Radar, Coast Defence, Mark I, or CD Mk. I, was a radar system used by the British Army to detect ships, E-boats and U-boats during World War II. An Army-wide renaming exercise in 1943 made these the CD Number 1 Mark I, which confused things as unrelated sets were also made part of Number 1 with different Marks.
The initial concept was to provide early warning of German shipping in the English Channel or attempting to approach one of the UK's many ports. While being tested in July 1939 at the radar development site at Bawdsey Manor, operators noticed returns on their displays from the guns being test fired at Brackenbury Battery outside Harwich. The splashes of the shells in the water created blips that lasted long enough for the operators to measure their bearing and range. An adaptation with higher accuracy for directly guiding the Army's coastal artillery units was developed as Radar, Coast Artillery, Mark I, or CA Mk. I, but not put into operation. During early use, the name CD or CA was applied depending on the location of the radar, not the specific model installed.
During early testing it was noticed that the sets were also ideal for detecting low-flying aircraft. This was a significant problem for the Royal Air Force (RAF)'s Chain Home (CH) radars, which had difficulty with anything flying below about . The Germans learned of this when they began mine laying operations in British ports and the Thames estuary, and began flying low-altitude intruder flights to take advantage of this limitation. In response, the RAF took over many of the CD sets being constructed and deployed them as Chain Home Low (CHL) at CH sites so they could share the existing telephone networks. It was some time before enough sets were built that the CHL needs were filled and wider deployment of CD and CA could begin.
The introduction of the cavity magnetron in 1940 led to new models of almost all existing radar sets, including new CD and CA systems. Work on the Mark I in the CA role was cancelled and a new CD/CA system based on the Royal Navy's Type 271 radar began to be deployed around 1941. Once again, the RAF interrupted this process when began attacks at altitudes as low as , and these were deployed under the name Chain Home Extra Low (CHEL) along likely approach routes. These attacks largely ended by the end of the year and CD and CA was able to fully expand with the new sets.
Early work
In October 1936, having learned of the formation of the Air Ministry Experimental Station (AMES) at Bawdsey Manor, the British Army's scoured their ranks looking for radio experts that would be able to help develop radars for Army use. Ironically, two of the researchers, W. A. S. Butement and P. E. Pollard, had been the first to suggest developing radar to detect ships in the English Channel in 1930, but the War Office proved uninterested at the time. Now, with the news that the Air Ministry had succeeded in detecting aircraft, they sent Butement and Pollard to Bawdsey to form the Military Applications Section, referred to universally as the Army Cell. They were given the task of using radar to accurately measure the range to a selected aircraft, a major problem for existing anti-aircraft artillery operations.
Around the same time, Henry Tizard grew concerned about the German reaction to the Chain Home radar. He concluded they would turn to night bombing, where the radar would still provide early warning of their approach, but the fighter aircraft would not be able to see their targets. His memo led Robert Watson-Watt to set up a new group within AMES to develop a radar that could be fit inside an aircraft to allow them to track down their targets. The resulting experimental system was being tested in 1937 when the operators noticed odd returns while flying near the shore of the Channel. They eventually realized these were the piers and cranes at the Harwich docks miles south of them. Shipping also appeared, but the team was unable to test this very well as their Handley Page Heyford was forbidden to fly over water.
To further test the concept, Robert Watson-Watt provided the team with two Avro Ansons that were able to fly out over the North Sea from nearby RAF Martlesham Heath. The testing system was crude; a small dipole antenna was held by hand outside one of the escape hatches and rotated looking for the peak signal, indicating the antenna was aligned with the target ship. This was not easy as the signal naturally fluctuated. The first successes were in August 1937.
Test systems The Army Cell had already been asked about ship detection in April 1938 and they responded in May that they believed they could have a set in operation within a year of getting more help. It was not until October 1938 that such help arrived and work began.
At the time, they were exploring two concepts. The first was a set operating at 1.2 m wavelength that had initially been developed by Edward George Bowen's team for airborne use, and was designed to use several Yagi antennas ganged together to improve directionality. Experiments demonstrated it was difficult to keep the antennas tuned to the required degree. The second concept operated at the extremely short wavelength, for the time, of 18 cm and used parabolic reflectors to produce very good directionality. Development of this model dragged on and it was still not ready for testing in February 1939 when the rapid development of a third concept suggested development of the other two be shut down.
