Air Navigation Systems

Chapter 3. The Beginnings of Directional Radio

Techniques for Air Navigation, 1910—1940
Brian Kendal

This paper is the third chapter of a series on Air Navigation Systems during the period from
the early oceanic flights and the inception of commercial aviation to the introduction of INS
in civil aircraft. Some initial comments on the paper are included in the Forum section of this
issue of the Journal. Further comments would be welcomed by the author and the editor.

1. I N T R O D U C T I O N . When first introduced to the principles of aeronautical

radio navigation, the student is invariably surprised by the wide selection of aids
available. Not unnaturally, the assumption is made that these are the result of
recent development. Few people realize that the basic principles of most modern
navigation systems were developed long before the first radio equipment had been
installed in an aircraft. This paper discusses the development of directional radio
systems.
2.1. EARLY C O M M U N I C A T I O N S . Probably the first radio communication was
achieved by Professor D. E. Hughes in 1879. His experiments were conducted
in his own home in Great Portland Street, but on later occasions he walked up
and down the street with a telephone receiver in his ear, hearing signals up to
£00 yards from the transmitter. For these experiments he used an induction coil
for a transmitter and a microphonic joint with a telephone earpiece for a
receiver.
Over the next few years, many scientists, including such eminent names as
Oliver Lodge and Heinrich Hertz, worked on electromagnetic radiation.
However, as scientists, they were interested only in measuring the physical
properties of this phenomenon. For example, in 1887/8 Hertz first proved that
electromagnetic radiation was polarized and then embarked on a series of
experiments to determine the refractive index of pitch to electromagnetic waves,
measuring this as 1-69 compared with a value between I-J and i-6 for similar
substances when transmitting light. Despite all this experimentation, it was left
to Marconi to realize that electromagnetic radiation provided a practical means
of communication. In 189^ he developed a reliable means of detecting
electromagnetic radiation using a coherer of his own invention and later that year
embarked on a series of demonstrations on Salisbury Plain. For these, he used a
wavelength of about 1 metre, utilizing parabolic reflectors behind both
transmitter and receiver. It is interesting to note that this equipment still
survives and is conserved by the Marconi Company in their private museum.
Within a short time, however, Marconi's interest turned to longer wavelengths

3'3
314 BRIAN KENDAL VOL. 43

due to the relative ease with which they could be generated, and it was not until
twenty years later that he returned to the metric wavelengths.
Also in 1895, a naval officer, Captain H. B. Jackson, working independently,
succeeded in establishing ship-to-shore communication by wireless, but for
military reasons his work was carried out in secret and the results were never
published.

