Origin and Development of the Roesch Talbots



Browse pages
Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


Current page


(Continued from November issue)

The 90 Model

The performance of the new car was found to be so good that one of Talbot’s enterprising agents, Arthur Fox, of Fox and Nicholl, Ltd., suggested to the firm that such cars would have a good chance in competitions and that they should enter a team of three cars in the Double Twelve-hour race at Brooklands in May, 1930. Roesch did not believe the car good enough for racing, but saw at once an opportunity to further develop the car and advertise its reliability and efficiency as never before. He immediately set about getting improved performance. The engine had been designed to stand Diesel compressions for future developments and this provision was immediately put to the test much sooner than anticipated. Without making any other change in the engine than fitting a domed crown to the two-piece piston and grinding down the cylinder head to get a 10 to 1 compression ratio, the maximum power output was increased to 93 b.h.p. at 4,500 r.p.m.; 90 b.h.p. were available at 5,000 r.p.m. The standard single updraft Zenith carburetter was used with a larger choke and jets, no change was made to the Delco Remy ignition, but special Champion plugs were found necessary. The performance was satisfactory, but the exhaust valve was the limiting factor to engine life. The shape of the valve head was modified and tests revealed that the engine could maintain full power for over twenty-four hours. The cars were then thoroughly tested at Brooklands which was without equal for finding any weakness as it was the most destructive racing circuit of all.

The chassis was fitted with a 25-gallon petrol tank designed to give a streamline tail to the four-seater touring body. Thus air resistance had been reduced and the weight of the car was kept as low as possible despite the heavy chassis frame and axles designed to carry the more usual capacious saloon body.

Thus was born the new model which Roesch called the 90.

On the day of the race the three cream-coloured touring Talbots were the sensation of the event (Motor, May 13th, 1930) and surprised everybody by their silent, effortless and reliable running amongst the fastest racing cars of the day. They did not, however, finish the race for on the eleventh hour of the first day two of the cars running closely in formation touched one another and both crashed with fatal result. The third car driven by the Hon. Max Aitken was therefore withdrawn as a result of this tragedy.

Le Mans 1930

But next month the Le Mans 24-hour race was to be the opportunity for two of these low-priced comfortable four-seater cars to show their worth.

Amongst the starters were a supercharged Alfa-Romeo, six Bentleys, three of them supercharged 4½-litres, one Bugatti, one supercharged Lea-Francis, one supercharged Mercedes, and two Stutz.

After a day-and-night struggle on the old circuit, harder than the modern one without the Pontliene hairpin corner, the race ended with a resounding British victory. Two 6½-litre Bentleys, driven by Barnato and Kidston and by Clement and Watney, came first and second and the two 2¼-litre Talbots, driven by Lewis and Eaton, and by Hindmarsh and Rose-Richards, came in third and fourth, the leading Talbot having the distinction of being classed first in the Biennial Cup for efficiency.

This wonderful result was obtained with a standard vehicle of 9 ft. 6 in. wheelbase and a 4 ft. 7 in. track and an engine half the size of others in the race.

The power unit was not only remarkable for its high output, flexibility and silence, but more so for its diminutive size, extreme simplicity and neatness, which left no doubt that this engine was designed for large production with none of the expensive features which were deemed to be essential then to achieve success in competition.

With push-rod-operated overhead valves disposed vertically and in line and a single updraft carburetter, the engine performance was obtained by simply increasing crankshaft speed and thermal and mechanical efficiencies without the complications and cost of overhead camshafts, multi carburetters or a supercharger.

Here was, before its time, the engine formula for today’s lightest, cheapest and most popular cars. It had been realised by individual thought bent on solving problems in the most economical ways.

Only the utmost care given to the smallest detail and an enthusiastic team working in complete harmony could have secured at the first attempt the efficiency which Charles Faroux, the famous French authority, described in the Continental press as “absolutely stupefying and which had never before been seen.”

The Motor (July 1st, 1930) claimed the victory to be a triumph for Clement Talbot Ltd. and a personal one for Georges Roesch. In the following Irish International Grand Prix, Ulster Tourist Trophy and the Brooklands 500-mile races these Talbots were first, second and third in their class, all three cars averaging over 103 miles per hour at Brooklands. To finish this successful year the following British and International records were secured at Brooklands on October 23rd: 200 kilos, 200 miles, 500 kilos, 3 hours, 500 miles, 1,000 kilos and 6 hours at 104.77 m.p.h.

