大象传媒

The station at Freetown was well sited about 400ft above the battery, and much cooler than down on the coast. Then known as the White-Man`s Grave, people now actually go to Freetown on holiday.

The aerials were two side by side 7-foot parabaloids, with separate transmitter and receiver wave-guides facing into them. The surface watching equipment had a beam width of about 5 deg, with range 20,000 yards, and was no good for gun-laying. Also, to lay guns, an accuracy of 5 mins of arc was also necessary.

The Rx had a strobe, a short bright-up pulse the function of which I never could see. It was controlled by a manually operated potentiometer. I have a feeling that the Navy (the 271 was Navy) had slave displays on their ships, and the strobe indicated to the slaves which of possibly many was the target of interest.

At 400 feet, good long-range echoes could be obtained, much beyond normal, by delaying the time-base trigger with the strobe pulse effectively doubling the surface-watch range. Since this only involved putting a Pye plug in a different socket, I never felt guilty of illegal modification.

On one occasion the Navy depth charged the harbour all night long to catch what we all thought was a submarine seen on the radar heading for the boom entrance, a tiny echo travelling at 10 knots at about 7,000 yards, and disappearing at that short range 鈥榠n the grass鈥� (receiver noise looked like agitated grass growing on a flat lawn).

The following day we realised this was from a collier 186+ miles away, coming in from Newfoundland, whose echo was being caught on the subsequent scan (no wobbulator on 271). We would not previously have had an explanation, but we turned up a TRE (Telecommunications Research Establishment) bulletin, reporting from India that signals could come on a second scan as the result of multiple reflections between ground and an inversion layer, and they had some from a mountain!

Fall-of-shot

9.2-inch shells make geysers of water about 80 foot high, lasting several seconds, and on stations fitted with a PPI (Plan-position Indicator) would give useful correction inserted manually into the predictor, which was also fed with temperature, humidity and shell nose profile (HRS). It also gave a warning to watchers near the end of the predicted time-of-flight which would be about 20 secs, at extreme ranges.

Mortar spotting

This was an application I have never seen or read of, but after the war I worked with an ex-mechanic who, up front with the infantry, had both operated and serviced one. He told me the gear was portable in 60lb packs and could rapidly identify the origin of mortar shells.

Resolution

The accuracy of a gunlaying radar is a function of its resolution both in bearing and range. I thought the set in Freetown was pretty good but I have no doubt that my mind would boggle at what can be achieved today.

Bearing

The aerial lobe may be wide and lacking in resolution, but if the focus of the paraboloid were alternated left/right, say 5 deg, the height of the two displayed bearing signals can be balanced by turning the aerials and the median setting will give a great deal more accurate bearing reading.

This was achieved on the Coast Artillery GL by having a forked wave-guide into the receiver dish, and alternating an ionisation blockage in the two wave-guide mouths, done by having two paralleled beehive neon lamps in each wave-guide and alternating the d.c. firing of each neon pair with each transmitter pulse.

The ionised neon gas proved to be a pretty effective block to any microwaves trying to get down the wave-guide. On our radar, I could separate the masts on a beached wreck transversely placed at 2,000 yards, which would represent a bearing discrimination of about 5 minutes of arc. Neeson says 'Freya' was very accurate and the aerial was rocked over a narrow angle, but he doesn鈥檛 enlarge. I can only assume that target amplitude was recorded on the limits of the rock angle and the aerial rotated until the two measurements were equal.

Range

Accuracy was a function of the linearity of the range time base and the resolution with which it could be read. On the CA GL the TB was a spiral of about 4 turns on a 12-inch television type tube fitted both with electrostatic deflection plates and magnetic deflection coils.

A trimmer capacitor across the deflection plates and deflection coils wired in parallel made a tuned circuit whose periodicity was equivalent to the transmit/reflect time over 5,000 yards. It was shock excited and allowed to decay naturally. The spot described a spiral inwards as the voltage and current decayed naturally and was terminated after about 5 revs.

