PA0SE

pg 10a INTRO to German WW2 Radio equipment
(Radio Bygones No65, June/July 2000)


German World War II Radio Equipment
By Dick Rollema PA0SE

This article is based upon two papers: 'Some Aspects of Unconventional Engineering during Interbellum - as particularly in Germany to enhance the specifications of their communication apparatus' and 'Receiver and transmitter development in Germany 1920-1945', both written by Arthur O. Bauer, PA0AOB, president of the Foundation Centre for German Communication and related Technology 1920-1945. The B/W photographs were taken at Arthur's beautiful collection. His help is gratefully acknowledged (The colour pictures are taken by LA8AK).

HOW IT BEGAN
Development of radio equipment in Germany at first went along the same lines as in other countries. Components were mounted on wooden baseboards and behind a front panel of Bakelite or similar material. More expensive sets featured a cabinet, also made of wood. Electrical shielding was hardly used and coils were of the open ('air cored') type. For a reasonable Q coils had to be of large size and were sometimes wound with 'litz wire', made of a bundle of individually insulated strands. Also special winding techniques were employed to reduce self-capacitance of the coils.
By the end of the twenties a different approach was initiated in Germany. There were several important developments that made this possible.

NEW CERAMICS
Ceramic insulation had already been used for many years but only for DC and low frequency AC.
For frequencies above 1 MHz dielectric losses in the material were too high. This was
changed in the second half of the 1920s when German firm Hermsdorf-Schomburg-Isolatoren-Gesellschaft,
also known as Hescho, marketed new types of ceramic material with very low loss at RF, due to the absence of iron.
At first Hescho had problems in measuring the loss. But the solution came from young physicists Dr Rohde and
Dr Schwarz who developed suitable test equipment. That was the start of their successful firm
'Physikalisch-technisches Entwicklungslabor Dr Rohde & Dr Schwarz'.
After the war well known as 'Rohde & Schwarz' (R&S). Their first export order
came from Britain in 1934 for a 'tan delta measuring device'.
After 1933/1934 low loss titanium dioxide capacitors with controlled temperature constant became available and these were especially suitable for frequency stabilisation of free-running oscillators, as will be seen later.



Temperature stabilization block of capacitors (EZ6) is shown on uppper left side of the picture)

Lo40K39a and Lo40K39d temperature compensations are somewhat different.

Ceramics also made possible very good coils, manufactured by vaporising and burning a heavy copper and/or silver layer onto a ceramic cylinder. Parts of the layer were then cut away such that a helix remained, forming the coil winding.
Stable coils with high Q could thus be mass-produced. The construction also resulted in a very low temperature coefficient, equal to that of the ceramic material and not of the metal of the winding. Siemens reported temperature coefficient of such coils was up to 200 times lower than that of the best conventionally made coils.



Ceramics also made possible very good coils, manufactured by vaporising and burning
copper and/or silver layer onto a ceramic cylinder. Parts of the layer were then cut away
such that a heliz remained.


IRON DUST CORES
Coils with iron dust cores were already used by the telephone industry in the early 1920s, for instance as Pupin line loading coils, but they were not suitable for frequencies above about 10 kHz.
A great improvement was made by versatile inventor Hans Vogt who had the smart idea of using 'carbonyl iron'. The advantage of this material was that the iron dust particles were spherical and only between 2 and 6 micron in diameter, with a very thin oxide film on their surface. This avoided eddy currents and the associated losses. In the thirties the material became known in Britain and Germany by its trade name 'Ferrocart'.
It made possible compact coils of high quality that could be easily screened by a small can without affecting their Q.

