First implemented in 1928, LFR was a system of specialized low frequency radio stations that each projected 4 “beams” that interconnected to form a network of radio airways that pilots could follow in low visibility conditions by listening to auditory cues over their radio headsets. The beams were created by two overlaid signal patterns emitted by two pairs of antennae in an “X” configuration, one a Morse “A” (dot-dash) code, the other a “N” (dash-dot) code which synced along each leg to create a continuous tone when a pilot was “on the beam” or on course. If a pilot tracked off course, either the A or N code would be heard over the other, indicating the direction back to the beam. It was the first viable solution to a problem that had long dogged aviation since its first days.
As soon as the first aviators stumbled into fog, it was recognized that “blind flying” or “instrument flight,” (the ability to pilot a plane through the densest weather without any reference to the horizon or ground) would be fundamental to reliable, scheduled air travel. By the early 1920’s gyroscopic instruments, first the turn and bank indicator and later the artificial horizon, would solve the first problem: giving pilots a sense of “up” in the spectacularly disorienting conditions within the clouds (this hard lesson was and is still learned the fatal way by untrained pilots). However, this technology could not guide pilots to their destination - more was needed.
Aviation and radio started their lives in parallel at the beginning of the 20th century. Early on it was realized that the latter technology could penetrate the weather and night sky to help pilots find their way. Experimental radio compasses appeared on aircraft as early as 1920 and as the name implied, they could point their way to beacons or even a commercial radio station. However, in the era of vacuum-tube radios the technology was complex, bulky, temperamental and expensive – only larger planes could carry the weight. Additionally, until World War II, most systems required that a pilot manually turn an antenna or turn the entire plane to find a signal – and even then it couldn’t immediately tell the pilot if he was moving toward or away on that bearing line (the “180° ambiguity” problem). If there was a continual crosswind, a pilot could actually fly a curved course to the beacon per the diagram below. “Tracking” at a slight angle to compensate could help, but as winds could fluctuate along a course, it was not precise. The ideal system would allow pilots to use cheap, light equipment to follow a fixed course, a highway in the air, which made it easier to fly a straight line with a crosswind, avoid hazards such as high terrain and allowed for better separation of two-way air traffic along a course.
The 1924 Transcontinental Airway System showed aviation’s clear potential. It’s hundreds of lighted tower beacons sped US Air Mail flights through the night and allowed for 2-day coast-to-coast delivery times. It was a technical marvel but proved absolutely useless against inclement weather. Finding a solution became paramount.
The answer that became the Low Frequency Radio Range system was first made workable by the Ford Motor Company. A patent was filed in 1928 by Eugene S. Donovan, the lead engineer on this effort. Earlier concepts for the system were developed in Germany in 1906 which were later experimented with by the US Bureau of Standards and Army Signal Corps. Henry Ford personally saw the future of aviation and wanted to part of it, stating in 1924 “it is now or never to get a hold of commercial flying and make a success of it.” He placed his considerable financial resources behind his ambitions, buying out Stout Metal Airplane, the company that would later make the famous Ford Tri-motor, and building a new state-of-the-art Ford Airport at Dearborn, just outside of Detroit.
In 1926, as part of this initiative, Ford engineers looked to perfect a reliable radio beacon to help them ferry auto parts between their Chicago and Dearborn plants using Tri-Motors during the brutal Midwest winter. The first station at Ford Airport was built and licensed by October 1926 and by February 1927 the system proved to very successful. Citing the transformative impact it would have on aviation and the public welfare, and considering perhaps the probable growth in sales for its aviation division, Ford declined to collect royalties from US Government for using the invention.
Progress was very quick once the 1926 Air Commerce Act empowered the US Department of Commerce's new Bureau of Air Commerce to establish a system of radio navigation aids along the “Federal Airways.” They, in turn, enlisted the US Bureau of Standards to find suitable technology. Experimental work commenced at their College Park, MD station in December 1926, and test runs began between additional demonstration stations at Hadley Field NJ, and Cleveland, OH from December 1927 through February 1928. Right after that point, the Bellefonte, PA station went on-air as the first commercially available Federal range station. It was joined by the two earlier demonstration stations to establish the first “Radio Airway” between New York and Cleveland in November 1928. This grew to 9 stations by 1929. On September 24th of that year, James Doolittle, who would go on lead the famous raid on Tokyo in 1941, successfully took off, circled and landed a plane with a tarp-covered cockpit using a radio range and another specialized radio beam for landing, proving once and for all that flight solely on instruments was possible. Coast-to-coast service from New York to San Francisco was achieved in 1931. By 1935 the US Mainland was crossed by a network of 117 stations, and by the end of World War II over 400 were in place with a couple hundred more across the globe.