Maximum electrostatic deflection and maximum magnetic deflection alternated every quarter cycle of natural periodicity. It will be appreciated that without decay the spot would have described a circle. The effective TB length was then I believe about 7 feet, giving good resolution, and each rev represented 5,000 yards (?).

Signals deflected the spot radially outwards from the spiral by means of a deflection electrode, about the size of a penny, placed centrally on the outside face of the crt.(Many a dirty dig of the signal voltage I had from it too). Calibration was easy. The deflection circuit was tuned so that 1,000-yard crystal markers (6.1348MHz?) were radial on the time base. The operator manually set a radial marker onto the leading edge of the target blip.

The duration of the transmitter pulse was determined by an open ended delay line (about 6 stages of 5,000v 0.1uFd linked with coils wound on SRBP tube formers) which was allowed to charge up relatively slowly and then was discharged by a thyratron in series with a pulse transformer.

The line started off charged up to a certain suitable voltage (V) and when the thyratron fired, the pulse transformer with its secondary load (which was the magnetron) was equivalent to the Zo (characteristic impedance) of the delay line. Being matched, the line discharged without ringing. On firing, the line voltage immediately was halved (impedance matched) and a negative step of 1/2V appeared on the beginning of the delay line leaving 1/2V on the line which was maintained until the step had travelled along the delay line and been reflected back from its open circuit end.

After a time equivalent to twice the transmission time of the line (2T) the negative step arrived back at the front end of the line reducing its voltage by the remaining 1/2V to zero. A pulse of 1/2V had then appeared across the transformer for a time 2T secs and been transformed to the 15Kv necessary to operate the magnetron. The sharpness of this pulse was important in that the shorter it was, the sooner the display could be looking at a new or alternative target (resolution).

However the real limit resided in the bandwidth of the receiver which had to follow the rise and fall of the transmitted pulse. From memory, as radar improved, the 1飦璖 pulse dropped to O.1飦璖 but this could only have been achieved with an improvement to the Rx bandwidth.

Valves

I would need access to a much more extensive library than our local one to read up on these sophisticated devices but I have seen some disagreements in the literature about them. My understanding has always been that a rhumbatron was a metal tuning device, a klystron was an oscillator valve in which a rhumbatron was incorporated, a Sutton tube was a CW oscillator of the klystron type, utilising a single rhumbatron and a reflecting electrode, while the common T/R Cell was a gas-filled device incorporating a rhumbatron and an ionising electrode.

Batt tells us that the klystron started with two rhumbatrons until fitted with the reflector - the only version I knew.

Magnetron

This was a multi-anode diode with a cathode axial to the short cylinder forming the anodes. It was subjected to an enormous permanent magnet field parallel with the anodes and central cathode such that electrons coming off the cathode pursued a sharply curved path around the cathode.

It will be appreciated that transit times would influence the time of arrival of the electron to an anode, so that the device was voltage sensitive in respect of to which anode an electron finally arrived, and that if a.c. of differing polarities were on the anodes (ie they were across a tuned circuit), then this could determine whether an electron was captured by a particular anode, or whether it continued its curved path round the cathode, and was captured by the other anode because it momentarily had the higher attraction voltage. Such selective capture was what maintained the oscillation.

Batt describes these special devices well, but I guess it helps understanding if I also add my three-pennyworth. If you imagine a short conductor shaped like a circlip, it could be considered that the open part of the circlip, the lips, were capacitative to one another, and that the arc of the circlip could be considered as an inductor. The two things formed a tuned circuit.

With very high frequency currents travelling only on the surface of a metal assembly, it follows that the configuration of a conductor is affected only by its surface, and that a hole drilled close to the edge of a thick sheet of metal such that it broke through the edge of the sheet has the shape, considered as a conductor, of a circlip, and could operate as a tuned circuit. Such a resonant cavity in the thick sheet could be connected in series with other cavities resonant to the same frequency, by sharing a lip (anode) with a neighbour, in such a way that the anodes formed a circle round a central axial cathode. This arrangement is then what was the best kept, and certainly the most important, secret of the war - the multicavity magnetron.

