How it all began (Pt.1)


  • Republished here, in 1987 Peter Russell wrote a number of articles for RCM&E covering the history of radio control developments.  

There can be few who enjoy the present state of R/C, with radio gear that is sophisticated, reliable and cheap, easily built models with good performance, a large range of supporting fittings and gadgets, who have much idea of how this Utopian state of affairs came about.

To cover the history of R/C fully would be far too long and tedious for an article such as this, for there were a number of periods when no apparent progress was being made. But an attempt to pick out the more interesting ‘high spots’ might reveal some ‘little-known-facts’ about our favourite hobby.

The story starts much earlier than you might think, long before successful powered model aeroplanes appeared. For instance, ‘Straight and Level’ reader R. P. Meredith of the North Birmingham Model Power Boat Club sent a cutting from a national daily newspaper, the ‘Daily Sketch’ dated June 1910, which describes how two engineers, Messrs. Healey and Roberts had been giving demonstrations at Dagenham Lake, Essex, “of their power to steer electrically propelled airships, boats and torpedoes by wireless waves”. Illustrations showed ‘the operator standing by his induction coil on the bank and guiding a launch about the lake’ and ‘Mr. Healey’s model airship which was controlled in its movements about a shed’.

During World War I there were experiments in England to produce radio controlled flying bombs, masterminded by prolific scientific experimenter and writer Professor A. M. Low. Some indication of the importance of these projects is given by the fact that, as well as the Royal Aircraft Establishment, de Havilland and Sopwith were also involved. As far as can be ascertained, none of the projects were successful and the same can be said of experiments in the USA and Germany.

In the twenties, ‘Jane's pocket book of RPV’s’ tells how RAE finally got an R/C aerodyne to fly more or less successfully and how three Fairey IIIF's were modified to fly by R/C. This involved increasing the dihedral to ten degrees ‘to enhance stability’ though we modellers know that the main reason would be to give yaw/roll coupling. The Fairey ‘Queens’ as they were known had simplified controls rather like our own ‘rudder only’ models.

At the same time, Major Raymond Phillips was touring the country giving displays with his electrically propelled and controlled model airship. Flown in large halls and cinemas, it used a spark transmission system (hence the German term ‘Funk-gerät’ - ‘spark apparatus’) radiating a vast amount of energy right across the spectrum.

Major Phillips airship and control gear was demonstrated on many occassions in the 1920s.

The controls were selected by a sequential system and must have been the first application of control 'puffers' now used on such things as Harriers and Space Shuttles, though in the Major's model separate motors and propellers were used for the ‘up-down’ and ‘left-right’ functions.

R/C modelling as we now know it started in 1934 when Hollywood actor Reginald Denny, a keen modeller and model manufacturer, produced a nine foot model of a type that would be instantly familiar to today’s vintage fraternity. He offered it to the Army as a practice target for AA gunners but it crashed on its demonstration flight. Later, a simpler, welded-steel tube affair with a five channel tone filter radio - a type later perpetuated in Graupner’s ‘Varioton’ equipment - was much more successful and was built on a vast scale. The controls were rudder and elevator, two channels for each, with the fifth channel operating the engine cut-off and with a 24 foot recovery parachute deployment, though the drone could be landed normally on its wheels and frequently was. As the government contracts rolled in, the plant was expanded to become Radioplane Inc. and one of its employees was one Marilyn Monroe, working in the wing assembly section, later to become the famous actress. Not a lot of people know that! Reginald Denny was once a member of the St. Albans Club - and not a lot of people know that either!

From the point of view of the amateur aeromodeller, articles were appearing in Model Airplane News as early as 1936, when glider enthusiast Ross Hull produced a glider with the earliest known application of the escapement principle for sequential rudder-only control. In the interests of simplicity and reliability, the escapement was powered by a twisted rubber band, a system that was to become well known 20 years later. The model made more than a hundred flights but not without some fifteen crack-ups. However, the number of crashes due to radio control failure could be counted on the fingers of one hand.

The following year saw the introduction of a contest category specifically for R/C models at the American Nationals and this is when things started moving. Chester Lanzo, one of the leading lights of the very successful Cleveland (Ohio) aeromodelling group, and a personal friend of the writer, won this first event with a model almost identical to the Lanzo Record Breaker popular with today’s vintage fans. Chester once confided that the prize was really for models carrying radio equipment because the control bit was somewhat unconvincing!