The LFR system was considered an "indispensable" godsend in its era – for the first time, pilots could reliably navigate long distances without seeing the ground. Its development received front page coverage in the press and popular media as a major technical achievement.
The crux of LFR was simple. The airways were established by the "beams" projected from hundreds of stations, typically near a significant intersection or airfield. Pilots would listen to their signals by tuning into stations with simple, inexpensive AM radio sets in the cockpit with headphones, about the same size and cost of small tabletop radios from the era. Uncle Sam provided and maintained all of the heavy, expensive equipment on the ground. LFR was simultaneously the first nationwide radio communication network that could reach aircraft in flight. As covered in the next section, this ability was first used to broadcast regular weather reports to aviators and then to help establish the US air traffic control system.
Each station had two diagonally opposed sets of vertical antennae that transmitted a figure 8 directional pattern of radio waves at a 90° angle to each other per the diagram above creating four sectors or quadrants. Two antennae emitted a Morse code dash-dot tone, the “N” signal, into the two yellow quadrants and the other pair emitted the dot-dash tone, the “A” signal, into the two opposite blue quadrants. Note that each signal was an exact negative of each other, the noiseless gaps in one signal were synced with the dot/dash sounds in the other. When the pilot was equally placed between the towers on the line that separated the quadrants, whether it be one mile away or fifty, both signals would precisely intermesh into a steady, continuous 1,020 hertz tone. These four lines were the “legs” or “beams,” the course a pilot would fly along. When “flying a beam,” if a pilot heard that simple monotone, they knew they were on-course, or “on the beam”, and that is how that term entered the American lexicon in the mid-20th century. If the pilot was closer to the A side or the N side, that tone would become dominant, and the pilot would make the necessary course corrections. These beams were roughly 3° wide and could extend from 50 to over 100 miles from the station.
If a pilot was on the edge of the beam in the “twilight” zone, one signal would faintly emerge above the tone and the pilot knew to make a slight correction. When there was two-way air traffic along a beam, standard practice was for opposing aircraft to fly the right twilight edge of this beam to ensure proper separation. Farther away from the beam, both signals (one much louder) could be easily distinguished in the “bi-signal” zone. Deeper into a quadrant only one signal would be heard, the “pure quadrant” or “clear” area – especially at the “bisector,” a line in the middle of each quadrant equidistant from the beams at either side. These variations were subtle but could provide clues to help a pilot navigate to a beam.
Every 30 seconds a two-letter Morse code station identifier (lasting 8 seconds) was repeated first on the N side and then on A side in Morse code to allow the pilot to confirm that they were listening to the right station. For example Los Angeles was “LA”, Chicago “CG”, etc. Starting in the 1940's, some station identifiers were changed to three letters (for example, “LA” became “LAX”) due to the increasing number of overall radio facilities.
Recordings of an actual range station can be found on the next page
A pilot could tell that the plane was approaching a range station when the volume in the headphones steadily increased. When a pilot was over the range station, as the radio waves from like A or N antennae cancelled each other out above the station, both signals would suddenly die off in a “cone of silence” then, just as quickly, resume which became proof-positive that a pilot was directly over the station. Additionally, the A and N sides would reverse. Later on, stations added a “Z” marker, a second upward pointed radio beam that would activate a beacon light on the instrument panel and a 3,000 Hz tone in the headset to more positively confirm the crossing.
Flying between ranges, the system could not tell you the exact distance from any station. As such, the ranges were complimented by additional similar “fan markers”, which as the name implied, had a wide 3 by 12-mile elliptical or bone-shaped beam spread perpendicular to an airway to mark particular waypoints. These beacon lights would flash 1 to 4 times per their compass direction from a reference station (1 for north, 2 for east, etc.), or could also be heard aurally, to make a positive ID. This basic system survives today in the form of the inner marker beacons used on instrument landing system approaches to airports.
