EARLY RADIO NAV / 1926 - 1950

---

This site focuses primarily on the history of the Low Frequency Radio Range (LFR). However, we believe it's also useful to provide the context of the many other major radio navigation systems that were its contemporaries throughout the world. This page provides a general overview of each system, maps of known station locations, and references to other sites run by dedicated subject matter experts that can provide more in-depth information. These references are based upon our personal research on the web, if others believe we should add other sources, by all means, please contact me.

Inspired by a 1907 patent by German engineer Otto Scheller, the US deployed LFR as the first widespread aerial radio navigation system in 1927 as a superior alternative to the then inherent limitations of the Radio Compass. Within a few years, Germany’s Lorenz A.G. developed landing and long-range navigation systems that used the same interlocking beam principle. As Europe veered closer to World War II, both Axis and Allied powers sought to gain a strategic upper hand with improved technology. The Lorenz company would upgrade the power of its landing aid to steer Luftwaffe bombers over Britain, and used it as the basis for a multibeam system that became Sonne / Consol. The Allied forces applied earlier theory of hyperbolic navigation systems to create Gee, LORAN and Decca.  

These three latter systems proved very effective and were widely adapted for civilian use during the Post War period. The 1960’s saw the addition of Omega, the first worldwide system, and the Russian Chayka and Alpha systems. VOR (after the interim VAR system), covered in detail on the “Its Fate” page, would ultimately become the international standard for aerial navigation, but LORAN, Decca and Omega would persist until the turn of the 21st century until they were finally displaced by Satellite Navigation. 

Comparison of Systems by Total Number of Stations: In compiling this information, we tallied the total number of ground stations for each system, listed below in descending order. This list includes only long-range stations and excludes instrument landing aids:

• VOR                                          ≈3,000 Stations, 66.3 %
• LFR                                            ≈ 880 Stations, 19.5%
• Decca                                      ≈ 230 Stations, 5.5%
• LORAN-A & C                     155 Stations, 3.4%
• VAR Stations                        ≈ 120 Stations (70 US, plus ≈50 in Australia), 2.6%
• Gee                                            ≈ 75 Stations, 1.7%
• Chayka                                    24 Stations, 0.5%
• Lorenz (long range)          15 Stations (Australia Aeradio), 0.3%
• Sonne / Consol                   14 Stations, 0.3% 
• Omega                                     9 Stations, 0.2%
• Alpha                                        5 Stations, 0.1%
• TOTAL                                     ≈4,524 Stations, 100.0%    

Thus, it appears the Low Frequency Radio Range was second only to VOR in terms of being the most prolific radio navigation system by total number of stations, more than all of the other systems outside of VOR combined.

Germany’s Lorenz Beam (named after its company) was, in essence, a narrow single aural “beam” that used two overlapping 33.33Mhz signals, one of dots and the other of dashes, that merged to create an on-course "equisignal" tone. Lorenz engineer Ernst Kramer derived it from an earlier 1917 patent by Otto Scheller that also informed the design of LFR. Lorenz actually transmitted a single signal on a center 4m vertical dipole antenna, closely flanked by two identical dipole reflectors which would shape its narrow beam. A cam switch would activate relays that would alternately break or short each reflector dipole in the middle, rendering it less effective, slightly deflecting the beam in that direction. By alternating these breaks, the beam was “waggled” briefly 1/8 of a second left for the dot lobe, then longer right for the dash lobe, in a one-second cycle. A pilot on the centerline heard both orientations equally, forming the audible 1000 Hz equisignal. These beams actually extended from the front and back sides of the antennae array, giving it a reverse sensing but usable “back course,” much like some modern approaches. 

Lorenz started its life as a blind landing aid. Located at the far departure end of the runway, the single 120 to 500-watt transmitter and keying equipment were accommodated in a compact 10’ x 15’ shack. The aerials mounted just outside projected their beam up to 30km (18 miles) downrange. Aircraft typically approached and lined up with the beam at about 250m (820’) above ground. An outer marker beacon 3km (1.8 miles) from the end of the runway notified the pilot to start their descent, and an inner marker at 300m (1,000’) indicated that landing was imminent. The outer marker emitted two audible 700 Hz dashes every 2 seconds, the inner emitted 6 1,700Hz dots per second. Each 5-watt beacon transmitter fit under the approach in a small, aptly named “doghouse” with a dipole raised on a short mast above. This size also lent the system to portable, truck-based installations. Although based on Scheller’s patent, its design was also influenced by the earlier US Diamond-Dunmore system tested between 1929 and 1932. 

