A Brief History of RAF Defford


Geoffrey Negus, who was Chairman of Air-Britain, the International Association of Aviation Historian & Enthusiasts, wrote a useful summary of the history of flying and radar development at Defford in the mid-1980s, which was more recently the subject of a posting on the Air-Britain Ab-Ix web-site.

This document, with the kind permission of Geoffrey Negus, is reproduced here in two parts – the first part, the history of Defford follows.

In giving his permission, Geoffrey Negus wrote: “Bob – No problem at all – but somehow or other you need to acknowledge that this was only ‘work in progress’ from the 1980s – there are bound to be omissions and errors”.  More recent studies may indeed suggest some changes, but nevertheless the document is seen as such a useful and succinct summary, it is here reproduced almost in its entirety as written.


RAF Defford – General

Geoffrey Negus, Air Britain Ab-Ix, Nov. 23, 2005

These are my notes on this topic, made in the 1980s, doubtless imperfect but I hope of interest to enough people to justify such a long posting.

Defford opened as satellite to 23OTU (Pershore) in September 1941. At first it only operated in daylight, using rudimentary facilities. On 29 March 1942 Wellington X9629 stalled and crashed when its port engine failed to respond during an overshoot procedure. On 27 April Wellington R1618 was unable to get airborne on its take-off run because the throttle lever was not tight enough and it swung off the runway. Wellington T2851caught fire after practice bombs were accidentally dropped on 4 May. The OTU moved out on 18 May 1942.

The airfield was taken over by the Ministry of Aircraft Production for use by the Telecommunications Flying Unit (TFU), within 10 Group. The TFU was dedicated to providing realistic experimental and test facilities to the Telecommunications Research Establishment, which during the Second World War pioneered new radio and radar aids.

The story of the TRE is fascinating, but first its origins and the background to its move to Worcestershire should be explained.

The TRE, formerly the Air Ministry Research Establishment, was formed at the end of 1940. Originally at Bawdsey, then Dundee with an establishment at St Athan, the unit moved to Worth Matravers, near Swanage in Dorset, in early 1940. Its expose position and the regular attentions of German bombers caused it later to move into Swanage itself. Flying facilities were provided at Christchurch, Hampshire, and later on a bigger scale at Hurn.

One of the TRE’s first achievements was the development of the GEE navigational device, which enabled aircraft navigators easily and quickly to pinpoint their position. GEE had the special virtue of not requiring the aircraft to transmit a signal and thereby betray its position to the enemy, but suffered from limited range (350 miles) and the risk that the Germans could jam it. The first GEE installations into Bomber Command aircraft were made in the summer of 1941. Scientists (in particular Bernard Lovell) at the TRE wanted to enable aircraft to be navigable using a device that could `read’ the terrain they were flying over, and not depend on vulnerable radio signals from ground stations.

However, they received little encouragement because Bomber Command was vastly over-estimating the damage it was inflicting on Germany and saw little need to improve bombing accuracy. This complacency was shattered by an analysis of photographs from recent raids, conducted by a War Cabinet official at the prompting of the Prime Minister’s Scientific Adviser, Lord Cherwell. The survey revealed that only one in three aircraft recorded as attacking their target got within five miles. In New Moon periods or over thick haze the proportion fell to one in fifteen. The report caused Churchill to make the improvement of Bomber Command’s ability to find its targets a top priority.

As a consequence, development of RDF (Radio Direction Finding) aids by the TRE was expanded and liaison between the scientists, civil servants and Bomber Command improved. The newly-appointed Senior Air Staff Officer at Bomber Command, Air Vice Marshal R H M S Saundby, was one of the few in his Command who had previously believed that the bombs were not reaching their intended targets. Saundby formed a RDF Department at Bomber Command Headquarters at High Wycombe and selected Wing Cdr Dudley Saward to lead it. At last there was direct contact between the boffins and the airmen who actually used the equipment.

As early as 1939 the Air Ministry Research Establishment was investigating Air Interception RDF for night fighters, known as AI. Early work was based on using a wavelength of 11/2 metres. Dr E G Bowen at Dundee had recognised that AI might be adapted for terrain-identification, but the pressing need then was for systems to deal with bombers and U-boats at night. Lovell later examined the potential of using 5/10-metre equipment, using an Anson, and also modified 11/2-metre AI apparatus. This suffered interference from ground returns, which led to using centimetre wavelengths which could be transmitted in very narrow beams. A prototype magnetron made by Birmingham University produced promising results, which resulted in AI Mk.VII, which was fitted tooperational night fighters in late 1941.

