The Early Days of Radar in the UK

Notes for talks

by

Dr W.H. Penley

This is a fairly comprehensive summary of the early days of

Penley Radar Archives

Copyright:
For educational purposes only, other use may require agreement of author or other provider.
Further information: www.penleyradararchives.org.uk/copyright

Document ref:
www.penleyradararchives.org.uk/documents/penley/early_radar/ 15th October 2002
 THE EARLY DAYS OF RADAR in the U.K.

Notes by Dr W.H. Penley (11 Oct 93 Rev1) for talks to local Organisations about the
significance of the R and D work done in Purbeck for the RAF from May '40 - May '42

INTRODUCTION

"Radar" has become a household word; it conjures up all kinds of ideas - Speed Traps -
Weather forecasting - Air Traffic Control - Collision Warning at Sea - Space Travel -
Defence against Ballistic Missiles - Smart Weapons - and much more. However, it seems to
be little known or appreciated that some of the most far-reaching developments were carried
out in PURBECK. This is a great pity, as what was achieved, particularly at Worth
Matravers, Langton and Swanage, from May 1940 to May 1942, played a critical role in
avoiding our defeat in WWII and then continued to swing the balance of War in our favour
throughout its various stages, until the Allies were ultimately successful. Subsequently it also
led directly to most of the present uses we now take for granted.

I am glad to have this opportunity to tell you a little about this amazing Saga, in which I was
very fortunate to have been involved. I shall, of course, only touch lightly on the vast range
of activities that grew out of the initial events which started it all off and shall try to avoid
technical matters, except in the very broadest terms, even though they were at the very heart
of what was being done.

In these days when you all hear about computers, lasers, smart bombs, and all the rest of the
miracle of electronically operated devices, it is difficult to imagine what things were like in
the 1930s before WWII. It subsequently transpired that the basic ideas for Radar type
equipments had been put forward earlier in several other Countries, and also had been the
subject of a secret Patent taken out by the Admiralty in 1928, and some work in the Army
Signals Experimental Establishment in 1931. However, in 1934, nothing had been done in
this Country to develop the ideas for military use, and, of course, what was being done in
other countries had been kept very secret.

How RDF/RADAR Development started in the UK

To start the Saga then, you must visualise the situation in 1934. In 1922, Mr. A. P. Rowe had
joined the small RAF laboratory at the Imperial College, headed by Dr Wimperis, and had
trained as a meteorologist. A few years later Dr Wimperis became the first Director of
Scientific Research in the Air Ministry with Mr. A. P. Rowe as his junior Personal Assistant,
or in Rowe's terms, "dogs-body". The position of civilian scientists in the other Services was
very subservient to the Naval and Military personnel and though, in the Air Ministry, which
depended entirely on modern technology, these attitudes were not so strong, the civilians
there too had to be very careful not to overstep the mark!

Mr Rowe recalls that there was a continuing attitude of "no war for at least ten years", and a
general view that "the bomber will always get through", which had been enunciated by Mr
Baldwin. Professor Lindeman (Lord Cherwell) was pressing strongly for the development

The Early Days of Radar - R&D for the RAF 1936-42 WH Penley 11 Oct 93 Rev1 Page 1
of balloon barrages and aerial mines to combat the bombers, and work was going ahead to
improve searchlights and listening devices. Rowe says that he visited a site on the Romney
Marshes to see a tremendous concrete acoustic mirror 200 ft long and 25ft high arranged with
sensitive microphones at its focus to pick up sounds from aircraft approaching over the sea,
together with 30ft diameter steerable mirrors to determine direction. His visit was in
preparation for a demonstration to Air Marshall Dowding who was to inspect it. While he
was walking round the site with the Director Dr. Tucker, a milk cart came rattling through.
Tucker stopped the milkman and said, "you won't be here this afternoon will you?" as the
noise completely wrecked the operation of his equipment! These devices eventually
achieved about 15 miles detection ranges but no positional data and so were hopelessly
ineffective, even when working.

During 1934 a large-scale exercise had been held to test the defences of Great Britain. Mock
raids were carried out on London, with the Air Ministry, the Houses of Parliament and
Buckingham Palace being especially targeted. In spite of the fact that the targets and the
routes of the bombers were known in advance, well over half the bombers reached their
targets with little or no opposition. In the opinion of the umpires the Houses of Parliament,
the Air Ministry and other key Establishments were eliminated!!

Rowe, even though the Directorate was excluded from research into radio and armaments,
was so concerned about the situation that he collected together all the files he could find on
the subject of Air Defence. There were 53 of them. What he read appalled him so much that
he wrote a formal minute to Wimperis to say that if we became involved in a major war
within the next ten years we would lose it unless something new could be discovered to
change the situation! He proposed that the best advisors obtainable should review the whole
situation and see whether any new initiatives could be found.

Setting up the Tizard Committee

Wimperis agreed and on 12th November 1934 wrote an historic minute to the Secretary of
State and all the senior people in the Air Ministry, on the lines suggested by Rowe. It was
quickly agreed to set up a committee as Wimperis had recommended, under the chairmanship
of Tizard, (Sir Henry). The other members were AV Hill, PMG Blackett and Wimperis with
Rowe as the Secretary.

The idea of using rays to kill or disable people or machines was very popular at that time, and
many submissions were made by crackpot inventors to say that they had "black boxes" which
would do what ever was required. Their bluff was called by offering £1000 to anyone who
could kill a sheep at 100 yards with such a device, with no questions asked about the inside
of the box. The mortality of sheep remained unaffected!

To start things off, Wimperis consulted Professor AV Hill about the possible effects of radio
transmissions on the human body and the energy levels needed to cause physiological
damage and Hill had estimated the levels that he judged would be necessary. With this
information Wimperis wrote to Mr. Watson-Watt (Sir Robert) who was the Superintendent of
the National Physical Laboratory Radio Research Station at Slough, to ask his views on the

The Early Days of Radar - R&D for the RAF 1936-42 WH Penley 11 Oct 93 Rev1 Page 2
possibility of developing a radio "Death Ray" with enough energy to melt metal or
incapacitate an aircraft pilot.

Watson-Watt passed the query on to one of his senior staff - AF Wilkins, who did the sums
and reported that there was no possibility of achieving significant destructive effects at a
distance with foreseeable radio techniques. However, he indicated that there ought to be
enough energy reflected from an aircraft to enable it to be detected at useful ranges.

This paper was presented to the first meeting of the Committee on 28th January 1935. They
were very interested and instructed Rowe to get quantitative estimates for detection. On 6th
February wrote to Watson-Watt to ask him to provide these quantitative for the possibility of
detecting aircraft as distinct from their destruction. The further calculations Wilkins made
gave favourable results and Watson-Watt wrote his now famous memorandum in which he
proposed "by the extension of known means" a system of radio-location using a pulse/echo
technique. This was presented to the Committee in draft form on 12th February and after its
very favourable reception Wimperis, on 15th February, proposed to Air Marshall Dowding,
who was Air Member for Supply and Research, that £10,000 should be expended to
investigate this new method of detection. Dowding was very interested but asked for a simple
practical demonstration to show feasibility before committing scarce funds to the project.

The Daventry Experiment

For this demonstration, Watson-Watt and Wilkins decided to make use of transmissions from
the powerful BBC short-wave station at Daventry, which broadcast 10kW of radio power on
a wavelength of 49 metres in a beam about 30 degrees wide and inclined up at about 10
degrees. The plan was to fly a Heyford bomber up and down this beam and attempt to
measure the amount of power reflected from it at various ranges. Radio receiving equipment
was quickly installed in a van and this was positioned, late on 25th February, in a field south
of Weedon about seven miles away from the transmitter. The receiving aerial was arranged
so that signals coming from the direction of the transmitter could be reduced to a suitably low
level which would not saturate the receiver, whilst retaining good sensitivity for signals from
the direction of the bomber. The strength of the combined signal after detection by the
receiver was to be shown on a cathode ray tube.

Next morning, on 26th February, Watson-Watt and A. P. Rowe joined Wilkins, and the
Heyford bomber was flown at heights of 6,000ft down to 1,000 ft along the beam. The
results were immediate and conclusive - "beats" between the direct and reflected signals were
detected up to distances of over eight miles. This confirmed the predictions Wilkins had
made. Watson Watt was so impressed by the results that he turned to Rowe and said:
"Britain has become an island again!”

AM Dowding was now convinced that urgent action should be taken to exploit this effect. It
was designated MOST SECRET and he obtained the £10,000 requested for the urgent
investigation of these ideas. Watson-Watt was authorised to form a small team from his NPL
staff at Datchet and chose A. F. Wilkins and E.G. Bowen as his senior assistants. They
decided that it was essential to carry out their detection experiments over the sea and on 1st
March Watson-Watt visited and approved a site at Orford Ness in Suffolk for the work.

The Early Days of Radar - R&D for the RAF 1936-42 WH Penley 11 Oct 93 Rev1 Page 3
Design work started, using the experience the team had built up in the Radio Research Labs,
particularly in investigating radio interference from thunderstorms.