Finally, another helpful aid was that, in the US, true north from the station always lied in an N quadrant, which meant that there would generally always be north and south N quadrants, and east and west A quadrants. In Canada this reference line was shifted 45° west to 315°, which in theory would allow the 4 beams to better align to the cardinal compass points but created northwest and southeast N quadrants, and northeast and southwest A quadrants.
LFR is certainly best known as a radio navigation system. However, its dual capacity as an aviation communication network wasn’t lost upon its inventors, first to make flying safer by broadcasting the latest weather conditions to aircraft on their journeys and then later as a founding piece of the nation's air traffic control system, and is an often-overlooked story.
Early experiments with push-button “radiotelegraphs” confirmed that it was simply unrealistic to expect pilots to simultaneously fly and transceive long messages in Morse code. As not every aircraft could carry a radio operator it became clear that aviation demanded voice communications. Fortunately, the compact radio sets developed for LFR already included this new capability, which meant right from the start every pilot that used the system could also hear a voice broadcast. After their promise was shown during the First World War, the two-way “radiotelephone” also became practical for aviation by 1930. Additionally, the teletype had matured by this time into nationwide networks where volumes of information could be disseminated as quickly as one could type, from stock quotes to the latest news. As the Bureau of Air Commerce rolled out LFR along the Federal Airways after 1928, it repurposed the older airmail radio stations it inherited, originally intended for just station-to-station ground use, as Airway Radio Stations (after 1936, Airway Communication Stations or ACS) and used all these developments to give LFR two more critical capabilities.
The first was the regular broadcast of current and forecasted weather reports from the ACS over all LFR stations to any aircraft on their part of the network, typically every 30 to 60 minutes, 24/7. This was made possible by the National Weather Bureau which collected dispatches from its army of weather observers and issued comprehensive national forecasts via teletype to all points every 6 hours around the clock, a cadence that continues today. To this day, aspiring pilots must memorize the standardized abbreviations originally invented for teletype still used in aviation weather reports. The broadcast format evolved over time but a sample from 1936 would start “This is Airway Communications Station Burbank Broadcasting.” It would then progress through each LFR station in succession: “Burbank, Burbank: Overcast. Ceiling 1,000. Visibility 7 miles. North Pass Open. Daggett, Daggett: Clear and Unlimited” and so on, concluding with “Burbank Radio Range is Now Being Resumed.” Finally, pilots could get a “look ahead” for their journey.
Second, for the planes so equipped with radiotelephones (which was, at first, only larger commercial and military aircraft) LFR also provided the critical two-way link to ACS. In the beginning, pilots could request updates, relay messages and seek assistance in emergencies – but this ability helped spur an even greater transformation. In 1920 London’s Croydon Airport set up the first “control tower” to manage its air traffic by signals and radio. Cleveland’s airport was the first in the US to be similarly towered and 20 other cities soon followed suit. In 1935, the Bureau of Air Commerce urged airlines to create their own control “centers” to coordinate increasing air traffic in busy regions between airports. In 1936, it took charge and merged these into its new national Air Traffic Control (ATC) system using LFR as its communication backbone along the 22,000 miles of airways its beams now defined. Now all aircraft within two-way radio range of an LFR station, even those on remote stretches of the airways far from any airport tower, could be contacted and controlled, greatly improving safety and reliability.
Working around the clock and coordinating via teletype, center controllers could ensure aircraft separation by assigning specific “time offs”, routes and altitudes. Flight plans were reviewed, adjusted as needed, approved and forwarded to other ACS on the filed route. Well before computers and radar, controllers chalked up flights on dispatch blackboards, and tracked their progress by moving “shrimp boat” markers, each clipped with a note card of flight details, on their table maps. It is interesting to note that in this era pilots did not have direct contact with ATC, which functioned more as a behind the scenes clearinghouse until the 1950’s. Further ATC instructions were relayed to ACS who, in turn, communicated them via radio through the LFR stations to aircraft en route. If required to manage traffic “flow” ATC could request aircraft circle and “hold” or alter course. Pilots radioed backwards through this same chain their position reports and requested plan deviations. 1936 also brought the requirement that all aircraft flying on instruments on Federal Airways (which was nearly all scheduled air service by this point) be under positive ATC control through two-way radios and have proper instrumentation; however, the use of receive-only LFR radios by small planes would continue until well into the postwar era.
