This page has been dedicated for submissions from people who were formerly employed by the Decca Navigator Company.
NOTE: TO ALL FORMER DECCA EMPOYEES IN THE UK
There is an informal meeting of ex-Decca staff at the Prince of Wales pub in Shere, Surrey, UK for lunch every first Thursday in the month. We get ex-Navigator, ex-Survey and ex-Radar people along. Anybody transitting the UK is very welcome. Contact me, Walter Blanchard, firstname.lastname@example.org for further info.
Also, I've started writing an unofficial history of the Decca Navigator Company using a lot of material sent me after the reunion we had when the UK chains were closed down in 2000. Anybody who can contribute anything please let me know, particularly about overseas operations. Photos specially welcome.
Thanks, Walter (Decca Navigator 1957-1971; Decca Survey 1976-1983)
DECCA MAIN-CHAIN MEMORIES
by David .S. Jones
Dept: Chain Implementation Dept., Mew Malden, UK.
Responsibilities: Field engineering services for Navigator Chain stations.
World-wide Installation, commissioning, system upgrades, etc. of ground based Decca LF transmitter sites.
Moved to newly formed Racal-Decca Avionics Ground Systems group responsible for ILS/DVOR/NDB deployment and commissioning. Between 1980 and 1983 was still on-call as support for main-chain special projects, system re-tuning, etc. Sales engineer with Racal Avionics Ground System until 1988.
At the time of my joining, the company was headquartered in London and had production facilities in New Malden, Hersham and Maidenhead, all south west of London. Engineering and main production was at New Malden. Here also, the navigation charts were plotted and evaluated prior to a chain's service. System corrections for terrain effects were calculated and passed on to operations by this group. Hersham developed and manufactured the airborne Doppler navigation systems. Maidenhead produced the Decca system antenna coils, their wooden support frames and many accessories. It also re-furbished older equipment returned from closed down stations.
Also in the group, the company operated a flight test facility at London's Biggin Hill airport where a Bell helicopter and a fixed wing twin prop was used for airborne system trials. This facility survived into the late-80ís under the Racal name and operated a Jetstream 31 turbo-prop and BA125 twin jet. A second facility was located at Heathrow Airport for airborne systems installation and support. The chain staffing and operations group operated from a converted country mansion at Stevenage, 50 miles north of London. They managed the UK regional transmitter sites as well as those overseas in Australia, South Africa, the Persian Gulf and I think Germany. This involved the provision of long term on-site staff and supplies. Again, some older equipment was also kept and re-furbished at the Stevenage location. Chains in Scandinavia may have been supported by them, but I can't be sure.
The group training facility at Brixham, Devon was also home to a small, specialist support group that focused on antenna tuning and motor/generator systems. They were used principally for UK/European activities. At this time (early 1970ís), the Decca group had many production and research facilities in the London area and some not listed here may well have been used to support navigator system main-chain activities at various times.
One interesting feature of the product line was its resilience. A lot of the equipment was recovered from chains that had closed down and was re-furbished for later use. Equipment from the Bahamas, Vietnam and other defunct chains was returned to inventory and found life in some of the later UK chain installations.
At the time of my joining (1974), the main transmission station equipment was the 820 style. This was a valve (tube) design of receiver, oscillator and transmitter equipment, all housed in rugged, gray painted metal racks, each about 7 feet tall. Each station housed three, identical receiver/oscillator racks and a fourth control rack. This control rack went by the acronym RASME or RAMME for Remote Automatic Slave Monitoring System or Remote Automatic Master Monitoring Equipment. All three receiver/oscillators were running in time sync with each other and one was selected for duty via the RAMME/RASME. The RAMME/RASME used a control receiver and individual alarm monitoring circuits from each rack to decide the on-air duty status. It provided a two out of three vote and scalable alarm severity monitoring to automatically switch out a suspect unit. When warranted, the system alarm conditions could be overridden from this unit and the station manually held on-air. It could also be used to manually remove a rack from use during maintenance. All logic sensing was carried out via relays and rather like an early telephone exchange, when it worked, it worked well.
