Sutherland

Home

 

THE EVOLUTION OF RADAR – A PERSONAL VIEW

 

As you will all know, this year is the 100th anniversary of the birth of Marconi and therefore as a Marconi man I am very proud, both for myself and for the Company, that you have chosen me to be your President. Now I don't expect many of you - other than our very distinguished colleague, Eric Payne - will have heard Marconi speak, so I thought as an introduction you might like to hear his voice. I have an interesting transcription of his speech made from the yacht Elettra in the Mediterranean, and transmitted to Australia by short wave on the occasion of the opening of the Sydney Exhibition of 1930.

 

I would now like to turn the clock back, even further, to June 20th 1922 when Marconi addressed a joint meeting of the American Institute of Electrical Engineers and the Institute of Radio Engineers, and received the latter's Medal of Honour. That was one of several occasions on which Marconi predicted the development and utilisation of radar as we know it to-day. Quoting from the published proceedings of that meeting, Marconi said:

 

"In some of my tests I have noticed the effects of reflection and deflection of these waves by metallic objects miles away. It seems to me that it should be possible to design apparatus by means of which a ship could radiate or project a divergent beam of these rays in any desired direction, which rays, if coming across a metallic object, such as another steamer or ship, would be reflected back to a receiver screened from the local transmitter on the sending ship and thereby immediately reveal the presence and bearing of the other ship in fog or thick weather. One further great advantage of such an arrangement would be "that it would he able to give warning of the presence and bearing of ships even should these ships be unprovided with any kind of radio....."

 

It was a long time before these ideas were really put into practice, and it was about 1935 before any real progress was made in radar - or R.D.F. as it was known in the early days - indeed the name Radar was officially adopted on this side of the Atlantic about the middle of 1943, and I remember a signal from Admiralty to all ships to say that in future R.D.F. was to be known as Radar.

 

The earliest work grew out of the experiments performed at the Radio Research Board on ionospheric sounding in the very early 1930's. Short bursts of radio waves – pulses - were transmitted overhead upwards from a fixed aerial and the reflection was received on the ground in a nearby receiver. The time of transit was measured using a cathode ray oscillograph and the distance was calculated. The timing was very simple; as the pulse was sent out, the spot started to move across the screen, and as a reflected signal was received, this was used to deflect the spot vertically in a "blip" on the screen and mark the position of the ionosphere. The oscillograph screen was calibrated using an accurate oscillator to give the speed of the trace and thus the distance of the echo. The day an aeroplane flew through the beam, and reflected energy back to the receiver and caused a "blip" on the screen corresponding to its range, that was the day practical radar was born. Indeed, it was the late Sir Robert Watson-Watt, who I understand has actually addressed this Society on the subject, who foresaw the possibilities and exploited them.

 

I would like now to develop the story of radar from those earliest days to the present along two parallel tracks - firstly, the techniques which have been developed over the last forty years since those first experiments, and secondly, the spread of operational use of radar over the same period.

 

This must obviously be a personal view of radar.

 

SLIDE – Block Diagram

 

I thought first we should consider radar in its component parts - the aerials, the transmitter-receivers, the means of extraction and presentation of information to the user etc. The earliest radars were, of course, just these elements in their most rudimentary form. As we go along I would like to consider some of the engineering factors concerned, not in a deeply technical or theoretical way, but in a manner which I hope will make it interesting also to engineers outside the electronics field. First, let us consider transmitters and receivers. Most radars operate, as did the very earliest Watson-Watt measurements, on pulse modulation. We generate repeated pulses of radio frequency lasting for a few microseconds, that is a few millionths of a second. It is interesting to observe that a pulse of this kind when examined in such a way as to reveal its true frequency appears not as a single frequency but as a spectrum of frequencies centred on the basic frequency of the transmitter. Indeed, if we wish to analyse the spectrum, as you will see in the slide, then this is the Fourier transform of the original pulse and has the characteristic sine theta-over-theta shape.