The 3-litre 105 Model

This first season’s racing experience showed that however efficient the 2¼ non-supercharged engine was it was handicapped by having to compete in the 3-litre class, particularly with a chassis built for a large saloon body. So to improve the power-weight ratio Roesch designed the 3-litre engine which he called the 105.

As the new unit was to be interchangeable with the other two smaller units in the same chassis it had to be very compact. With improved foundry practice the same spacing for the cylinders was used as on the 90. The bore was fixed at 75 mm. and the stroke was increased to 112 mm. In order to provide larger valves in the cylinder heads they remained vertical, but were staggered. In this way it was found possible to shorten the exhaust port to a minimum and appreciably reduce the hot surface exposed to water cooling and to ensure that the coolant on entering the underside of the head from the cylinder would be forced under the port of each exhaust valve and right round its seat. Thermo-couples fitted just underneath the seat showed that its temperature never exceeded 104 deg. C. The three siamese inlet ports were consequently longer, with a large smooth curve starting at the valve seat and emerging on the same side of the head, but above the exhaust ports.

The valve gear was further improved to operate if need be faster than the 90. The camshaft was first case-hardened and ground and to increase further the load capacity of the cams the latter were hard chromium plated. A very light tappet of small diameter could then be used to operate the large valves through thin solid and barrel-shaped push-rods and short inlet and long exhaust rockers. These very light and stiff rockers instead of articulating on a knife edge were doing so on an adjustable ball-ended threaded stud. As the push-rod was also ball ended, the rocker was located in two directions by having at the valve-stem end a fork drilled to take an articulating pin provided with a flat in the middle to rest on the valve stem and be located sideways by it. This method ensured against any possible misalignment.

Each valve had three concentric springs retained by a one-millimeter thick washer stiffened by six radial webs. This washer was retained on the valve stem by a short split conical collar. This webbed washer may sound a very difficult part to make, but Mr. Roesch had it finished-stamped in a coining die. Both valves were of a shallow tulip shape designed to avoid concentrated stresses anywhere. To this effect both valve stems were reduced in diameter on part of their length. The material was the same as on the 90.

The connecting-rods were of Swedish nickel chrome-steel of very thin but stiff H-section also finished-stamped in a coining die and lightly polished. Not only was this method considerably cheaper than machining the H-section, but the hard surface skin left un-disturbed was more resistant to fatigue. The seven-bearing crankshaft was similar in design to that of the 75 and 90 type, except that the main bearing diameter was increased from 60 to 64 mm. and the big-end from 43 to 48 mm. In most other respects the engine was similar to the 14/45 and 75-90.

The compression ratio was fixed at 6.5 to 1 and the engine gave a maximum of 100 h.p. at 4,500 r.p.m. for ordinary use. For competition the compression ratio was raised to 10.2 to 1 and the maximum power rose to 140 h.p. at 4,900 r.p.m. The dynamotor was of the same dimensions as the one fitted to the 90 and to increase its starting torque 24 volts were used. The battery consisted of two 12-volt units coupled in parallel. A relay switch was operated by the starter button on the instrument board to couple the two batteries in series for starting. A radiator thermometer and two-tone horn were added.

The same clutch was used as on the 75 and 90 models with the pressure plate springs adjusted to a higher pressure to transmit the increased torque. The same gearbox was also used to begin with. A silent third type was designed a little later for use on all models with a set of helical gears at each end of the box and engaged by dogs for third and top. The standard ratios were 3.4, 1.89, 1.36 to 1 with 4 or 4.363 to 1 in the rear axle. For racing, the silent third box had 2.466, 1.476 and 1.166 to 1 ratios and the axle ratios were either 4, 3.8, 3.66, 3.461 or 3.307 to 1. Tyres varied from 18 in. to 21 in. by 5.25 in. to 6 in. sections. The steering and back axle were also common to the 75 model.

The capacity of the brakes was increased by casting a finned aluminium ring round the thin brake drum pressed in high carbon steel. This new practice greatly increased heat dissipation and appears to be coming into favour at present. Steering-column control of the shock-absorbers was also introduced.

This new 105 car, the engine design of which had been started in December 1930, was ready early in May of the following year for the then most gruelling race in the world, the Double Twelve-hour at Brooklands organised by the J.C.C.

