7

Radar at work

During the year 1938 radar was transformed from an experimental layout into an operational system capable of being used in war.

In August there was an important change in the management of Bawdsey. Watson-Watt was promoted to the post of Director of Communications, Air Ministry, and his place as Superintendent of Bawdsey was taken by A. P. Rowe.

Although not professing to be an electronics specialist, Rowe was a first-class organiser with an unusual flair for analysing problems and finding the right people to answer them.

He soon realised that many of the trials completed at Bawdsey had been carried out under almost ideal conditions such as would be found in a laboratory. The injection of practical circumstances into exercises showed up marked deficiencies. For instance, attempts to intercept K.L.M. and Lufthansa airliners in the Thames Estuary proved complete failures. If radar was to work it had to be tough and reliable. Dr. E. C. Williams, at that time a young research worker, was given the job of setting-up radar countermeasures designed to make operation of radar as difficult as possible.

Using a modified diathermy set installed in a Sunderland flying-boat, Williams conducted the first electronic countermeasures flight. This showed that ‘jamming’ could make the cathode ray tube on the ground almost unreadable with dancing lights blotting out signals from aircraft.

An interim solution was found using coloured filters to show up the trace on the screen, but the real answer came later with development by Professor Merton of the long afterglow. With this the echo of the aircraft remained visible for some time and the flickering from jamming could be separated.

July 1938 saw the addition of Great Bromley and Dover radars, thus completing the first five-station chain in time for the highly important air defence exercises on August 5th–7th. The stations were not entirely ready as two were uncalibrated in direction and height-finding. Scientists from Bawdsey were posted to each unit to supervise operation.

The backscreen of sensing relays which cut out the radar stations’ view to the rear and avoided confusion between landward and seaward echoes had not been completed. Despite this and poor weather the results were very promising. The only type of aircraft which consistently eluded the watchers at the cathode ray tubes was the low-flier which penetrated below the cover of radar beams. It became clear that some new type of equipment would be required to deal with this.

A Bawdsey memorandum of August 12th noted that ‘it is apparent that filtering is still an art rather than a science’.

Discussing the current situation of radar and the communication requirements revealed by the exercise, a report from Bawdsey ran on August 30th

It is understood that:

1 There will be one map in the filter room covering the whole of the east and south coasts.

2 There will be three Fighter Groups.

3 Information is to be told to groups and thence broadcast simultaneously to sectors.

4 Experience with Biggin Hill has shown that sectors require information accurately and speedily at a rate of one plot per minute per raid.

5 R.D.F. stations will be required by the filter room controller to ‘bring in’ raids …

Information required by the three groups will be obtained by at least 15 R.D.F. stations, the Observers at which, in times of high raid density, will tell plots at a high rate. This information is to be filtered and passed accurately and speedily to groups and sectors simultaneously. This means that on the average when stations are all ‘bringing in’ raids, the group teller will have to tell the information received from 5 R.D.F. stations and will therefore have to tell information at five times the rate of the R.D.F. observers. …

Arguing backwards from the sector, a sector requires information accurately and speedily at the rate of one plot per minute per raid. If each sector can handle even four simultaneous raids, plots must be received at the rate of four per minute.

No. 11 Fighter Group has six sectors, Nos. 12 and 13 have three each. Hence filter room must pass plots at the rate of 12 per minute to Nos. 12 and 13 Groups, and at 24 per minute to No. 11 if all sectors are to be able to work to capacity. These figures include only the plots required by the sectors for interception and take, no account of more advance information concerning distant approaching raids, which must be given to fighter group to enable orders for interception to be passed to sectors. … It is estimated that, with the proposed method of handling the information, at least two and possibly three lines will be required between filter room, groups and sectors. The actual requirements may vary between groups.

Close on the heels of the exercises came the Munich crisis of September 1938, the five stations being put on continuous watch at 2.30 p.m. on the 26th. They gave warning of attack on London from Suffolk to east Kent at a range of about eighty miles from the coast.

The threat of war caused the building of chain stations to be accelerated and at a meeting at the Air Ministry on October 6th it was decided that ‘the R.D.F. chain be hastened so as to be complete by April 1st, 1939’. Compulsory purchase replaced the normal negotiation procedures for sites, and wooden towers were used where steel was not available. The building work went on night and day, and the new stations were under the supervision of No. 2 Installation Unit. This was formed at No. 1 M.U. Kidbrooke under Squadron Leader Rose, transferred from the Royal Aircraft Establishment.

