Naval History Homepage and Site Search

  World War 2 at Sea


Links to Main World War 2 Pages:
- Royal Navy Organisation
- Casualties - killed, died, missing
- 1,000 Warship Service Histories
- Convoy Escort Movements
- Battles, Major Warship Losses
- Naval & Military Campaigns
- Navies Daily, 1939-42
- Admiralty War Diaries
Aircraft carrier HMS Argus, one of the first British warships to carry out trials with HF/DF (Navy Photos/Mark Teadham) return to Lt-Cdr Mason's researches
or  World War 2, 1939-1945

 by Geoffrey B Mason, Lieutenant Commander, RN (Rtd) (c) 1992



Land and shipboard high-frequency radio direction finding (HF D/F or Huff Duff), is now generally recognised as being with radar and code-breaking, a primary factor in the allied victory in the Battle of the Atlantic against the German and Italian submarine onslaught against shipping. Perhaps this is evenly balanced? However little has been published about the background of the work carried out after 1918 to provide this capability. This Paper sets out to record some of the various steps taken at HM Signal School, Portsmouth to make this possible. It seems essential that some details must be given of the transfer of effort from the improvement of direction finding on the lower frequency transmissions then in use, to the provision of the same service on signals over 2 MHz. Theoretical explanations and detailed technical descriptions have been kept to a minimum as this type of information is available in the references.



The Inter-War Years


This Paper sets out to record some of the various steps taken at HM Signal School, Portsmouth to make this possible, despite the limited resources and lack of appreciation of the value of this operational capability. It  does however seem essential that some details must be given of the transfer of effort from the improvement of direction finding using the radio transmissions on lower frequencies, then in use, to the provision of the same service on signals over 2 MHz. 1.


Use of radio equipment for communication was well established before the outbreak of war in 1914. Part of the development of this capability had shown that it was possible to determine with some accuracy, the relative bearing of a radio transmitter whose signals on lower frequencies were being received. However further investigation by HM Signal school before the end of hostilities in 1918 had not provided any significant results.2  Despite the lack  priority given to other work, such as production of suitable electronic valves for use in radio transmitters on shorter wavelengths, work was continued to further develop equipment for direction finding at higher frequency transmissions.


It had been established that the ‘Bellini-Tosi’ system was best suited to use on board ships. The most important element in any Radio Direction Finding equipment is the design of the aerial system used. Between 1920 and 1930 several different designs were used. These ranged from use of two frame coils fitted within the bridge structure of large ships which were later replaced by two large loops. One was of rectangular shape in the fore and aft plane and the other of triangular shape was fitted athwartships. Both were sited in positions as high as possible in the foremast structure, rigged at right angles and bisecting each other. The total area covered by these some what extempore arrangements was about 2,500 Sq. Ft to suit reception of long wave signals at any distance. It was recognised that effect of adjacent land masses on transmitter sites and also that of atmospheric conditions at Sun rise and sunset had to be considered. 1


The various alternatives using different type of shaped loops can be identified on photographs of many contemporary warships including HM Battleship WARSPITE 3. In early trials, some warships these large aerials were at lower level and fitted between the funnels or in one funnelled ships between the funnel and bridge structure. The beam loop was suspended from the triatic stay and connected to stump masts or booms. This arrangement presented difficulties in the physical handling of the portable lengths of wire which had to be kept rigid using bottle screws2. Operational efficiency was affected by the effects of radiation from the main transmitting aerial and bearing errors were also caused by proximity to fixed structures as well as the movement of boats or gun mountings. Different sizes at varying heights in and on both masts may have been used in some ships for conduct of trials3.


The main advantage of fixed rigid loops was to avoid the problems of rotating such a large assembly. It also would suit reception of signals transmitted from positions over a long distance. Although the aerial circuit could be tuned this would require adjustment for each received signal and introduce large errors since the exact size of each loop would not be identical. Use of tuned aerial circuits was not continued.


Signals received by aerials of this design would be of maximum strength when the loops were pointing directly towards the position of the transmitter and be at their minimum when pointing 90o  away from its actual direction1.T his had to established by turning the ships through a full circle which was inevitably a long process and could involve several time-consuming operations involving significant man-power. It follows that by use of two loops there are two positions in which maximum and minimum signals from the bearing of the transmitter site are received. In order to overcome this ambiguity the two loops are connected to a Sensefinder1. This is an angle measuring device, known as a ‘goniometer’, fed by a separate input of the same signal.  Accuracy of the determination of the actual bearing of the received signal depends on there being no mutual coupling between the loops and that the received aerial current fed from the loops produces a uniform magnetic field1.