Central mast

Umbrella aerial
Stay
with insulators
at each
end Directional
aerial

— Outer ring
of masts

Transmitting station
Fig. r. The Telefunken Compass (Reproduced from 'Wireless Direction Finding'
by R. Keen, 3rd edition 1938)
NO. 3 DIRECTIONAL RADIO, 19IO-I94.O 3IJ
In succeeding years, Marconi gradually increased the range of his transmissions,
culminating in December 1901 when signals from the transmitting station at
Poldhu in Cornwall were heard in St Johns, Newfoundland.
2.2. THE ROTATING BEACON. In the early part of the century, Marconi was
not alone in the development of wireless transmission and, by 190^, he and
several other experimenters had expended considerable effort in the investigation
of the properties of the classic inverted ' L' aerial for both transmission and
reception, claiming exceptional directive properties when the length of the
horizontal portion considerably exceeded that of the vertical. Developing from
this principle, the following year Marconi patented an aerial system comprising
a number of inverted ' L' aerials evenly spaced around a receiver. By selecting
the aerial receiving the strongest signal, the approximate direction of the
transmitting station could be determined.
Within twelve months (in 1907), the German firm Telefunken had introduced
a similar principle in the transmitting idiom. This consisted of a transmitter
which initially radiated a prearranged ' start' signal to a central omnidirectional
radiator, followed by one-second transmissions on each of thirty-two aerials
spaced radially around the central aerial at each of the points of the compass. A
navigator wishing to use the beacon merely had to start a stop-watch on hearing
the ' start' signal and stop it when the signal reached maximum strength. For the
convenience of observers, a special watch was issued. This had a hand which
made one revolution in thirty-two seconds and the face was calibrated in'the
points of the compass. Beacons of this type were installed at Kleve in western
Germany and Tonder in Schleswig-Holstein during the winter of 1917—18 to aid
the navigation of both Zeppelins and the heavy bombers which were superseding
them. It is hard to believe that the accuracy of such a system was high, but in
191 £ the German engineer Zenneck, writing in his book Wireless Telegraphy,
claimed that an accuracy of four or five degrees was possible. Whether or not this
accuracy was actually achieved, the Telefunken equipment must be considered
the forerunner of all modern rotating beacons, for it had two characteristics
which would soon be seen as essential for air transport navigation systems. Firstly,
as the aircraft did not radiate any signal during the determination of the bearing,
there was no limit to the number of users and, secondly, the quality of the
airborne installation could not affect the accuracy of the bearing, thus eliminating
inaccuracies due to quadrantal or any other error of the aircraft equipment.
2.3. THE COURSE SETTER. In 1907 a patent was granted to Otto Scheller of
the Lorenz Company for a method of delineating a straight course by means of
a single radio installation. This comprised two intersecting horizontal aerials,
each energized such that a null appeared at right angles to the wire. If separately
identifiable signals were then radiated from each aerial, a receiver located on a
line bisecting the angle between the aerials would receive both signals at equal
intensity. Moving to one or other side of that line resulted in one signal
predominating. To provide suitable identification, Scheller suggested that the
morse letter ' A ' should be radiated from one aerial, interlocked with the letter
' N ' from the other. Thus, hearing either ' A ' or ' N ' would indicate an ofF-
course approach, merging into a steady tone as the centre line was approached.
3 i6 BRIAN KENDAL VOL. 43

Direction of zero Direction of zero

radiation from radiation from
'N' aerial 'A' aerial
A

A' radiation

/
Fig. 2. Scheller's Equi-signal Course Setter of 1907 with interlocked A/N
transmission. (Reproduced from 'Wireless Direction Finding' by R. Keen, 3rd
edition, 1938)

Further development was done on this principle during the First World War
in Germany, using wavelengths in the order of joo metres. Unfortunately, these
tests were abandoned due to conflicting results being obtained in the air, possibly
due to polarization errors resulting from the use of long trailing aerials in the
aircraft, although quite accurate indications were received on the ground.
2.4. SMALL L O O P AERIAL SYSTEMS. One of the more famous demonstrations
by Hertz was that a loop of wire containing a spark gap could accept sufficient
energy to cause a spark to pass across the gap when brought adjacent to the
transmitter. If, however, the orientation of the loop were changed, at the same
mean distance, the spark would no longer be produced.
By the early 1900s, the properties of radiation of various aerials were being
investigated and, extending from Hertz's demonstrations, J. H. Round in the
UK carried out a number of experiments with small loop aerials. Demonstrating
first the figure-of-eight polar diagram of the simple loop, he then combined this
with an open aerial to provide a cardioid pattern. These experiments, however,
were eventually abandoned due mainly to the insensitivity of the detector in use.
NO. 3 DIRECTIONAL RADIO, 19IO-I94O 317