The BBC transmitter had sent out the radio waves continuously, but Watson-Watt’s proposal
was to transmit short bursts so that the distance to a target could be found by measuring he
time taken for each pulse to travel to a target, be reflected, and return to a receiver. A
cathode ray tube was to be used to display both the transmitted and received pulses and show
the time delay between them. Radio waves travel at 186,000 miles a second, so take 10.75
millionths of a second to go and return over each mile of distance to a target.

Move to Orford Ness

Design and construction of equipment was quickly put in hand and the small team of
Wilkins, Bowen and a few others moved to the hastily provided accommodation at Orford
Ness on 13th May 1935. Detection ranges of 17 miles were quickly obtained and were
rapidly increased to 30 and then 40 miles by July. The aerials were mounted on three pairs of
75ft wooden lattice masts, one pair was sited near the building used for transmitter work, and
the other four masts were set at the corners of a rectangle within which the receiver building
was situated. It was decided to refer to this activity as "Radio Direction Finding" or "R.D.F."
This was already a well-known technique, and so, if it should by accident happen to get
referred to, it should not excite public interest. (The term RADAR was introduced during the
War)

In the initial discussions it had been decided that the "Tizard" Committee should be set up
within the Air Ministry rather than as a part of the machinery of the Committee of Imperial
Defence. The Admiralty and War Office were immediately informed, and Wimperis was
instructed to tell their Scientific Directors personally what was proposed. He was to find out
what sort of help they would be able to give, ascertain whether they were studying similar
problems, and see whether they might wish to put forward suggestions for more direct
collaboration. These initial contacts were made, and in July with the clear success of the early
trials, requests for assistance were made to the Admiralty and War Office. In addition, some
recruitment had been authorised, and as a result more staff began to arrive.

The Admiralty Signal School was the only source of silica transmitting valves and undertook
to make them for Watson-Watt. They concluded that it would be better to do their own
equipment development work at HM Signal School rather than join the team at Orford Ness,
because of the very special engineering problems presented by the fitting of this kind of
equipment on board ship. The Board of Admiralty gave approval in September 1935 and a
team was set up under R.F. Yeo at HM Signal School in October.

In July Col. Worlledge of the Royal Engineer and Signals Board visited Watson-Watt, and as
a result of this meeting, Dr Paris was subsequently attached to the team, to report on the
suitability of the new techniques for Army use. However, it was not until October 1936 that
he was officially detached from the Army Air Defence Experimental Establishment (ADEE)
as head of a small Army Cell to concentrate on Army related needs under the direction of
Watson-Watt. The two Scientific Officers in his team, Butement and Pollard, were from the

The Early Days of Radar - R&D for the RAF 1936-42 WH Penley 11 Oct 93 Rev1 Page 4
Signals Experimental Establishment where they had proposed a system for ship detection in
1931!

The experimental work was initially carried out using a wavelength of 50 metres as this, at
the time, was considered likely to give the best reflections from aircraft. Interference with
commercial radio traffic was severe and 25 metres was tried - then 12 metres, new
transmitters receivers and aerials being required for each of these wavelengths.

To enable our defending fighters to be directed to the best positions to engage the enemy it
was essential to get positional information. So, in addition to achieving the best possible
ranges of detection, bearing and height were also needed.

To find the angle of elevation of a target it was decided to attempt to measure the time
difference between signals received by aerials placed at different heights, by determining the
difference in phase. Equipment was set up and an apparently successful demonstration was
given on 21st September. However, when the effects of the reflections of the signals from
the ground were considered carefully, it became clear that a comparison of the signal
strengths received by the aerials, rather than the phase difference, would be a much more
satisfactory method, and this was subsequently used.

To achieve the longer detection ranges needed (over 100 miles) it was calculated that much
greater transmitter powers would be needed together with higher aerials. It was judged that
to avoid saturation problems, it would be necessary to site each receiver several miles away
from a transmitter. The scheme envisaged, therefore, was to have a central transmitter which
“flood-lit” the area in front out to about 100 miles and to place receivers on each side about
15 miles away. The map position of a target was to be calculated by using the time
differences between receiving the reflected and direct signals at each receiver.

Watson-Watt’s proposal for the Chain

On this basis, in September 1935, Watson-Watt submitted proposals for a chain of stations to
be erected round the coast to provide warning of attack and positional information which
would enable defensive fighters to be "scrambled" and be told where to engage the attackers,
and thus avoid the need to maintain "standing patrols". In this historic report he discussed the
effects of height of target, height of aerials, wavelength of the transmissions, power
transmitted and receiver sensitivity on the detection ranges possible. He concluded that using
a wavelength of about 10 metres and aerial heights of under 500ft he could meet the RAF's
requirements. He proposed that a full scale station to demonstrate this should be built at once,
to be followed, if successful, by a group of stations to cover the Thames estuary and then by a
final chain covering the south and east coasts.

The move to Bawdsey

The research and design work now needed to be stepped up. To provide extra space and a
suitable site for the separation of the transmitter and receiver equipments, Bawdsey Manor,
about 10 miles down the coast from Orford Ness was proposed. On 16th September 1935 the
Air Defence Committee of the Committee for Imperial Defence sanctioned its purchase as the

The Early Days of Radar - R&D for the RAF 1936-42 WH Penley 11 Oct 93 Rev1 Page 5
centre for research work and also as the headquarters for the organisation of a chain of
stations.

Approval to construct 5 stations

On 24th October 1935 the Deputy Chief of the Air Staff held the first conference on the
problems of using RDF and on 19th December the Treasury sanctioned a scheme to provide a
chain of five stations covering the approaches to London. Each station was to have a central
transmitter site with receiver sites in each side. Sites were chosen and surveyed by Wilkins,
approved by Watson-Watt and construction put in hand. They were to have 240ft high
wooden towers for the receivers and a similar tower for the transmitter.

A 250 ft guyed lattice mast was erected at Bawdsey and receiver aerials installed in February
1936. Early in March ranges of 75 miles were obtained which Watson-Watt was able to
report to a meeting on 13th March, set up to consider whether to proceed with the giant
acoustic mirror project. This was cancelled! Two further 250ft guyed masts were erected at
Orford Ness to support a six element transmitter array, and performance was still further
improved.

In May 1936 Bawdsey Manor became the centre for the work, and very soon after this a
solution was found to the receiver saturation problem so that transmitters and receivers could
be operated in reasonably close proximity. Watson-Watt then devised a method for
measuring the target bearing by using a pair of crossed dipoles for the receiver aerial.
Wilkins showed that two pairs of crossed dipoles, mounted at different heights, could also be
used for height finding, and so complete stations could now built on one site and the
confusions of attempting to find the map positions of many targets by complex range
calculations avoided. This was very significant as the problem of identifying and calculating
the positions of many targets by range cutting methods would have been impracticable then
and would be extremely difficult even with modern day computing methods.

Sites for the first group of stations had been chosen on the original basis of separated
transmitters and receivers, and work had already started. To save time, it was decided to use
these sites for the new design of stations, even though it might have been possible to find a
better disposition. Industry was now playing a major part in the design and the construction
work for these operational stations, under the overall supervision of the staff at Bawdsey. The
first five stations were to be at Dover, Dunkirk, Canewdon, Bawdsey and High Street in
Suffolk.

Tizard was sure that aircraft positional information would become available much sooner
than was generally thought. He convinced the RAF that theoretical and experimental studies
of fighter control and interception techniques should start immediately. A study group was
therefore set up at Biggin Hill and arrangements made for trials to start as soon as there was a
fully operating RDF station.

Bawdsey was the first station to become operational and plans were made to test the new
interception procedures, which had been worked out theoretically. To avoid any possibility
of complicity, conscious or otherwise, it was decided to intercept a completely unsuspecting

The Early Days of Radar - R&D for the RAF 1936-42 WH Penley 11 Oct 93 Rev1 Page 6
aircraft rather than use another RAF aircraft as the target. Bawdsey had been plotting civil
airliners flying into Croydon and so, on the trial day, a KLM flight was selected and a perfect
interception was made. Our fighters continued on course to counter the impression that they
had come out to meet it. To avoid suspicion no further interceptions of civil aircraft were
made. On the strength of the results being obtained Watson-Watt was able persuade the
Government to earmark ten million pounds for a coastal chain of stations, an unheard of sum
in those days!

Soon after this successful demonstration, the stations at Canewdon and Dover were
completed. This revealed that the sites were sufficiently close to give considerable overlap in
their cover, and because of the relatively low accuracy of the positional data derived, the
reports from adjacent stations on the same aircraft could be mistaken for several different
ones. It was clear that if this information were fed straight to a central situation map
considerable confusion might result. It was therefore decided to introduce a preliminary map
on which all information received (from the Observer Corps and the fighter D/F system as
well as from RDF) was plotted and assessed by experienced observers. They had to produce a
weighted average and as far as possible determine the nature of each "plot" - i.e., friendly,
hostile, number of aircraft, etc.