In October 1941 Lovell and his colleague Dr P I Dee realised that the angle at which they were using centimetre AI on high ground above Swanage, pointing in the direction of the Isle of Wight, was not dissimilar to the angle from an aircraft when looking at features on the ground at long range. Dee immediately arranged for two of his assistants to fly Blenheim V6000 with a helical scanning AI operating on 10 centimetres, with the scanner giving a depressed forward angle of view. This historic experiment, on 1 November, indicated that a device that could show towns, rivers, coastlines, lakes, even military camps and aerodromes, was within grasp. Wing Cdr Saward was told of the discovery in early December. At that time it was called BN (Blind Navigation) but later it became H2S. Though still very crude, it operated on the principle that different surfaces reflect back very high frequency radio waves in varying degree. Coastlines and inland waterways showed up especially clearly, towns less so. The scope for making H2S more sophisticated was immense – the aircraft’s heading could be indicated, also range finding facilitiesand the ability to alter the area shown by the picture.

The next battle facing the TRE scientists was to be allowed to use the magnetron rather than the klystron transmitting valve in H2S. The magnetron was much more powerful but was the secret of the success of the AI Mk.VII, and it was feared that if it was used in bombers over enemy-held territory it would be captured and turned against the RAF. Magnetrons were also used in gun-laying devices and searchlight control and was likely to be the basis for improved ASV (Air to Surface Vessel). The power of the transmitting valve greatly influenced the definition of the images.

On the night of Friday 27 February 1942 a force of about 120 men was dropped from twelve Whitleys to capture vital components of a German Würzburg radar installation at Bruneval near Le Havre in France. The equipment was examined by the TRE at Swanage. But the success of the raid prompted fears that the Germans might mount a reprisal attack on the TRE and obtain information on new types of British radar. So, in the latter half of May 1942, the TRE moved to Great Malvern to take over Malvern College, and the TFU moved from Hurn to Defford.

At the time of the move Halifax V9977, which had been delivered to the TFU at Hurn on 27 March 1942, was being used for H2S development. Its first flight with H2S was on 16 April. On Sunday 7 June, the aircraft caught fire at 500 feet and crashed six or eight miles south-west of Ross-on-Wye, killing six brilliant scientists and their crew of five and destroying the one and only magnetron H2S. This was a disaster not only to the TRE but the entire bomber offensive against Germany. Only a week before the crash, on the night of 30-31 May, Cologne had been devastated using the TRE-developed GEE, but it would be only a matter of time before the Germans jammed it. If the newly developed accuracy of bombing was to be maintained, H2S had to be made operational quickly.

Largely thanks to Bernard Lovell, though, within weeks replacement apparatus was ready to be fitted to Halifax R9490, which made its first flight with H2S on 2 June, and HalifaxW7711, which was delivered to the TFU on 26 June to replace V9977.

It was during this frantic period, on 19 June, that the new Commander-in-Chief of Bomber Command, Arthur Harris, appointed Gp Capt (later Air Vice Marshal) D C T Bennett to head the new `Pathfinder Force’. Bennett spent most of July at Defford test-flying both klystron and magnetron H2S systems. The flights with the klystron failed to get results from more than twelve miles, while the first flights with the magnetron showed Birmingham at nearly thirty miles’ range. RAF Bomber Command supported the plea of TRE scientists to use the magnetron and on 15 July the go-ahead for developing magnetron H2S for use on operations over Germany was given, and all work on the klystron version stopped.

Bennett wrote in his autobiography, `Defford had the vague idea that aeroplanes, even in wartime, had to be wrapped in cotton wool, seldom flown, but in the glorious name of “inspection” repeatedly pulled to pieces and put together again. In short, they had the most old-fashioned pre-war RAF conception of maintenance, and no idea whatever of getting on with the job. The fact that I required them to test-fly at all hours of the day and night shook them to the core. Their own maintenance personnel very soon fell down on the job, so I signalled to Bomber Command to let me have some men from 4 Group who were accustomed to maintaining Halifaxes’ (Pathfinder, Fredk Muller 1958).