 

SLIDES – Pulses, Spectrum sketch

 

I would like to examine the factors which determine the choice of the wavelength/frequency of the radar transmitter. The shorter wavelengths will give a narrower beam and therefore better discrimination with aerials of manageable size. However, their  performance in bad weather will be worse; Rayleigh's law states that the reflection from a conducting sphere is inversely proportionally to the fourth power of the wavelength and thus unwanted reflections from rain drops are less the longer the wavelength. In practice, in the early days of radar there was not a great deal of choice in wavelengths anyway because of the difficulty of generating power at the shorter wavelengths, and the first radars used conventional power valves which had been developed for broadcasting and long-range communications, and therefore the band of frequencies available ranged from 40 Mc/s to at most a few hundred Mc/s. There was then a tremendous breakthrough in the invention of the cavity magnetron in the early months of the war which made the use of radar at centimetric wavelengths possible. It was remarkable the speed with which a laboratory result was translated into hardware and equipment designed, produced and introduced into service at sea and in-shore stations.

 

The magnetron worked as follows - the "block" consisted of a number of resonant cavities of very high Q; there was a central cylindrical cathode which emitted electrons, and a high voltage applied to draw the electrons from the cathode to the block. At the same time a very strong fixed magnetic field was applied along the axis of the cathode; the combined effect is to move the electrons in a curved path, as seen in the slide. When the field and voltage are right, the effect is to extract power at D C from the electron current by exciting the resonant circuits in such a way as to build up fantastically high peak power at radio frequencies. It is as though the oscillations in these cavities were a swing, and the majority of the electrons hit the swing just at the right time to make it go higher and higher.

 

Actual Magnetron: This is what a magnetron actually looks like and the slide shows some early examples.

 

SLIDES – Sketch, Photograph

 

Later came the Klystron which was capable of use, both as a power amplifier and as an oscillator.

 

SLIDES – Diagrams- klystrons, Photograph

 

Again, this derived its amplification or its oscillation power from the transfer of power from the electron beam to resonant cavities. In this case it was by bunching of the electrons in the beam, a process called velocity modulation, which enables the power to be developed. In low power applications, a tube was devised which used a single resonant cavity by reflecting the velocity modulated beam back through the same cavity. This was called the "Reflex Klystron" or Sutton Tube. Thus recapping, we have the special characteristics of a radar transmitter pulse shown in the slide. This shows the definition of peak power, mean power, pulse width and pulse repetition, frequency of transmission.

 

SLIDES – HF pulses, Block diagram of modulator

 

In order to drive the output stage of the radar, it is necessary to generate DC pulses with amplitude of many thousands of volts and an instantaneous current of many amperes. This presented, and still presents, fascinating engineering problems. The pulse must be carefully shaped so that it has an accurate rate of rise, and a flat top and extremely good stability of timing. The problems to be solved in the generation and control of high voltage, in screening, in cooling and in ensuring mechanical and electrical stability are manifold. In general basic elements of a modulator are now a high voltage rectifier to provide the DC power, a modulator switch which is triggered to initiate the leading edge of the pulse, a pulse forming network which you will see in the slide, and which, in essence, is a synthesized part of a transmission line which determines the shape and duration of the pulse, a pulse transformer which takes the pulse up to the final required voltage. It may be interesting to see a few pictures of transmitters through the ages.

 

SLIDES – Type 79, 271, SR1000, 84, 992Q, S2020, S2011

 

The use of these extremely high frequencies posed particular problems in receivers; at tens or even low hundreds of megacycles, special low-noise amplifier valves were being developed; disc triodes and miniature triodes and tetrodes etc. At thousands of megacycles there was no possibility to use conventional means and we reverted to the crystal and cat's whisker, but in a novel form.

 

SLIDES – Crystal diagram, photogrpah

 

The crystal was used as a frequency mixer in a super-heterodyne receiver. By mixing the signal frequency with a local oscillator frequency we got an intermediate frequency in a manageable band. Indeed, one very important and interesting existing design of wideband amplifier at 45 Mc/s for pre-war television made an ideal IF amplifier for the first centimetric radar

.

SLIDE – IF strip

 

Today, there are enormous improvements - the ability of a receiver to pick out the wanted signal from a background of noise - (known in engineering terms as the "noise factor") - is vital to radar performance.

 

SLIDE – spectrum diagram

 

Looking again at the spectrum - the receiver must have sufficient bandwidth to cover the envelope of the spectrum but not be so wide that it introduces too much noise. Thus we have a means of generating a signal and picking it up when it is reflected from the target. But what about forming the radar beam? We could talk for a week on this subject and still not exhaust it, but I will try and pick a few key points out in ten minutes.