To avoid any confusion with an airport control “Tower,” an Airway Radio Station or later ACS called up by a pilot through an LFR station was referred to as a “Radio,” e.g. “Oakland Radio”, “La Guardia Radio” etc. During World War II, the UC Army Air Force set up its own separate “Army Airways” Communications Stations (AACS), but eventually these were merged back into a single national system in 1958. As air traffic continued to rapidly grow, ACS responsibilities were devolved: radar equipped “Approach” or terminal control centers were set up to manage arrivals and departures from the busiest airports and the “Centers” now communicated directly with the aircraft in between. In 1960 the ACS were renamed Flight Services Stations (FSS) to reflect their remaining briefing and advisory responsibilities. However, nearly a century later, when a pilot calls up these stations from the air, the convention “Radio” is still used.
The stations were fairly large as the longer wavelengths of low frequency radio required large antennae. First generation Loop Stations were square shaped and were roughly 300’ on each side (2-acre site area). They had 40’ wooden masts at each corner that supported elongated “A” and “N” loop antennae in an “X” configuration across the site to create the signal quadrants and beams. Second generation Adcock Stations emerged after 1933, shown above and described in more detail below, and were about 600’ to 800’ on a side (8 to 15 acre site area). These more advanced stations had 120' to 135’ tall riveted steel antennae towers or “mast radiators” at the corners of a 425' square, spaced 600’ diagonally, that performed the same function. Fortunately for the Bureau of Air Commerce, land was cheap back then; especially for the mostly rural station sites. Federal land agents would commonly appeal to the owner’s sense of patriotic duty to secure a lease or purchase agreement.
The system used Low-Frequency radio (190 kHz to 400 kHz, and up to 536 kHz for military ranges), just below the AM Radio broadcast range. As 300 kHz and above is considered “Medium-Frequency” some sources referred to the LF Range as the “Low-Frequency / Medium-Frequency” or LF/MF Range. These frequencies have long wavelengths that tend to travel further over the curvature of the earth and reflect off of the ionosphere, e.g. the “skywaves” that make global shortwave radio possible. Longer ranges were possible with "night effect" when the ionosphere is most reflective. For this reason, most stations needed only 50 to 150 watts of transmitter output power (equivalent to an incandescent household lightbulb) to be heard over 100 miles away, although some stations had up to 1,500 watts of power. As discussed later on, this frequency was also susceptible to other negative effects but this low power requirement made the station apparatus relatively economical and easy to set up, and only a cheap, lightweight AM receiver was needed on each aircraft, which greatly aided its adoption.
With the loop type antennae used at the first series of stations, transmission or reception is strongest in the direction of the loop and is weakest at the “null” point 90° to its plane. This principal works to transmit radio signals from a loop in a specific direction or, in the case of a radio compass, to align it with the source of a signal. As applied to these LFR stations, it caused the A and N loop to emit two opposed signal lobes which appeared from above as crossed figure-eight patterns forming the 4 quadrants. The beams formed where these signal lobes where in balance along the beams, as both A and N signals would interlock at equal volume along this line. The signals of each identical A or N antenna pair were transmitted 180° out of phase, which meant that when these antennae were seen as exactly equidistant, as would be case if one was directly above the station or along one of the bisector lines, they would cancel each other out – forming the “cone of silence” and the “pure quadrant” or “clear” zones that better defined the quadrants.
Loop stations were easy to set up, which would prove advantageous in certain situations; but required that horizontal wires be strung hundreds of feet between masts where they and their beams were exposed to the effects of wind. More importantly, it was soon discovered that the long flat sides of the loops sent significant amounts of horizontally polarized radio energy upward, creating excessive skywaves that bounced back off the ionosphere and could interfere with the station’s own signals beyond a 30-mile radius, especially at night.
In 1919 British Engineer Frank Adcock determined that two electrically connected but physically separate pairs of rigid vertical antennae acted as a more effective version of a loop antenna. In 1933, US Bureau of Standards scientist Harry Diamond realized this antenna form, lacking the troublesome horizontal lines, would eliminate the skywave problem by transmitting all of the radio energy outwards as vertically polarized waves. The Bureau of Air Commerce, not wanting to reference the name of a British inventor, initially called these new stations the “T-L” or Transmission Line” type. However, common sense would prevail and “Adcock Range” became the common term. 28 stations would be so upgraded by year’s end and they would be dominant type going forward. However, loop stations would still be used for many medium and lower power sites as they were simply more economical with their smaller size and wooden masts, especially during the steel shortage of World War II.