On 10 August the Germans began to jam GEE successfully, at least over territory they occupied. An anti-jamming device restored the usefulness of GEE to a limited degree, but it was obvious that H2S development had to be hastened. Operational trials by 1418 Flight (Gransden Lodge) began at the end of September and by the end of 1942 H2S-equipped Halifaxes were being flown by 7 and 35 Squadrons. The first operations over Germany with H2S were in early 1943.

Alongside this work was the development of ASV for Coastal Command, particularly to enable U-boats to be detected when they surfaced at night to re-charge their batteries. The earlier 1 1/2-metre ASVs still being used by Coastal Command could be detected by listening devices that enabled the U-boats to dive before attacks from the air could be pressed home. The new 10-cm ASV not only gave a better image but was also more difficult to detect. Official emphasis on developing H2S for naval applications undoubtedly slowed its wide-scale introduction into Bomber Command. The new ASV was fitted to Wellingtons based at Chivenor, Devon, in January and February 1943, to become operational on 1 March. The first sinking of a U-boat by a H2S-equipped aircraft was on 17 March. By May Allied shipping losses were falling dramatically.

Naval work was carried out on Swordfish, Martlet, Barracuda and Firefly aircraft and co-operational was maintained with the Coastal Command Development Unit. A Royal Navy section was attached to the TFU for trials purposes.

On the night of 30-31 January 1943 H2S-equipped Stirlings and Halifaxes marked the target in Hamburg for the Main Force of bombers. The crews were enthusiastic about the range and definition provided by the H2S sets. Further raids in early February on Cologne, Hamburg and Turin confirmed the effectiveness of H2S. Berlin was bombed for the first time using H2S on 1 March 1943. However, operational use also showed its shortcomings, particularly when operating at altitudes above 20,000 feet. In early 1943 work began on a 3-cm device that showed topographical detail more clearly, to be known asX-band H2S. This equipment would also give Coastal Command the ability to switch to another frequency immediately there were indications that the U-boats were listening on 10-cm, a likely eventuality as bombers with 10-cm H2S were being lost over German-held territory.

The new 3-cm H2X was experimentally flown in Stirling N3724on 11 March 1943, but this equipment unfortunately presented its own problems – the returns at 5,000 feet were much better than at 10,000 feet. Saward flew Lancaster ED350 from Defford on 24 June to judge the latest version of H2X, which was distinctly better than 10-cm H2S for detailed ground representation. Meanwhile work on the 10-cm H2S aerial system was improving its performance at higher altitudes.

The Americans had condemned H2S as a useless and dangerous system in 1942, but by May 1943 had developed their own version of it and had the audacity to want it fitted to Bomber Command aircraft. However, their version was totally inappropriate to the needs of Bomber Command. At two high-level meetings in London on 7 June 1943 it was agreed to let the TRE continue with its programme but let the Americans work on X-band H2S for ASV. Fortress 42-5793 came to Defford, followed by other B-17s from the USAAF Pathfinder base at Alconbury.

Another device to improve bombing accuracy, OBOE, was developed in the spring of 1943, but it was limited by the fact that only one aircraft could be guided by two ground stations at a time and its range was only 300 miles. However, during the shorter summer nights when targets closer to home were preferred, OBOE was used with devastating effect on cities in the Ruhr valley.

Nevertheless losses due to German night fighters were mounting. They often attacked from beneath and behind the target bombers. An AI device, Monica, was supposed to warn crews when another aircraft was in their vicinity, but it did not tell them where the intruder was, or its closing speed, and continually picked up friendly aircraft in the same bomber stream. This led many crews to switch Monica off, finding it moreof a nuisance than a help.

The TRE was working on AGLT (Air Gun Layer Turret) – RSF controlled turrets which would automatically aim and fire at unseen enemy aircraft, but these would not be ready for operations for at least another year. Lancaster KB805 was used on AGLT tests.

An idea, nothing short of a brainwave, hit Bernard Lovell as to how to provide a stop-gap defence. It was to use the principle of H2S, but instead of scanning below the horizon so as to pick up details of the surrounding ground terrain, to aim it above the horizon. Any aircraft within this cone would normally not be seen because its signal was swamped by ground returns – after all, H2S was not designed to show aircraft. Lovell now proposed to use a second receiver-indicator calibrated to use the part of the transmitter pulse from when it left the aircraft to just before the first ground return, which was that directly below the aircraft. So if the bomber cruised at 18,000 feet, the second H2S set would detect anything nearer the aircraft, but below its altitude, less than 18,000 feet away. Not only that, it would show the bearing of the aircraft, its distance and speed, enabling the operator to distinguish it from friendly aircraft, which would tend to keep station in the stream. It would also act as a useful collision indicator and warn if the aircraft was slipping out of the main bomber stream.