Tested at Berlin’s Tempelhof airport in 1932, the system was commercially introduced by German flag carrier Lufthansa in 1934 and was later adopted at dozens of airports worldwide. In keeping with Scheller’s patent the system was aural, but visual indicators were soon added that utilized a vertical needle that rhythmically “kicked” left or right toward the beam, with additional neon lamps that marked passage over inner and outer marker beacons, rhythmically flashing to their corresponding dots and dashes. Crude altitude guidance was also provided by a signal strength meter that slowly dropped to zero the closer the aircraft approached the runway. However, as it measured the inverse square of the signal, the resulting glide slope was actually parabolic, steeper at the start and shallower at the runway threshold – not as easy to fly as the constant glide slope of modern systems.   

Lorenz was indeed German, but the company was actually owned by the American concern International Telephone and Telegraph (ITT) and its design would ultimately shape the development the modern Instrument Landing System (ILS) by the US during World War II. The British Royal Air Force also coopted the design as the Standard Beam Approach (SBA) system, but using Morse “A” and “N” letters for the left and right sides instead of dots and dashes like LFR. They widely installed and used SBA well into the postwar period.  

Lorenz’s utility was initially expanded beyond a landing system when Australia adopted it as its first long-range navigation system, and the very first Very High Frequency (VHF) navigation system in the world. Driven by 400-watt transmitters, the system elevated the dipole array on 25m towers boosting its range up to 120 miles. Each station’s front and back course alignments could be bent up to 30° to better align with airways. Civil aviation authorities first experimented with an LFR station in Sydney in 1936, but ultimately preferred Lorenz as it higher radio frequency was more immune to static and the system’s potential for visual indicators which LFR totally lacked. After the 1938 crash of the DC-2 Kyeema the Lorenz subsidiary Standard Telephone and Cable rapidly implemented an “Aeradio” network in the populous southeast that grew to 15 stations. The system remained in use until 1953 when it was replaced by the Visual Aural Radio Range (VAR), which was very similar with its two aural and two visual beams. 

Lorenz would go on to play another, more menacing role when Nazi Germany redeveloped it into the Knickebein (“Crooked Leg”) bombing aid during the 1939 “Battle of the Beams." Using large, rotating rectangular antenna arrays 300’ wide and 100’ tall, the system intersected two powerful 3,000-watt Lorenz beams with a 750-mile range over targets in England from three sites across the Channel and North Sea in Occupied Europe. These were later augmented by 9 smaller half-size units that used standard Lorenz transmitters that had a 180-to-260-mile range. Bombers flew along one beam and crews released their payloads when they crossed and heard the equisignal of the other. The more accurate X-Gerät (“X Apparatus”) version added two more intersecting beams ahead of the release point to alert the bomber crew that the target was approaching and to automatically arm the bomb drop. Both systems were accurate enough to send bombs down the centerline of factories, but fortunately the British learned to jam them by 1941. 

Lorenz, primarily in the form of SBA, was used in the postwar period but became obsolete in the presence of VAR, VOR and modern ILS - the last system was shut down in 1960. 

Map & Additional Resources: Frank Dörenberg maintains an excellent website, Radio Air Navigation through World War II, that has great history and technical descriptions of early radio navigation systems, including Lorenz, Knickebein and many others. The Australian Civil Aviation Historical Society & Airways Museum website at https://airwaysmuseum.com/ contains detailed information on Lorenz’s use in that country, but unfortunately little could be found in the way of maps. A public domain database (no cited author) provided the location of the 12 Knickebein stations, and we authored an Australia Lorenz database by locating the airfields the beacons were reportedly placed at, but these are generally rough locations. We hope that additional information will be received over time to further refine these.

The Sonne system (German for “sun" and short for Elektra-sonnen) was also developed by Kramer and implemented by Nazi Germany at the start of World War II in occupied Europe and neutral Spain. The British learned to surreptitiously use it and named it Consol (Latin for “by the sun”) and its 1,000-mile range and ease of use would be long endearing features. 