The following year saw the Good brothers, Walt and Bill, win with a model now displayed in the Aerospace section of the Smithsonian Institution in Washington. Walt was already well-known and successful in the free-flight sphere and had won a number of contests with his 6ft Brown powered Guff. One of his prizes was lunch with Reginald Denny! The 8ft R/C model was on similar lines to the Guff, with the fuselage based on Charlie Grant's then highly regarded ‘low-centre-of-lateral-area’ theory in which the fuselage side areas were arranged to give spiral stability, hence the rather queer shape. The wings reverted to straight dihedral and used the Grant X-8 wing section, a feature of proven merit as opposed to the ‘Low CLA’ theory which was somewhat less convincing.

The writer’s replica, built some 25 years ago to the ‘words and music’ in a 1941 ‘Air Trails’ magazine has done a vast amount of flying and has never given any need to be modified in any way, except in the case of the undercart location which was altered to remove ground looping tendencies. This was in line with the Goods own mod, incorporated in 1946.

The radio gear was, of course, all home made stuff, using mostly normal radio components of the time. The transmitter was a push-pull affair feeding a big dipole antenna and powered by a rotary generator. The receiver was a one valve super-regenerative type operating a polarised relay, which worked on a current change of no more than one or two milliamps. This operated in turn the sequential rubber powered escapement that deflected the rudder. To save weight the valve base was removed and the exposed filament, grid and anode wires were soldered direct into the receiver circuit. One of the big snags with these state-of-the-art radios was that they had to have a high voltage supply for the anode circuits - usually about 45 volts for the receiver and anything up to 180 volts for the transmitter - hence the rotary generator. Nevertheless, with a lot of ingenuity, the Goods got the airborne R/C weight down to little over one pound, a high proportion of which was the high tension battery. Slightly unusual by later standards was the fact that both the escapement and its rubber band power unit were mounted inside the fin.

For the 1938 Nationals winning flight, the tailplane got knocked out of adjustment during the take off, so, like Chester Lanzo’s winning flight of the previous year, there was some doubt whether the radio was really controlling it. At the 1939 Nats, however, there was no doubt, and the Goods put on what must have been just about the first really convincing display of R/C flying in public. With Bill on the transmitter the model flew around, did some figures of eight round designated markers and landed quite close to the transmitter. Modeller - as opposed to ‘scientists’ - radio controlled flying had arrived.

The outbreak of the second ‘Great War’ might seem a convenient point to make a break in this treatise, but from a technical point of view this is not the case. Although the Americans had another two and a bit years of peace, and two more Nationals, there seems to have been no great progress in R/ C until they too suddenly found themselves pitched into the conflict and they, like us, had very little time for aeromodelling.

So the story resumes in 1948, more or less where it stopped in 1939. With the great advances in radio and electronic technology, almost all modellers thought that real radio control - reliable and sophisticated - would be the universal mode of aeromodelling when peace returned. But it was not to be. Very little of the technology that produced VHF R/T, radar and so on could be economically adapted to ‘our’ sort of radio and most of us were very depressed to find that aeromodelling was very much “as you were”.

An eight reed relay by Martin Pfeil, Germany. Adjustment was made by bending back the contacts.

The Good Brothers dusted off their R/C model, sometimes referred to as ‘Big Guff’ (though I once called my replica ‘The Goodship’ and the name stuck) and won the R/C event at a post-war American Nationals. Jim Walker appeared with a multi-channel model, which featured what must have been
the first steerable nose wheel, but details of actual flight performance are sparse and sometimes contradictory.

By 1950 there were several commercial R/C units offered for sale in England, mostly single channel, not vastly different from the Goods’ original conception. They all worked - sometimes.

R/C was featured at the British Nationals for the first time in 1950 but on the day of the contest a gale was blowing. I’m sure winner ‘Chuck’ Doughty will not think me unkind when I suggest that the repeated loops performed by his ‘Stentorian’ as it blew away were not exactly as scheduled.

By this time many modellers were beginning to think that single-channel was not what they really wanted and the Kingston firm ‘Electronic Developments’, mainly known for its range of no-nonsense diesel engines, offered for sale a multi channel ‘reed’ set which enabled modellers to get what is now regarded as minimum control - rudder-ele-throt - as we say.

These reed sets had an electromagnetic gadget resembling a comb with progressively shorter teeth under which an electro-magnet was fed with different tones developed in the transmitter. When carefully ‘audio tuned’ each tone caused a selected ‘tooth’ or reed to vibrate. This made a connection to a relay, which in turn operated the control servo. The servos were often home-made using the almost universal ‘Mighty Midget’ motor made by Victory Industries of Guildford and many of us had our first practical and moderately reliable flights with this set of gear.

Chester Lanzo's 1937 US Nats winner.

The Americans too were developing reed gear and by the mid fifties, with the arrival of the Bonner ‘Duramite’ servo, R/C had developed from the pursuit of a few isolated experiments to an increasingly popular segment of the modelling hobby. However, the real ‘breakthrough’ in R/C came with the development of the transistor (deleting once and for all that troublesome high voltage battery).