2.5. THE BELLINI TOSI SYSTEM. As early as 1899, S. G. Brown had claimed
directive properties for two aerials connected to the spark coils of an induction
coil and spaced half a wavelength apart. Some three years later this principle was
adapted by J. Stone in the United States as a direction finder but, considering the
wavelengths in use at the time and that his proposal was to rotate the aerial about
its mid point, the suggestion was hardly practical.
A scheme to avoid rotating the spaced aerials of Brown and others was
developed by E. Bellini and A. Tosi in 1907, who placed two fixed pairs of open
aerials at right angles from which the direction of the incoming wave was
determined by a goniometer. A goniometer consists of two multi-turn coils
mounted concentrically at right angles to each other, one of which is connected
to each pair of aerials. A third (search) coil, which can be rotated, is mounted
inside the others and connected to the detector. On reception of a signal, the
search coil is rotated to a position where the output from the coils connected to
the aerials cancel; that is, a null is obtained, this indicating the relative bearing
of the incoming signal.
The Bellini—Tosi system was further modified in 1912 when C. E. Prince
converted the four open aerials to two closed loops. However, Bellini later
claimed to have used closed loops before that date.
Sense determination was achieved by one of two methods. The first, developed
by Marconi for their early shipboard installations, inductively coupled the
common connection of the stator coils to the input of the receiver via a variable
resistor, thus additionally using the loop aerial system as a sense aerial, the
resistor serving the double function of adjusting both phase and amplitude of the
induced current. The alternative system used a separate open aerial for
development of the cardioid polar diagram for sense determination. Whichever
method was used, the bearing was invariably taken using the figure-of-eight polar
diagram, the reasons for this being: (a) that the minima for the figure-of-eight
diagram are much sharper; (b) the necessity for accurate balancing of the
cardioid diagram is avoided.
In this period too, the fledgling aviation industry was developing, but the two
sciences did not join forces until 29 August 1910, when the Canadian aviator,
J. A. D. McCurdy transmitted a signal from his Curtiss biplane flying at joo feet,
to an operator on the ground, half a mile distant, at Sheep's Head Bay racetrack
near New York. A few months later the experiment was repeated by a British
aviator, Robert Loraine, using Marconi equipment over Salisbury Plain.
3. A P P L I C A T I O N TO AIRCRAFT N A V I G A T I O N . From the foregoing it will
be seen that, by 191 2, the basic principles of four major navigation systems had
been derived: (1) the small loop aerial which would later be used in airborne
direction finding (D/F) systems; (2) the Scheller Course Setter system using
interlocking signals which was the forerunner of the Lorenz and SBA systems and
ultimately ILS; (3) the Telefunken Rotating Beacon which began a line of
development that culminated in VOR; and (4) the Bellini-Tosi D/F system for MF
and HF which provided navigational assistance until after the Second World War
and its transmitting equivalent, the MF Radio Range, which was the major
method of delineating airways until the mid 1950s.
3 I8 BRIAN KENDAL VOL. 43

By the time, therefore, that aircraft had developed sufficiently for aviation to
be considered a practical means of transport, the radio industry had,
unknowingly, developed the basic principles of all the navigational aids which
would later be used in aviation until the introduction of hyperbolic aids and radar
in the 1940s. Even after that date, they would still play a major role in navigation
and airspace management.
4.1. THE ' W I R E L E S S L I G H T H o u s E S ' . From 1916 onwards, Marconi became
involved with ' Short Wave Directional Wireless' in a series of trials which were
held variously at Hendon and Caernarvon. From these, he developed the
'Wireless Lighthouse' which was installed in Inchkeith Island in 1921. In the
design of this equipment, probably the most important fact realized by Marconi
was that only by raising the frequency to the VHF spectrum could a sufficiently
sharp beam be generated to give worthwhile accuracy. Furthermore, the design
of the keying arrangements was such that it was impossible for the beacon to
radiate inaccurate information and the use of a stop-watch was also rendered
unnecessary.
Operating on a wavelength of 6-2 metres, the Inchkeith beacon comprised two
back-to-back paraboloids of 13 metres aperture, each fed from a separate
transmitter. Around the base of the aerial on a large ring, segments were
arranged in the form of morse code letters which operated a contact breaker to
key the transmitters as the aerial revolved, transmitting a letter at each bearing.
Between these, the station identification letters were radiated. The aerial array
rotated at a speed of one revolution per two minutes, thus enabling the navigator
to take a bearing once per minute. On board the receiving ship, once the signal
had been tuned, the wireless operator reduced the gain of his receiver until only
one or two letters could be heard, these corresponding to the relative bearing
between beacon and ship.
The success of this beacon was such that a further beacon was installed at South
Foreland in 1926. This had a similar radiation pattern but used a broadside aerial
array, 76 feet long and 30 feet high, designed by Franklin. Further work on
similar beacons was carried out by the Air Ministry and the Radio Research
Board, this culminating in beacons working on similar principles (but using
different frequencies and keying characteristics) at Spithead and Orfordness on
the Suffolk coast. It is interesting to note that some ten years later the latter site
was also used for the first British radar development. These were essentially
marine beacons but were obviously usable by aircraft. However, within the
aviation context, all these beacons had the disadvantage that it took at least a
minute to make a bearing determination, during which time the aircraft may well
have travelled two or three miles.
4.2. RADIO RANGE TO VOR. In addition to the 'Course Setter' mentioned
earlier, Scheller also proposed a system in which the interlocking signals were
transmitted from two pairs of vertical aerials, each of which produced a figure-
of-eight polar diagram, thus delineating four tracks to or from the transmitting
station. He also showed that the relative position of the tracks could be altered
by varying the relative power to the aerial pairs. In 1916 he filed a patent which
showed that the tracks could, within limits, be rotated bodily by the use of a
goniometer without the necessity of physically moving the aerials.
NO. 3 DIRECTIONAL RADIO, I9IO-I94O 3 19
In 1924, Scheller's equi-signal principle was taken up in the United States,
where it was applied to a system which transmitted from crossed loop aerials,
producing four symmetrical courses. The frequencies used were in the MF
Beacon band. This was known as the 'Radio Range' beacon.