Sonne combined the Lorenz Beam and the 1929 British Orfordness experimental beacon, which rotated a loop antenna at 1RPM that emitted a continuous tone: it would key the letter “V” (3 dots and dash) the moment it passed north and the observer obtained a bearing by timing the passage of the null. Sonne’s stations radiated an array of 22 rotating equisignals arranged in two equal opposed “fans” about 130° wide on each side of the station. Charts would show these fans, each subdivided by bold lines into 11 stationary pie shaped sectors averaging 12° wide, which were further divided into 60 increments. A repeating 60-second cycle began with a 20-second omnidirectional timing pulse, followed by the Morse code station identifier, then over a 30-second “keying cycle” each equisignal would slowly sweep its assigned sector from one side to the other. The user would first hear a series of dots every ½ second that would then blend momentarily into the equisignal tone at the moment of passage and then fade into ½ second dashes - the total number of dashes and dots always summed to 60. The cycle would then repeat, for some stations these time increments were doubled in a 120-second cycle. 

Like LFR, a user only needed a simple AM receiver. If one was so equipped, a radio compass could use the initial timing pulse to provide a handy bearing to determine which sector they were in; however, this could also be ascertained by the last known position, dead reckoning or other means. From there, a user just needed to count the dots from the start of the keying cycle to the moment the equisignal was heard. A chart or table would then convert this count into a bearing accurate to within ½° to ¼° (1° with night effects) - many charts dispensed even with this step by directly numbering bearing lines with the dot count. Ideally, two or three counts would be averaged for accuracy, and a second station could then establish a fix.

Although simple to use, Consol had a complex signal pattern. Its transmitter continuously fed a 250-350khz signal to a 100m center antenna, which was flanked by two identical outer antennae spaced 3 wavelengths (about 1.8 miles) away - all three were arranged on the same axis. During the keying cycle, the same signal alternated between the outside antennae which, in conjunction with the center signal, created two heart-shaped “cardioid” dot and dash patterns. Interference effects within the three-wavelength gap rippled each cardioid into a starfish-like collapsed hyperbolic pattern where the perimeter was separated into distinct spikes or lobes by deep nulls. These lobes were, in effect, narrow beams 7.5° wide - one set carrying dots and the other dashes. Further, one keyed signal was phase shifted ¼ of a wave forward, while the other was retarded ¼ of a wave, which correspondingly rotated one set of beams around the station slightly toward one of the outer antennae, and shifting the other set of beams an equal amount in the opposite direction.

This made each dot and dash pattern a mirror image of each other along the antennae axis, where the dot and dash beams were evenly interlaced at each side per the diagram above. The equisignals formed at the midpoints between where both beams were at equal intensity. During the keying cycle the initial phase shift was slowly reversed: the advanced signal was retarded by ½ wave and retarded signal advanced ½ wave. This 180° phase flip caused each signal pattern to gradually morph into each other’s mirror image. The dot and dash beams exchanged positions, crossing over each other, ending up at the starting position of their counterparts. As they migrated around the station's center, so did their corresponding equisignals which swept across each sector. 

Some sources use the term “rotating” to describe how the beams interacted during the phase change. This term should be applied carefully as it can imply the two patterns counterrotated over each other which is not the case: the beams on both sides of the longitudinal axis for each pattern rotated toward one end, like oars on opposite sides of a rowboat.

Kramer developed Sonne’s precursor Elektra in 1938 as a long-distance multi-beam navigation system for the German Wehrmacht that improved upon LFR. Sonne built upon this by adding the sweeping equisignals, allowing users off a beam to still determine their location. Three stations were operational by 1940, with more added during the war. The British detected it early on and by 1944 they obtained complete system details from a captured U-boat. They determined that it offered better coverage in many parts of Europe than its own Gee system (see below). The RAF quickly printed and distributed classified navigation charts under the “Consol” codename. Its operations became dependent on Consol enough that it saw to the resupply of vital parts to a critical station in Lugo, Spain that fell off the air near war’s end with the Nazi retreat. 

Postwar, 11 stations would be deployed across Europe and three stations were built in the US under the ”Consolan” brand which saved land by alternating the continuous and keyed signals between just two antennae. Consol's simplicity and range kept it in service in for 50 years, it’s last station finally going off-air in 1991 at the dawn of GPS. 

Map & Additional Resources: Sonne and Consol proved a bit more difficult to research. There is some good information on the technology online (see this link for one example, “Radio Navigation Systems for Aviation and Maritime Use”, W. Bauss, Pergamon Press, 1963); however, there appears to be precious little on the stations themselves, possibly because of their origins during the former Third Reich. However, some enterprising souls on Wikipedia worked out 14 locations of former stations – this, along with some other research, was combined into a database that drives the map above.