The start was aptly described by The Motor in the following terms: “Among the 48 cars which set off into the rain, the three pale-green Talbots were far and away the most impressive. Their getaway was a wonderful and beautiful thing. Hardly had the starter dropped his flag when the cars pulled away from the pits, swinging out like half a battle squadron, forming up into line ahead as they disappeared into the mist.”

The Earl of March and C. S. Staniland won a wonderful and well-deserved victory in a 746-c.c. M.G. Midget at 65.62 m.p.h. The three Talbots were first, second and third in their class and the leading one, driven by T. Rose-Richards and John Cobb, proved the fastest car in the race with an average of 79.29 m.p.h.

Then followed the Le Mans race on the old circuit, in which the 105, driven by Rose-Richards and Saunders-Davis came third, at an average of 73.46 m.p.h. and second on formula. In August H. E. Symons won a Coupe des Glaciers in the Alpine Trial. His was the only Talbot entered. In the Ulster Trophy race on the difficult Ards circuit Brian Lewis came in fourth behind three supercharged cars in a magnificent non-stop run at 77.13 m.p.h. Hindmarsh averaged 72.93 m.p.h.

Then came the Brooklands 500-mile race in which the Talbot 105, driven by Brian Lewis and A. O. Saunders-Davis, finished second at an average of 112.93 m.p.h. The three Talbots were first, second and third in their class and won the Team Prize.

In 1932 three 105 Talbots ran in the British 1,000-mile race at Brooklands. Saunders-Davis finished second at 95.43 m.p.h., John Cobb and Brian Lewis were fourth at 93.86 m.p.h. and Rose-Richards seventh at 91.23 m.p.h. The Talbots were first, second and third in their class and won the Team Prize.

The Le Mans race proved an unlucky one, for on the way there the car in which Georges Roesch was a passenger hit a tree at about 60 m.p.h. and this prevented him from seeing the practice. On the day of the race the only Talbot entered proved to be out of tune and considerable time was lost in adjustments, after which Brian Lewis and Rose-Richards, driving as brilliantly as ever, finished third behind two supercharged Alfa-Romeos.

In the Alpine Trial of the same year, the only team of three 105 Talbot open touring cars entered by Fox and Nicholl, Ltd., and driven by Brian Lewis, Tim Rose-Richards and Norman Garrard, won the coveted Coupe des Alpes in all classes above 2,000 c.c.— the Team Prize which had only once been accorded to cars made in Britain, Rolls-Royce before the first world war, an interval of nearly 20 years. It was also the first time that a British team had ever completed the trial without the loss of a mark.

Although the three Talbots entered in the 1934 Alpine Trial again all finished without the loss of a mark, 1932 was virtually the end of a most successful period of competition which had established Talbot cars as second to none in efficiency and reliability.

Various Models

We have seen how the 14/45 chassis was adapted to give the 75 and 90 models and the latter the 105 with long and short wheelbase of 10 ft. and 9 ft. 6 in. with high and low radiators. The 14/45 was also made with 9 ft. 6 in. wheelbase as the “Scout ” model which later became the 65.

A long wheelbase 3-litre with lower compression and high radiator was called the 95 and a variety of bodies were fitted on all these chassis.

Just as with chassis design Roesch had his ideas on bodies and believed in great visibility. The front pillars were cast in aluminium so as to be exceedingly thin, the sliding roof opened the car right from the top of the windscreen, the six lights were high and the rear window, was exceptionally large. All these features are in great demand today and are coming back to us from overseas.

In 1929 the London County Council wanted a very low-loading ambulance and Mr. Roesch designed a new rear axle to give a 4 in. lower floor. This was done by using a double helical gear reduction for each rear wheel thus permitting the use of an axle 4 in. lower than the orthodox one. This axle was an instant success and the type became the well-known London ambulance which did yeoman service from 1930 right through the Blitz until 1950 when they were replaced by the present Daimlers. The 95 limousine was fitted with the same axle and was unique for its flat floor at almost kerb height, its superb roadholding and absence of roll.

New Transmission

The most interesting change on the later cars was their new transmission evolved to match the highly efficient engines.

The Wilson epicyclic gearbox had already been introduced in this country, and General Motors had developed the synchromesh gear change. The question was, which of these two systems was worth adopting? The synchromesh had the great advantage of being cheap to make and another step in the development of the orthodox gearbox.