The Decca chain was set up in a master/slave configuration. The master station provided the phase and timing reference for the three slaves. Phase stability was the basis of the system accuracy and many factors could corrupt the relationship. Distance between sites, topography between sites and nighttime skywave were all external factors. Local conditions included equipment temperature variations, voltage fluctuations, component tolerance drift and changes in ground plane conditions.
Each receiver/oscillator had a master clock that provided timing functions to the rack. This clock was a mechanical unit with a circular array of reed relays over which a magnetic arm swept. These reed relays provided the timing control functions for the rest of the electronics. Five individual, frequency linked channels were generated from the oscillator and these were the phase-locked multi-pulse outputs to the transmitters. A unique, variable coil device, the goniometer, was used to set calibrated phase path adjustments to each output channel. In the case of the Slave stations, a goniometer was fitted to the receiver unit which set the station's fixed phase relationship with the Master.
Outputs from the racks went via the RAMME/RASME to the transmitters. These were individual, 1200 watt, modular based units, one per frequency. A single transmitter comprised a 6 foot tall rack with six plug-in power amp modules and two plug-in driver units. The power amps used a type 807 tube, six per amp. All modules could be removed while live but caution was recommended. Each driver fed one half of each power amp. Transmitter output was fed to a tank circuit that provided matching to the antenna coil house feeder cables, typically 600-800ft long. The coil house, located at the base of the site antenna, housed the large, open weave coils that fed energy to the antenna.
At the coil house, each feeder coax terminated into a dedicated, adjustable matching unit that in turn fed a coupling circuit on the antenna coils. These coupling windings were on the face of each antenna coils. To maintain system operating efficiency, all the transmission line segments were adjusted to give optimum energy transfer.
A multi-pulse (five frequency system), coil set had five discreet, open weave coils made of litz wire, all mounted on a hardwood frame. Each coil was about 4 foot diameter and about 1foot wide. They were spaced about 18 inches apart on the frame and the center coil was connected to the antenna itself. Each was a tuned circuit with it's own high voltage, oil filled capacitor bank below. These capacitors were metal cased with a glass insulator top, finished with a 1 inch threaded stud connector. Each weighed about forty pounds and stood perhaps 18-20 inches tall.
The antenna coil was terminated at one end in a copper tube, 1 inch diameter that went out through an insulator window to the antenna, while the other connected via it's capacitor bank and thermal ammeter to the ground plane. This coil had a variable winding in the center which was moved by a cord loop and adjacent pulley wheel. The inside of the coil house was covered in several continuous bands of copper sheet, all bonded to the ground plane radial wires. The living quarters of many Decca stations often had elaborate bars constructed by the engineers and covered in polished copper sheet, salvaged from past installations.
Several designs of transmitting antenna were used during the life of the Decca system. Some were top loaded for increased efficiency, others not. Generally, the antenna mast was a 300 foot guyed structure made of triangular metal sections. It stood on a single insulator with spark gap and was supported with insulated steel guy lines. Some early sites had masts of 500-600 foot with top load wire supports. These high efficiency masts were mainly used for stations providing transatlantic coverage. All had obstruction lights fitted and any volunteers to climb and change the bulbs would always be welcome.
A towable, trailer mounted coil house was developed in 1975/76 that could be taken to a UK regional site while the main coil house/antenna system was undergoing repair. This would be connected to a lightweight temporary mast erected adjacent to the main antenna and after tuning would take over the station duty. In this way, repairs and maintenance on the main mast could be carried out with minimal disruption of service. Antenna insulators required cleaning or replacement after several years exposure to the elements on a coastal headland and painting of the main mast was also carried out.
Tuning the coil system was a lengthy process in normally unpleasant conditions, (working in the coil house in extreme heat or cold). If the station was operational then time was a priority. Many tuning sessions were carried out at night when cooler temperatures prevailed. After first disconnecting the station feeder, each color frequency of the chain was injected at low level into the coupling coils using a special signal generator. The antenna coil was set to resonate by adjusting the coupling distance between the other coils while keeping the adjustable winding in the mid-range of its travel. Where required, the antenna coil was shortened one or more turns to achieve an acceptable tuning point.