This intermediate step was called “filtering”. It was carried out in a "filter room" and the
results were passed to the General Situation Map in Fighter Command so that the AOC could
decide the action to be taken. Each Squadron then had to be given precise instructions by a
Controller watching the "plot". Clearly they could not all be sitting round the General
Situation Map at HQ Fighter Command, so the relevant parts had to be relayed back to lower
levels of command. From there the flight orders were transmitted, initially by telephone, but
once airborne, by Radio. As you may imagine, a lot of work was done to determine the most
effective form for these various maps, and the display of other essential information. A large
flat table map marked with "grid squares", on which "plotters" placed and moved blocks to
represent groups of aircraft, was usually in the centre of the Operations Room. Other maps
and displays giving more general information, and specific data on fighter readiness, were on
the walls. Senior officers and the aircraft Controllers could view all this from a balcony.

Fighter Command was set up in 1936 and A.M. Dowding was made the first AOC in C. He
therefore remained directly responsible for the implementation of the Biggin Hill studies and
for the RAF training unit that had been set up at Bawdsey. He took a very direct interest in
these developments and because of his close involvement in the development of RDF this
had a major bearing on our success in the "Battle of Britain" in 1940.

In his memorandum, Watson-Watt had explained the propagational effects that would reduce
the detection ranges possible for targets at lower heights. He had indicated that the use of
shorter wavelengths would improve matters if adequate transmitter power and receiver
sensitivity could be obtained. A wavelength of 26 metres had been chosen for the initial
development as the power available below this wavelength tailed off very quickly. However,
Watson-Watt realised that even though his proposal to use wavelengths between 10 and 15
metres for the main development might satisfy the RAF's initial requirement to give warning
ranges of over 100 miles, there remained many other promising and important applications,
especially Naval and Military, for which shorter wavelengths were essential. Experiments

The Early Days of Radar - R&D for the RAF 1936-42 WH Penley 11 Oct 93 Rev1 Page 7
were therefore carried out at wavelengths down to 50cm and suggestions were made for the
use of beamed aerials rather than the "flood-lighting" provided by the initial chain proposals.
The concept of having a "RDF Lighthouse" with a continuously rotating beam to pick up
targets and our own aircraft was put forward by E.G. Bowen but its realisation was then
beyond technical possibility.

The start of Airborne RDF

Meanwhile, Tizard had become convinced from the Biggin Hill trials, that even though
effective interceptions should be obtainable by day against the mass raids expected if
hostilities commenced with Germany, the relatively crude positional data which could be
obtained from the CH stations would not provide the accuracy necessary to direct our fighters
in poor visibility or at night. He therefore pressed for the development of RDF equipment to
go into fighters so they could find and engage targets when placed within a few miles of
them.

In September 1936 a staff of three, led by E.G. Bowen, who had been investigating shorter
wavelengths, started working on the development of airborne RDF. This investigation was
called RDF2 to distinguish it from the ground based equipment programme that was called
RDF1. Initial trials were done using a wavelength of 6 metres with a large (TV) transmitter
on the ground and a receiver and the large aerial required in the aircraft. (This arrangement
was called RDF 1a). A range of 10 miles was obtained but it was clear that much lower
wavelengths were needed to reduce the size of the aerial and equipment and that, if possible,
the transmitter should be in the aircraft, not at a distance on the ground.

The experimental work was therefore continued on higher frequencies, for airborne use.
Bowen favoured a wavelength of 10cm but the transmitter power and receiver sensitivity
possible were totally inadequate. He therefore built a crude equipment working on a 1.25
metre wavelength with both transmitter and receiver installed in an Anson. This was first
tested in the summer of 1937. Aircraft and ships were detected with this equipment, and in
demonstrations that autumn, HMS Rodney and HMS Courageous were located, in vile
weather conditions, at ranges in excess of 5 miles. Development of equipment for both of
these functions now called AI (Air Interception) and ASV (Air to Surface Vessel)
respectively now went ahead with priority given to AI. A wavelength of 1.5 metres was
chosen as considerably higher transmitter power and reasonable receiver sensitivity could be
obtained.

The start of Army RDF

The small Army team that had been sent to Bawdsey towards the end of 1936 had been put to
work with various members of the resident team on the most pressing problems of the day.
This was part of the learning process, and was intended to encourage a free and easy
exchange of ideas. The policy was quickly to prove most fortunate. The Bawdsey Research
Station staff now had to concentrate on the development of the main Chain Stations with their
associated reporting and command system and also on the development of equipment for
aircraft use, with all the major new problems involved. The Army team therefore

The Early Days of Radar - R&D for the RAF 1936-42 WH Penley 11 Oct 93 Rev1 Page 8
concentrated on equipment to aid anti-aircraft guns, search light control and coastal defence
against ships.

By the end of 1937 Pollard had a prototype anti-aircraft gun ranging equipment working
which, after further development, went into service in 1939. Butement was placed in charge
of a Coastal Defence sub-section and after seeing tests of Bowen's AI equipment on the roof
at Bawdsey decided to use this as the basis of his CD set. He was able to use a lot of the
techniques developed for the AI equipment but was not restrained to the small size of the
aerials allowable on aircraft. Instead of Yagi type aerials (like present day television aerials)
he opted to use large broadside arrays, about 24ft wide and 10ft high which gave much
sharper beams and a higher gain.

The Start of Naval RDF

At the Admiralty Signal School, they decided that the flood-lighting system being used for
the CH stations would give serious blind spots in cover, and that to operate over the full 360
degrees, beamed rotating aerials at mast head level were required. Experiments were carried
out at various wavelengths over 1936/7 and finally a wavelength of about 7 metres was
chosen as this was judged to be the longest for which useful aerials could be made which
could be fitted at the masthead. Two were required for transmitter and receiver and these had
to be rotated in step. Development went ahead and the first installation was in HMS Sheffield
by August 1938 and the second in HMS Rodney by October. Meanwhile, work was also in
hand to develop equipment for locating surface vessels and for improving range finding for
the main guns.

Extension of the Chain

By the end of 1937 the International situation was deteriorating rapidly. Plans to extend the
Chain were hastily made and a final design for the main stations was prepared. It was
intended to operate on four wavebands as a safeguard against enemy jamming or accidental
interference. Four 360ft steel towers were to be used at each station for the transmitting
aerials and four 240ft wooden towers for the receivers. Progress in erecting these stations was
disappointingly slow especially for the steel towers, and as the situation continued to worsen
stations were put into operation with just wooden towers. In some areas, using mobile
stations with portable 70ft towers that had been designed for use overseas by the Army
provided emergency cover.

Though remarkable progress had been made by 1938, it became clear to Tizard and Watson-
Watt that in the event of War there would be a tremendous demand for scientific effort to
cope with the vast range of scientific and technical problems that were arising. Watson-Watt
moved to London to deal more effectively with policy matters and in May Rowe took over
from him at Bawdsey. In addition to the limited amount of recruiting that was underway for
Bawdsey, a list was prepared of suitable people who were at Universities. Tizard approached
Cockroft (Sir John) to see whether some of the eminent researchers in the Cavendish
Laboratory at Cambridge might be persuaded to help. He told him a little about RDF, and
explained that in the event of War "nursemaids" would be needed to make the equipment
work well and keep it going. He particularly invited him to consider what might be done to

The Early Days of Radar - R&D for the RAF 1936-42 WH Penley 11 Oct 93 Rev1 Page 9
produce large transmitter power at very short wavelengths. Little happened as the Admiralty
was handling valve development.

In August 1938 the RAF carried out an exercise to test the air defences. This clearly showed
the need for better low cover, as Watson-Watt had originally predicted, but no specific
requirement was stated until June 1939 when the RAF asked for equipments based on the 1.5
metre coastal defence set being developed by Butement. The Army team was therefore
expanded to provide 24 CHL (Chain Home Low) equipments for the RAF in addition to their
CD programme. The CHL stations were to consist of a pair of broadside aerials 28ft wide by
10ft high on searchlight turntables mounted on pairs of 20ft high wooden gantries over huts
for separate transmitters and receivers. “Binders” used crude bicycle chain drives to turn the
aerials and had to keep them in step to within 2 degrees. After detecting a target they had to
inch the receiver aerial round to point straight at the target so that the direction could be read
off and, with range given on a CRT, enable the map position to be plotted. During this
development Wilkins and Larnder tried to establish the RAF's requirements. It was agreed
that the need was to detect aircraft flying at 500ft at least 50 miles away. The original CD set
achieved about 35 miles but Wilkins calculated that, with improved transmitters becoming
available, 50 miles could be achieved it the aerials were sited 200 ft above sea level. The
RAF wanted the CHLs to be placed at CH stations, but many of these were inland or on low
sites, so mounting the CHL aerials on towers was proposed. (This was my first task after
joining the team!)

