 Kcs., the frequency already being transmitted (in addition to 7.50 Mcs.), and she should have been able to get bearings that would have led her to the ship. Unfortunately, she was unable to hear signals from the Itasca on 3105 Kcs., although the ship was hearing her well. It was thus impossible for the Itasca and Earhart to coordinate their actions.
THE AIR-TO-SURFACE COMMUNICATION PROBLEM: A report by Guinea Airways Ltd. shows that Earhart’s radio gear was checked at Lae by one of its wireless operators, H. J. Balfour, and found satisfactory. Good two-way communication was maintained during a 30-minute test hop at Lae, although a roughness in the transmitted voice signal made Earhart difficult to understand. Balfour told her that her speech might be more intelligible if she spoke in a higher pitch while transmitting.
After the flight left Lae for Howland, two-way communication with Lae was maintained until about 0720 2 July Greenwich Mean Time (GMT) [now Universal Coordinated Time], when she shifted to her 3105 Kcs. night frequency. Several times throughout the night she was heard broadcasting at the prearranged times by stations on Nauru Island and the Itasca, but little of her transmissions were intelligible. Nauru, and later the Itasca, called her numerous times, but there is no indication that she heard any of the calls. At 1515 GMT, the Itasca picked up Earhart calling to say she would listen on 3105 Kcs. on the hour and half-hour. At 1744 GMT she asked the Itasca for a bearing, to be taken then and given to her on the hour. She then whistled into the microphone on 3105 kcs. to create a signal on which the bearing could be taken. The DF operator on Howland heard this signal but was unable to get a bearing. He remarked that the signal had very little carrier and seemed over-modulated. The plane made no response to numerous calls from the Itasca at this time.
At 1815 GMT Earhart again asked the Itasca for a bearing. She wanted it taken then and reported to her in a half hour (at 1845 GMT), and she whistled into the microphone to provide a signal; she said they were about 100 miles out. Again the Howland DF heard her signal but was unable to get a bearing, and again Earhart made no response to numerous calls from the Itasca. At 1912 GMT, Earhart transmitted the following to the Itasca on voice radio:
WE MUST BE ON YOU NOW BUT CANNOT SEE YOU. RUNNING OUT OF GAS. ONLY ONE HALF HOUR LEFT. BEEN UNABLE TO REACH YOU BY RADIO. WE ARE FLYING AT ONE THOUSAND FEET.
The Itasca was on the correct frequency and putting out strong signals at the time—even San Francisco picked them up. In turn, the aircraft’s signals were very strong when the Itasca picked up her transmission; it was obvious that the aircraft’s fixed antenna and its feeder to the transmitter were still intact. Thus Earhart’s transmission “BEEN UNABLE TO REACH YOU BY RADIO” clearly indicates that her receiving system had failed, probably early in the flight. Beyond that there was no clue as to the nature of the failure—but the clue was not long in coming.
After twice failing to obtain a bearing from the Howland DF on 3105 Kcs., Earhart tried to home on the Itasca radio beacon using the aircraft’s direction finder. At 1925 GMT she broadcast to the Itasca:
WE ARE CIRCLING BUT CANNOT HEAR YOU. GO AHEAD ON 7500 NOW OR ON THE SCHEDULE TIME ON HALF HOUR.
By “7500” she was referring to 7500 Kcs., the radio beacon frequency she had specified for the Itasca. The ship complied immediately and transmitted the specified beacon signal—Morse “A”— on 7500 Kcs. The transmitter had no voice capability, so it was impossible to talk to the plane on that frequency. Earhart responded at once on 3105 Kcs., saying:
WE RECEIVED YOUR SIGNALS ON SEVENTY FIVE BUT UNABLE TO GET A MINIMUM. PLEASE TAKE BEARINGS ON US AND ANSWER THREE FIVE NAUGHT FIVE (3105 intended) WITH VOICE.
This was followed by a series of long dashes. No bearing was taken and there was no reply to the Itasca’s subsequent transmission.