Perhaps the most complex part of the process was the adjustment of the rejection circuits in the coils. This used a movable, flexible link winding on the face of each coil to achieve an acceptable rejection ratio of unwanted signal. By a process of continuously injecting each frequency on each winding and measuring the out of band response after bending the wire link slightly, could the ratio be set.
The two rear coils of the set were mounted on a separate frame that slid along the top surface of the main coil trestle using glass balls. This frame was linked by a rod to the center coil adjuster wheel and formed the main tuning adjustment for the station. By turning the center coil wheel and moving the pair of adjacent coils in or out, the antenna was kept tuned to most weather conditions. Later systems used an automatic sensor unit and motor drive on this mechanism to keep the station on tune. The end of the antenna coil was passed through a small transformer before it was bonded to the ground plane. This transformer provided a phase loop feedback circuit to the receiver system in addition to antenna current monitoring.
Due to the high voltage RF field present at the base of the antenna, all the parts of the coil system were made of non-metallic materials. All fixing screws were non-ferrous whilst nylon screws were used for coil frame construction. Heavy duty copper strip was used to bond all metallic equipment to the main antenna ground plain. Extensive ground bonding was also carried at the transmitter/receiver site where copper strip was bonded to all metallic frames and equipment. Some sites also had metal roofed buildings that were bonded to the main ground plane. Any possibility of induced fields causing system errors was taken very seriously. It became a subject all by itself.
Critical to the system performance was the need to maintain continuous service and dedicated power supplies were the key. All stations ran from a heavy duty 220v DC lead acid battery system that was float charged by dual chargers.
Some sites used a motor/generator system to maintain emergency power, others a large UPS. The motor/generator was a 220v DC motor, shaft coupled to a 240v AC motor and a 220v DC generator. Primary AC power ran the AC motor under normal conditions and the 220v DC generator output fed the equipment. When main AC power failed, the DC motor was fed from the battery and shaft speed was maintained during the break. System power was now provided by the DC generator on the shaft set.
Depending upon the location, the transmitter site would have one, two or perhaps four diesel generator sets. For locations with a reliable local power supply, a single diesel plant was sufficient. Where no local power existed a full set of four diesel generator units was installed along with large capacity fuel tanks. These generators ran continuously and were cycled through load duty changes every few days. Prior to a diesel change over, the load would be transferred to the standby unit via the switchboard or would be reduced by shutting down non critical systems. Main electronics would be switched to inverter or battery set while the new generator was started and tested. Soon after, the main electronics was switched over to the new plant. Full load was added and the old plant shut down. This was normally the time that the new engine decided to die. A fully automatic switchboard had control of the power systems and if the main feed was cut it would start backup power and switch over. With remote manned sites it was preferred to cycle plant as a routine.
The 820 style main-chain equipment was superseded by the 1880 style in the mid-1970ís. This was a hybrid design that used tubes for the RF receiver stages and discreet logic devices for switching and control. Again three racks comprised the main receiver/oscillator units with individual, stand alone transistor based transmitters. Tank circuits were now an integral part of the transmitter design. A modified Mk12 marine receiver unit was fitted into a single cabinet and was used as the system monitor. Some chains used an on-site monitor receiver while others located the receiver several miles away, out of the antenna field and fed back to the station via a microwave link.
Power requirements were now substantially less and the use of relays was replaced by semi-conductor logic circuits. Cabinets were finished in light blue and one of the most striking features for system operators was the lack of relay noise within the equipment room. With the 820 equipment, dozens of relays were being switched in a continuous, clock driven synchronized cycle that, when all was well, took on the feeling of a comforting, system heartbeat. At the 1880 sites, silence reigned. These newer systems were deployed in Australia, South Africa, Holland, Spain, Japan and India, to name a few. Antenna coil systems were the same as the 820 generation.
Some additional features were added to the 1880 generation of equipment that allowed a simple alarm code to be broadcast from the station transmitters and detected by a central monitoring point. This used the guard portion of the station's 8.2f transmission to modulate a series of breaks, each one representing a status bit. By using this status link, the remote site could be left unattended for extensive periods. A parity count bit was used to improve detection accuracy. In some locations this coding and detection technique worked well but since it relied on brief LF level changes it could not be said to be totally reliable. Australia used this system as did South Africa.