Adcock ranges also solved another nuisance: with loop stations the range signal had to be shut down to allow the voice or weather broadcast. Pilots who had the ability to transmit sometimes had to ask the operator to suspend a report if they were in the middle of an approach or other critical phase. Nearly all Adcock stations added a dedicated fifth center antenna and a “simultaneous transmitter” that could handle the voice communication separately to allow the range signals to run uninterrupted. As a result, such Adcock ranges were also referred to as “Simultaneous” ranges.
A transmitter building, hut or “blockhouse” lied at the center of each station: a generally utilitarian structure approximately 500 sf in area that housed the transmitters (16’ x 32’ was a common size). They were built to standardized plans but construction type varied by what was common to a region, from cast-in-place concrete, masonry block, wood frame to metal construction. It housed two dual-redundant transmitter units that allowed for backup and switchover for routine maintenance. An automatic telegraph key triggered by revolving, notched Bakelite disks (akin to a music box) continuously keyed the dots and dashes for the signals and station identification. The signals were then sent through a radio goniometer, a set of adjustable nested coils, that could rotate the entire beam pattern as needed to best align with the airways (this device was developed by Italian engineers Ettore Bellini and Alessandro Tosi in 1907 for the reverse application of finding the direction of a radio signal). If the power failed to the site, assuming that it wasn’t a more rural site that already generated its own power, an automatic standby generator was provided in a separate room or a nearby detached hut along with a battery rack sufficient to briefly sustain power to the transmitters as the generator started up.
For loop stations, antennae were basically cable loops fed from a center mast next to blockhouse, supported by glass insulators from wooden utility poles driven into the ground. This certainly greatly sped erection time which was a definite wartime advantage. Later loop stations also had Z markers which took the form of a simple dipole antenna with the ends stretched to a third set of opposite masts between the loops.
Adcock antenna (technically, “mast radiators”) were more impressive 120’ to 135’ tall 4-sided vertical lattice trusses typically made of prefabricated sections of riveted angle iron. Each were topped with a red anti-collision light that also made visual identification of the station possible on clear nights. This height was actually less than ideal as antennae should be at least 1/8 wavelength or around 400’ for LFR, but it was compromise made due to cost and the proximity of many stations to airports. The self-supporting towers also lacked guy wires that could cause interference. The antennae towers stood on top of squat, truncated pyramidal bases of heavier steel members that transferred the wind forces (which could be significant at exposed sites) to a reinforced concrete foundation. Each antenna was mounted on this lower structure by four ceramic insulators that electrically isolated it from the earth. Each had a rain shield that prevented water from running over its exterior, which could short circuit the antenna and impact beam alignment (this was a particular problem in deserts when storms wetted the conductive alkaline dust that coated exposed surfaces). The antennae were fed by shielded coaxial cables, that ran from the blockhouse either underground or via ground level supports, to minimize signal leakage and stray skywaves. These were connected via a “tuning unit” at the base, an adjustable signal transformer that helped optimize signal output and ensured that it was synchronized across all four antennae.
Mast radiators need to have a good electrical ground (calibrated through the tuning unit) so the earth’s surface can act as a “ground plane” to effectively reflect any downcast signal back into the sky. Many sites (especially in desert regions) had dry, rocky soil that was not electrically conductive enough to provide this. In this situation the base of each Adcock antennae was also fitted with a “counterpoise,” a square wire mesh or round cable disk elevated about 8’ off the earth to serve this function. At other sites, they also mitigated the impacts that vegetation growth, groundwater changes and nearby tides could otherwise have on the signal pattern.