Towards the end of 1938, after "Munich", Cockroft visited Bawdsey where he was told what
had already been done on RDF and was shown the experimental and development work
going on. In the spring of 1939 he took a group of the leaders of the Cavendish to Bawdsey,
and the senior "old hands", Wilkins, Bowen, Larnder, Whelpton and Williams gave them a
comprehensive exposition on the situation. Shortly after this they were asked to consider
ways of jamming or confusing RDF systems. It was agreed that Dr W.B. Lewis should join
Bawdsey straight away to take charge of the research. Arrangements were also made for
parties of physicists, totalling 80, to be introduced to RDF by spending a month on Chain
stations, starting on 1st September.

Evacuation of Bawdsey

Concern that the Germans might be aware of the work going on at Bawdsey was heightened
by a visit from the Graf Zeppelin. This, for almost a whole day in August, hovered 20 miles
out to sea and also flew inland over the coast. It was (correctly) thought to be trying to
monitor the RDF transmissions, and it was assumed that if the Germans knew Bawdsey's
function it would make it one of their earliest bombing targets. In the few days before the
declaration of war, therefore, most of the BRS staff was evacuated to the Teacher's Training
College in Dundee. Bowen had been given the task of providing a squadron of AI
equipments before the end of 1939 and the first MkI was delivered to the RAF in August. To
carry out the installation, Bowen's airborne equipment team went to Scone near Perth - a
grass airfield without runways - and then to St. Athan, a maintenance aerodrome in South
Wales where they were bombed by a JU88! It was very fortunate for Lovell (Sir Bernard) and
Hodgkin (Sir Alan) that the bomb didn't go off! The equipment was flight tested in
December.

The Early Days of Radar - R&D for the RAF 1936-42 WH Penley 11 Oct 93 Rev1 Page 10
The Army team was moved to Christchurch to a site on the cliffs at Steamer Point, very
suitable for the work on CD equipment.

On 14th November, six weeks after war was declared, U-boat U-47 sank HMS Royal Oak in
Scapa Flow. Admiral Sir James Somerville, who had seen a demonstration of Butement's CD
set tracking a surfaced submarine at Bawdsey, asked for three of these equipments to be
made immediately to help protect that vital area. Cockroft was asked to undertake this task.
He drove a lorry to Bawdsey and collected equipment which had been left in the stores there,
took it to Cambridge, and mobilised a team from the Cavendish to do the design and
construction. These sets were called CDU's (U for U-boat). After a short time Cockroft and
his team moved to Christchurch to continue the work. The first station was operating on site
early in December and the other two before the end of February. Performance on submarines
was up to 25 miles, and on aircraft up to 70 miles. They detected low flying aircraft at much
greater ranges that the CH stations.

Early in this period the first magnetic mines had appeared off the East Coast and were
causing many wrecks. Cockroft was asked to provide two stations for the Thames Estuary.
The first of the CHLs was erected at Foreness Point and was operating by 1st December.
There Cockroft saw W/Cdr Pretty (Sir Walter) control the first interception of a low flying
mine laying aircraft. As a result it was agreed at a meeting at Aerial House on 19th
December, that the use of single stations for this task was better than tracking target and
fighter with separate equipments. This had a profound influence on the development of
Ground Controlled Interception techniques (GCI). Air Ministry now required many more
CHL stations and instructed the BRS team at Dundee now called the Air Ministry Research
Establishment (AMRE), to take over the task from the Army team.

For the AMRE team Dundee was most unsuitable, the work had been disrupted, and the
distance from London prevented the close contacts with the RAF which had been so
important. However, three of us completed the installation of an experimental CHL on the
200ft platform of the 360ft steel tower of the CH station at Douglas Wood and showed that it
gave good detection of low flying aircraft as had been predicted. Another team under RA
Smith was able to design and test changes to the CH transmitter aerials and achieved a
valuable increase in the range of detection possible. We were all greatly relieved when we
heard we were to move to a new site which was being constructed for us near the CH Station
being set up at Worth Matravers. Most of those in Dundee packed up and moved to Worth
on 8th May, and were joined by many of Bowen's aircraft equipment team. Quite a
significant number from both teams moved to HQ, or joined the RAF, to help deal with all
the new technical tasks and problems presented.

The move to Worth Matravers

At Worth Matravers, towers and huts, some with earth barriers for protection against air
attack, were being erected in a field next to Renscombe Farm and the rudimentary CH station
that was already in place. Huts for offices, workshops, drawing office and experimental labs
were being provided - a vast improvement on Dundee and St Athan. Arrangements were

The Early Days of Radar - R&D for the RAF 1936-42 WH Penley 11 Oct 93 Rev1 Page 11
made for us to use the airfield at Christchurch for equipment installation and experimental
flying until Hurn was ready.

We were just coming to the end of the "Phoney War" and daylight raids were building up.
Construction of the RDF Chain, both the main CH stations and the new CHLs based on
Cockroft's design, was going ahead apace, but there were many technical and operating
snags. Our prime task was to improve performance and make it more consistent. In
particular the CHL stations were not performing as hoped. The task of keeping two aerials
pointing in the same direction within a couple of degrees seemed too difficult and the
requirement for two turntables at each station was slowing the rate of construction.

Many different aerial designs were tried in the attempt to operate the transmitting and
receiving aerials on the same turntable but, though giving some improvement, none gave the
performance we were seeking. There was no way of checking the performance of a station
other than getting aircraft to make "calibration" flights. With the increasing number of
stations, the small number of aircraft that could be made available and the risks of enemy
action over the sea, this was clearly very unsatisfactory. A lot of effort was put into devising
test equipment to enable effective setting up to be achieved without aircraft. We found that
the performance of the CHL erected in the main CH compound near Renscombe Farm was no
better than the CH for low flying targets, so we had an experimental set erected at St
Aldhems Head. This gave very good performance and, as a result, a RAF operational CHL
station was quickly installed there.

At Dundee, Dummer and Franklin had been constructing a "radial time base" CRT display to
show the map position of targets. Lewis had visualised this being used with an RDF set with
a sharp beam rotating like a light house. He had given the task of developing a suitable RDF
equipment to a team of senior research scientists who had just joined, and they were
struggling to do the job on a wavelength of 50cm. Soon after we got to Worth, Dummer was
ready to try out his display but the 50cm equipment was not yet giving adequate
performance. I was therefore asked to set up a 1.5m CHL type equipment for the trials and a
very successful demonstration was given to AM Sir Phillip Joubert, who was very impressed.
The Plan Position Indicator (PPI) was born! Much development of our basic RDF
equipments to make use of this new display technique was clearly needed, but France had
been over-run, and the evacuation of our Forces from Dunkirk shocked us all. We had to
concentrate immediately on whatever was needed for the immediate defence against
invasion. Daytime attacks were building up and it was essential to ensure that the best use
could be made of our relatively small force of fighters. We stopped our research work at
Worth and went out to the CHL stations being set up to do what we could to ensure that they
gave the best possible performance. We converted many of them for single turntable
operation, put in various technical improvements, checked the performance, found that
several were in poor positions and for these chose better sites which would give good low
level detection.

During this time the daylight blitz was at its peak. The CH chain with the CHL additions
and the complex but very effective control and reporting system that had evolved enabled the
Hurricane and Spitfire Squadrons to stay on the ground until the raids were approaching.
They were then directed to the most favourable positions to attack - often from a higher

The Early Days of Radar - R&D for the RAF 1936-42 WH Penley 11 Oct 93 Rev1 Page 12
altitude and "out of the sun" which gave us a tremendous advantage. Losses of about 10%
were inflicted on the enemy, who could not sustain this, and so, as Tizard had predicted,
turned their main effort to night attack. It was quickly found that the Chain system was
unable to direct our fighters against the scattered targets with the accuracy needed for the
crude AI to pick them up. The loss inflicted on the enemy was about 1% - mainly by AA
fire!

Our operational research group had been using the "inland" CHL equipment on the main CH
site at Worth to develop more accurate methods of directing aircraft. One project they put a
lot of effort into the task of dispensing the Long Aerial Mines in the path of incoming raids.
Professor Lindeman was very keen on this concept. It was visualised that if an aircraft flew
into a long length of steel wire trailing below a small parachute, the drag of the parachute
would pull the wire up across the wing, and bring an explosive charge fitted to the wire into
contact with it. This "LAM" concept, whether dispensed from aircraft or by rockets, did not
give attractive results.

Their other main project was to find a way of using the high accuracy that it should be
possible to obtain from a CHL type beam station to direct a fighter to the best position from
which to engage an incoming bomber. Many schemes were tried using the split-beam
plotting techniques with various mechanical navigation computing instruments. These
certainly did better than could be achieved with the main CH station but were not good
enough for operational use.

In August 1940, just after we had returned from our trip round the CHL stations, we heard
that special equipment for Ground Controlled Interception (GCI) suitable for inland use, and
capable of being moved easily between sites, was to be evolved. The design was to be based
on the Worth trials and what had been achieved at Foreness and take into account a
suggestion that using low aerials in a saucer shaped site with higher ground about half a mile
away might greatly reduce the level of ground returns which severely degraded overland
tracking from standard CHL stations with gantry mounted aerials. It had been decided to use
CHL aerials mounted on cabins designed by the Army team for their Gun Laying sets and the
Army team at Christchurch would be making the aerial trailers. The two aerials would still
have to be swung to and fro and kept in step, but a PPI would be fitted as well as a standard
"A" scope.