Three kinds of stations were laid down at this stage, ‘Advance’, ‘Intermediate’ and ‘Final’. Advance sites had wooden huts, seventy-foot or ninety-foot towers, mobile or experimental equipment and no height finding.

Intermediate sites had 240-foot towers, experimental or mobile equipment, improved aerials and height finding.

On the Final stations buildings were protected, the aerials were contained on 350-foot steel towers for transmission and 240-foot wood towers for receiving. Range, direction and height finding were provided together with high-power transmitters and anti-jamming devices. All main equipment was duplicated and four wavebands were to be provided to avoid interference.

In the event only the 10–13.5-metre wavebands were used and the other three abandoned.

The M.B. (mobile base) radar became of increasing importance for Advance and Intermediate sites, for overseas requirements and for replacing chain stations damaged by air attack. The waveband chosen for the mobile stations was under ten metres, and to maintain the range a powerful transmitter, the M.B.2, was used.

Work developing mobiles progressed while Bawdsey was heavily committed on fixed R.D.F. and was solving the many problems of gun-laying, shipborne and airborne interception radars.

Immediately after the Munich crisis it was decided that the filter room must be moved to Fighter Command at Bentley Priory. Accordingly Squadron Leader Hart with several technicians built a rudimentary filter room in what is now the bar at the Priory.

A rough outline of south-eastern England was made in chalk on the floor and five telephone jacks and a wooden table for the map were installed. Within five days, on November 8th, 1938, the Fighter Command filter room was in use. The operators were drawn from those trained by Hart at Bawdsey, and in charge was Group Captain Rudd, then a flight sergeant. While filtering became the responsibility of Fighter Command the operational research behind it went on at Bawdsey. On December 19th, 1938, a provisional scheme was laid down for identification of raids appearing on the operations table.

In the instructions issued from Bawdsey raids first reported by the radar organisation were numbered by the filter room controller and told on to all concerned. On reception raid numbers were treated as provisional until they reached the ‘action line’ a few miles from the coast, or had been identified as friendly.

Raids undetected by radar and first reported by the Observer Corps were given a number by the controller, Fighter Command, although by 1940 this arrangement had been altered so that Observer Corps groups numbered raids originating in their territory, prefixed by their own code letter. Finally it was decided to use round plotting counters for friendly bombers, square counters for coastal aircraft, and triangular for civil machines.

Radar was fast becoming an integral part of the fighter interception system, but a great deal of work remained to be done before the whole organisation could work smoothly under war conditions. One of the many problems to be dealt with was that of estimating numbers from the radar blips. A good deal of experience had been gained in counting aircraft in numbers up to nine, but not above. To remedy this five small exercises were held by Bawdsey in November and December 1938. Formations of 11, 12, 17, 18, and 24 Blenheims were provided by No. 2 Group Bomber Command, performing a standard flight from the North Hinder Light Vessel to Orfordness or Bawdsey. On an ‘unknown’ run of 12 aircraft two observers estimated between 9 and 11, one between 10 and 12, and one exactly 12.

With the flight of 24 Blenheims on November 22nd, the response on the cathode ray tube was eight miles wide. The five distinct components in the total echo were each taken as denoting three or more machines. The civilian operators on the exercise little thought that within eighteen months W.A.A.F. and R.A.F. operators would be assessing multiple raids of 100+ with a high degree of accuracy.

A series of controlled interception experiments known as ‘lambs’ was made in October and November following an earlier one in April when an Anson intercepted a flying-boat—although this could hardly be called representative of fighting conditions. On October 14th a wireless telegraph control test was carried out. One Blenheim was plotted by four radar stations. This was followed by a similar flight on the 24th. Conditions on November 2nd were realistic, with much local flying off Felixstowe, which made selection of the ‘Lamb’ Blenheim difficult. Hart’s report on the run stated ‘by sheer luck the correct one was selected and interception was effected at 5,000 feet, twenty miles from the coast’.

An operational research run on November 24th proved less successful. Three Ansons of 220 Squadron were used, but detailed analysis of the run could not be made as unfortunately the photographer at the cathode ray tube turned the film the wrong way round.

Exercises continued throughout 1939 as the chain was extended, but results were patchy and continuous research was necessary to eradicate faults. In the coverage experiments in May affecting Stoke Holy Cross and West Beckham, Norfolk, Stenigot and Staxton Wold, Lincs., and Danby Beacon, Yorks., the report from Bawdsey, dated May 12th, read:

The general comment is that these stations under review are erratic; sometimes they give good results and sometimes bad ones. It is suggested that the cause of complete lack of performance be investigated immediately. While Staxton is so bad there is a huge gap in the coverage of R.D.F. on the Yorkshire coast.