 

When radars operated at frequencies of 40 or 90 200 Mc/s, i.e. wavelengths of a metre or two, the aerials looked very like television aerials to-day. Some were able to be rotated at the ship’s masthead, or on "mattress" structures, but others both in aeroplanes and small ships were fixed and the target was detected by swinging the whole craft to the direction of the target.

 

Slides – 965, 281

 

I have a set of slides of the first radar ever to go to sea in HMS SHEFFIELD, just before the war. It was Type 79 operating on 40 Mc/ s, 70 kW Peak, 50 pps, Pulse length 8-30 Microseconds. It was designed to give about 15 minutes warning of air attack and indeed achieved that throughout the war.

 

SLIDES – Type 79

 

With "the advent of centimetric wavelengths, the whole design of aerials changed and became quasi-optical, and aerials began to look more like searchlights. For the first time one has the concept of a primary feed - or point source - feeding a reflector like the bulb of a car headlight. This technique is still widely used to-day. The wider the aerial, the narrower the beam. Early aerials were the same size in each direction and gave a pencil-beam; obviously in different application the beam needed to be shaped. In a ground radar one might want to cover up to say 20° vertically but have a beam about 1° in the horizontal plane. Thus aerials of different dimensions in the two planes, and with more complicated feeds, were necessary. I will mention this in connection with operational applications later.

 

SLIDES – 282, 271, 273

 

Looking at the engineering aspects of aerial design - these are threefold - firstly, the mechanical design of massive structures for rigidity and survival in the total environment. In the model here you see different approaches to this problem.

                                                                                               

Secondly, the electro-mechanical aspects of the turning gear, the data take-offs, the movement of individual parts of the aerial etc. and finally the radio frequency design of the aerial itself. Looking basically at this as a problem in physics, we consider an aperture illuminated with radio frequency energy with a particular distribution of phase and amplitude. Normally this will be a flat or uniform phase front and the amplitude can be controlled. We get back to our old friend the Fourier transform - the ultimate aerial pattern is the Fourier transform of the illumination, and with uniform phase and amplitude we get our old favourite sine theta-over-theta shape

 

SLIDE – polar diagram

 

This of course means that we get an aerial pattern with more than one peak - or with sidelobes, and the aerial designer aims to cut these down to a negligible level.

 

SLIDE – polar diagram improved

 

This is done by shaping the primary pattern in amplitude, whilst maintaining phase across  a uniform/front. The real skill in the aerial designer is to get the nearest to perfection in a compromise between the conflicting requirements.

 

SLIDE - Block Diagrams - Common Aerial, Duplexer

 

You will see around us various types of aerial. In the early days the transmitter and receiver aerials were quite separate; a particular break-through came during the war when by various techniques it became possible using initially forms of spark gap but, later, more sophisticated devices called duplexers, to operate efficiently using the same aerial for transmitting and receiving. I have described pulse radar so far - there are other techniques for example, using the Doppler effect, but time really doesn't permit me to go into that at this stage.

 

So now we have the aerials, the transmitters and receivers - how is the information going to be used?

 

SLIDE – “A” trace

 

In the earliest days, the same techniques as ionospheric sounding were used and the operator would point the beam at the target, looking for the maximum 'echo' and then read the range off on the cathode ray tube scale. This scale was calibrated using an oscillator to generate 'pips' corresponding to 1000 yards or mile intervals. The pulse would travel 1000 yards and back in 6.102 microseconds, and this corresponded to an oscillator of about 163 kc/s.

 

SLIDE – PPI trace

 

The next improvement was to use the echo to brighten the cathode ray tube rather than to move the spot, and then to rotate the trace about the centre, in exact synchronism with the aerial. Thus the bright spots, corresponding to target positions appeared in a true map or Plan Position Indicator. By having a tube with a long persistence - a phosphor which continued to glow - a picture was sustained from one revolution of the aerial to the next.