Earhart obviously picked up the Itasca’s 7500 Kcs. beacon signals on the aircraft’s loop antenna, because she reported being “Unable to get a minimum,” (the indication of a bearing) and she would not have expected to get a miniumum except with a loop antenna. That she heard the signal indicates her receiver was functioning on at least one band. It was uncommon for only a single band to fail; usually, if one failed, they all failed, and so it is quite likely that the receiver was also functioning on the frequency band containing 3105 Kcs. Under existing conditions, Earhart should have been able to hear both signals on the loop and on the fixed antenna. She did hear 7500 Kcs. on the loop, where signals went directly from loop to receiver, but she did not hear 3105 Kcs. on the fixed antenna, where the incoming signals had to pass through the send-receive relay before reaching the receiver. It is probable, therefore, that the relay had been damaged by lightning or static discharge so that the contacts were not closing properly on the receive side, thus leaving the receiver without an antenna.
No more requests for a bearing were heard. At 2013 GMT Earhart came up on 3105 Kcs., gave a line of position, and said she was shifting to 6210 Kcs.; that was the last time the Itasca heard signals from the plane.
Had Earhart been more familiar with her radio gear and manipulated the antenna selector switch on the receiver to transmit on the fixed antenna, but receive on the loop, she probably would have established two-way communication with Itasca. She apparently did not attempt it.
THE HOWLAND ISLAND RADIO DIRECTION FINDER: Earhart obviously had misconceptions concerning the radio direction finder on Howland. She apparently thought it was a functional equivalent of the Pan American Adcock systems that had furnished her bearings from her 3105 Kcs. signals during the Alameda-Honolulu flight, and she expected that the DF station would be monitoring her signals and it would take a bearing when she asked the Itasca for one. The bearing would be passed to the ship, which would send it to her on the next schedule. This explains why she repeatedly asked the Itasca for bearings on 3105 Kcs. She did not expect the ship to take the bearings with its own DF gear—she was counting on the Howland Island DF.
I think there was some basis for her misconception. After changing to an east-about route, and while the long Lae-Howland leg was being studied, Earhart and Noonan suggested to the Coast Guard that a radio direction finder be set up on Howland. According to an unpublished manuscript by the late Captain Laurance F. Safford, U.S. Navy, (Retired), it was Richard Black, scheduled to go to Howland in the Itasca, who arranged for the Howland DF. Apparently reacting to Noonan’s suggestion, he recommended to George P. Putnam, Earhart’s husband and business manager, that they borrow a high-frequency radio direction finder from the Navy. Subsequently Black, assisted by Lieutenant Daniel A. Cooper of the Army Air Corps, also going to Howland in the Itasca, obtained the desired apparatus from a Navy patrol plane at Pearl Harbor and took it to Howland, where it was jury-rigged to provide a temporary DF capability. An Itasca radioman operated it.
According to Captain Safford, the apparatus was “. . . a 24-volt aircraft type of loop-direction-finder similar to the one installed in Miss Earhart’s plane—possibly its twin.” Black later described it to author Fred G. Goerner as an “experimental model of some of the direction finders we used in the war.” It may have been one of the three experimental receivers built by Bendix, and thus a twin to Earhart’s. Cooper, who helped Black obtain the DF gear, wrote in his official report: “It is true that an airplane direction finder capable of working 3105 KC had been borrowed from the Navy just prior to sailing. This was set up on Howland mainly as a standby in case the ship’s direction finder on 500 KC should go out.”
This clearly shows that Cooper, Black, and Putnam believed that inasmuch as the frequency range of the receiver included 3105 Kcs., it would be able to take bearings on that frequency. Putnam communicated frequently with Earhart and certainly would have kept her apprised of developments regarding the Howland DF; when he told her (while she was in Darwin) that the Itasca reported the DF had been installed on Howland, she had good reason to believe that en route to Howland she would be provided with bearings taken on her 3105 Kcs. signals just as they had been provided her by PAA on the Alameda-Honolulu flight. She was wrong. The apparatus undoubtedly was an excellent receiver and was capable of receiving a wide array of frequencies well above Earhart’s 3105 Kcs. For direction finding, however, it used a simple rotatable loop type antenna, which, because of the very nature of radio wave propagation, is incapable of obtaining meaningful bearings over significant distances on frequencies higher than about 1800 Kcs. On higher frequencies, signals can be heard but no steady null or “minimum” (which indicates the bearing) can be obtained. It should have been no surprise then that the Howland DF was unable to get bearings on the plane. The operator complained that Earhart did not transmit signals long enough for him to take a bearing, but this was irrelevant; longer transmissions would not have helped.