Hyperbolic Navigation also emerged during the war. It used the time difference between an initial signal pulse sent by a “master” station that was responded to by another pulse from two or more or “slave” stations located hundreds of miles apart along a defined “chain,” after a short, fixed time delay, to determine a location. Mathematically, a line that has a constant amount of signal delay between two points forms a hyperbola, hence the origin of the name. The greater the time difference, the tighter it curves towards the closer station. 

For these systems, dozens of such “lines of position” or LOP’s were plotted on aerial and marine navigation charts for each chain, typically in increments of one-millionth of a second or microseconds (μs to not be confused with ms for milliseconds). An indicator would show the time difference between the master and at least two slave stations (or a pair of stations on another chain) and the user could then determine which two LOP'S they were on. The intersection marked their location. LOP’s could also be checked against depth soundings, bearings to other landmarks or beacons, etc. to establish location. 

Interestingly, the math behind hyperbolic navigation originated in World War I to pinpoint German artillery fire. Sound recordings were made at different locations and the observed time differences between each cannon’s reports were analyzed to work out their locations. By 1931 German scientist Meint Harms realized it could serve as the basis of a radio navigation system and filed a patent by 1932 which interestingly, was never apparently utilized by Nazi Germany in the impending war. However, technology still needed to provide suitable oscillators to regulate the system timing, as a well as means for detecting and measuring the nearly infinitesimal time differences.       

British scientist Robert J. Dippy developed Gee (or GEE) as the first hyperbolic system, named after the Grid formed by its LOP’s. It was originally conceived in 1937 as an instrument landing system that notably first used a cathode ray tube (CRT) as its indicator. However, after its German bombing missions suffered devastating daytime losses with poor accuracy, the RAF switched its raids to the cover of night and Gee was rapidly adapted into a suitable radio navigation system for this role by 1942. It operated at 20-90 Mhz and had a range of a few hundred miles and accuracy of a fraction of mile. Gee chains were generally arranged in a “Y” configuration with a central “A” master and “B”, “C” and “D” slaves. The 300kw transmitters, timed by new and precise crystal oscillators, were small enough to fit onto mobile truck units that rapidly deployed on advancing fronts after D-day. Permanent stations used antennae towers in the 240’ range. 

To solve the signal measurement problem, at core of each Gee receiver was a small, circular 4” CRT oscilloscope which traced the received signals left to right, where the horizontal “X” direction was an increment of time controlled by the display's “sweep” speed adjusted by the operator and the vertical “Y” direction reflected the signal’s amplitude. Even though Gee's signals lasted mere microseconds, by repeatedly tracing the same sequence of signals in place on the screen hundreds of times a second in accordance to the set sweep speed, their patterns and any differences would become visible and measurable.

Gee transmitters issued 6μs pulses in the following sequence: in the first cycle, the “A” station would emit a double pulse, the “B” slave would respond, then the “D” station would end the cycle with a double pulse. In the second cycle the “A” station would emit a single pulse, the “C” slave would instead respond and the “D” station would again end in a double pulse. The next two cycles were identical except that the first “A” station pulse was single, the second double. For the "A", "B" and "C" stations, these cycles lasted 4 ms for 250 cycles a second. However, the "D" station actually ran at 6ms or 166.66 cycles a second, causing it to drop from two-thirds of these cycles leaving only the “B” and “C” station to respond. The result was that the receiver could pick a set of cycles that showed only (1) “A” and “B”, (2) “A” and “C”, or (3) “A”, “B” and “D” or (4) “A”, “C” and “D.” This gave the operator great flexibility as to what set of slaves and LOP's they could use: typically, the selection was made on the basis of signal strength and ensuring the LOP’s were as close to right angles as possible for better accuracy. 

In this generalized example, we’ll assume the user wishes to compare “A”, “B” and “D” plus “A”, “C” and “D.” The operator would set the unit to receive the selected cycles then stack the signal trace of the cycles with the “B” slave above the trace of the cycles with the “C” slave. Both traces were bookended by the “A” and “D” double pulses. The second alternating “A” pulse would cause it to rapidly blink across all cycles as the “A Ghost”, making it easy for the operator to identify. Like adjusting the speed of a strobe light to freeze the motion of a spinning wheel, the operator altered the sweep speed so the “A” and “D” pulses of both sweeps were frozen over each other, evenly aligned. A signal generator then superimposed imposed evenly spaced pips 66 μs or one “Gee Unit” apart and the operator could read the time difference. It was possible to obtain a fix, based on the delay times to all three slaves, within 10 seconds.