 

The tricks which can now be played with a modern display are beyond the wildest dreams of the users of thirty years ago. The position of the target can be extracted instantaneously by electronic circuits with much greater sensitivity than the human being, and very substantial data written on the tube electrically, and all the operational functions carried out by computer - the human operator merely being there to supervise. I will not dwell on these developments at this stage - I would like you to see a demonstration afterwards which will illustrate some of the possibilities. Meanwhile I have some very fascinating film - the radar aerial will rotate once every 15 seconds or less. If now we photograph the display on the basis of one frame per revolution and play this out at 16 frames per second, we can project an enormously speeded up radar picture.

 

FILM

 

Speaking personally, I have earned my living out of radar for thirty of the last thirty-three years, in virtually every possible capacity, so perhaps you will excuse me if I may dwell for a minute or two on the commercial aspects of radar. Although the British Government are an important customer, and a significant influence in radar for air defence, for naval and military systems, for air traffic control, for the police etc., nevertheless/the export market has been vital, and this industry has always been a very significant contributor to our balance of payments. Many of the radars I have talked about so far, and indeed the whole sequence I am about to show you, have been developed with Company funding, and over 50% of the output has been exported for two decades. Thus very careful planning of the product is vital, to be competitive with other countries, many of whom are graced with much more help from their government in their export marketing than we ourselves have received, or are likely to receive. Our development philosophy is to produce and keep up to date an attractive and competitive range of equipment capable of satisfying the majority of world markets.

I would now like, to talk about a few of the operational aspects of radar - these cover warships, ground air defence, air traffic control, meteorology, airborne radar, radars for the Army, for guided weapons etc. Perhaps the easiest way to give a brief sketch of all these aspects is to look at the models and pictures and show the different aspects of radar design for different applications. From both the engineering and the commercial point of view, the key factors are -

-                                          cost effectiveness

-                                          reliability/ease of maintenance

-                                          operational flexibility

In radar for military applications the designer has an additional problem in that the enemy, through jamming, will try to deny information to the user

 

MODELS

 

Let us look at some different radar types.

 

A specialised aerial for air traffic control S654 designed to discriminate against ground clutter.

 

Another aerial, primarily for air traffic control, S264, operating at the longer wavelength of 50 cms, for good discrimination against weather.

 

These both give plan position; to measure height, S269, we have a typical "nodding" height-finder - the aerial is pointed in the direction of each target in turn and "nods" to measure the accurate angle of elevation. These slides show operational use of these aerials.

 

SLIDES – S650, S269, S654, ST801

 

When we want to maintain accurate positional information continuously on a target, we use a pencil-beam aerial, which automatically acquires the target and locks on to its position; this is known as a "tracker"

So far we have looked at static or fixed radars. There are many operational uses for mobile and transportable radars which can be driven across country, carried by transport aircraft or lifted by helicopter.

 

SLIDES – Anvil cabin, S600 convoy, S600 surveillance, S60 heightfinder

 

Radars have also been devised which measure height range and bearing on all target simultaneously - these are called 3-D radars.

 

SLIDE 40T2

 

In the example shown in the slide, this is done by "stacked beam" techniques. In effect there are 12 radars in one, and by comparing the signals in each beam, the height of each target can be acquired. From the sublime to the ridiculous - here in my hand is a complete radar - it has power supplies, aerial, transmitter, receiver and a means of presentation. I will finish by showing you a very sophisticated radar with many facilities, weighing 3 kilograms - 7 pounds -  which one man can carry and which will give excellent surveillance, and accurate ranging performance on a crawling man or a moving vehicle at substantial distances.

 

Finally, it is most important to put all these various functions and pieces of equipment together in the most economical and efficient way; this is known as System Engineering. It involves studying the operational requirement - establishing the right equipment to do the job - defining interfaces between all the elements of the system, and above all producing specifications by which it can be tested and shown to be performing the job it was defined to do. Furthermore, to be effective in business in the world there is a multitude of ancillary functions to be performed to give the customer a total turnkey service. The provision of buildings, power supplies and other utilities, such as water, fuel, etc., domestic accommodation, roads, security, are important in a total project. Furthermore, one must provide the customer and his men with installation and commissioning services, documentation, full test gear and spare parts complements, training of personnel, maintenance and support in all technical areas.

 

Well that concludes a lightning survey of the evolution of radar; I have certainly left out more than I included, and vast areas of technology and application have been overlooked.

 

I would be very happy to answer any questions, and then would like to offer to anyone interested a demonstration of the modern display system.

 

Sutherland

Home