A few details for the more technically inclined: an Adcock station’s center tower actually emitted a continuous omni-directional signal at the nominal station frequency (let say it was 200 kHz), and this carrier wave could be modulated for the voice and weather broadcasts. As it was continuous it could also be used as a homing beacon by a radio compass by those planes so equipped. The N and A towers actually broadcasted their signals at that frequency plus 1.020 kHz (so 201.02 kHz); the 1,020 hertz tone the pilot heard for the dots and dashes. When both signals were converted into sound waves by the radio’s electronics, like a piano chord, they blended and produced the audible, out of phase “beat” frequency of 1,020 Hz. Later radio receivers had a bandpass filter that allowed the pilot to switch the receiver to hear the tone only, voice only (as the human voice is lower between 300 Hz and 800 Hz) or both. Many aviation receivers from the era also had an “automatic volume control” (AVC) which made it easier to listen to conventional AM stations as they drifted in and out of range; however, handbooks earnestly warned pilots to always disable this feature when using the range so the pilot could properly hear changes in the signals.
Nearly all stations ran unmanned, one of the first “lights out” facilities of the 20th century, illuminated only by the faint warm glow of vacuum tubes within the transmitters accompanied by the sounds of whirring cooling fans and the never ending clicking of the automatic key. Range station operators were actually located at the Airway Communication Stations many miles away (often at airport terminals) that handled the weather broadcasts and radio communications for several stations through dedicated telephone landlines. Through the same circuits they could control each one remotely by using a rotary telephone-like device that could send 16 preset commands, ranging from turning signals off to allow a voice broadcast, to turning on the anti-collision lights on the radio towers at night. Generally, only a daily inspection visit was needed. A few remote stations, such as the Donner Pass range (located on frequently snowbound 7,135’ ridge) and some Aleutian Island sites, did have a live-in station keeper but this was, by far, the exception.
Especially given the fact vacuum tubes were prone to occasional burnout, there was a rigorous schedule of inspection, cleaning and preventative maintenance. Numerous ground monitoring stations, a fleet of inspection aircraft and the end-user pilots themselves were always checking for deviations. Any departure from normal would be investigated immediately and a Notice to Airman (NOTAM) would be dispatched via teletype to all ground facilities to warn pilots. Later stations also incorporated an early form of automatic internal fault detection. The station would monitor its own radio output and if beam alignment drifted or an automatic key stuck, the station would transmit a series of three U’s (dot-dot-dash) between station identifiers telling the pilot to be especially wary using this station.
Although LFR was certainly not problem free its issues were at least predictable and, to a large extent, manageable. Overall, the system was thought to be generally reliable for its era. It was rare for LFR to be implicated in aircraft incidents but it occurred from time to time. In April 1936, TWA first faulted the Pittsburgh range for the nearby fatal crash of its DC-1 on Flight 1; however, it was found be in satisfactory working order – pilot error and schedule pressure were later blamed. Nor was LFR to blame for the United Flight 6 DC-3 crash off Pt. Reyes, CA in November 1938, but an unusually reflective ionosphere overwhelmed the pilot’s headphones with the signals of distant stations, resulting in a navigation error and fuel exhaustion. However, in November 1940, early A.M. snow shorted out one of the antennae insulators of the Salt Lake City range which long had a troublesome operating history due to local atmospheric and terrain effects. This pushed its north beam 17° eastwards off course, sending United Flight 16’s DC-3 into a nearby mountain with no survivors.
Stations were grouped into five classes based on type and power as shown on radio facility charts and other tables. The lower power of medium range stations produced less interference due to reflected signals from terrain - an important consideration at mountainous sites:
These designations would then have the prefix “S” added to it confirm that that station was Simultaneous and “B” if it had weather broadcasts. The suffix “W” would be added if it lacked voice capability and “Z” if it had a Z marker. Thus a “SBMRAZ” station was a Simultaneous medium-powered Adcock range, equipped with a Z marker, that provided weather broadcasts. Classifications for supporting stations included:
In terms of cost, the May 15, 1939 “Air Commerce Bulletin” gave the 1938 cost of a new simultaneous Adcock radio range station as $44,000 or $813,000 in 2020 Dollars. A group of 3 medium and low power ranges also from that year cost $87,000 or roughly $29,000 per site - $536,000 in 2020 Dollars. These figures did not include the original development costs, the cost of infrastructure (power, telephone and road access) needed to support the sites or the equipment in the Airway Communication Stations, as well as ongoing maintenance and operational costs which a 1928 report stated alone was roughly half of each station’s initial cost per annum. When multiplied across the 600+ US stations that existed over the 45-year lifespan of LFR the overall program cost would certainly be in the billions in today’s dollars.
These were all the major pieces, let's see how well it did (and sometimes didn't) work...