Alarm code transmissions had also been used by the 820 type equipment and were based on a simple sequence of power level fluctuations. This was a single alarm condition, triggered by any change in RAMME/RASME status. A small battery powered, specially made receiver unit was used to detect this change. The receiver was a single channel unit, housed a 7-inch metal cube and could be used within several miles of the station by the duty staff. It allowed the watch keeper (note the nautical terms used for station staff!), to leave the site as needed.
Prior to commencing full operation, any new chain was put through an extensive monitoring exercise requiring numerous fixed position readings to be taken for a period of time. These were carried out for 24 hours a day for several weeks and at numerous definable locations within the coverage area. Typically a lighthouse would be used, also survey markers on hills or other well defined locations. Often the installation team for a chain would be replaced by a monitoring and operations team prior to final handover. During this period, all operational functions were set up, spares organized, documentation completed and last minute modifications carried out. It was also a chance to put a final coat of paint on the site and clean things up.
Shifts of staff would log receiver readings and plot standard variations, all of which was fed back to the charting division for study. In this way, the predicted chain pattern could be compared to these fixed locations and system errors calculated. Adjustments were then made to station goniometer settings after which the chain was declared open for use. Periodic updates were also sent to active chains based on long term analysis of the fixed readings.
In the mid Ď70ís, some sites were retro fitted with a very stable, rubidium standard oscillator unit made by Hewlett-Packard. This was intended to take over from the existing crystal oven oscillators used in the 820 equipment and to counter the effect of skywave drift. They were phase locked to the main equipment and were used to keep the station to a highly stable reference when master signal lock was impaired by skywave. It was hoped that by minimizing the system skywave drift it would halve the user corruption effects. These systems were not however, widely deployed.
By the early 1980ís, the patent or license had lapsed on the basic receiver design and a new, rival model appeared on the market from a company in Europe. This threatened to erode the traditional rental market for receivers and Decca tested out changes to the multi-pulse timing sequence hoping to add subtle changes that would corrupt the new receiver. This was only a stop gap solution to the eventual deployment of cost effective GPS and Loran receivers. Once these became widespread, the use of Decca equipment shrank exponentially. The rest, as they say, is history.
The standard marine receiver units were the Mark 5, then the Mark 12 and finally the Mark 21. There were similar generations of airborne receiver equipment during this period, each with a range of ancillary units for map display or multi-input navigation processing.
The Mark 5 comprised an external power unit, bulkhead mounted receiver unit and decometer display unit. All electronics were tube based. The power unit was often a rotary converter module. Chain selection on the receiver was via side mounted dial and only whole number chains could be selected. The receiver unit was a single electronics chassis with a metal cover. The decometer bowl (as the display was known) was a rugged cast metal case with glass face lid. This lid could be opened for setting zero and lanes. Ideally suited to life aboard a fishing boat, the decometer bowl could be swivel mounted anywhere in the bridge. All parts were finished in hammer finish black enamel.
The Mk 12 again was a three unit solution but now had the system controls on the face of the decometer bowl and it could tune into the sub-chain frequencies (5A, 5B, 5C, etc.). Chain selection was on the display face and zone indications were now shown (the 0.2f decoded and used for zone identification). The electronics were again tube based and the system was finished in a dark gray enamel. The receiver had hinge down modules covered by a removable case that facilitated service on the sub-units. Both the Mk 5 and Mk 12 shared a similar antenna, the 18 foot fiberglass whip.
The Mk 21 unit was a transistor based design that housed everything in a single display/electronics case. About 18 inches cubed, the unit had three rectangular analog decometers and a neon lane display. A facia mounted control changed the dial illumination and also glowed amber in sequence with the received RF signal. Flickers in this lamp signified the breaks in each transmission cycle. The lower half of the front face was a hinged cover that gave access to the power on, chain selection and reference zero controls. Inside, large circuit boards slotted into a motherboard style of assembly. The unit could be powered by 12v dc or any one of a number of alternative sources. It was rugged and easily set up for mobile monitoring applications. It came with a short, stubby encased antenna unit connected via 50 feet of wire. The Mk 21 was finished in a dark gray/graphite color with the hinge panel cover in cream.