Fig. 3. The Marconi Rotating Beam Transmitter at South Foreland, 1926. This was the
second to be installed and, unlike the first at Inchkeith Island, used a broadside array
designed by C. S. Franklin. (Photo reproduced from 'Wireless Direction Finding' by
R. Keen, 3rd edition, 1938)

In the late 1920s, further development by US Army engineers resulted in a

system in which a modified goniometer with two rotor coils at right angles
energized the crossed loop aerials, whilst the angle between the stator coils was
variable, the latter being energized by transmitters keyed with the interlocking
morse code characters. By varying the angle between the rotor coils, the relative
amplitude and phase of the signal in each of the loops could be varied, thus
modifying the beacon polar diagram and consequently the bearings of the
equisignal zones, and allowing the four courses to be aligned in almost any desired
direction.
In the period up to and immediately after the Second World War, this became
the standard radio navigational aid in the United States. During the late wartime
period, several were also installed in the United Kingdom, and subsequently the
coverage of the US chains was extended down the Brazilian coast and European
coverage as far south as Africa. These were only removed from service when
replaced by VOR in the 1950s.
However, propagation problems were inherent in the use of medium
frequencies and consequently, in 1936/7, the US Department of Commerce
initiated development work into the use, firstly, of a two course beacon using the
Lorenz principle and then a four course beacon using two pairs of vertical
dipoles, all of which operated on 63 MHz. Initially, problems were experienced
320 BRIAN KENDAL VOL. 43

with these beacons due to polarization difficulties but, when solved, excellent
straight courses were obtained. However, additional problems appeared under
abnormal propagation conditions which led to the operational frequency being
increased to i2j MHz for a further series of trials. At the increased frequency
one of the most serious problems was that of signal strength variations,
particularly when flying over mountainous territory. A similar effect was also
noticed by aircraft on the ground due to reflections from aircraft circling the
airfield. Further tests were made in which the interlocking morse signals were
replaced by two interlocked audio tones which facilitated presentation on an
instrument rather than by ear. Additionally, other tests were made on four
course beacons in which two courses were indicated by interlocked morse signals
and the others by 90 and 150 Hz tones.
Even if these experiments had been completely successful, the result was still
only the delineation of four possible courses to the beacon and, as a consequence,
consideration was then given to an updated version of the rotating beacon
principle.
In 1936, the US Civil Aeronautics Administration proposed a rotating beacon,
operating on a frequency of 6-£ MHz, in which the pattern rotated at 60 Hz. This
was achieved by developing an aerial system whose horizontal polar diagram was
a limacon (approximately heart shaped) which could be rotated by feeding the RF
power differentially to the various elements of the aerial system. This was
achieved by use of a capacitative goniometer rotating at 60 Hz. To provide a
'start' signal, the beam was momentarily interrupted as the maximum passed
through magnetic north.
In the aircraft, the output of the receiver was fed to a cathode ray tube
displaying a circular trace and calibrated on its periphery with a compass scale.
The timebase was synchronized such that the spot moved around the screen in
the opposite sense to that transmitted by the beacon and was at the 360 degree
position as the received signal reached maximum strength. When the transmission
was interrupted as it passed through north, a 'blip' appeared on the CRT trace.
Due to the relationship between the rotation of the trace and the signal strength
received, this corresponded to the relative bearing between beacon and aircraft.
By the late 1930s, VHF was being increasingly used and the decision was soon
made to abandon the use of 6-5 MHz and transfer operations to the 125 MHz
band. About this time also it was realized that the momentary break ' start signal'
was not satisfactory and a further 60 Hz modulation, radiated from an
omnidirectional aerial, was introduced as a reference. This was an FM
modulation on an AM subcarrier similar to that used in VOR today. At the same
time, the Cathode Ray Tube indication in the aircraft was abandoned and
replaced by a phase meter which gave a more positive indication. A second
display system was also introduced which gave the pilot a fly-left/fly-right
indication from a pre-selected course. Development continued through the
wartime period and afterwards, during which the frequency was again changed
to the 112-0 to 117-9 MHz band and the rotational speed and FM modulation
frequencies were reduced from 60 to 30 Hz. Although there are now a number
of different techniques for producing the required radiation pattern, this is the
NO. 3 D I R E C T I O N A L R A D I O , I9I0— 1 9 4 0 321