We saw earlier that the first twenty years or so of radio control, in modellers’ terms, made little apparent progress and that the immense leap forward in radio and electronic technology during the second great war did not seem to be transferable to the modelling scene. There were several reasons for this:

First was the fact that compared with the established modelling regimes, free-flight and, after the war, the immense craze for control-line flying, there were really very few active R/C modellers. And many of those soon became disenchanted with the sheer unreliability of most early gear and decided that two wires between the model and the driver was a lot better than lots of bits of wire soldered to valves, relays and several batteries.

Second, there was the belief, widely held in ‘full-size’ circles, that it would prove impossible to ‘hand fly’ remotely piloted vehicles and that the latter would have to be flown by an automatic pilot with the radio link used to programme it. Most of the full-size radio-controlled aeroplanes and drones then flying worked on this principle. That is, the auto pilot did the flying and the ground controller simply told the auto pilot what he wanted. This was to a certain extent perpetuated in modelling where inherent stability took on the role of auto pilot while the control - almost invariably limited to a very de-sensitised rudder – ‘interfered’ with the inherent instability. Even when it worked, it was clearly not everybody's idea of what radio control should be.

An idea of the state of model R/C in the immediate post-war years was the often heard ironic joke: First modeller: “Is that model radio-controlled?”. Second modeller: “No, it's just naturally unstable.”

By the early fifties things began to move. Nearly everybody was using one channel only, often with ‘self’ built gear, but there were many enterprising attempts to extract more control from the single ‘on-off’ channel. One such was the ‘Ruddevator’ in which a surface looking like a rudder tab was pivoted about its CG in such a way that on ‘no signal’ the surface rotated in the air stream, rather like a small windmill immediately behind the fin.

The electro-magnetic actuator was rather like the almost universal electro mechanical escapement normally used to operate the rudder tab on a sequential scheme. But instead of directly operating the tab, it was arranged so that it could stop the whirling vane in one of four positions corresponding to:

1. Left rudder,
2. Up elevator,
3. Right rudder,
4. Down elevator and so on.

The humble escapement itself came in for some concentrated development too. First came the ‘selective escapement’ with the stops arranged so that one press always gave ‘left’ and ‘press-release-press’ always gave ‘right’.

On release, the control always reverted to neutral. A further development was to add a ‘cascaded’ escapement, triggered by the main rudder control. In this a ‘quick blip’ on the transmitter button caused the main escapement to perform a complete cycle but as long as the back contact of the relay was back in position before the cycle was complete a second escapement would be triggered. This would be a two position affair usually rigged to give throttle open and throttle shut. With a bit of practice quite a lot of performance could be extracted from this system.

A completely different approach was in the various pulse proportional systems some modellers favoured. In its simplest form this consisted of a rudder tab spring loaded to one side and an electro-magnetic device or electric motor - the ubiquitous Mighty Midget was the favourite - to pull it the other way. By pulsing the transmitter button at, say, four per second, a reasonably straight, if slightly ‘wiggly’, flight could be achieved. Releasing the button gave one turn, holding it down turned it the other way. Sounds crude, but it worked - in a fashion.

The next step was to produce the pulses mechanically and this was usually done by the system used in the Henschel 293 guided missile - the one that sank the surrendering battleship ‘Roma’ in 1943. In this system a rotating drum was partly covered by a tapered insulating material such that one end of the drum was fully conducting, the other end fully non-conducting and in the middle, 50-50. Intermediate positions produced a ratio of conduct/non-conduct proportional to the position of the pick-up wiper. With a suitable ‘pull one way’ spring-load, true proportional rudder control was achieved. However, if the rudder was moved by a rotating crank it was a fairly simple matter to couple an elevator tab to the crank as well. Then, if a 50-50 signal to no signal ratio were sent, the controls would both be neutral. As before, varying the ratio gave proportional rudder deflections but by speeding up the pulse, the crank would be oscillating nearer the bottom of its throw, giving ‘down’. Slowing down the pulse rate caused the crank to rotate its full switch, which gave (pulsed) ‘up’ elevator.

A Deans' reed unit for orbit and other sets. Adjustment is by screws, coil resistance was 3000 Ohms.

The system worked OK but the big problem, as far as the guided missile was concerned, was the difficulty the operator had in keeping track of it as it dwindled to a tiny speck. A flare in the tail gave positional information but no attitude ditto. So, as modellers down the ages have discovered, “If you can’t see it, you can’t fly it!”