In 1942, the British introduced Oboe, an early form of Distance Measuring Equipment (DME), that used the time of flight of pulses between a ground station and an aircraft transponder to determine distance. An aircraft flew a wide arc a fixed distance from a “cat” station, which was intersected with another arc at a given distance from a second “mouse” station over specific target. Like Gee, the two ground station operators would compare the difference between stacked ground and transponder signals using 12” CRT’s, some of the largest made up to that time, to enlarge the traces for accurate measure. Dot and dash tones were transmitted to the pilots for guidance. It was accurate to within 200’ or about 6x better than the advanced Norden Bombsight. However, only one aircraft could use a station at a time.

By 1943, a Gee-H capability was added that modified Gee to reverse this problem and make a single ground station a “transponder” that replied back to numerous aircraft. This was done by adding a “broadcaster” to each aircraft, which could transmit 100 pulses per second. Each plane slightly varied this rate to create its own unique “jittering” repetition frequency. Ground Gee receivers were modified to receive and rebroadcast any pulse they received to a wide area. Once the desired distance from the ground station was set, the operator would stack his broadcast signal over the received signals on the scope and set its sweep to match his aircraft’s unique pulse rate. On the scope, that aircraft’s pulses would remain stationary, while those of other aircraft would still track across the screen and be disregarded. The stationary pulses would remain aligned above each other if the correct distance was maintained, but began to drift if the aircraft moved off course. Gee H could service up to 80 aircraft in a coordinated bombing raid, and proved difficult to jam. 

By War’s end over 60,000 Gee receivers would be manufactured. Gee was one of several British innovations of the so-called “Wizard War” where radio waves and electronics had had the same life and death impact as bullets and bombs, as certainly was the case with the Battle of the Beams. These inventions were quickly adopted by the other Allied powers and would profoundly shape the outcome of World War II and much of 20th century life. As will be seen, Gee was the first of several hyperbolic navigation systems that would remain dominant until the arrival of GPS. Gee stations were often collocated with Chain Home stations, the first radar defense system invented by Scotsman Robert Watson-Watt, who supervised Dippy’s laboratory. It used cavity magnetrons, the first stable, compact and powerful source of microwaves that enabled modern radar. Although now obsolete for this purpose, a billion such magnetrons still power the microwave ovens in our kitchens.

After the War, once transient Gee stations were rearranged in a series of seven or so permanent chains that covered Western Europe. It remained a familiar and popular option for the RAF and many of their former aircrews that now piloted civilian airliners. However, Gee’s technology stagnated in the postwar period, and fell woefully behind its competitors by the time its last station went off air in 1970. 

Map & Additional Resources: The UK has a number of sites that document World War II sites of archaeological interest, including former Gee Sites, such as SecretScotland.org.uk, Canmore.org.uk and others. In terms of mapping the system, these groups have done a great job locating these heritage sites in the UK; however, coverage and accurate locations become much spottier in continental Europe, especially for the sometimes-transient wartime stations. Wikipedia maintains the best online listing of station locations, and with a few augmentations, this data was used to create a database that drives the map above. However, outside of the UK this map just marks the nominal town or geographic feature for each station and probably many stations and perhaps a chain or two still need be identified and shown. As always, if anyone has any better information, updates or a database, please reach out to me.

In 1940, Alfred Loomis (who had the unique trifecta of being a wealthy physicist, banker and philanthropist) started a Microwave Committee that became MIT's Radiation Lab (or “Rad Lab”) that started development on a US “Project 3” hyperbolic navigation system that soon caught the US Army Air Force and Navy’s attention. Test sites were set up at two abandoned eastern Coast Guard stations. However, in 1942 the Army Air Force became aware of Britain’s nearly identical Gee system and that it was already set for full scale production. With the Rad Lab’s blessings, it quickly abandoned Project 3 and adopted it.

Instead, with Dippy’s assistance and US Navy’s encouragement, the Rad Lab developed a Long Range Navigation version or LORAN by year's end to extend coverage for the Allies' Atlantic shipping and aircraft convoys by 1943, and for the Pacific region by 1944. It operated at a lower frequency of 1.75 to 1.95 Mhz that used skywave propagation to boost its daytime range to 700-miles which extended up to 1,400 miles at night. This came at the expense of accuracy which dropped to several miles or 1% of the distance from the station, but this was sufficient for long distance navigation. 