DIE-CAST FRAMES
After the radio on a wooden baseboard the use of a metal chassis became popular, and it remained like that till the appearance of the printed circuit board.
Early in the 1930s C. Lorenz company (since May 1930 owned by ITT, after WWII known as Standard Elektric Lorenz until the end of the 1980s; now owned by French Alcatel) was looking for more rational construction techniques. Until then the chassis moved all the way from the metal workshop up to the testing station at the end of the production line. So only a few people could work on it at the same time. Lorenz wanted a construction consisting of separate modules which in the final stage of the production line could be bolted or clicked together. Such modules could be manufactured at the most suitable production site, even being tested and calibrated there.
The use of die-casting (Spritzguss) was soon found to be the answer. For their application, involving small and relatively complicated shapes and very tight tolerances, Lorenz sought support from outside. That was found at the firm of Mahle, manufacturer of aluminium pistons for internal combustion engines. According to an advertisement in a German technical magazine of 1938, Mahle could produce die-cast samples of 0.5-2000 gram and tolerances of a few hundredths of a millimetre!
The alloy used became known as 'Elektron' and consisted of about 8.5-9.5% Al; 0.5% Zn; 0.2% Si; 0.2% Mn and the rest Mg. It had a specific gravity of only 1.8.
During the war, shortage of aluminium forced the industry to replace the lightweight Elektron by an alloy based on zinc. After mid 1943 the German Army received more and more equipment made of the much heavier zinc alloy. Only the Luftwaffe (Air Force), the best equipped of the three forces, retained the high quality Elektron for their radios.

LIMITED NUMBER OF VALVES
The German air force, army and navy were trying to standardise their components as much as possible. A problem was the large number of different valves used in communications equipment. It was therefore decided to force the industry to integrate and co-ordinate their activities on commercial and military projects. This resulted in a completely new generation of radio receiver and transmitter valves. Also the number of different types was greatly reduced. Several military receivers used only one type of valve in all stages. This was the universal pentode RV12P2000 (Photo 1). R stood for valve (Rohre); V = low power, 12 = heater voltage; P = pentode, 2000 = amplification factor (mu). More than 16 million were produced during WWII. The valve is inserted upside down into its holder, which completely surrounds it. To withdraw the valve a knob, visible on Photo 1, is screwed into the bottom.
In the later to be described aircraft radio FUG 10, as well as in many other transmitters, pentode RL12P35 with 35 watt anode dissipation was used. Details of this valve can be found in 'The German 807s? The RL12P35 and its derivatives' by Alan Davies, RB49, October/November 1997.
The author still uses three of these valves in parallel in the final amplifier of his home-made amateur single sideband transmitter, built in 1967. Though in frequent use, over all those years only two of them had to be replaced, in one case the cause being an open contact at the antenna connector, resulting in severe overloading of the valves. The RL12P35 was designed in 1935 and never meant for linear amplification but the valve nevertheless performs very well in this mode. Between them the three develop some 130W peak envelope power with very low distortion. The author presently is constructing a new transceiver for the amateur bands 10-160 metres. This will be transistorised except for the final amplifier which again has three RL12P35's in parallel!



Universal pentode RV12P2000 (RV2,4P700), socket, and the special type "AF100"


