Digital Technologies: Digital Fire Control
Digital Fire Control (1947 – 1960)
A fire control device for anti-aircraft artillery is a calculator that determines the probable trajectory of an enemy aircraft from a series of positional observations. Subsequently, this device calculates the point on this probable trajectory at which a missile must be fired for the missile to hit the aircraft. The data regarding the position of the aircraft is received by radar in the form of angles and distances. The direction in which the missile has to be fired has to be transmitted to the artillery in the form of angular dimensions. It was natural to build such a calculator in a so-called analogue form. The Hengelo-based N.V. Hollandse Signaalapparaten (HSA) was already very successful in constructing such fire control devices before WW II.
Digital computing technology is in its infancy
In 1947, research was conducted at the Physics Laboratory into digital computing technology, a technique still in its infancy. Initially, the digital technique offered little chance of application on a larger scale because of the many electron tubes required. In 1951, a small digital calculator was constructed that could add two numbers provided the resulting sum was less than sixteen. The whole laboratory staff wanted to have a look. According to Van Soest, we could now do digital calculations. Therefore, the laboratory could develop a digital fire control system for the Armed Forces.
Determining the correct position of underwater targets (DICO)
In 1952, the Laboratory developed the Depth Corrector (DICO) calculator device to determine the correct direction and depth of underwater targets (submarines). When calculating the depth, the influence of the acoustic refraction due to the sound speed gradients in the water has to be accounted for. Sound transport in water is not linear but is influenced by temperature differences, differences in salinity and pressure according to the Snellius laws. Temperature gradients and salinity were measured in advance at different depths. Their values were entered as parameters on the DICO.
DICO used electron tube technology: 20-bit wide computations using 370 ECC32 electron tubes and 50 voltage reference tubes (CV2213). These consumed 1,500 Watts (120 Amp). The DICO had five 20-bit words of magnetic core memory and three registers. Conversion of the chart angle and elevation parameters was ‘pre-programmed’ as digital numerical values that were stored as digital values in an optically readable metal code disc (1953). A multiplication took 850 microseconds.
An ECC32 electron tube contains two triodes. Money was saved by using just one filament for the two triodes in the electron tube. But if one triode fails, the tube becomes useless. Some two to three triodes failed each day. As all triodes were functionally necessary, an ECC32 with a failing triode was replaced by a new tube. After a year, all the tubes had been replaced. An employee collected the still half-working tubes for private projects …
In June 1954, DICO was successfully tested aboard the HNLMS Marnix (D807). Between November 1955 and February 1956, DICO was used aboard the destroyer HNLMS Gelderland (D811). Some adaptations followed. In 1956, Hollandse Signaal Apparaten started to make a production series of the DICO amongst others for the German Navy.
Digital fire control for torpedoes
The fire control radar, which determines the direction of a target (ship, aircraft) provides this information in the form of two angles, namely the rotation in the horizontal plane and the rotation in the vertical plane. Likewise, the gun is supplied with the firing direction in the form of two angles. To be able to perform the required calculations, the angles have to be translated to and from digital numbers. Therefore, a conversion is required twice: first analogue-digital, and after the calculation of the shooting data, digital-analogue again. This disadvantage, however, would have the advantage that digital computing technology could better maintain the required accuracy in the many calculations required for fire control. In addition, where desired, one calculator would be able to calculate firing data for more than one cannon simultaneously.
In the 1950s, fire control using analogue technology was reaching its limits, making the desired greater accuracy unattainable. Something that could be achieved with digital technology.
In 1954, the laboratory was convinced that it was capable of constructing a digitally operating fire control because solutions had been found for:
- the electronics, the memory, and the operation;
- the analogue-digital conversion of angle data;
- the digital filters for trajectory prediction;
- the storing of firing tables and the calculation method for the firing data; and later: the servo guidance of axis positions from the computing device.
The first analogue-digital conversion used large photographic glass plates that were obtained from (Agfa) Gevaert in Antwerp. Photographic binary numbers in the form of small black and translucent blocks were placed around the edge of the plate. These numbers could be read with a series of very small homemade photocells. A sinus table was encoded in binary digits. This allowed trigonometric data to be entered into the calculation unit when such discs were connected to the rotating axes of the radar and guns. This part of the fire control equipment we hoped to build was essential to our success.
The Royal Netherlands Army, however, indicated that it was not interested in a digital fire control system: “After all, the mechanical fire control systems of the NV Hollandsche Signaal Apparaten (HSA) worked flawlessly and the conversion from the radar rotation to the computer is considered cumbersome“. HSA gave the same response.
Admiral Pinke, a well-known Dutch WW II commander operating in the Pacific Ocean and former submarine commander, recently became head of the Royal Navy’s Scientific Research Office. He wanted a demonstration of digital fire control, preferably with something that looked like guiding a torpedo. That demonstration could not fail.
The Director was informed in England about a new type of steerable torpedoes that the Royal Netherlands Navy intended to purchase. He also witnessed the launch of training torpedoes aboard a submarine.