A goniometer comprises two fixed Field Coils, connected to the two loop aerials and a separate rotating Search Coil fed from a separate Sense aerial. This Sense signal will not vary and is 90o  out of phase with that from the loops. As result of this, the combined signal sent to the main receiving circuits is modified by movement of the rotating component


There are two positions when the received signal is at maximum during one complete rotation. When turned clockwise the true bearing as indicated on the Bearing Indicator Dial is being received. However due to the combination of the Sense aerial input with that from the two loops, the reciprocal bearing is identified when the signal strength increases (See Figure 21 in Reference 1).


Transmissions on frequencies about 600 KHz were a more complex matter because of the effect due to movement of layers in the ionosphere. Use of a super-heterodyne receiver was more efficient at higher frequencies. Shore trials at a special establishment exclusively for aerial design were carried out first. During trials at sea using one of the loop aerials to instead of a separate Sense Aerial proved unsatisfactory compared with results obtained when a completely separate wire aerial was used.1. Amongst the factors considered during these early trials was the effect of coincident wireless transmissions on trials ships when direction finding was being carried out and these were forbidden. It was also found that during reception of signals on higher frequencies, significant bearing errors were caused by ship’s rigging, ship structure and the movement of gun mountings or boats. In addition it was discovered that the physical length of the ship also had an influence on the accuracy of bearing errors on frequencies whose wavelength was related to the received signals. It was therefore evident that improvement had to be made to the aerial design in order to reduce bearing errors although this could not be eliminated on some frequencies2


After 1930 a single rotating loop aerial of circular design were installed in some warships3 (HMS BARHAM) instead of these large loops. It was then realised that smaller ‘diamond shaped’ rotating doubled loops would be more successful on higher frequencies than the some what extempore fittings previously used on large ships. Clearly they would also be much more suitable for fitting on smaller vessel, especially destroyers. This proved to be most significant step and the fixed Frame coil design used had two rotatable smaller loops and became a standard fit in TRIBAL Class destroyers on build before September 19393. The final design used one of two different designs of aerial units, identified as Frame Coils S16 and S17. These were of different sizes, that of S16 being 3ft square and S17 of 4ft 6” square. The size of aerial used depended on the structure of structure of particular ships mast. The aerial unit was mounted on a pole mast fitted at the top of foremast of the ship. Both Frame Coil and Pole-mast could be struck using a separate beam to lower or raise the complete assembly when necessary to allow passage under a bridge or for replacement.


Much more attention was paid to further development of H/F DF after 1931 and trials were carried out in HM Cruiser CONCORD using a rotating aerial fitted on the roof of the spotting top sited at the top of the foremast. Results using high frequencies revealed errors in bearing varied on various frequencies were due to ship’s rigging, and the length of the ship. In addition, efficient earth connections had to be used for the aerial cabling. It was essential to site the receiving equipment as near as possible to the aerial loops and the receiving equipment was fitted on the mast structure for the installation in HMS CONCORD2.


Trials were carried out on an aircraft carrier, probably HMS ARGUS in 1938 for further investigation. As this vessel had no masts or rigging it was possible to establish the effects of factors other than re-radiation from these. By this time an improved design of goniometer to suit use on higher frequencies had been developed. Installation of D/F equipment had been completed on many other warships before the outbreak of war. HM Battlecruiser RENOWN had fixed loops; HM Battleship BARHAM retained fixed loops of the mainmast. as well as the fixed loops on the mainmast.  HM Cruiser AURORA also had two rigid loops whilst HM Cruiser MANCHESTER was fitted with a rotating Frame Coil similar to the TRIBAL Class destroyers. Soon after the outbreak of war HM Cruisers EURYALUS and BONAVENTURE had rotating frame coil aerials3. In all these ships, existing receiving units, modified to accept separate inputs from two loops, were used with Sense Aerial Tuners, Heterodyne Receivers and Amplifiers. The receiving circuits of the these items did not incorporate provision for elimination of ambiguous bearing indication when used for higher frequencies because of the errors caused by structure and other factors. However all low frequency receiver units used a goniometer.