The first equipment was installed at Durrington, near the Poling CH station, on 18th October
1940 but the urgency was such that the RAF had decided that six more were needed before
Christmas, and work on these had been put in hand before any testing had been carried out.

Duckworth and I went to Durrington to see how the equipment was performing. We found
that the plotting rate was too slow for accurate control and that height information from the
CH at Poling was ineffective. There were also gaps in cover - targets being lost at critical
times. Nevertheless, it provided better control than the Chain. Gradual improvements were
made, step by step, by changes to the equipment. The receiver aerial was divided into upper
and lower sections at about 7.5ft and 12.5ft and the "split" switch used to switch between
them so that the different vertical polar diagrams could be used for elevation determination.
The transmitter aerial was sectioned in the same way and arrangements made to feed these

The Early Days of Radar - R&D for the RAF 1936-42 WH Penley 11 Oct 93 Rev1 Page 13
sections either in phase or in antiphase to alter the vertical pattern. These two changes
effectively overcame the serious gaps in cover and enabled some rudimentary height finding
to be achieved, but plotting rate and height accuracy were still inadequate.

Meanwhile at Worth our team was struggling to overcome the many shortcomings of the
CHL technology. We wanted to use the same aerial for transmitter and receiver, and to do
away with the twisting feeders and cables so that aerials could be rotated continuously. We
also needed an effective way of determining target height. The 50cm "Lighthouse" set was
still a long way off so Lewis asked me to design a highly mobile equipment on 1.5 metres
which would include all our new technology. This was first called the Ground Rotating
Beam or "GRB", and later "The Field Gun". While this was being developed the team got
satisfactory design answers for common aerial working and rotating joints for both
transmitter and receiver. We also showed that height finding could be satisfactorily achieved
by switching between aerials with an electronic switch that worked at the repetition rate.

I was asked to design 50cm aerials to be mounted on the rear side of the CHL aerials at
Dover, for an experimental station. The idea was that, as we now knew the Germans had
RDF equipment, we might avoid jamming if we used the same waveband.

Goodier and Wiblin were using early ASV equipment to make a very small "Pack Set",
which could be easily transported by air. This was for use in the protection of overseas
airfields.

In October, because of the night blitz, the RAF was pressing us to make major improvements
on the Durrington type equipments. It was agreed that we would have a free hand to make
whatever changes we wished to the station to be installed at Sopley. Work on the "GRB"
was slowed so that the technological improvements we were making could also be applied
quickly at Sopley as well as to the CHLs. Like Durrington, Sopley had a two aerial system
and John Duckworth was given the job of incorporating our new technology as soon as
possible. Installation of this station started on Boxing Day with some of the improvements,
and work continued, to produce the continuously rotating, single common T & R aerial, with
height finding and gap filling, which was our aim.

Critical technical innovations were made such as the rotating coil magnetic PPI (Tutchings),
a means of height finding without stopping the rotation of the aerial (Tutchings and
Hopkinson), and the spark-gap method of using the same aerial for transmission and
reception (Banwell and Lees). The first enemy aircraft shot down from Sopley was on 5th
March and when HM the King visited the station on 7th May, he saw a successful
engagement and also witnessed the enemy aircraft dive to the ground in flames. During the
final stagers of the interception the Controller Sqn/Ldr Brown told everyone, including the
King to “SHUT UP!”

Another operational GCI was set up at Sturminster Marshall and further development work
was carried out there as well as at Sopley.

Most of the successful changes (except common T & R ?) were introduced quickly into the
other GCIs, with the result that our Beaufighters could now be directed consistently to where

The Early Days of Radar - R&D for the RAF 1936-42 WH Penley 11 Oct 93 Rev1 Page 14
they could pick up enemy targets with AI MkIV - an improved version of the 1.5 metre
equipment. Our success rate increased rapidly, and reached about 10% during March/April.
This, for the time being, effectively countered the night blitz. However the handling capacity
of our defences was inadequate to cope with massed night attacks, which we knew could
overwhelm them. The AI, because of massive signals reflected from the ground, could not
pick up targets much beyond two miles. The use of Electro-mechanical switches for gap
filling meant that each GCI could only do one interception at a time. Performance also
rapidly tailed off as the attacking aircraft came in lower - a very worrying situation.

For the "GRB", my team and I had tried many mechanical and electronic methods for
switching the transmitter between aerials at high speed without success. When Dr Westcott
showed us that, theoretically, a variable capacitance would switch a feeder network. I
immediately designed an experimental capacity switch, to handle the high voltages involved.
It rotated at 50 revs a second and sounded like an air-raid siren! However it worked very
well, and so the BTH Company were contracted to design and produce one to my basic
design. Transmitter and receiver could be switched between aerials a thousand times a
second.This later enabled us to evolve a very versatile GCI for installation on permanent
sites. This used three aerial heights to give better low cover and could handle many
interceptions simultaneously. The aerial rotated continuously at 6 rpm and the operations
room was nicknamed the “Happidrome”. Installation of these was just about starting as we
left Swanage in May 1942.

Fan Beam Height Finding

The problem of getting accurate height information on a target was a long standing one and
the idea of using a fan beam which could be wagged up and down in elevation, had seemed
attractive. To do this Banwell and Bacon devised a system of putting aerial elements all the
way up a 200ft tower to give a beam width of about 1-degree. This could be wagged up and
down with a phase shifter – the groups of dipoles were also tilted in maintain a good beam
shape. As soon as the spark-gap common aerial system was devised several of these Variable
Elevation Beam (VEB) equipments were constructed and put onto CH sites.

Towards the end of May 1941 the work on 50cm was showing more promise. I was asked to
design aerials for mounting on CHL 20ft gantries. Lewis suggested that height finding might
be possible by using a 10cm equipment! This was my first introduction to this possibility.
Duckworth and Bacon moved to be near the centimetric teams, and we initially used the new
techniques to do model experiments on aerial designs for 50cm equipments. In October,
when the "GRB" was cancelled, I was given full responsibility for all the 50cm
developments. In addition to the "Radio Lighthouse" now called AMES Type 11, I started
work on a 50cm VEB. From model experiments on 10cm we decided on a wire-mesh
reflector dish about 30ft high by 10ft wide. This gave a 2 degree fan beam which could be
moved up an down slowly and scanned over + and - 5 degrees quickly by moving the dipole
feed up and down to give a good "paint" of the target return on the display. By then
Duckworth and Bacon had found that, instead of using complete paraboloidal reflectors, a
section of a parabolic cylinder covered with flat plates to form a "cheese" shape would make
a satisfactory aerial for operation on 10cm. They used "cheeses" about 15ft by 2ft for both a

The Early Days of Radar - R&D for the RAF 1936-42 WH Penley 11 Oct 93 Rev1 Page 15
height finding "VEB" like equipment and plan position equipment. These were called AMES
types 13 and 14.

All these basic equipments were greatly developed and played a very important part in the
invasion. They were brought together in various combinations to make best use of their
differing capabilities. On 50cm we were able to develop a system which would eliminate
interfering ground returns and so provide very effective over-land operation.

Centimetric Equipment & the Magnetron

When Bowen had started working on airborne RDF he had said that a wavelength of about
10cm or less was really needed, but there was very little power available and receiver
sensitivity was poor. However, Randall and Boot working under Oliphant at Birmingham
University had been asked to see whether they could generate much higher powers for us at
these low wavelengths. They achieved a phenomenal "break-through" by devising the cavity
magnetron. On 21st February 1940 the first experimental model was switched on and gave
around a kilowatt of power - several orders of magnitude up on what had previously been
obtained. The GEC were given the task of designing a sealed version and the first of these
arrived at TRE in July.

Since about 200 of us had arrived at Worth in May, there had been a large influx of top-level
research scientists to TRE. Even though he was under extreme pressure to put all his effort
onto immediate tasks, Rowe agreed with Lewis that a team should be given the job of
developing and exploiting the new possibilities. They decided to say as little about it as
possible for the time being, so it was kept very much "under wraps". Lovell says that he got
the first echoes from an aircraft with a crude set-up which used separate 3ft diameter
paraboloidal aerials for transmitter and receiver in August, and this was quickly increased to
about 6 miles. It was clear that a lot of work would be have to be done to devise the
components needed for a working system, and the Admiralty Signal School people, who, of
course knew about the valve work, were kept fully in the picture. This was very fortunate as
they immediately decided to develop operational equipment for use at sea and the basic parts
of this were subsequently used for ground RDF too.