VOR system (VHF Omni Range) which is used today as the standard short range
navigational aid throughout the world.
4.3. AIRFIELD APPROACH SYSTEMS. While these developments had been
proceeding in the United States, Europe had not been inactive, for, realizing the
problems affecting Scheller's Course Setter when operating on medium
frequencies, by the early 1930s the Lorenz Company had developed a version
using vertical polarization on a frequency of about 33 MHz. This provided an
accurate azimuth guidance beam to assist aircraft on their landing approach path.
Furthermore, vertical guidance was provided by a signal strength meter which
enabled the pilot to fly a signal strength contour of the vertical radiation pattern.
Distance-to-run information was provided by three marker beacons along the
approach path operating on a frequency of 38 MHz.
This proved to be extremely successful and was further developed by the
Standard Company in the United Kingdom. Their product (Standard Beam
Approach or SB A) became the standard R.A.F. approach aid during the war and
remained in service both in the service and in civilian airports until superseded
by the Instrument Landing System (ILS) in the late 1950s.

R, R2

Fig. 4. Aerial system and Polar diagram of the Lorenz azimuth system.
(Reproduced from 'Wireless Direction Finding' by R. Keen, 3rd edition, 1938)

Not content with developing a highly successful landing approach aid, Lorenz
then proceeded, in secret, to increase the power and narrow the beam still
further into the bombing aid which was known as Knickebein. In this, beams
were radiated from two widely-separated stations. The first beam was aligned to
give the bombing aircraft guidance to the target, at which point it was intersected
by the second to indicate that the destination had been reached. This system was
322 BRIAN KENDAL VOL. 43
highly successful and was used during major raids on many British cities until
the beams were discovered and jammed by the Royal Air Force.
Several other bombing systems were developed by the Luftwaffe, all basically
using the Lorenz principle. However, alerted to the use of beam systems, the
British authorities were quick to take jamming action and the systems only
enjoyed limited success.

Fig. j . The Lorenz azimuth guidance beacon at Gatwick. (Photo reproduced from
'Wireless Direction Finding' by R. Keen, 3rd edition, 1938)

It had long been realized that a visual indication of course would be far easier
to interpret than an aural, and many attempts were made to give a meter
indication from signals originally intended for aural interpretation, but few were
successful. It was not until the slow keying intended for aural interpretation was
abandoned and replaced by interlocked audio frequencies that visual display
NO. 3 D I R E C T I O N A L R A D I O , I9IO—1940 323