Track tracings of Luftwaffe raids on Southern England on the morning of September 15th, 1940, as shown on the operations table at Fighter Command, Stanmore. The tracings show the location and time of the plots and raid progress as reported by the radar stations (via the Stanmore Filter Room) and the Royal Observer Corps. Raid serial numbers are shown in a circle at the beginning of the track. Several tracks are prefixed with the letter ‘X’ denoting a plot not positively identified as hostile

By this time the French armed forces were showing a keen interest in radar as their own detection system, D.E.M., was of little use. A mission arrived from Paris in April 1939, and on May 23rd was shown the underground filter room which had been brought into use three days previously. Two officers from each of the French services attended courses in radar in June, and plans were drawn up for C.H. radar stations to be erected at vital points in France, including Bordeaux. Nothing ever came of this, but the French Air Ministry gave manufacturers details of the equipment, although specifically requested not to by the British Government.

It is only surprising that when the Germans overran France they did not make good use of the information on the British radar system which was freely available there.

The war clouds were gathering, and on Good Friday 1939 the chain of radar stations known as ‘Chain Home’ or C.H. started a twenty-four-hour watch which was to continue until the end of the war.

The operational research work on control and interception carried out by such Bawdsey scientists as Messrs. G. A. Roberts, E. C. Wilhams and H. Larnder had not gone unnoticed. Park, then an Air Commodore and Senior Air Staff Officer Fighter Command, arranged with Rowe that the three men should, at the outbreak of war, be transferred to Bentley Priory as an operational research section, Rowe having already teamed them as such. This was to be the beginning of a vital contribution to ultimate victory and a principle adopted later by all the services.

Roberts himself had developed the answer to one of the most pressing problems of reading and passing radar information. In the 1937 and subsequent air exercises the radar station staff used a mechanical converter invented by Mr. L. H. Bainbridge-Bell. With this they had to convert range and bearing of an aircraft seen from the station into a position on the standard grid map used by the R.A.F. The calculations were involved and wasted valuable time.

The first large formation seen on the Bawdsey cathode ray tube during a counting exercise on November 22nd, 1938. The trace shows twenty-four Blenheim aircraft in line astern flying in from the North Sea. By August 1940 radar plotters were dealing regularly with formations of 100+

Roberts hit on the idea of using rotary switches and relays with metal fingers ‘wiping’ over contact plates to perform the functions of the calculator automatically when a button was pushed. For the prototype unit he obtained a number of Siemens switches destined for the obsolete army sound locator operations room in London.

The electrical calculator was a simple form of computer which greatly increased the efficiency and speed of radar operation. Inevitably, because of its comparative functions, Roberts’s brainchild became known as the ‘Fruit Machine’.

A further Bawdsey invention was I.F.F. (Identification Friend or Foe). This involved the use of a small transmitter in R.A.F. aircraft which gave a distinctive character to the blip observed in the radar station. In this way enemy aircraft could be distinguished from friendly machines as they approached the coast. I.F.F. became available during the first year of the war, but in the early stages gave rise to problems. Bomber crews were not told the purpose of the transmitter in their aircraft. They were loath to use it in case the Germans homed on to the signals.

The first taste of operational work came to the radar chain in May 1939 with the flight of Graf Zeppelin. This was closely followed by the German air exercises carried out by Luftflotte 2. One lunchtime the watchers at Bawdsey were amazed to see a big formation on the tube fifty to seventy miles away and heading in for the coast. The cry went up ‘The Germans are coming!’ and everyone crowded into the hut to see. Fighters were at that time under orders from Dowding not to fly over the sea and to intercept only if the enemy came within three miles. The observers gazed, fascinated, as the formation, estimated at about fifty, closed up to seven miles and then turned away, back to Germany.

Work went on unceasingly on the sites to get stations operating and calibrated in time for radar’s biggest test so far, the R.A.F. annual exercises held from August 6th to 8th.

For the first time radar plots were married to those of the Observer Corps and warning was given to the Observer operations rooms of aircraft approaching from the sea. ‘Lost Property Offices’ were set up at Observer groups to show any discrepancies between radar and visual plotting as the Corps was allowed to produce tracks only on aircraft picked up by radar. One such ‘Lost Property Office’ was run by Mr. G. A. Roberts and Mr. E. Fennessy, another Bawdsey research scientist (now Group Captain Fennessy).