Centimetric AI

For the TRE team the task was to convert the very large, crude, experimental set-up, kept
working by scientists, into small reliable airborne equipment which Service personnel could
operate. It was judged essential to use the same paraboloidal aerial for transmitter and
receiver and that this should be installed in the aircraft nose in a plastic dome. A dish of
about 28" diameter gave a beam width of 12 degrees. A fighter needed to be able to detect
targets over most of the forward hemisphere, so it was clear that this beam would have to be
scanned by moving the dish. Experiments were done with both spiral and helical methods,
and the former, evolved by Alan Hodgkin, was chosen for development, as it seemed simpler
to manufacture. An experimental equipment was installed in a Blenheim at Christchurch in
April 1941. Performance predictions were confirmed, 4 miles detection range being obtained
down to very low altitudes. This was called AI Mk VII and a production run of 12 was
ordered immediately, and then another 100. The initial sets began to arrive in the autumn and

The Early Days of Radar - R&D for the RAF 1936-42 WH Penley 11 Oct 93 Rev1 Page 16
the main production early in 1942. These sets, with the improved GCI, accounted for over
100 enemy shot down in the very reduced attacks made during the early part of 1942. An
updated design - AI Mk VIII - was brought out early in 1942. A further design (AI Mk IX)
with scanning and "lock-follow" functions to allow "blind firing" attacks to be made was
started and completed after the war, while the USA developed a helical scan equipment (AI
Mk 10) which was very successful in the later stages.

ASV Development

After the autumn demonstration in 1937 when HMS Rodney and HMS Courageous were
located, the development of ASV went ahead in parallel with AI, most of the techniques,
except for the aerials and displays, being very similar. Three ideas were worked on - forward
looking, sideways looking and all-round looking. These required very different aerials and
displays. The biggest aerials could be mounted along the length of the fuselage and gave a
sideways looking beam. In a trial of this equipment between 10th and 12th May 1938, when
units of the Home Fleet passed from Spithead to Portland, photographic records were taken
of the display as the sideways looking beam swept past the Fleet. The aircraft returned and
took off several times and the films were shown to those in command. Detection ranges of
30 miles were achieved on capital ships, and up to 15 miles on smaller ones of about 5,000
tons. Also coastlines were clearly recorded.

However, Coastal Command decided to have the forward looking equipment first as they
hoped that this would allow aircraft to home onto ships for direct attack. Intensive work on
ASV only started after the delivery of the first AI equipments in August 1939 as these had
priority. Trials of an experimental version took place at Gosport in December 1939 against
surfaced submarines. Signals returned form the sea were very strong and when flying at
5,000ft detection was only possible between 4.5 and 5.5 miles. When flying at 200ft
detection was between 0.5 and 3.5 miles. Ranges on big ships were up to 30 miles.

Deliveries of 200 for installation in Sunderland Flying Boats and Hudsons started in January
1940. They were used until the end of the year to search for enemy shipping and for anti-
submarine patrols. They did not help much as the enemy shipping at this time was hugging
the Norwegian and Dutch coast lines and great difficulty was experienced in separating the
echoes from ships and those from the many islands. Its main use was for rendezvousing with
convoys (which were usually out of position) in poor visibility, and to give warning of
coastlines to avoid the cliffs of Norway and the mountains of Scotland. A responder beacon
had been installed at Leuchars in April 1940 to enable aircraft to find their way back in bad
conditions. Many more were subsequently provided for operational use to help aircraft find
their way back to aerodromes in bad conditions, for finding convoys and for air/sea rescue.
These aids to navigation were highly regarded and so kept interest alive.

The Royal Aircraft Establishment (RAE) was given the task of engineering the main units for
full production and to give higher performance and good reliability. They began to come off
production towards the end of 1940, and were called "ASV MkII". These units were
subsequently made in many parts of the world, including the USA, and were used in many
different equipments. In addition to aircraft installations these included air-transportable pack

The Early Days of Radar - R&D for the RAF 1936-42 WH Penley 11 Oct 93 Rev1 Page 17
sets, search light control equipments, etc., and eventually, in gliders used in the invasion, to
provide immediate warning of air attack after they had landed.

For ASV longer ranges were wanted, particularly on submarines, so the sideways looking
system was reconsidered. The initial production of the MkII boxes was used for forward-
looking sets but sideways looking LRASV sets were quickly introduced in Whitleys,
Wellingtons, Sunderlands and Catelinas, and by April 1941 Coastal Command was operating
with 50 sideways looking and 60 simple forward looking sets. From 2000ft the LRASV
detected surfaced submarines beyond 3 miles out to 8 miles; 60 miles on coastlines; 40 miles
on larger ships; and 20 miles on destroyers. The forward looking sets were not much better
than the early sets, and better aerials were fitted which brought the range to about 80% of the
LRASV. To fill the central gap, combined installations were introduced, with a means of
switching between aerials that greatly improved the search and attack capability. These
equipments were used in patrol and reconnaissance and particularly on the colossal task of
protecting the shipping routes to the UK. One detected the escaping Scharnhorst and
Gneisenau on 12th February 1942.

Their greatest contribution was in the early phases of the battle of the Atlantic. The Germans
were using the French Atlantic ports they had captured in 1940 to provide bases for their U-
boats. Coastal Command's major task was to intercept and destroy these U-boats as they
passed to and from their bases. Initially ASV greatly increased detection and attack rates by
day, but the Germans quickly began surfacing at night to charge their batteries. The ASVs
could detect them at night, but the aircrews were unable to make visual contact unless the
boat could be silhouetted in the moonlight. A visual was essential because of the large
number of French fishing boats. Experiments were quickly instigated to provide
illumination. Pyrotechnics were unsatisfactory because of release height and position and the
difficult aircraft manoeuvre required to bring the suspect between the flare and the aircraft.
Most promising was a 20" diameter naval searchlight - the Leigh Light - mounted on a
retractable turret and remotely controlled in azimuth and elevation. Trials in the Summer and
Autumn of 1941 showed that an ASV homing run on a target followed by exposure of the
light at a range of about one mile should be successful. Extensive training of the whole
aircrew was essential to form a skilled team.

Use of this system started in the Bay in June 1942. It was very successful and U-boats
reverted to daytime surfacing. U-boat kills rose rapidly but fell in the autumn, when the
Germans introduced effective warning receivers to tell them of the approach of an ASV
aircraft. Many schemes for overcoming this setback were considered such as floodlighting
the whole area from highflying aircraft with powerful transmitters on the ASV band. But,
clearly, a much more efficient ASV was needed. Fortunately, an attempt was already being
made to use the new centimetric waveband, and trials of an experimental equipment had
started in March 1942. This used a rotating beam scanner and a PPI and had given four times
the detection range on submarines.

It was decided that the first installation should be in Sunderland Flying Boats, but this
presented considerable problems as two synchronised scanners mounted under the wings with
complex switching were required to provide all round cover. Because of the urgency, Dee
and Lovell suggested that a modified version of the ground mapping equipment they were

The Early Days of Radar - R&D for the RAF 1936-42 WH Penley 11 Oct 93 Rev1 Page 18
developing (H2S) should be used. They produced a modified design very quickly and Lovell
and his team personally modified six equipments and took them down to Chivenor in North
Devon. The reluctant Station Commander was ordered to fit and fly modified Wellingtons
over the Bay against the U-boats. This operation started in March l943, and by the end of
April the use of these half dozen sets had resulted in the sinking of well over 20 U-boats.
Such was the importance of this programme that Sir Robert Renwick said he would
immediately report any trouble we had to the Prime Minister! The Germans did not guess the
frequency change for a long time, even though an H2S fitted bomber had fallen into their
hands in March. The U-boat Command became aware in September - and then quickly
introduced warning receivers on the new band. The initial design was ineffective and was
soon replaced with another type, yet the crews were convinced that something else was
causing their large losses. Their morale became very low and never really recovered.

Our shipping losses had risen from about 40,000 tons in November 1941 to a catastrophic
average loss of nearly 500,000 tons per month for the whole of 1942. The initial use of ASV
gave a dip to 200,000 tons in January 1943, but losses rose again to 300,000. When the
centimetric equipment became operational there was a rapid drop to less than 20,000 tons a
month from June 1943 onwards. The Battle of the Atlantic had been won and the way for the
USA to supply troops and equipment for the invasion was opened!

Equipment operating on much shorter wavelengths and countermeasures to confuse any new
listening receivers produced by the enemy were developed to counter any resurgence of the
threat. These maintained our ascendancy.

Sunday Soviets

Soon after we arrived in the Swanage area Mr. Rowe, who knew all the senior people in the
RAF and The Air Ministry as a result of his period in London, invited senior individuals to
visit Swanage at weekends to see and discuss what we were doing. The attraction of a day or
two by the sea in the Grosvenor Hotel may have helped, but this quickly evolved into what
Rowe called his Sunday Soviets. At these, C's on C with their senior staff would come and
discuss their problems with senior members of the Establishment, and would accept the
complete freedom of comment and remark that marked the discussions. Junior staff working
on projects would often be called in to discuss technical and operational details. The mutual
confidence and close co-operation built up by these free and completely informal exchanges
played a major part in harnessing scientific and technical capability to the war objectives and
in bringing scientific objectivity to the study of operational matters.