The third set was an adaptation of the Air Ministry's 1.5 m sets for aircraft, developed directly from the hand-held sets in the Anson. To improve directionality, the system used a "split" antenna, today known as lobe switching, in which two antennas are placed close together in front of a common reflector. A switch rapidly connects the transmitter to one antenna and then the other, while adding a small electrical delay to one of the two antennas on reception. The result is that two blips appear on the display for every target, and the one that is more closely aligned with the antenna will be longer. By rotating the entire antenna until the two blips are the same length, the system can accurately track the angle of the target. Testing in May demonstrated that angles on the order of 15 minutes of arc, a quarter of a degree, were possible. A high mounting allowed ships of to be detected at .
During these tests, the first examples of anomalous propagation were seen, with detections on sailing boats at distances that would normally be below the radar horizon. Another important result was that the sets detected low-flying aircraft on several occasions, including one seaplane flying at altitude at distance, well below the lowest angles that Chain Home could achieve. The Air Ministry researchers immediately ordered a second set in mobile form that could be moved around the countryside to test its performance from various altitudes on sea-side cliffs.
On 20 June 1939, Winston Churchill visited Bawdsey, where he was shown the experimental CD unit, along with CH. In July, the experimental set at Bawdsey happened to be operating when a coast defence artillery station at Brackenbury Battery protecting the Harwich docks was test firing their BL 9.2-inch guns. To the radar operator's surprise, the splashes of the shells in the water produced strong blips on the displays and they lasted more than long enough to take measurements of their range and bearing. From this point the concept of Coast Artillery radar was developed in parallel with CD, mostly based on the same hardware. During a demonstration to Army artillery technical staff in August, they saw not only the splashes, but even the individual rounds on their way through the air.
The major difference in the CA role was introduced by Butement around the same time, the spiral time base. Most radars of the era used a A-scope type display, where the blip was drawn horizontally across the face of the cathode ray tube (CRT), with a maximum length of about . If the operator might be expected to measure the blip to an accuracy of a millimetre, and the display was showing range, that was an accuracy of about , too low for gunnery. Butement's system moved the blip in a spiral motion such that the range to the target was a combination of the distance from the centre of the tube as well as the angle around it. This produced a line that was up to long, dramatically improving range accuracy. If the same 1 mm accuracy was measured on this display, that represented a distance of only 15 metres.
CD Mark I In July 1939 a development contract was sent out to build prototypes of production versions based on the results of these tests, with the goal being to release the final design in October. Both the Air Ministry and Army were part of the contract, as the RAF planned to use the system under the name Chain Home Low (CHL). The only major difference between CHL and CD was that the former later included a plan position indicator (PPI) which made it easier to track fast-moving targets. This was not as useful in the CD role where the targets were much slower moving.
The declaration of war on 3 September resulted in pre-war plans being put into immediate effect, which included the movement of the radar teams away from their exposed location on the Channel coast. The Army Cell was ordered to move to the Christchurch area, which had been heavily fortified due to its location near Bournemouth and Southampton. The CD team was sent to a shoreside location with practically no infrastructure other than a few pre-fab buildings. The team spent the next days involved in everything from running electrical wires to digging sewer lines. Despite this, they were able to restart development of the production sets soon after arriving.
Within days another demonstration was arranged, this time for Admiral James Somerville, with the specific intent of testing the system for use against submarines. Butement hastily returned to Bawdsey and got the original system operational in time for to pass by in two days. For this demonstration, John Cockcroft was invited to attend as well.
As they were demonstrating the system to track the submarine, an RAF aircraft, apparently thinking it was a German U-boat, began attacking it. Somerville personally warned them off with "effective naval phraseology" and the test went on to successfully prove the radar could detect submarines as well as ships. As Cockroft described it:
Somerville's own recording of the event is somewhat more circumspect:
Work was aided by the arrival of a team of researchers from Cockcroft's Cavendish Laboratory at Cambridge University. They took it upon themselves to disassemble the Bawdsey unit and bring it to Cambridge, where Pye Radio engineers were brought in to begin producing copies.
Immediate deployment On 14 October the Royal Navy suffered an enormously embarrassing setback when U-47 snuck into the harbour at Scapa Flow and sank the battleship HMS Royal Oak. This led to demands for an immediate deployment of the CD systems in the anti-submarine role, under the name CDU.
Three sets, one of these possibly being the Bawdsey prototype, were shipped north on 27 October 1939 in an emergency-speed deployment to Fair Isle and Sumburgh in the Shetland Islands. Days later it was realized that a required part had not made the voyage, and a Navy message asked for it to be delivered simply to "Cockcroft, Edinburgh". Hours of phone calls eventually solved the mystery of who this might be, and he quickly left with the needed part. Sumburgh was operational in the first week in December, but the conditions on Fair Isle were significantly more difficult. The first set was operational there on 16 February, and the second on the 23rd. All of these stations soon proved themselves highly valuable in detecting aircraft attacking the fleet.