However, getting back to the pulse proportional business, modellers of the fifties were quick to realise that the Hs293 system was practically tailor-made for ‘our’ purposes. And soon modellers all over the place were sticking triangles of sticky tape onto emptied out pen-cell cases and using a Mighty Midget to drive the drum on the transmitter and another to drive the crank in the model.

The name Charles Ryall immediately springs to mind in this connection. He was perhaps the most successful advocate of this system and put on many convincing public displays with it. Since it was impossible to completely eliminate the effect of the continuously pulsing controls, some wag once remarked, “Look out here comes the galloping major!” Which no doubt was the basis of the universally adopted title of ‘galloping ghost’, or for the more flippant, ‘perambulating poltergeist’.

A large section of the ‘interested parties’ considered all this development as a dead end and decided that multi-channel
‘full house’ control was what was needed.

Early experiments with the audio tone system had come up with a lot of difficulties but these were gradually developed out. The usual scheme used multi-tone (usually eight or ten) generators in the transmitter which could be selected at will - only one at a time to start with, however - and transmitted to the aeroplane’s radio which incorporated an electro magnet over which was positioned the appropriate number of different sized metal reed contacts, each of which responded to one of the tones.

The microscopic AC current passed through the reed contact was used to operate a sensitive relay, which in turn operated the final control actuator. This was nearly always some form of electric motor driven ‘servo’. (There were one or two notable exceptions such as the Stegmaier brothers who won some of the first R/C World Championships with a reed system operating pneumatic actuators).

The reed banks were early brought to high reliability and the arrival on the scene of the Bonner ‘Duramite’ servo took care of the actual push and pull part of the system. But many, the writer included, found that after the initial shine had been taken off the relay contact, these started to give trouble. Fifty flights or so was typical of the trouble-free period after which burnishing of the points with a bit of clean shim steel would restore normal operation for a bit. But inevitably, after a few more flights, the seemingly impossible situation of two bits of metal coming into contact yet failing to pass a quite small electric current would cause another good model to bite the dust.

An early Futaba leaflet showing the set up for both escapement and servo type controls.

No doubt the relays could have been developed but Bonner came up with ‘a better mousetrap’ in that the small AC current from the reed bank was fed into an electronic amplifier built into the same ‘Duramite’ servo and generated enough juice to operate the servo direct. This turned out to be the complete cure.

While all this was going on, other important developments were taking place. Transistors had been around for some time but early applications had been a bit cautious, often mixing transistors with thermionic valves in the same circuit. This seemed a bit pointless because the big advantage of the semiconductor was its ability to operate on low voltages. Up to that point the usual battery complement in an R/C model was formidable, typically consisting of a ‘U11’ cell (1 ½ oz.) to light up the filaments of the thermionic valves, the ‘A’ supply; a ‘B’ or ‘high tension’ battery, comprising two or three deaf-aid batteries to give 45 or 67½ volts and finally four ‘C’ pen cells to work the actuators. A total weight of between eight and ten ounces - quite a handicap, even if you ignore the failure possibilities of all the extra bits of wire and connectors involved.

This untidy situation was rectified in two ways. One, ‘All Transistor’ sets were developed to work the receiver on four pen-cells and these could also power the servos. Second, rechargeable alkaline cells were miniaturised to a size not much bigger than a £1 coin and four of these proved to be the complete answer to the reliable airborne power supply problem.

The introduction of the ‘Transmite’, as the transistorised ‘Duramite’ was called, required the addition of a further cell (i.e. five in all) to bias the amplifier, but these five were still a lot lighter and more compact than the old arrangement and, above all, much more reliable.

So it came about that, by about 1963, the amount of control and reliability that we had dreamed about for years was finally realised. The writer had two sets at this time, an American ‘F&M’ ten channel (rudder-ele-throt-ail-eletrim) and an O.S. six (rudder-ele-throt: these controls were all non-proportional ‘bang-bang’ and so needed two channels for each control) which over several hundred flights never malfunctioned - 100% reliability and you can't better that!

At the same time pilot skills had improved out of all recognition. As early as 1968, using all home made gear, valves, A B C batteries and all, Chris Olsen and his buddy Stewart Uwins (later to became Mr. Skyleader) were putting on scintillating displays of aerobatics, the first real radio control many of us had ever seen.

This finally disposed of the myth that you needed an auto-pilot to do the actual flying of your radio controlled aircraft. It also disposed of the often-heard criticism that you couldn’t fly accurately with ‘bang-bang’ controls, and that true proportional control was essential. Chris Olsen continued to do well in international competition using ‘bang-bang’ well into the proportional era, which started in the mid-sixties. He was often quoted as saying that he could get proportional control by pulsing his thumb - much simpler than the electronics needed to do it automatically! However, full proportional control was bound to come…

(Continued in Pt.2).