The only tuning possible was determined at manufacture by the basic dimensions of the unit cavity with only very slight frequency sensitivity to the applied voltage. Since this was varied primarily to tune for maximum output, the device was, to all intents and purposes, of fixed frequency. Although the magnetrons I knew consumed only an average power of about 40W, during the short transmission, the peak power was in the order of 1/4 million watts - an immensely powerful transmission.

This type of valve is well described by Reg Batt pp50/2 and a photo appears in Nissen and Cockerill's book 鈥榃inning the Radar War鈥�, to me, the best of the books I鈥檝e read to date. This device was the major technical advance we had on the Germans, and I have all these years later read that if anyone with the knowledge of the workings of the magnetron been taken prisoner, a hit-squad would have been sent in with the express purpose of silencing him.

The Klystron (Sutton tube)

Continuing the circlip analogy, if we have a multiplicity of circlips stacked above one another and bent round in a circle with the lips inwards, we would have a doughnut shape with a hole through the middle. Each of the circlips alone would form a tuned circuit, and the fact that many of them can be put in parallel does not much alter the fundamental frequency of the whole. After all, two similar inductors in parallel halve the value of 'L' but double the value of their capacity 'C' so that the L/C ratio and hence the frequency, is preserved.

Such a configuration represents the rhumbatron. This arrangement became the tuned circuit of the local oscillator (Sutton tube) in the receiver. The frequency could be altered slightly to adapt to the magnetron frequency, by means of screw in plugs round the outer edge of the doughnut formed.

Since the inside surfaces of the rhumbatron are carrying the oscillatory currents and the path is being shortened by the screw in plugs, the frequency of some of the 鈥榗irclips鈥� is altered, and because they are all in parallel, to a diminished extent, so too is the frequency of the whole.

The valve (Sutton tube) consisted of an electron gun like that in a c.r.t., which passed an electron beam through the glass-encapsulated lips of a rhumbatron. The beam approached a negatively polarised reflecting electrode, which turned it back to the rhumbatron, where it was collected on the nearer lip. The beam in passing through the rhumbatron was chopped into parcels of electrons by the oscillatory signal on the lips, which alternately speeded and retarded them so that the flow was bunched. At the right voltage this gave the correct phase to these current pulses to maintain a low power C.W. oscillation, which was used as the local oscillator for a crystal diode mixer in the superheterodyne receiver.

The klystron did not appear to have been used as a pulsed transmitter in service, in spite of the efforts the Americans were making along these lines before we gave them the magnetron.

Common T/R cell

This they said at the time was pinched from the Germans, and was the objective of the Bruneval raid, but it now seems highly unlikely, in view of the frequency limitation. It was used in single paraboloid stations like 271Q. It consisted of a gas filled (neon?) glass envelope, enclosing the lips of a rhumbatron (called a 'soft rhumbatron'), and an electrode carrying a potential such that a discharge produced an ionised glow between the electrode and one lip of the rhumbatron.

It also had a pick up loop into the rhumbatron cavity (physically just a short circuit to the coaxial feeder). Since the glow did not extend to both lips, the rhumbatron was free to oscillate and behave like a tuned circuit for the receiver, but when the massive transmitter pulse came down the wave-guide, the signal voltage on the rhumbatron lips was sufficient to extend the glow to both lips and thereby short circuit the rhumbatron and protect the receiver with its sensitive mixer crystal input.

The local oscillator was fed via a capacitative probe to the crystal (which was simply mounted in series with the IF input circuit, the loop crossing the wave-guide thus also picking up RF signals coming down the wave-guide). The LO signal rectified by the crystal was read on a meter as an indication that the Sutton tube was working. Knowledge of previous readings indicated current performance.

Power supplies

These were a real hotchpotch, starting with 230V from a diesel generator which was used for the display. Because the transmitter and receiver were Naval types, a transformer and mercury arc rectifier system took this to 110V d.c., which in turn fed a rotary converter, giving the 180V 500c/s they required. RAF inspired to keep transformer iron weight down. Army would either have local 230V 50c/s, or generate their own - as we did. Incredible but typical of the times!