LORAN was the first radio navigation system that spanned entire oceans where clouds could deny the use of celestial navigation for days on end and storms made dead reckoning less reliable. A top wartime secret, it immediately proved vital to the Allies’ efforts to move massive quantities of men and materiel around the world. This was especially true in the Pacific where “island hopping” depended on precise point-to-point navigation, and it was far faster and easier to train LORAN operators than more classically trained navigators. By war’s end when the system was finally unveiled to the public, 75,000 receivers were in use. Later, after its successor LORAN-C was rolled out after 1957, the original system was referred to as “Standard LORAN” or more commonly, LORAN-A.

LORAN chains had one master and between 2 to 5 slaves, arranged more in a line instead of Gee’s “Y” configuration. LORAN stations operated on 4 different channels (1.75, 1.85, 1.90 and 1.95 MHz) and each master sent a continuous stream of 40μs pulses at three “basic” Pulse Repetition Intervals (PRI) designated “High”, “Low” or “Slow” (30,000, 40,000 and 50,000μs) to which the slaves responded to after a short, fixed "coding" delay. In addition, 8 “specific” settings lowered the actual PRI in 100 μs steps below this base value. Accordingly, each master / slave pair was given a 3-character ID: the first number was the channel, the second letter the basic PRI and the third number the specific PRI rate: 96 combinations were possible. In addition, separate monitoring stations were set up at strategic points outside the chains to continually oversee the system's performance. If a station’s signal strayed outside minimum parameters, that pair’s transmitters would be set to turn on for 2 seconds, then cycle off for 2 seconds, causing the signal to blink, warning users.

Operators needed a few weeks of training to learn how to manipulate the controls on their greenish CRT scopes. Generally, the first step was to select the correct chain frequency, then set the basic and fine PRI. The set would sync its sweep speed with the PRI, and the blips of other stations on same frequency but of differing PRI's would drift across the screen and be ignored. Similar to Gee, the user stacked the master “A” trace and slave “B” trace above each other as they were constantly repeated. At the slowest setting, the sweep speed was further adjusted until the offset between the signal blips was became visible. LORAN receivers allowed the operator to highlight a portion of each trace with a “pedestal” – the selected part of the trace would be bumped slightly upward, elevating the signal blips on what appeared to a small table. At higher sweep settings, the chosen portions of both signals would be greatly magnified and overlaid to match their peaks to ensure alignment to a common basepoint. Then the operator would select an appropriate sweep speed to superimpose a microsecond “pip” scale to measure the total delay.

These steps were simple in theory but complex in actual practice. Gee relied on the sharp ground waves of its shorter range, higher frequency signals that were “clean” and easy for operators to read. In contrast, LORAN deliberately used skywaves to achieve its longer range. However, they were much noisier and often received as multiple “hops” or reflections bounced back from under the ionosphere’s “E” and “F” layers. Operators needed to interpret their often distorted and muddy waveforms and select the correct signals. But with skill, a position could be worked out in a few minutes. Dippy had the foresight to make both Gee and LORAN receiver units identical in size to allow easy interchangeability between aircraft traversing regions covered by different systems.     

As LORAN was highly classified, the stations were designated only by single letters on navigation charts to minimize their location information in the event these materials were ever compromised. This secrecy was imperative: LORAN stations could not be heavily fortified and were considered expendable. A sustained attack from a curious enemy could likely only be repelled long enough to ensure the destruction of the equipment. It is unclear if the Axis were ever aware of LORAN, though they may have noticed Allied fleets started to unerringly find their way through the worst weather. Station ID’s were also not transmitted – the operator used the channel and PRI to confirm they had the correct pair. For economy, some stations served dual roles, e.g. as a master for one chain but a slave for another. 

The pulse transmissions of their dual and even triple-redundant transmitters, ranging in power from 100 to 1,000 Kw, were regulated by timers with highly stable crystal oscillators accurate to within 1 second per decade (11.4μs per day). They issued precisely timed trigger pulses that set the cadence of the PRI at the master station, or to exactly retard the slave response to a master station after the set coding delay. Timers had built-in receivers that listened for the distant station with an attenuator circuit that cut off its antenna feed any time the local transmitter was active to ensure its more powerful signal didn’t overload it. Four redundant timers were provided at each site, a commentary on their criticality as well as the temperamental nature of the era’s vacuum tube electronics - especially in the humid, salt laden air of the tropics. As the “heartbeat” of the system, timers had several oscilloscopes showing incoming and outgoing signals constantly monitored by station crew around the clock, ready to switch to a standby unit at any sign of an anomaly. They were housed in RF shielded rooms to ensure stray signals didn’t interfere with their operation.