These receivers could be linked to other, related Decca products to provide an integrated navigation system. The marine autopilot, track-plotter, TANS aircraft navigation computer and other products all added value to the basic system concept.
In conjunction with the Dutch, a Mk 21-based special differential receiver was designed that was used for Europort marine pilots. This receiver used two chain fixes to produce a very accurate pattern in the congested waters of this Dutch port. They were one of the early operators of VLCC (very large crude carriers) in the narrow confines of the North Sea and safety was paramount.
Also remember that the Decca Survey company was formed to support oil exploration activities and used several, low power, transportable radio position fixing systems developed by the company. These included the Decca Hyperfix and Superfix.
The description of life on a transmitter station published on your site brings out the fact that many of the resident station staff knew little about Deccaís corporate changes or developments. Their culture was one of enforced solitude and a routine dedicated to keeping the station on the air at all cost.
Indeed, there was almost a resentment by station staff when company engineers would visit their station to try out a new piece of equipment. A new face arriving with a box of electronics that needed connecting to a vital circuit was not a welcome sight. With some justification, they did not want amateurs playing with the station and often a territory argument was resolved at the last minute. Many station staff had been with Decca for years and had long memories of a past crisis.
The political machinations of the Decca Group and its subsequent acquisition by Racal were probably lost on most of them. It is clear that without Racalís intervention, Decca would have fallen apart and broken up. When Racal acquired Decca there were a lot of parts to the group. It covered marine navigation, marine radar, airborne radar, e/w sensors, oil exploration surveying, marine systems integration, air navigation systems plus the commercial activities of music production and radio/TV set manufacturing.
It is ironic that the station staff interviewed on your site documents believed that Racal was keen to acquire Decca for its marine radar business since this sector pretty much collapsed soon after under intense foreign competition. In addition, the marine sector was already suffering from a decline by the late 70ís due to oversupply and the use of cheaper foreign vessels. Racal had historically been a military supplier but soon after acquiring Decca undertook a major campaign to build public awareness in preparation for its move into telecommunications. From a standing start as the alternative UK cellular carrier behind the might of British Telecom, it went on to build Vodafone into a powerhouse company that subsequently split from the parent.
The Racal group went through many changes in the mid-late 90ís before splitting off many parts to focus on core activities.
David S. Jones, (formerly of Surrey, England).
Litra Manufacturing Inc.
AN ENGINEER FROM FROM THE UK
I was employed by the Decca Navigator Company as an installation and commissioning engineer for the main chain stations from 1974 to 1977. My first task was to dismantle the "Dectra" transmitters which had been used to enable Great Circle navigation across the Atlantic before advent of inertial navigation. Those two UK stations were located at Stirling and the Isle of Whithorn. Apparently the Dectra system suffered from phase-lock problems across the Atlantic hence the reason for its demise.
I had the distinction of commissioning the last of the thermionic valve stations. This chain was located in the Outer Hebrides off Scotland's Altantic coast and I got the transmitters on the the air just three days before Christmas 1976. These were the last UK stations to be built. Since the equipment had been bought back from the United States, the gear had been fitted with integrated-circuit master timing clocks and at the master station on the Isle of Barra had rubidium frequency control. After this installation, work dropped off quickly and I was for a time, in the labs at Burlington Road New Malden, London where Bill O'Brien was still working on new ideas!
Last year (2000) I found to my chagrin that I had missed the shut down of the Decca system and the closing conference event. The head of Decca main chains contacted me after I had made an enquiry to Racal-Decca. He was sitting in the dismantled transmitter building at Puckeridge (English chain master) writing notes to take down all the masts and dispose of the whole system. I have tried to get hold of some old transmitters but apparently they have all been skipped. All is not lost. A museum display of a complete transmitter and various receivers has been established in Great Yarmouth, U.K.
Douglas Sim E-mail: doug.sim(at)btinternet.com
Back to Decca Intro Page