D/F in Submarines


The trials after 1930 included work on submarines which particularly relied on reception of long wave transmissions. Initially, frame aerials sited within the conning tower structure were used and it was found that these good results on low frequencies. Subsequent trials were carried out using a two fixed loops attached to the periscope but this was replaced by fixed loops fitted on the hull of the boat aft of the conning tower. The main aerial used for the main transmitter was used for sense finding3.



Developments after 1939


Before the outbreak of war in 1939 it had been decided that D/F aerials used for frequencies above 2MHz must be sited at the highest point possible to minimise errors due to the surroundings. A new design of Bellini-Tosi aerial using two rigid loops was required with the Sense aerial positioned at exact centre of  its framework.  A counterpoise ‘earth’ was incorporated in the design of the aerial unit4. By introducing an adjustable capacitance it was possible to reduce amount of current induced in the Sense Aerial by the support mast thus reducing the amount of unnecessary content provided to the receiver for sense finding.


This was a most vital discovery and enabled an operationally effective system identified as FH1 to be produced in HM Signal School.2 During March 1941, the first RN HF D/F Outfit was installed in HM Destroyer HESPERUS as Outfit FH1 after the completion of trials on the new Frame Coil S25 in cruisers4. Further trials had to be arranged in convoy escorts with priority for fitting given to those deployed for ant-submarine defence. However the selection of position introduced conflict between ship designers and users which was hard to resolve satisfactorily. In both cases, choice of the site to be used had to take due account of the arcs of fire of weapons and the positions of other wireless aerials. Mandatory requirements associated with topside stability and electronic interferences, in the early stages of development, made it impossible to fit both radar and Huff Duff on the same ship2. Most escorts had the unit fitted on the mainmast aft in order to enable a surface warning radar set to be fitted on the foremast. In post war years the aerial was generally fitted on a pole mast forward, above the radar aerial. An improved version of FH1 was introduced during July 1941, identified as Outfit FH3 and installed in HM Destroyers GURKHA and LANCE2.  These equipments all used the ‘aural-null’ method to establish the true bearing of signals but work was in hand to introduce a method of visual presentation instead. This would enable less skilled operators to rapidly establish and interpret the true bearing of signals received. Twenty five escorts and rescue ships had been fitted with FH3 by January 1942.




Distance from the transmitter was impossible to determine accurately, but operators soon learned to distinguish HF ground waves from sky waves. Since ground waves could only be detected 12 to 14 miles from the transmitter, FH3 operators knew when an intercepted signal represented a dangerously close U-boat. The FH3 incorporated the B21B receiver which had a frequency coverage from 1 to 20 Mc.  By the end of January 1942 25 escorts and some rescue ships were fitted with an improved version.


The improved version, FH4, used a cathode ray tube display and was fitted in HM Cutter CULVER, the ex US Navy Coastguard Cutter USS MENDOSA during October 1941 Production of the FH4 Outfit was developed and produced by Plessey Company which included a ‘twin channel’ receiver developed by the company.


In this improved equipment each of the loops were connected to separate identical amplifying receiver units which could be balanced for phase shift and gain in order to provide outputs to the deflecting plates of the cathode ray tube.. This ensured that the resulting display would show when any phase differences existed in the signals being received from the two loops. If ‘in phase’ a single line will be displayed, but if the phase from one loop differs an elliptical trace will be shown., A Sense aerial is incorporated in the design of the Frame Aerial, S25B as shown in the diagram4 and ensures that the bearing display is ‘true’5. The visual display of signals made it easier to distinguish ground wave signals since these were much stronger and the length of the displayed single line would be greater and indicate the range of the position from which the transmission was being made. This was a vital factor in successful defence against submarine attack as U-Boat transmission were of very short duration and visual display provided in FH4 was easier for an experienced operator to assess. Although Sky wave signals reflected from ionospheric layers were of lower strength and made bearing and range determination unreliable. Ground wave signals at ranges up to 12 miles made both factors much easier to determine since the signals were much stronger. A Test facility is included in FH4 design to balance the operation of the two receiver systems.



right - HD/DF aerial believed on destroyer escort USS Francis M Robinson DE220 (US Naval Historical Centre)


The first of 30 production models was available in March 1942 and fitted on board HM Destroyer LEAMINGTON (Ex USS TWIGGS.) and later that year on a few other convoy escorts and rescue ships. This outfit because a standard fit in all RN vessels requiring a direction finding capability and was in use for the rest of the 21st century.2. During service this equipment also proved invaluable for other use than convoy defence. It was used for tactical purposes to establish positions of ships in a Task Group which were required to change position during a current operation. For example, during the Home Fleet operation in December 1943 to intercept the German battle cruiser SCHARNHORST HF D/F was very effectively used to engage the enemy ship and also cover the passage of vital supply convoy. It also provided a more reliable method of establishing ship’s position from Shore Beacons when in poor weather.