The wartime construction of the stations was, in itself, also a major technical feat. To provide LORAN’s intended coverage over the oceans, most stations had to be located at inhospitably remote corners of the earth, hundreds of miles from civilization, such as at Pacific atolls, the Aleutian Islands, Greenland, etc. Even continental stations were far from urban areas. The Coast Guard dispatched engineering battalions to quickly and secretly erect 72 stations worldwide by war’s end. Each site required the delivery of several hundred tons of construction equipment, Quonset huts, radios, generators and other building materials to mostly unpopulated locales lacking any preexisting transportation facilities. In keeping with their utility and economy, early stations had relatively short 75’ to 135’ antennae array, some supported by wooden poles or masts, with extensive underground radial grounding wires emanating from their bases.

Over time, Quonsets gave way to more permanent facilities, some with full wave height 625’ antennae towers. Nearly all stations were designed as entirely self-sufficient compounds, generating their own power and water, able to sustain their personnel for months on end with only occasional mail / supply drop-offs at a nearby pier or dedicated-airstrip. Their young crew compliments, averaging 15 persons or so, were managed by the US Coast Guard along with their international counterparts. Despite efforts to provide comfortable quarters, good food and other creature comforts, many “Coasties” posted at LORAN stations found their one-year tours of duty incredibly isolating. However, nearly all beam in pride of their service and the memorable life experiences gained. Over its 73-year life, their efforts maintained LORAN and its heir LORAN-C at 99.7% uptime through 2010. 

Map & Additional Resources: The website Loran-History.info posted by Bill Dietz and Joe Jester, is probably the most comprehensive online source for LORAN. Set up by former Coast Guardsmen and women that served the system, it contains detailed history on the technology, personnel and provides a thorough profile of each station. The site offers a LORAN Station GIS database authored by Michael Greene that was uploaded, without alteration, to drive the map display that shows both original LORAN and later LORAN-C stations (covered later) above. As LORAN existed into the late 20th and 21st century, it was well documented in the first GIS databases and indeed the locations in this data appear to be spot-on in aerial images. Another overlay layer compiled from Wikipedia data adds non-US stations not included in the Loran History data set, as well as later Russian Chayka stations LORAN-C coordinated with.

Additionally, The Coast Guard at War IV LORAN Volume II by Historical Section Public Information Division U.S. Coast Guard Headquarters, 1946 has is an excellent history of the wartime deployment of the system, found here. The film “LORAN” for Navigation / US Coast Guard / US Navy Training Film MG-1861, Smithsonian Institution, ca 1946 is another great source, found here.

The Decca Navigator was another hyperbolic system conceived by American engineer William J. O’Brien in 1936, but passed over by the US military as “too complex” and the UK’s team then in the process of developing Gee. However, the Admiralty was intrigued with its accuracy after O’Brien’s friend, the chief engineer of its namesake British record label, presented it to them in 1941. They authorized testing in 1942, and in 1944 on the eve of D-Day, Decca was first used by minesweepers to precisely clear channels in advance of the invasion. It used low frequency waves between 70kHz and 129kHz and had a 400-mile daytime range which reduced to 200 miles due to night effects, reasonable for the era. This, along with its easy-to-use dials that constantly indicated the user’s LOP, made Decca popular for decades.   

Like Gee, Decca chains had 4 stations: a center master plus red, green and purple slaves arranged in a “Y” configuration – incidentally, the same colors used on Gee charts to differentiate the three sets of LOP’s between each master and slave. Decca referred to these corresponding color-coded LOP’s as “lanes.” However, with Decca, master and slaves were phase locked to each other, transmitting their signals in sync and, instead of relying on timing, the system read the phase difference between the master and slave signals. As they passed each other in space they created a hyperbolic pattern of standing waves that formed maxima (antinode) lines ½ wavelength apart where both signals' peaks and valleys summed in perfect phase. These demarcated lanes a few hundred yards wide at the base line between stations.

A navigator had three, easy-to-read clocklike “decometers" for each color that would express any partial lane position in hundredths, e.g. “Red 18.47.” Lane numbers were grouped accordingly to avoid ambiguity: Red was 0 to 23 (24 lanes), Green 30 to 47 (18 lanes) and Purple 50 to 79 (30 lanes). Each set of numbered lanes constituted a zone, ordered alphabetically A through J. For all colors, zones were about 10.6km (6.6 miles) wide at the base line. The master and each slave frequency were multiples of a single frequency (f) of roughly 14.2kHz. The master transmitted at 6f, Red at 8f, Green 9f and Purple 5f. The frequencies were then multiplied in the receiver so phase discriminators could compare both the master and each slave signals at an equal whole integer value, 18f for green, 24f for red and 30f for purple – the phase difference drove the decometers. Thus, a zone was also effectively ½ wavelength of (f), and these integer values divided each color zone into the corresponding number of lanes. 