AIRCRAFT RADIO EQUIPMENT FuG X (= FuG 10)
In August 1936 the Reichsluftfahrtministerium (German Air Ministry) issued new specifications for radio equipment to be used on board aircraft. It had to cover long wave, short wave and even VHF in the future. It was to be standard equipment for more than a decade. The specifications were very tight: Frequency should stay within 3 x l0 E-4 of the set value over a temperature range of -50 to +50C and voltage supply variations between 22 and 29 volts. Evaluation of a set in the USA showed frequency stability to be even better: within 0.001 over a temperature range of -30 to +50C. Whether power supply voltage was also varied is not clear, but the American report stated ". . . This is considered very good stability ... Both transmitter and receiver units operate satisfactorily under conditions of vibration."
Both Lorenz and Telefunken competed for the contract but Lorenz eventually got it.
The FuG 10 (FUnk Gerat 10), as it was designated, exemplifies the principles set out earlier and we will therefore have a closer look at it.
The following list shows the incredible pace at which the project was completed. But then Germany was preparing for the Third Reich . . .
1936, August: Specification issued by the Reichsluftfahrtministerium RLM.
1937, February: Two prototypes, in conventional construction, handed over.
1937, February: First flight trials in a JU 52 aircraft (the one with corrugated sheet fuselage and three engines).
1937, July: Two die-cast prototypes available for trials.
1937, July: Two Lorenz sets competing against sets by Telefunken.
1937, August: RLM decides to place the order with Lorenz.
1937, December: Lorenz supplies eight complete installations to the Luftwaffe.
1938, January: Mass production started.
1939, September: All long distance Luftwaffe aircraft equipped with FuG 10.
Up to the date of Germany's defeat 50,000 systems had been delivered, comprising a total of 300,000 units.
The frequency stability requirement could have been met easily by using quartz crystal control. But Germany had no quartz of its own so that had to be imported. This was against the orders of Hermann Goering who was not only commander of the Luftwaffe but had also been given the task of making Germany self-supporting (autarky). As a result quartz crystals were only used for frequency calibration and in IF filters. Less than a million crystals were produced in Germany, against 30 million in the USA alone over the period 1941-1945!
So frequency control in FuG 10 had to be by a free-running oscillator. But the working conditions could hardly been more unfavourable. The transmitters of the FuG 10 were of the MOPA type (Master Oscillator Power Amplifier = VFO/PA). In the final amplifiers two RL12P35 pentodes were in parallel. The oscillator used the same 35W pentode in order to provide sufficient driving power for the final amplifier (Photo 2). A good solution from a logistic point of view, hut a technological nightmare for the development engineers. Almost all German transmitters used grid block keying of all stages. This made full break-in possible. Therefore the VK) stage would only warm up during transmissions. The stage never reached a stable temperature, due to the intermittent operation. The internal dimensions of the valves in the VFO and final amplifier, and with them the inter-electrode capacitances, were constantly changing, resulting in frequency drift.
The recently developed ceramic capacitors, having controlled temperature coefficient, opened the way to counter the frequency drift. But they could not follow changes in environmental temperature immediately due to their thermal lag. This problem was solved by increasing the surface. Where, for instance, a 100pF capacitor was required ten capacitors of 10pF in parallel were used and mounted at strategic locations within the transmitter. However, this did not help in the case of the inter-electrode capacitances of the valves that changed quickly when the Morse key was operated, but Lorenz also countered this problem. There were available not only capacitors with a controlled temperature coefficient but also ones with a controlled loss resistance. Capacitors of this kind were included in the tuned circuit of the oscillator. The RF current flowing through them during transmission heated the capacitors in step with the valves and so frequency drift was sufficiently eliminated.
The low air pressure in a high-flying aircraft can cause sparking in the variable capacitors of a transmitter. This can be avoided by increasing the spacing between the plates, but this also increases the dimensions of the capacitor. This was not a viable solution for the FuG 10 transmitters, covering 300-600kHz for the long wave (S10L) and 3000-6000kHz for the short wave transmitter (S10K), packed in a cabinet of only 210 x 220 x 220mm and delivering 70-80 watts to the antenna. Here the invention of the dust iron cores by Hans Vogt brought the solution. Instead of varying the capacitance of the tuning circuits these were fixed and the inductance varied by using variometers. Inside the sphere-shaped coils on their ceramic formers were cores of sintered dust iron. The variometers of the master oscillator and final amplifier were ganged and tuned by a single knob.
The receivers of the FuG 10 installation were also housed in cubes with dimensions 200 x 200 x 190mm. They were high-grade superheterodynes with 11 valves of the universal pentode type RV12P2000.
An important aspect of the FuG 10 was its serviceability. Complete units and/or parts of it could be changed in very little time by minimally trained technicians. All modules could be fixed to the mounting frames by turning two fasteners through 90 degrees. The units - receivers, transmitters and all other parts - were automatically connected via flat cables, similar to the ones now found in computers, to the junction box, fixed to the fuselage.
For ground station use the units were fitted to a frame that can be seen in Photographs 3 and 4.
RB

More about FuG10 transmitters on page 30e


Receivers, transmitters, remote control box for the antenne tuning unit and
other units can be easily removed from the frame by loosening two fasteners.
Electrical connections are made via multipole plugs and sockets. [FuG10].




Some socket connections




Die-cast boxes were also applied in BC-equipment just after WWII, here is
shown Telefunken UKW1C "VHF add-on". Click on the picture for more details.


e/m


BACK

Last update: 2005.01.28