Post-flight signals.
• Nauru. On 3 July (GMT date) an operator at Nauru radio station VKT sent the following “wire note” to RCA radio station KPH at San Francisco, with the request that it be passed to the Itasca:
VOICE HEARD FAIRLY STRONG SIGS STRENGTH TO S3 0843 0854 GMT 48.31 METERS [6210 Kcs] SPEECH NOT INTERPRETED OWING BAD MODULATION OR SPEAKER SHOUTING INTO MICROPHONE BUT VOICE SIMILAR TO THAT EMITTED FROM PLANE IN FLIGHT LAST NIGHT WITH EXCEPTION NO HUM ON PLANE IN BACKGROUND.
The Nauru operator was a professional wireless operator, well qualified to judge the quality of radio signals. He had heard some of Earhart’s transmissions the night before and was familiar with the sound of her voice and of the cockpit background noise. That he was able to recognize the voice but was unable to understand what was being said, and his diagnosis of probable overmodulation, jibe with the reports of the wireless operator at Lae and the DF operator at Howland. He had nothing to gain by fabricating information. Given this—and because Earhart probably was the only woman in that part of the world transmitting voice signals on 6210 KHz, there is a strong case for ascribing the signals to Earhart’s plane. Since more than 12 hours had elapsed between the time the Itasca last heard the plane and the time the Nauru operator intercepted the signals, the aircraft certainly was no longer in flight. The absence of the “hum” (engine noise) in the intercept tends to confirm this.
• Pan American Airways. Shortly after Earhart became overdue at Howland, the Coast Guard requested PAA assistance in the search. The stations at Mokapu Point, Midway, and Wake almost immediately began to monitor the plane’s frequencies consistent with available personnel, and were prepared to take bearings on any signals reasonably believed to be coming from the plane. The airline established a special radio circuit linking the three stations. Numerous weak signals were heard but nothing of interest was picked up until 5 July (GMT). The following is extracted from a report made by the Radio Operator-in-Charge at the Wake Island station, R. M. Hansen:
“At 0948 a phone signal of good intensity and well modulated by a voice but wavering badly suddenly came on 3105. While the carrier frequency of this signal did not appear to vary appreciably, its strength did vary in an unusually erratic manner and at 0950, the carrier strength fell off to QSA2 [2 on a scale of 0 to 5] with the wavering more noticeable than ever. At 0952, it went off completely. . . .At 1212 [GMT 5 July] I opened the DF guard on 3105 KC. At 1223 a very unsteady voice-modulated carrier was observed on 3105 KC appx. This transmission lasted until 1236. I was able to get an approximate bearing of 144 degrees. In spite of the extreme eccentricity of this signal during the entire length of the transmission, the splits were definite and pretty fair. . . . After 1 obtained the observed bearing, I advised Midway to listen for the signal (couldn’t raise Honolulu). He apparently did not hear it. This signal started in at a carrier strength of QSA5 and at 1236, when the transmission stopped it had gradually petered out to QSA2 during the intervals when it was audible. The characteristics of this signal were identical with those of the signal heard the previous night (0948 GMT) except that at DF the complete periods of no signal occurred during shorter intervals. . . . While no identification call letters were distinguished in either case, I was positive at that time that this was KHAQQ [Earhart’s aircraft call letters]. At this date I am still of this opinion.”
• Midway. At 0638 5 July (GMT), the station heard a signal having the same characteristics, and almost certainly the same station. The operator computed a quick bearing of 201° True, but the signal was not audible long enough to take a really good bearing and the 201° figure was labeled “approximate.”