GEE

Great importance was attached to bombing, particularly by Prof. Lindeman (Lord Cherwell)
who became Churchill's personal scientific advisor, and, of course, by AM Harris C-in-C
Bomber Command. Photo-reconnaissance had shown that astro-navigation, dead reckoning
and the use of navigational beacons was not getting our bombers to their targets, and Harris
raised this problem at one of Rowe's Sunday Soviets. He was surprised to learn that a
possible solution had been put forward by RJ Dippy at Bawdsey, but had not been developed
then because of the urgency to complete the defensive Chain round the country. Next day

The Early Days of Radar - R&D for the RAF 1936-42 WH Penley 11 Oct 93 Rev1 Page 19
Dippy was set up with a supporting team to develop the scheme, and a contractor was quickly
appointed to put the designs into production.

In this system a master station and two slave stations transmitted pulses in synchronism. A
receiver in an aircraft measured the difference in the times of arrival of each pair of pulses
and this enabled the aircraft position to be plotted. The system was called Gee, was
introduced into service in March 1942. It proved of inestimable value throughout the whole
of the war as a navigational aid which helped to marshal our heavy bomber raids and helped
returning bombers find their way home safely. It was, however, not sufficiently accurate for
precision bombing, was limited to about 300 miles range and could be jammed.

OBOE

Another bombing system grew out of a proposal made in the autumn of 1940 to use a CHL
station fitted with a crystal controlled calibrator for more accurate range measurement to
position a bomber accurately over a target... Trials gave promising results and in January
1941 experiments were done in which dot and dash signals were sent to the bomber to
indicate its deviation from a central split line for the beam so that the pilot could correct this
to fly directly over the chosen target. The time for bomb release was to be given from the
ground station. The system of azimuth control was called "Howler Chaser" but one of the
team said it sounded like an Oboe, so from then on it was referred to as "Oboe". In the only
flight trial recorded, the aircraft reached a point a mile away from the planned target.

In February 1941, after for some time being unconvinced about their existence, the RAF
wanted to bomb the German beam transmitters on the Cherbourg peninsula. It was decided
to fly back up the German beam and to use range from an “OBOE” CHL station to fix the
bombing time. A transponder (modified "IFF" set) was used to enhance the return from the
bomber and overcome any likely jamming. Two sorties were carried out but it was not known
whether significant damage had been inflicted.

In March a new team lead by A.H. Reeves and F.E. Jones was formed to pursue this
application, and the name "Oboe" persisted even though the techniques to be used were
considerably different. It was decided to use very accurate range measurement from two
stations, and to arrange for an aircraft to fly at a constant range from one station on the arc of
a circle over the target. The range from the other station would determine the bombing point.
This would be sited to give good positional accuracy. In December, before the full "Oboe"
system had been developed, it was decided to attempt to bomb the Scharnhorst and
Gneisenau. Great accuracy was called for and so the RAF rigged up a very sharp Lorenz type
beam using components from the Beam Approach System and TRE provided an "Oboe"
accurate ranging equipment. Raids with over 30 Stirlings were made without loss and this to
some extent overcame the fears that the final period of straight and level flying would invite
large losses. Schemes were devised to overcome the low handling capacity. One pair of
ground stations was required for each bomber for the last 10 minutes of the bombing run so
multiple frequency and modulation systems were worked on. But finally the introduction of
good target markers that were relatively light and could be delivered by the Mosquito aircraft
determined the operational role. The Mosquitos, which were just being introduced, were

The Early Days of Radar - R&D for the RAF 1936-42 WH Penley 11 Oct 93 Rev1 Page 20
very fast and could fly over the targets at above 30,000ft, so were relatively safe. This
combination then provided a very effective target marking system, and was brought into
operational use in December 1942. Further developments to operate on centimetric
wavelengths, which were expected to be reasonably immune to jamming, were put in hand as
well as methods of using repeater aircraft to extend the operational range.

H2S

In the early summer of 1941 a review of the economic situation in Germany resulted in
Bomber Command's attention being concentrated on the destruction of German Industrial
targets. Cherwell, who was very much in support of bombing, got an analysis made of our
night attacks. It concluded that, whereas 75% of pilots claimed they had hit the target, they
in fact had failed to strike within five miles of it. This meant that the effectiveness of our
bombing campaign was very poor.

On October 26th 1941 Cherwell came to one of Rowe's Sunday Soviets and told him that it
was essential for the RAF to have a self contained device in bombers which would enable
them to navigate and strike targets deep in Germany. Dee remembered that the early
centimetric equipment, in trailers at Leeson overlooking Swanage Bay, was picking up
echoes from buildings in Swanage as well as from the Isle of Wight. He got O'Kane and
Hensby to set up the helical scan experimental AI, installed in a Blenheim, so that the beam
was depressed at a constant angle of 10 degrees. They got airborne within a week and on 1st
November 1941 they filmed returns, shown on a crude "A scope" display, from Salisbury,
Warminster and military installations on Salisbury plain.

Dee rushed the film into Rowe's office even before it was dry, and Rowe exclaimed, "This
will win the War!" The results were discussed with the Secretary of State for Air on
December 23rd 1941. He ordered that six special flights should be made to determine if the
echoes always related to specific targets. These flights were ultimately carried out and
provided the confirmation required, but, meanwhile on December 29th, Lovell, who was
helping to develop the centimetric AI, was summoned to see Rowe who instructed him to set
up a team to develop a centimetric navigation and bombing equipment, for use in bombers!
Lovell protested to no avail, and so started on the task. The trial had been with a forward-
looking equipment, and all round cover was clearly needed. Production of the Halifax
Bomber was just starting so Lovell went to see Handley Page on 4th January to get a Perspex
cupola mounted under it to house a scanner. This was strongly resisted but Cherwell
immediately reported it to Churchill and an aircraft fitted with a cupola arrived at Hurn on
27th March! A three-foot wide scanner was designed and installed, and using the basic
transmitter and receiver units designed for the AI, the first successful airborne trial was made
on 16th April 1942. This gave ranges of about 5 miles on towns from a height of 8,000ft.

Much development was needed to provide the range of 15 miles from 15,000ft altitude,
which the RAF needed. This was made more difficult by the prohibition imposed, at an early
stage, on the use of magnetrons over enemy territory. Intensive development of klystrons was
undertaken, but even though much higher powers were obtained and other difficulties
overcome, the performance achieved did not reach the minimum set by the RAF for effective
operational use. The programme was then hit by an overwhelming tragedy just after our

The Early Days of Radar - R&D for the RAF 1936-42 WH Penley 11 Oct 93 Rev1 Page 21
move to Malvern. On 7th June the Halifax crashed in South Wales, killing five of the small
team and wrecking the equipment. Nevertheless, the pressure of demand increased and
Churchill, after reviewing the situation on 3rd July, ordered an all-out effort to be made to fit
two squadrons of heavy bombers with H2S by October. Then on 15th July the Secretary of
State ruled that magnetrons could be used over enemy territory if the operational situation
justified the risk.

Development went ahead with gradually improving performance through many "Marks", and
at shorter and shorter wavelengths. H2S played a crucial role in Bomber Commands attacks.
These could now be made with high accuracy through 10/10ths cloud and at night.

Countermeasures

During in the development of the Chain, Tizard and Rowe had been very concerned about the
possible effect of jamming and other countermeasures upon the operation of RDF. Early on
Rowe had told E.K. Williams to get in an aircraft and "do his worst" to jam the system and
Cockroft had been invited to get the researchers at the Cavendish to consider the possibilities
of jamming. When war started we quickly became involved in helping to analyse the radio
signals emanating from the enemy, and to devise means of jamming or spoofing their
stations. This activity at TRE was brought together under the title "Countermeasures", with
Cockburn (Sir Robert) in the lead. RAE also set up a team. It soon became evident that
most of the developments in RDF and radio systems were vulnerable to interference,
deception and manipulation, and it was a rude shock to designers to discover how quickly
performance in the laboratory could be nullified in operation against a resourceful enemy.

From reconnaissance, intelligence and listening, we were already aware of German Radar -
early warning, tracking and airborne - but it was the German navigation and bombing beam
systems that first demanded our counter offensive. The first system, code-named
Knickebein, used beams that crossed over the target. The centres of the beams were defined
by the merging of dot and dash modulations into a continuous signal. Radiating additional
dashes that destroyed the accuracy easily fooled these. The Germans soon replaced this
system by one on a shorter wavelength that used a bombing computer in the aircraft. We
developed new jamming transmitters that neutralised this system too. A third, and much
more complex system, was introduced in the spring of 1941. It was countered by adding a
similar modulation transmitted from the Alexandra Palace Transmitter!