After setting the receiver to the desired Decca chain, a user first input their starting lane coordinates along each of the three-color lattices based upon their last known location or other means. As the journey advanced, the decometers automatically counted and tracked the current lane coordinates the receiver was in. There was no need to analyze traces on a dim tube: one simply identified the corresponding lane coordinates on the chart to work out their position to within 50m (180’) during the day to around 200m (700’) at night, increasing 4x at the limits of its range. Unlike other systems that had their signals helpfully extended by a more reflective ionosphere at night, the stronger skywaves instead degraded Decca’s critical phasing, and thus its nocturnal accuracy and range. Additionally, it suffered from precipitation static and sometimes the decometers "slipped” lanes. In first generation receivers, the user also needed to know their starting position to at least an accuracy of one lane, problematic for faster aircraft that could move through a specific lane in seconds. 

Decca first began to resolve this “lane ambiguity problem” after 1949 by adding a fourth “Lane Identification Meter” which told the operator the correct lane they were in within a zone. The 1f frequency was extracted as a direct beat frequency between the master and slave signals. The lane numbers for each of the three colors were printed around the circumference of the meter. A six pointed “vernier” would spin per the signal phase, one point of which would align over a second rotating 60° wide sector pointer, indicating the lane. The meter would cycle through the red, green and purple stations automatically in sequence every 60 seconds, in some meters the correct color dial would illuminate to confirm the current signal reading. Thus, a user only needed to know their starting location to a width of one zone, and this tool made it more obvious when lane slip occurred.

In 1962 Decca further improved on this by adopting the “Multi Pulse” system: this added an 8.2f “Orange” frequency that beat again the 8f frequency to provide a 5-zone overlay roughly 53km (33 miles) wide at the base line, further broadening the margin users, especially aviators, needed to know their initial position. Additionally, over a 20 second cycle, each station would, in turn, simultaneously broadcast all 4 frequencies at once for 0.45 seconds allowing more frequent signal comparison.

Decca stations were typically unmanned and transmitted a 640-watt signal to a 325’ mast with four 75’ booms that supported an umbrella type aerial array. Each of these masts typically had 90 subterranean grounding wires radiating outward at a distance equal to its height. Early stations had triple-redundant transmitters, but by the late 1940’s this was replaced with a single unit with several redundant modular components that ran in parallel. A room sized air induction coil was needed to property match the system impedance with this large array.  

Decca’s fame as a record label helped its navigation subsidiary build a profitable monopoly by leasing, not selling, its equipment. With the exception of early Lorenz systems, Decca was unique in being a privately held, commercial for-profit operation. Nevertheless, its ease of use and well perceived management gave it a loyal user base for decades, especially across the British Commonwealth. Postwar, 30,000 ships and 8,000 aircraft were users. Although Decca lacked LORAN’s transoceanic reach, it was widely used for marine and air navigation along coastal waters and inlets. By 1949, it also added the “Flight Log,” the first automatic mechanical, scrolling moving map display for aviation nearly a half century before GPS. It worked well for defined areas: e.g. routine flight plans, defining approaches to smaller fields (useful for executive jets), and helicopters found it useful for navigating cities and between offshore oil platforms. 

Decca would go on to face increasing competition from other systems, most notably the US VOR (to which it lost its bid to become the international radio navigation standard in 1958). Decca also had to compete against LORAN’s successor LORAN-C and the later Omega system, both of which illicitly “borrowed” certain aspects of Decca’s technology. However, Decca’s popularity would well endure through the end of the 20th century, when GPS finally forced its last station off-air in 2001.

Map & Additional Resources: On the Internet Jerry Proc (radio call sign VE3FAB) maintains a detailed set of pages on hyperbolic systems at http://www.jproc.ca/hyperbolic and as of 2023 probably maintains the most comprehensive archive of the Decca system, especially from the system’s “Golden Age” in the 1950’s and 60’s. He also painstakingly researched the station themselves and completed a fairly complete database of former sites, with some information sourced from Decca’s records. This was used to compile the database that drives the map above. The data varies greatly in quality, from very accurate around the British Isles to about 1° where information is very limited. For the LORAN-C stations, please refer to the map in the earlier LORAN section. 

Continued: Early Radio Nav / 1950 - 2000