• Honolulu (Mokapu Point). This station also heard the 3106 Kcs. “peculiar signal” several times. From 1523 to 1530 on 4 July (GMT), the station attempted to get a bearing; the signal was weak and shifting, and only a rough bearing was obtained. It was logged as 213° but was by implication a doubtful bearing. Sometime between 0630 and 1225 GMT another bearing was attempted. The log describes it thus: “Signals so weak that it was impossible to obtain even a fair check. Average seems to be around 215 degrees—very doubtful bearing.” It is obvious that the bearings from Honolulu were much inferior to those taken from Wake and Midway; they are useful mainly as indications that the unknown station continued to function.
Few paid any attention to these intercepts at the time because no one was aware that Earhart’s radio signals had been abnormal. Had it been known that she was having overmodulation problems more attention probably would have been given them because the wavering in the carrier strength is consistent with a varying degree of overmodulation rapidly increasing and decreasing carrier power. The gradual drop of signal strength from QSA5 to QSA2 over a span of 13 minutes is consistent with the further discharge of an already partially discharged storage battery power supply. The peculiar signals on 3105 KHz. heard by Wake, Midway, and Honolulu may very well have come from the Earhart plane, and it is likely that the radio bearing taken on those signals by Wake was accurate within a degree or so. The one from Midway may have had a slightly larger error.
FREDERICK J. NOONAN: From personal observation, the writer knows that as of late 1935 Noonan could send and receive plain language at slow speeds, around eight to ten words per minute. Recent research by Noonan biographer Michael A. Lang has revealed that circa 1931 Noonan held a Second Class Commercial Radio operator license issued by the Radio Division of the U.S. Department of Commerce. The license, which was valid for two years, certified that the holder was capable of: “Transmitting and sound reading at a speed of not less than sixteen words a minute Continental Morse in code groups and twenty words a minute in plain language”.
CONCLUSIONS: Earhart failed to reach Howland, because she was unable to use the electronic aids that had been set up to help her find the Island. Her inability to hear the Itasca on the communication channel precluded any possibility of receiving aid from the Howland DF. Therefore she was completely dependent upon bearings she could take on the Itasca beacon with her own DF.
When it became evident that she would get no help from the Howland DF, Earhart prepared to take bearings on the Itasca’s beacon. She tuned in the beacon on her DF and heard the signals clearly. When she tried to take a bearing, however, she was unsuccessful because she could not get a “minimum.” She had no idea why she could not get a bearing, nor did she know what to do to improve the situation. Lack of two-way communication with the Itasca prevented her from getting advice from the ship. Apparently, after a final unsuccessful attempt to have a bearing taken on her 3105 Kcs frequency, she gave up on radio navigation and left the area.
The direct cause of the flight’s failure was Earhart’s unwitting error in designating 7.50 Mcs. as the beacon frequency for the Itasca.
The probable cause for the antenna system failure was malfunctioning of the “send-receive” relay, located physically in the transmitter unit, which left the receiver without an antenna. The relay probably malfunctioned because of damage by lightning or heavy static discharge.
Were They for Real?
If the “peculiar signals” intercepted by Nauru and the PAA stations were in fact from the Earhart plane, then the following may be deduced from the radio signals:
• A successful landing had been made. The plane did not nose over or break up. However damage increased the mismatch between antenna and transmitter, which affected modulation enough to make voice signals unintelligible.
• The landing was not in the open sea. Had it been, salt water seepage would have disabled the transmitting gear in a relatively short time.
• Earhart survived the landing. She was heard by the Nauru operator long after the plane would have run out of gas.
• Noonan survived. A man’s voice was distinctly heard on the “peculiar signal” by Midway. It was unintelligible.
• Either Earhart or Noonan, or both, were alive and with the plane at least until 0948 5 July 5 1937 (GMT). The peculiar signals were last heard then.
• The peculiar signals probably were coming from the eastern or southeastern part of the Marshall Islands.
* Since 1937 the unit of measurement for radio frequencies has been changed from “cycles” to “Hertz" (Hz), consequently Megacycles (Mcs) and MegaHertz (MHz) will he used interchangeably, as will Kilocycles (Kcs) and KiloHertz (KHz).
Reply all
Reply to author
Forward
0 new messages