Photographs

While hostilities existed, any kind of photography would have put me into the glasshouse (military prison), and anyway film was completely unobtainable. However an associate nicked for me some 35mm film from Ordnance stores, and by dint of laying a length into the backing paper of a film already in a 620 camera, borrowed in exchange for developing the owner's film (chemicals were obtainable at a shop in Freetown), I was able to extend the markings across the paper for 127 spacing and block the viewfinder down to the smaller field.

This meant I could do my own negatives in the stifling heat of an empty 271 cabin - used as my darkroom. I have two negatives, not currently locatable, of the aerials of the Freetown station taken on completion. One is reasonably well framed but, of course, does not include me, the other, which does, was at the hands of some now unremembered person, who was unable to use my restricted viewfinder and who just pointed hopefully, getting the blokes in (happily) but chopping the aerials.

Richard has been successful in computerising the photos and merging the two together. The figures at the extremities are local workmen completing the roof. The local soldier in the middle is my personal assistant listed as an electrician - Osei from the Gold Coast.

I can't remember the names of the other white blokes, since they were not radar, but some were Artillery bombardier watch keepers, Signals line layers, and the bloke in the hat was a staff sergeant instrument mechanic (mag slips and oil-motors for the turning gear).

End of hostilities

At the end, and after my presence at the site was no longer vital, we were demonstrating the capability of the gun-laying radar to the GOC-in-C (West Africa) at about 15 miles, with two floats 100 yards apart, each with radar reflectors and towed by the Water Transport coy RASC and their German acquisition the 鈥楲emburgh鈥� (3 diesels). I was ordered to ensure nothing went wrong with the communications - No18 military back-pack radios, one of which I had already installed with an H-aerial on the 鈥楲emburgh鈥�.

The white radar operator bombardier was land-based and I was instructed to operate the radio on the launch. It was a total failure. We saw the flash of the guns, and some time later heard the bangs, but fall-of-shot - Zilch!

We had on board an Artillery Lieutenant who, to avoid embarrassment asked me to radio 鈥榃ater too choppy, necessary we return to harbour鈥�. Cobblers! The sergeant skipper of the 鈥楲embugh鈥� needed to shorten the tow, and to do that between floats he had to jump on the first one, which he missed, and came up spluttering. He just managed to climb aboard the first with a shark turning on its back for a bite.

The next day a highly irate Colonel ordered an investigation, and it transpired that after a manual exercise in gun-loading, just in case the Colonel ordered it, the pins locking the guns to the radar had been pulled and replaced 90 deg out of phase. This resulted in a very tart telephone call from the Forestry people, who complained about 9.2-inch shells whizzing over their heads.

Extra Notes on Batt

p.31/33 space is devoted to Barkhausen Kurz valves, which were only a lab phenomenon and as far as I know did not get into service.

p.48 - crystals did not appear to be sealed or need it for that matter. I have re-twiddled the cat's whisker on many crystals to improve back to front ratio, and then reset the locking screw.

p.59 - Yagi is best known for his aerial system used on SLC (Searchlight Control) and nowadays the design for almost every TV aerial.

Chapter 11 - H2S (sulphuretted hydrogen - with the smell of rotten eggs) is reputed to have got its name when someone asked what a user thought of this type of radar and got the reply 鈥業t stinks鈥�.

Radar at sea

This is the title of an interesting book by Derek Howse, which deals very thoroughly with the operation of such equipment but leaves a lot of technical questions unanswered.

Three more subjects are worthy of mention:-

鈥橺鈥� batteries

In my time at Portsmouth there was a 鈥榋鈥� battery on the front at Southsea. It consisted of an 8X8 array of rocket launchers. I don't know what the radar arrangements were, but with the limited height of the rockets it had to be primarily against low-flying aircraft such as those dropping mines in the Solent. They were highly successful, and seemed to catch an enemy with every other firing. 64 extremely noisy rockets went off together, but exploded in a cube at varying heights. I would imagine the shrapnel was very thick.