Our bombing of Germany was building up and our bombers needed protection against the
German Radars. Airborne noise jammers were developed for the early warning "Freyas" and
the tracking "Wurzburgs". But the most effective counter to centimetric equipments was
"Window" or "Chaff". In this, packets of half wave "dipoles" cut from metalised paper or
metal foil, when dispersed from aircraft, could simulate many other aircraft and confuse or
saturate the radar displays. The first trial was carried out using metal dipoles cut out by hand
by Mrs (now Lady) Curran. By March 1942 its effectiveness had been well demonstrated but
its use operationally was delayed for fear of similar reprisals until July 1943 when we had
helical scanned AI MkX equipments in service. Work on all aspects of countermeasures

The Early Days of Radar - R&D for the RAF 1936-42 WH Penley 11 Oct 93 Rev1 Page 22
intensified on both sides and reached its peak in the spoofing and jamming done for D-Day
and during the invasion.

IFF

A problem that became more and more significant as the war progressed was the need to
distinguish friend from foe without visual contact. Early equipments ("broody hen") had
been designed which would modify the signal returned from an aircraft target by using a
small receiver and transmitter (transponder). This could pick up the RDF pulse from a Chain
station and retransmit a lengthened version of it that would show up on the ground display.
With this arrangement different transponders were required for each RDF frequency - a
daunting proposition. It was therefore decided that a special frequency should be used for a
"universal" system, and a team was set up under BV Bowden (Lord Bowden of Chesterfield),
and IFF Mks I and II were developed. The team moved to Washington when close co-
operation with the USA started as a result of the "Tizard Mission" in September 1940.

The Tizard Mission to USA

By the middle of 1940 it had become clear to Tizard and many others, that though the UK
had some first class research teams we were woefully short of facilities for development and
production. He persuaded the War Cabinet that we should share our defence secrets with the
United States of America in the hope that they would provide help in the development and
production necessary to meet our war needs. He got blanket authority from Churchill to
disclose everything and a mixed civilian and service team was formed with Tizard in the
lead. Cockroft, Bowen and Woodward-Nutt were the other civilians and they were
accompanied by Col. F.C. Wallace - back from Dunkirk, Capt. H.W. Faulkner, RN, and Grp.
Capt. Pearce from Coastal Command.

The Mission arrived in Halifax on 6th September 1940 and Tizard decided to "lay all the
cards on the table" with no attempt to trade secret for secret. The Americans heard the
civilians speaking with authority on the equipments and systems and their Service friends
following with their practical experiences. Any initial doubts by the Americans were soon
dispelled and they responded with great speed and enthusiasm. The secrets of our RDF
developments including actual equipments (except for Gee because of jamming worries)
were all handed over, also many other items such as a jet engine and a proximity fuse. Our
trump card was undoubtedly an early production magnetron. This was left at Bell
Laboratories on 3rd October, and Bowen saw it working there five days later. It increased by
a factor of 1,000 the power available to them on the centimetric wavelengths they had been
trying to develop.

A very important unofficial meeting was at a weekend party for Cockroft and Bowen in
Alfred Loomis's house on 28th/29th September. Loomis was Chairman of the NRDC
Committee on microwaves and he had collected together small party of experimenters who
had been working on 10cm wavelengths. The magnetron opened new vistas for them and
Loomis proposed the formation of a microwave laboratory, based on the TRE model, to
concentrate on the new possibilities. It was agreed that night attack was the UK's direst peril
and the first objective should be a centimetric AI. The helical scan system described by

The Early Days of Radar - R&D for the RAF 1936-42 WH Penley 11 Oct 93 Rev1 Page 23
Bowen was considered attractive. Cockroft's suggestion of a centimetric Gun laying
equipment was also met with great interest.

The "Radiation Laboratory" was set up within a few weeks. It subsequently played a major
role in close co-operation with the UK research and development teams and a section of it
joined TRE. Both the AI and Gun-Laying equipments were developed and subsequently
played very significant roles in the later stages of the war.

Tizard returned to the UK on 2nd October and Cockroft took the Mission to Canada. He
proposed that the Canadians should put the British ASV into production, co-operate with the
Radiation Lab in the development of centimetric equipments, and themselves develop a
centimetric gun-laying radar to provide for blind firing. This they did, and the ASV was
available in North America at the time of Pearl Harbour.

Bowen stayed in the USA; played a major part in setting up the Radiation Laboratory and in
the arrangements for developing and producing versions of the equipments we had handed
over.

Support and PDS

Within TRE, many specialist teams were set up to handle basic investigations, develop
common components, test gear and circuits for general use, and to provide services such as
mathematical investigations and aerial design. These were to support the development of the
widening range of equipments. In addition, of course, there was a drawing office to help turn
concepts into manufacturable form, and work shops to make the experimental equipment.
Firms were brought in as early as possible to engineer experimental designs for production,
and as they learnt about the new techniques, to take responsibility for some initial design. All
these activities were supported by an administrative system, security and police and many
other activities such as “riggers” for erecting towers and aerials, a “heavy gang” to move
things about, cleaners, telephone service and a library to deal with technical books and
reports etc.

For the RAF it became clear that quite a small number of equipments made to exploit a new
concept could have a major effect on their operational success. These early equipments were
often little more than lab experimental models and members of the scientific staff would be
sent out with an equipment to set it up and make sure it operated when required! This
activity continued to be of vital help, even when improved designs had been made, and so a
special team was built up for this purpose. This was called "Post Design Services". A
special workshop "The Radio Manufacturing Unit" (RPU) was set up North of Bournemouth
by Sieger, who headed our design and workshop teams, to make pre-production quantities of
equipments to achieve the earliest possible introduction into Service.

Our rapid expansion with staff from Universities, the Post Office, the Services and elsewhere,
made things very tight at Worth and when we began to attract bombing and machine gunning
raids, two schools, Leeson House and Durnford in Langton Matravers were requisitioned.
Rowe and Lewis with the top administration moved to Durnford and the bulk of the
centimetre work to Leeson. It had been found that training in the use and maintenance of the

The Early Days of Radar - R&D for the RAF 1936-42 WH Penley 11 Oct 93 Rev1 Page 24
equipments in Service was a crucial part of making them operationally effective. Another
school, Forres in Swanage, was also requisitioned and became a training school for the
Service operators and maintenance staff as well as our own people. This was set up by
Ratcliffe, and when he went to do the same task for the Army at Petersham, was taken over
by Huxley (Sir Leonard). A mixed civilian and service teaching staff was set up. Section
Officer Renie ....... (Lady Adams) was one of the first WAAF instructors we trained to teach
there, and there was a constant flow of courses of all kinds.

To help with teaching and training, both at TRE and at Service Units, Dummer was building
up a team to design and develop simulators and training equipments. Many other tasks were
also being undertaken by small teams, such as message decoding, the development of radio
altimeters, airborne and ground beacons to respond to Radar signals and research into
methods of landing aircraft “blind”.

The Bruneval Raid

In February 1942 Don Priest, one of the early team, donned a Flt Lt uniform to go as the
technical brains of a raiding party tasked to bring back the most significant parts of a German
RDF station which had been photographed at Bruneval. He was ordered to remain on the
landing craft to reduce the risk of him being captured. Under his direction a lot of very
important pieces of the equipment were brought back to TRE. We were shocked at the high
quality of the engineering. How fortunate it was, as we found out after the war, that the
relationship between the German Forces and their scientists were those of master and servant,
and the views of the civilians on operational matters were not wanted! This was in sharp
contrast to the very close co-operation that had evolved between TRE and the RAF. Very
fortunately for us, the German equipment, thought excellently engineered, was not operated
in a way that would make full use of its undoubted potential.

The Move from Purbeck to Malvern

Shortly after this it became known that a Parachute Division had been assembled on the
Cherbourg peninsular and it was widely believed, by the Prime Minister and Cabinet, that
they intended to descend on TRE to capture us and our equipment. Night after night the
trailers holding the centimetric equipment being worked on at Leeson were driven out of
Purbeck and brought back the next morning! At the end of April the team leaders were called
together and told that we had to leave before the next full moon! Malvern College had been
chosen for our new home, and a few of us went straight up there to plan and arrange the
modifications - benches, power supplies, telephones etc. - to convert the buildings into
laboratories. Pickfords arrived in force; we packed up, and after one false start, left for
Malvern early in May.

About 200 people came in May 1940 and around 2,000 left in May 1942! In Malvern the
Establishment grew to nearly 4,000 and continued to play a crucial role in the conduct of the
war. Except for the initial work on the Chain our time in the Swanage area is judged to have
been the most significant and creative period in the history of RDF or RADAR as it is now
called. It played a major part in winning the battle against the night blitz and a crucial role in

The Early Days of Radar - R&D for the RAF 1936-42 WH Penley 11 Oct 93 Rev1 Page 25
the battle of the Atlantic and provided a sound foundation for the future developments that
played so important a role in the invasion on D-Day and in the battle to regain Europe.

Subsequent developments such as Air Traffic Control, Storm Warning, Weather Radar and
Guided Weapons together with a myriad of other electronic developments such as microchips
and computers grew out of the activity. The techniques were used by many of the Scientists
when they returned to their own research fields, to very great advantage. Several became
Nobel Prize Winners and achieved the highest eminence in the World of Science.

The Early Days of Radar - R&D for the RAF 1936-42 WH Penley 11 Oct 93 Rev1 Page 26