From the radar side the exercise was very successful, and vindicated four years of faith and hard work. I.F.F. was tried out for the first time, and as a result 25,000 sets were ordered for the R.A.F. and the French Air Force, although the latter were not delivered before the collapse in June 1940.

Opposite C.H.L., or Chain Home Low, a rotating aerial radar used in 1940 to search for low flying raiders. The C.H.L. network was built up in a very short period of time and the sets were based on a 1½ metre aerial array developed for coastal defence

A German photograph of the Dover Chain Home radar station taken with a 5-minute exposure on an infra-red camera at a distance of 22 miles. The various buildings silhouetted against the masts are, in fact, a long way in the foreground

The imminence of war meant that even greater scientific effort would be required at short notice. The Royal Society therefore classified the scientific manpower that could be utilised in Britain. Leading scientists from all over the country were shown over the military research establishments, including Bawdsey, and university graduates spent the summer vacations familiarising themselves with the work.

On September 3rd, 1939, the nation began to call on these brains, young and old alike. Many were surprised within six months to find themselves in uniform and commissioned, a number subsequently rising to senior rank.

The declaration of war meant the break-up of the Bawdsey team. The site was considered too vulnerable and it was feared that it might be the subject of an early heavy attack. One section of the team, under Dewhurst, went to Leighton Buzzard, Larnder, Williams and Roberts began operational research at Fighter Command, A. P. Rowe and others set up shop in Dundee, and the War Office section was housed at Christchurch.

The Leighton Buzzard establishment became No. 60 Group R.A.F. in March 1940 and took over all the radar stations which were officially known as ‘Air Ministry Experimental Stations’ or A.M.E.S.

Before Bawdsey was dismembered trials were in progress with coastal defence and gun-laying radars. The former, known as C.D., measured ships’ ranges with an accuracy of about twenty yards. This was achieved by producing new high-power valves, a special aerial array, and using one-and-a-half-metre wavelength. Much of the credit for this went to Mr. C. S. Wright, director of scientific research at the Admiralty.

The success of C.D. radar reached the ears of the Air Ministry. They immediately saw in it the answer to the problem of low-flying aircraft. Relabelled C.H.L. or Chain Home Low, the equipment was ordered on a top-priority basis.

Basically C.H.L. was a searchlight of rays from a rotating aerial, the beam being moved to search for aircraft. Thus it differed from the C.H. conception where fixed masts radiated a floodlight of beams over a wide area.

Professor (later Sir John) Cockcroft from Cambridge took over the C.H.L. programme in the autumn of 1939, cutting all red tape in the process. He requisitioned sites and ordered buildings without the usual preliminary form-filling. This resulted in C.H.L. cover being completed in time for the Battle of Britain, while the Air Ministry accounts department took years to catch up on the bills.

The first C.H.L. was opened on November 1st, 1939, and it gave a range of 100 miles using a Metrovick transmitter giving 100 kw. power to the aerials. C.H.L. was subsequently used for both low-flying aircraft and coastal shipping detection. Coloured counters were used on the filter-room table to denote C.H.L. plots.

As the C.H.L.s were only planned to complete the low cover from C.H. radars, they were linked by telephone with the C.H. stations and their plots passed on from there.

Throughout the first nine months of war the work of rebuilding and modernising C.H. stations, installing ‘fruit machines’, new aerials and anti-jamming devices progressed. Production of mobile sets was stepped up to meet emergencies.

Considerable difficulty was found in calibrating stations. Friction arose between Fighter Command and the scientists when stations were not operating while undergoing vital modifications and adjustments. Cierva autogyros were brought in as radar calibration aircraft. Trials of all types continued with Poling, near Arundel, used for interception tests, and Ventnor in the Isle of Wight available for long-range exercises.

There were many disappointments, particularly in the length of time taken to get radar plots through the system. One of the main bottlenecks was on the Fighter Command filter table. Early in 1940 three technical assistants with science degrees were given filter training and took up duties at Bentley Priory. The improvement in establishing tracks was immediate and the time lag reduced.

As a result the standards for filter officers were raised, and establishment made for commissioned filterers. Later, liaison officers from the various commands were posted to the filter room to aid in the identification of tracks.

When the Battle of Britain began there were twenty-one operational C.H. stations and thirty C.H.L.s. The only gaps were in the north-west of Scotland, the Bristol Channel and a portion of the Welsh coast, where neither type of radar was installed.

By dint of continuous training exercises and streamlining of techniques a very high state of efficiency was reached. The radar chain provided a long-range ‘picture’ of the air situation without which defence was almost impossible.