Shore stations


The development of suitable equipment for locating the position of U-Boats which proved to be essential requirement in meeting  the threat to convoys. The bearings of transmissions by U-Boats could be determined by two or more shore stations and passed to the convoy escort commander. This information enabled course of convoys to be changed to avoid the threat and for escorts fitted with HF D/F to search for transmissions and carry out search and destroy operations by following the bearings indicated by transmissions. Soon after the more important work by HM Signal School on ship equipment allowed it was decided that the most suitable aerial design  for shore station equipment would be an Adcock system. Vertical dipoles were used for detection of signals on both low and high frequency transmissions Initially wire aerials suspended from high masts were used for earlier outfits but later arrangements used smaller rotating aerials sometimes in mobile vans, After introduction of the S25 Frame Coil for FH4 this was used at shore stations and in mobile units for high frequency requirements.



HF  D/F Research and Development in allied countries


In France, pre-war R&D on Huff Duff was carried out at the Laboratoire Telephonique in Paris, a subsidiary of International Telephone and Telegraph (ITT).This work paralleled  that of the British, with whom the French shared electronic secrets until the Germans overran France in May 1940. Some of the engineers were able to get to the USA via Spain early in 1941. They were allowed to work on naval HF D/F equipment, but were not fully trusted.


The US. Navy, experimenting with its own shipboard Huff Duff models as well as the British FH3, was slow to embrace the new ideas. Like the Royal Navy, US warship captains preferred radar to ‘Huff Duff’. However by 1942 after comparative trials between use of FH3 and an a US designed outfit, a US design, Outfit DAQ was chosen. Over  four thousand outfits were produced and fitted in all new destroyers and escorts.


Once the development of satisfactory equipment for use at sea on high frequency transmissions had been completed it was possible to begin work on development of suitable outfits for use an Very High Frequencies (VHF). There had been a most significant increase in use of VHF equipment in aircraft and in escorts for communication during convoy defence and other operations. In aircraft carriers, immediate knowledge of positions of ships and aircraft was essential and made possible by the availability these new facilities. By 1945 direction finding outfits for use on VHF transmissions was available on some aircraft carriers deployed in the Far East and proved invaluable for navigation and the safety of aircraft during Fleet operations as well as in inclement weather. Several Dipole aerials arranged in a circular formation are used for these D/F outfits.



Post War Development


In the immediate post war years direction finding on Ultra High Frequencies (UHF) was successful. This band is now in general use for all inter-ship communications as well as for all aircraft radio communication requirements. As a natural progression similar progress has been made on the provision of equipment suitable for use on radar transmissions and is now widely used for many defence requirements. Since the much shorter wavelengths are used four individual ‘Horn Type’ aerial units are used for each different radio frequency band. These are installed in the Fore and Aft and Athwartships planes.





No better confirmation of the satisfactory development of HF D/F equipment in Great Britain before and during WW2 is the German admission after September 1945 that their own work on this particular research had been inadequate. It was admitted that during WW2 they had experienced great difficulty in locating the position of U-Boats. Pre war work in UK before September 1939 had been severely restricted by the resources provided by the Admiralty despite the recognition by a few senior officers that any work on this subject was an obvious essential to the conduct of war at sea..






2. AS WE WERE Unpublished account of the work of HM Signal School 1896-1946

    (Later Admiralty Signal and Radar Establishment.)

3. Photographs showing Aerial Units

4. Drawing of FH4 Frame Coil (S25).


Other Sources:


HITLER’S U-BOAT WAR by C Blair, Volume 1, Appendix 8 (1999)

SECRET WEAPON by KM Adams (1996)

RADAR AT SEA by D Howse (1993)




The Curator, Radar and Communications Museum, HMS COLLINGWOOD, Fareham, Hants.

return to Lt-Cdr Mason's researches

or to World War 2, 1939-1945

revised  5/12/10