By avoiding officialdom, Lovell managed to prepare the prototype within a fortnight. Saward flew with the device in Halifax BB360 from Defford on 27 May and tried it successfully in co-operation with a Mosquito night fighter. It was christened Fishpond, or officially Indicator Unit 182. By the beginning of July Fishpond was in full production, to be used from October. It saved hundreds of bomber crews’ lives.

At the end of May 1943 Hamburg was bombed on four consecutive nights by large forces – the Pathfinders all had 10-cm H2S and many aircraft in the Main Force did, as well. The effect on this important port, naval and industrial centre was catastrophic. The part played by H2S was immense, partly because the coastline showed well on the receiver sets. On 17-18 August Peenemunde, on the Baltic coast, the principal German rocket development centre, was largely obliterated by an H2S-equipped force, putting the V1 and V2 programmes back by months. At the end of August and beginning of September Berlin was raided on three nights, but the 10-cm sets could not cope adequately with the size of this inland target. Nevertheless, it was found that a greater proportion of H2S-equipped bombers found their targets than those not yet fitted with the device. The raid increased demands for 3-cm sets to be installed as soon as possible.

The problems in developing 3-cm H2S were many and Lovell and Saward again had to resort to subterfuge. By 13 November three Pathfinder force Lancasters, JB352, JB355 and JB365, had it installed. Six such aircraft were ready by 18 November for the first of a series of massive raids on Berlin lasting until February 1944, collectively called the Battle of Berlin. In general terms, the target was pinpointed by Pathfinders with 3-cm H2S, followed by the Main Force with 10-cm H2S and protected by Fishpond. The loss rate during the campaign was 5.4%, but more like 2.4% for H2S- and Fishpond-equipped aircraft. More than 5,000 acres of Berlin were razed by the raids. Other cities in deepest Germany were also attacked during this period.

The pause in the offensive against Germany while Bomber Command helped with preparations for D-day was used by the TRE to evolve further improvements to H2S, to make it easier to use under combat conditions. However, work on H2S slowed noticeably as the end of the war came into sight. Interest in the bombing offensive waned, there was official indecision on the priorities for TRE work, several of the Establishment’s scientists became ill, through exhaustion if anything, and while the rapport between Malvern and Bomber Command remained good, liaison with aircraft manufacturers and the Ministry of Aircraft Production became more difficult.

During the invasion of France a system called G-H was important, because it offered the accuracy of OBOE at ranges of up to 300 miles but could cope with up to eighty aircraft at a time.

It was used on Lancasters of 3 Group which did not have H2S, alongside the Monica tail warning device. Unfortunately Monica could be detected by German aircraft equipped with a device called Flensburg – so Monica was scrapped and Fishpond remained the principal warning device in the RAF.

On 25 June 1944 flight-testing of the H2S Mk.VI K-band (1¿-cm) began in Lancaster ND823. A six-foot scanner enclosed in a blister was tested from 6 July in Lancaster JB558. The Mk.IV H2S system, tested on Lancaster JA845, was found to give the best pictures yet obtained of both coastlines and towns, but the experiments with the six-foot scanner did not impress Bomber Command. The Mk.VI was earmarked for specialist army collaboration duties, for example detecting concentrations of tanks from the air, but Mk.IV became the firm requirement for Bomber Command. Saward has described Mk.IV as `the near-perfect blind navigation and bombing system’.

Dudley Saward, who liaised between the TRE and Bomber Command, is eminently qualified to judge the value of the TRE in the war. In his excellent biography of Bernard Lovell, he noted that accurate bombing of targets in Poland, Czechoslovakia, Austria and northern Italy depended on H2S, not to mention targets like Leipzig, Berlin and Peenemunde. The earlier development of ground radar and airborne AI for fighters was crucial in 1940-41. In the Atlantic theatre, ASV was the weapon that prevented the U-boats from strangling Britain. `The contribution of the scientists and engineers of TRE was of inestimable value’, wrote Saward, `and without it the war might have ended in a negotiated peace or even finished with the ignominious defeat of Britain’.