became practical. This was'developed in the United States during the immediate
pre-war period and introduced the concept of interlocked audio tones on a
common carrier to indicate off-course indication. Filters in the aircraft receiver
enabled comparison of the relative depths of the two modulations and thus the
alignment of the aircraft.
When applied to an approach and landing system, distance-to-run information
was provided by three marker beacons operating on a frequency of 7^ MHz along
the line of approach. Similar marker beacons were also used along airways to
indicate reporting points.
In the earliest equipment, the horizontal guidance was in the 108— 112 MHz
band and the vertical guidance on 92—96 MHz. During the development period,
however, the frequency of the vertical guidance was increased to 330 MHz. In
this equipment, the azimuth pattern was generated by feeding carrier and
sideband (CSB) and sideband only (SBO) signals in varying phase and amplitude
to an array of ten Alford loop aerials. For vertical guidance, a straight line
approach path was achieved by the interaction of the lobing patterns of two
aerials at different heights. The lower aerial was located at a height such that a
single vertical lobe was radiated. This was energized by a signal modulated at
90 Hz. The upper aerial was located at a height at which a multi-lobed vertical
radiation pattern was radiated such that the lowest lobe was below that of the
lower aerial. This aerial radiated a signal modulated at 150 Hz. The relative
heights of the aerials was arranged so that, at the required glide slope angle, the
strength of each modulation was equal, with 90 Hz (from the lower aerial)
predominating if the approaching aircraft was above the glide slope and 1 ^o Hz
(from the upper aerial) predominating if below.
In the aircraft, the signal was received by a simple receiver with the audio
output being passed through 90 and i^o Hz filters. The output from these was
compared and displayed on the horizontal pointer of a cross pointer meter, the
vertical pointer of which indicated the azimuth alignment. With subsequent
development, this (originally designated SCS 51 and now known as Instrument
Landing System) is essentially the system which is now in operation at all the
world's major airports.
4.4. G R O U N D BASED D / F SYSTEMS. Even by the early stages of the First
World War, direction finding using Bellini—Tosi systems on Medium and Low
Frequencies had been developed to a high degree of accuracy. Both belligerent
states had chains of stations along their coasts from which both friendly and
enemy communications were monitored. It is reputed that it was a two degree
shift in the bearing of traffic emanating from the German Navy which confirmed
that the German fleet had left port and caused the Royal Navy to put to sea from
Scapa Flow, culminating in the Battle of Jutland.
In addition to monitoring, the German network was also used for airship
navigation. Von Buttlar-Brandenfels, the only Zeppelin commander who flew
throughout the war, is reported to have stated that, in his experience, radio
navigation was far superior to celestial. This system used Bellini—Tosi installations
located at Borkum, Nordholtz, Sylt and Bruges. The airship requiring a position
fix would call a ground station and keep transmitting while D/F bearings were
324 BRIAN KENDAL VOL. 43
taken from the different sites, after which the position was plotted and
transmitted back.
The British airship service was used mainly for coastal patrols and, in general,
radio navigation was unnecessary. However, the 'fixer service' was available and
was used from time to time in inclement weather. With the advent of post-war
civil aviation, the utility of direction finding stations was realized and such an
installation was incorporated at Pulham in 1920. Further stations were
constructed at Lympne in 1924 and at Croydon Airport in 1928, these three
stations being operated as a chain, with the position of the aircraft being plotted
at Croydon and then relayed to the aircraft. Further chains were installed across
Europe to provide a position fixing service on all major air routes. From 1
January 1934, these chains operated on a frequency of 325 KHz with all aircraft
services being contained with the band 320-36^ KHz.

Fig. 6. Marconi Bellini-Tosi D/F receiver and aerodrome transmitter remote control
equipment at the Plough Lane site, Croydon, before the new buildings were erected
at Purley Way. (Photo courtesy of Marconi.)

In their initial form, most D/F installations used the Bellini—Tosi arrangement
with closed loop aerials. It was soon realized, however, that these systems
suffered severe errors at night, this being due to the horizontal component of the
incoming signal being received by the upper and lower horizontal sections of the
loop. Adcock therefore suggested that the loops should be replaced by pairs of
vertical aerials with the coupling to the equipment screened and buried.
NO. 3 DIRECTIONAL RADIO, I 9 I O— I 940 325
This system showed a considerable improvement with bearing errors, which at
times had exceeded 90 degrees with B—T aerial systems, being reduced by a
factor often or more, with over 90 per cent of bearings exhibiting errors of less
than three degrees. As a result of this experience, by the late 1930s virtually all
ground D/F installations had converted to Adcock aerial systems.

Fig. 7. The Handley Page HP42, Horatius, in front of Croydon control tower in 193J.
Surmounting the tower can be seen the gantries supporting the large Bellini-Tosi loops
for M.F. operation and a smaller pair of loops for H.F. can be seen on the parapet.
(Photo courtesy of Marconi.)