Proximity fuses

I have no personal experience of these, and there is little in the literature, but proximity fuses were obviously radar-controlled. The ability to detonate missiles above the ground was effective in damage to personnel by exploding above the ground, and blast damage to buildings was significantly greater than if the missile buried itself maybe 100 foot in the ground before exploding.

They were not used in Germany for security reasons, until the Rhine crossing. I have no knowledge of earlier use in the UK against aircraft, but see no reason why not. An anti-aircraft shell could be activated at a prescribed height, ie after a prescribed time as was normal, and then using its radar sensing, safe while the range was closing but detonating when the range became incremental.

Degaussing

Most aspects of the last war have had some sort of write-up but very little seems to be on record of the technique of degaussing. At the time, the gauss was the unit of magnetism.

The German magnetic mine was having some considerable success in the early stage of the war - due to the fact that it was impossible to prevent ships from becoming magnetised. A drastic temperature change in the presence of the earth's magnetic field was all that was necessary, and a moving magnet passing in the presence of a detector in the mine would cause it to blow. However, with coils passing round the ship both vertically and horizontally, it was possible to cancel the magnetic field of the ship.

However, first it was necessary to measure what this was, and it is in this area that I had some ground-floor experience. At this time circa 1940, 1941, as an Army sergeant I was teaching basic magnetism and electricity to fairly erudite non-commissioned Army staff at Portsmouth Municipal College, destined to become RADAR technicians.

One of my classes consisted entirely of vets (the Army having somewhat earlier given up horses). I was in cushy civilian billets in Southsea, the only Army contacts being my uniform with the red radar and 鈥極rdinance鈥� flashes on the sleeve, and the salute I gave for my money at the Marines' pay-parade on Fridays. I also had to march with the Marines to Church Parade on Sundays.

In the same billet was a civilian expert (Lenny by name), whose job it was to measure the magnetism. He had a little hut on the foreshore in the dockyard, with about 8 mirror galvanometers connected to coils on the sea floor in a row across the Solent.

The galvos each operated spot projectors, focused on separate 5-inch wide photographic paper recorders. This was very wasteful of paper since, at the most only two recorders showed any trace. The vessel being measured plodded up the Solent at about 5 knots. I remember there were 4 foot deep development tanks for the paper and, at a time when commercial printing was almost impossible, Lenny kindly invited me into the dockyard to look at his gear, and did an enlargement of our only wedding photo - a snap taken by someone with a Brownie and who had managed to get a roll of film.

I never delved into what Lenny did but assumed that given the knowledge of the state of the tide (depth of water and proximity to field of the coils), direction of deflection (polarity, N or S), number of deflections (single pole would be vertical polarisation, one in each direction would be horizontal), and given a standard speed for the ship, he could calculate the current and polarity required in the coils. He could then radio or Aldis the information to the ship and get a rerun, which would eventually show an acceptably low level of galvo deflection.

Finally, minesweeping was done, with a large airborne coil producing a magnetic field. Such aircraft have been seen many times on TV. The virtue of the aircraft was the speed with which the sweeper could get out of the way when the mine blew.

Return home

These notes would not be complete without mentioning that the radar at Wilberforce Spur was the end of my radar experience, although I went on to a career in electronic instrumentation until retirement (Airmec, then Advance, later Gould).

When my replacement finally arrived from the UK, six months beyond my normal tour period, I was then free to pursue my own ends over the two weeks that the trooper required to go to Cape Town and return. With so much spare time on my hands throughout my posting, I was able to indulge in tennis and swimming on Lumley Beach as long as the CA Regt. knew where I was and could do a quick recall.

Having had so much practice at tennis, I became fairly adept, and trounced the local opposition until we were visited by the West African Artillery Inspector, also with time on his hands and looking for a game.

鈥橲ee the radar mechanic鈥�, they said! He beat the pants off me! He saw how crestfallen I was, and later let drop that before the war he had been Kent County Champion. That restored my ego.

漏 Copyright of content contributed to this Archive rests with the author. Find out how you can use this.