Subsequently, the laboratory built a small digital calculator that could calculate the trajectory of a torpedo. The target ship became a cardboard model that was dragged across the floor of the laboratory restaurant. The torpedo was a light spot that also moved across the floor. Admiral Pinke was allowed to set every dial as he was used to on a submarine. He enthusiastically crawled after the light spot on the floor. This prototype led to the assignment at the Physics Laboratory to design and build a digital torpedo fire control system for the submarine HNLMS O27. The fire control had to be able to perform a path calculation for four torpedoes at the same time. This is because the space in the submarine was too small to place four fire control devices. This would not be possible with mechanical devices, demonstrating a clear advantage of the electronic fire control system.
This VUTOR assignment (1954) made it possible to build a laboratory model of a digital calculator for calculating the trajectory data of several torpedoes simultaneously. This offered the opportunity to apply the digital and analogue-digital techniques developed in previous years in practice.
For the time being, the VUTOR laboratory model was made with the then-usual electron tubes. in May 1954, the fire control was successfully tested onboard the submarine HNLMS O27 near Torquay, England.
The target ship was HNMS Marnix (F801). The exercise with the VUTOR went flawlessly. On the last day of the exercises, the submarine commander permitted the launch of an exercise torpedo. All the periscope information about the Marnix was processed by the VUTOR. The order to fire was given.
After returning to the port of Torquay, the TNO employees were taken to the HNLMS Marnix in a sloop in the early evening. The crew of the HNLMS Marnix still had pale faces. The exercise torpedo had not stayed at a depth of 10 metres but jumped from wave to wave straight towards the ship. The commander gave all the commands for dodging a torpedo. He succeeded, much to the relief of the crew. Even though an exercise torpedo is not loaded with explosives, it may seriously damage the ship.
In the end, the system with over 600 electro tubes, was considered too complex by the Royal Netherlands Navy to proceed with production and operationalisation of the system.
Meanwhile, in 1954, the first transistors appeared on the market. Before that, the transistors for experimental circuits were made in-house. The transistors were not very reliable yet. The researchers tried to make simple calculation components. Sometimes strange effects occurred. One morning a calculator unit worked flawlessly, that is, until the coffee break. After the break, the device stopped working properly. The next day the same phenomenon.
After some time it became clear that the first commercial transistors were sensitive to sunlight and acted as photocells. After the coffee break, the sun shone on the calculator unit… The next generation of transistors had a black lacquer layer removing this problem.
In 1956, the development of a digital torpedo control based on transistor technology started. Hollandse Signaal Apparaten (‘Signaal’), Huizen would develop the mechanical part. The project was completed in 1959. However, the Navy decided not to purchase the torpedo due to poor performance. As a result, the developed torpedo fire control was never operationalised.
DiPhySa – Digital Fire Control Device
In 1954, within the framework of the Mutual Weapons Development Project, the US government provided support for European countries for the development of new weapons systems. The Physics Laboratory TNO, in collaboration with the Dutch Royal Airforce, presented a project for the development of a digital fire control device. In 1955, this project was accepted under the description: ‘Development of an anti-aircraft fire-control equipment using digital computing techniques’. There was a stipulation that a Dutch company was to be part of the project. HSA decided to participate, with the result that on 3 April 1956, a contract was signed between the Dutch government, RVO-TNO and HSA, supported by the USA. This project was given the name ‘DiPhySa’, a name consisting of letters from the words ‘digital’, ‘Physics’ and ‘Signaalapparaten’. The Royal Netherlands Army was the contractor and the project was a collaboration between the Physics Laboratory and HSA.
In 1956, TNO started a study on the possibilities of using very small magnetic rings as memory in digital technologies with good results. Magnetic core memory was therefore used in the DiPhySa project.
The digital calculation section that processed the radar data was built by the Physics Laboratory TNO based on an earlier laboratory model of digital fire control. The calculator part that processed the firing data was built by HSA. The integration test took place in 1960. DiPhySa formed the base for the development by Signaalapparaten (HSA) of the L4/5 fire control (a.k.a. KL/MSS 3012) for the Royal Dutch Army.
International interest
TNO’s digital fire control technology attracted international attention. As early as April 1957, a delegation of the Physics Laboratory, consisting of three engineers, visited six American research establishments to explain the advantages of digital fire control, at the invitation of the Office Chief of Ordnance, US Army. They were given the opportunity to sound out their ideas with their US counterparts. In 1965, the Norwegian representative in the NATO Defence Research Directors group referred to project Diphysa as a successful example of a ‘hardware project of magnitude developed in a small country‘.
Acknowledgement
An important part of this text is based on
- Several writings by the late Prof. ir IJ. (Ijsbrand) Boxma.
- ir. Boxma (1958), “Digitale Rekentechnieken voor Vuurleidingsproblemen“, de Militaire Spectator nr 11, nov. 1958.
- Bram Leenhouts en het FEL-TNO: Herinneringen 1949-1987.