4.5. FIXED L O O P A I R B O R N E D / F . Airborne D/F is used for two purposes:

for homing onto a specific ground transmission or for determination of position
by comparison of the relative bearings of two or more ground stations. In the
former case either a fixed or adjustable loop may be used, but in the latter, a
rotating loop is essential.
In early airborne experiments a small loop within the fuselage was used for
both purposes but, due to the poor sensitivity of the receivers, inefficient ignition
screening and adjacent metal work, signal to noise ratios were poor and bearings
unreliable. Consideration was therefore given to a fixed loop fitted to the upper
and lower mainplanes around the fuselage. By this means, loop areas in the order
of 150 square feet could be achieved with consequent improvement in received
signal to noise ratio. With such installations, the ' null' was directly aligned with
the aircraft track, and thus the system could be readily used for homing. Should
a bearing be required during the en route phase of flight for positional fixing, the
326 BRIAN KENDAL VOL. 43
aircraft was diverted, with the original track being regained after the bearing had
been determined. With such a system, however, the accuracy of the bearing
depended on the definition of the ' null', which was frequently masked by
radiation from the inefficiently screened aircraft ignition system.

Auxiliary coils

Main coils

Fig. 8. Robinson D/F adapted to the wing coil system. (Reproduced from
'Wireless Direction Finding' by R. Keen, 3rd edition, 1938.)

This problem was addressed by J. Robinson of R.A.E. Farnborough who

introduced a second, smaller loop at a right angle to the wing coil and in the plane
of the fuselage. The bearing was taken on the latter coil (known as the main coil)
which, at the moment of taking the bearing, would be receiving a maximum
signal. This, however, was not well-defined. Arrangements were therefore
included to switch the wing coil (known as the auxiliary coil) in series with the
main coil. At this point, if the aircraft was accurately aligned, the auxiliary coil
would be on a null and would contribute no additional signal. If, however, the
aircraft was not accurately aligned, the auxiliary coil would make a contribution
and the received signal strength would vary. By reversing the phase of the
auxiliary coil, the direction of misalignment of the aircraft could be determined.
This system, although effective in minimizing problems due to ignition
interference, was capable of providing ambiguity, for it was possible for the
operator to inadvertently confuse a maximum on the auxiliary coil with that on
the main coil, thus effectively reversing the functions of the two coils and
producing a 90 degree ambiguity. To alleviate this, the Royal Air Force made a
further modification in which the auxiliary coil was switched in alternate phase
with the main coil and then alone, with the main coil being replaced with an
inductance to restore balance to the circuit. The switching was performed
automatically at the rate of one sequence per second. When on course, therefore,
the pilot heard two signals of equal intensity followed by a quiet period when the
wing coil alone was in use.
A further system for homing which gained popularity made use of a transverse
NO. 3 DIRECTIONAL RADIO, I9IO-I94O 327
loop in conjunction with an open aerial. This combination could provide a
cardioid polar diagram with the minimum either to the left or the right of the
aircraft depending on the relative phasing of the aerials. When on course, a
similar strength signal would be received on both aerials. Aerial switching could
be manual or automatic, but in the latter case, the switching was arranged to
provide an audible interlocked A/N indication similar to that given by
developments of Scheller's Course Setter.
Fore and aft
loops

Wing loop

[Link].
generator
for wireless

Fig. 9. The arrangement of the Bellini-Tosi loops on the Dornier Wai seaplane Plus
Ultra. (Reproduced from 'Wireless Direction Finding' by R. Keen, 3rd edition,
1938.)

The loop systems so far described have all been associated with homing
systems. However, installations of fixed loops for direction finding were
attempted in the early days after WWi. In these, large fore-and-aft and
athwartships loops were rigged for use with Bellini—Tosi systems. Probably the
earliest of these was fitted to the Handley Page V-1500 aircraft which was in
preparation for the first non-stop trans-Atlantic flight when forestalled by Alcock
and Brown in the Vickers Vimy. A very successful B—T system was fitted to the
Dornier Wai all-metal seaplane Plus Ultra, which was flown from Spain to the
Argentine. Developments in receiver sensitivity and ignition screening, however,
resulted in improvements in rotating loop systems which caused the development
of large loop systems to be neglected.
4.6. SMALL L O O P DF SYSTEMS. J. H. Round had abandoned development of
small loop systems in 1906 due to poor sensitivity of his detector systems.
However, by the early 1930s, receiver sensitivity had reached the point where
atmospheric noise was the limiting factor rather than any internal reason.
Reception using small loops was therefore a practical proposition. In general, the
328 BRIAN KENDAL VOL. 43

Fig. 10. The retractable D/F loop of an Imperial Airways 'Empire' Flying Boat.
(Photo courtesy of Marconi.)

loops were of the order of one foot in diameter and, in order to minimize
aerodynamic drag, were either fitted in a streamlined teardrop shaped housing or
made retractable when not in use. These were rotated either directly by an
operator or via a Bowden cable from the flight deck, depending on the
requirements of the operator. Within the receiver, the output of the loop was
NO. 3 DIRECTIONAL RADIO, I 9 I O~ I 940

combined with that from an open ' sense' aerial which enabled an unambiguous
cardioid polar diagram to be obtained. Unfortunately, however, due to the shape
of the ground plane formed by the aircraft fuselage, considerable bearing errors
were present which had to be measured on the ground and compensated for when
in operation. Perhaps the most extreme instance of this was in the large German
rigid airships whose length approximated to half a wavelength at the frequency
in use. Without correction, these errors could amount to in excess of 20 degrees,
but this was considerably reduced by the use of adjacent correcting loops similar
to those used in shipboard installations.
In addition to normal direction finding, these systems could be used as homing
aids in a similar manner to that described earlier.
Encouraged by the success of the manual small loop systems, it was perhaps
inevitable that efforts should be made to replace aural interpretation with meter
presentation and, by the late 1930s, to progress to a fully automated system. In
the latter case, the equipment employed a hunting loop of the switched cardioid
type. During direction finding operations the loop was controlled by a motor
which was caused to seek and hold the position of zero loop reception. This
system became known as the Radio Compass.
$. C O N C L U S I O N . In the immediate post WWi period, aircraft movements
were few and there was little, if any, need for positive air traffic control. In the
succeeding two decades, aircraft movements increased dramatically and
additionally, with the capability of flying in other than visual conditions, both a
system of navigational aids and a positive air traffic control system became a
necessity. This was first recognized in North America, where commercial flying
developed more rapidly than in Europe, and where a system of Airways was set
up. At first these were delineated by illuminated visual beacons, but the
deficiencies of these were soon obvious. By the mid-1920s Radio Ranges were
being introduced, and by the beginning of WW2 almost three hundred were in
service. The Radio Ranges were augmented by Non-Directional Beacons, a
facility which comprises a simple transmitter radiating a carrier modulated from
time to time by the facility callsign and which is used in conjunction with the
aircraft radio compass. These continue in use to the present day and are the most
prolific of all radio navigational aids. In the wartime and immediate post-war
period these systems were also introduced into Europe, where they remained the
standard short range navigational aids until superseded by VOR.
VOR, augmented by distance measuring equipment (DME), is now the
worldwide standard short range navigational system with little likelihood of being
replaced in the foreseeable future.

BIBLIOGRAPHY

Keen, R. (1938). Direction Finding. 3rd ed. Wireless World, Iliffe & Sons, London.
Keen, R. (1947). Direction Finding. 4th ed. Wireless World, Iliffe & Sons, London.
Hay Surgenor. (1938). Aircraft Radio, Pitman, London.
Editorial (June 1919). Direction Finding by Wireless. The Times Engineering Supplement.
Anonymous (c 1920). Note on the Position Regarding Direction Finding. Marconi Research Dept.
Anonymous (c 1920). Technical Data Sheet for Marconi Direction Finder, Type 6.
330 BRIAN KENDAL VOL. 43

Anonymous (c 1922). Note on Inchkeith Rotating Beacon. The Marconi Company.

Egan, M. (1923, Dec. IC). A Wireless Lighthouse. Wireless Review.
Anonymous (1921). Note on Revolving Beam Direction Finder. The Marconi Company.
Robinson, J. (1924, May 7). Wireless navigation. Proc. Royal Soc. of Arts.
Ventry, L. and Kolesnik, E. M. (1980). Airship Saga. Blandford.
Powell, C. (August 1986). Radio Navigation in the 1920s. Journal of the IERE.

KEY WORDS

r. History. 2. Air navigation. 3. Radio Navigation.