The Wilson epicyclic transmission was on the other hand a complete departure. It was then heavy and very costly, but it gave fingertip control and added possibilities for further development. In consequence Roesch decided to use the Wilson patents and design a lighter version of his own. It had been clear to him that, if weight was to be effectively reduced the best materials had to be used to the best advantage, and the utmost precision in manufacture had to be obtained to ensure that all the satellite gears should share equally the heavier load entailed.

All the internally-cut gears were finished after heat treatment of the nickel chrome-steel to give between 73 and 80 tons per sq. in. ultimate tensile. All the other gears were made of 5 per cent. nickel steel, case-hardened and ground, and ran on rollers. By reducing the number of articulated joints and their friction and also altering leverages in the operation of the foot-pedal gear change. it was found possible to considerably reduce the effort needed to press it down and eliminate all tendency to kick, whilst at the same time increasing the capacity of the hand brakes. The result was that the new design box proved capable of dealing with over three times the power for which the old design had been used. The gearbox was automatically lubricated from the engine by means of oil coming from the crankshaft and passing through the centre hole of the gearbox input shaft. To avoid any chance of seizure in case the car ran down a mountain pass with the engine stopped an oil pump in the form of a right- and left-hand thread was fitted to the output shaft and fed from a reservoir in the gearbox.

The new box was first applied without any main clutch, but owing to the cyclic variation in the engine rotation at low speed and the large number of gears in mesh, noise due to backlash could not be avoided when idling.

The Traffic Clutch

To eliminate this defect Roesch introduced his automatic traffic dutch operated by centrifugal action. It was composed of a finned-steel brake drum inside which two interconnected shoes would expand to grip the drum solidly when the engine had been accelerated above 900 r.p.m. A free wheel acting in a reverse direction enabled the engine to be started by pushing the car and to be utilised as a brake below the engaging speed of the clutch. The latter did not require any lubrication or any attention whatever. This simple foolproof mechanism gave a solid drive without any slip and automatically disconnected the transmission below 600 r.p.m. without inducing any drag.

With this automatic clutch the gearbox ceased to be lubricated by the engine crankshaft, the gland of which had given trouble. Instead, oil under pressure from the engine was led by a pipe to the rear of the gearbox and, through a bush, found its way to the centre hole drilled in both the output and input shafts. In this way the whole of the epicyclic gears and bearings were copiously lubricated from radial holes in these two shafts and no attention whatever was needed for the gearbox.

Automatic Preselection

In order to further simplify driving Roesch designed a very simple device to eliminate the need for preselection of the gears by hand when changing up.

Thus if neutral had been first selected by hand and the foot pedal pressed down the gears were in neutral but first gear had been automatically preselected at the same time. Without touching the preselector lever successive operations of the pedal would preselect and engage the gears up to top. This system proved very handy particularly when one was in a hurry, for with the accelerator down it was possible to accelerate through the gears without pauses and be certain that the gears were engaged in the right sequence automatically without the possibility of making a mistake, as sometimes happens when preselecting quickly by hand with all attention concentrated on the road.

A further improvement was made to enable third gear to be preselected automatically when changing into top. Thus, without moving the hands from the steering, by merely pressing the gear-engaging pedal it was possible to change up or down from third into top and into third again as desired. This fascinating device was only ready for production when the manufacture of the cars ceased.

The 110 Model, 3.3-Litre

The last model to be made was the 110 which was identical to the 95 and 105 except for a slightly larger engine with 80 mm. bore instead of 75 mm. and the same stroke of 112 mm., giving 3.3 litres. The smaller water passages between the cylinders were obtained by more accurate casting methods.

The saloon car as sold to the public had an engine giving a maximum of 123 b.h.p. at 4,500 r.p.m. with a single carburetter, and a maximum b.m.e.p. of 126 lb. per sq. in. at 1,500 r.p.m.

One of these engines fitted with a high-compression cylinder head and pistons was mounted in one of the Alpine Trial four-seater touring cars. Driven by W. M. Couper, this car won the Brooklands Jubilee Outer-Circuit race at 119.85 m.p.h. in 1938. The same driver put up a lap at 129.7 m.p.h. and a standing one at 104.85 m.p.h. The maximum speed was not far short of 140 m.p.h.— a remarkable performance for a car of standard construction 14 years ago.


Little has been said about this subject because design is the key to a firm’s prosperity and was particularly so in the case of Talbot without up-to-date equipment. But design and manufacture are in fact one and the same thing.

When Georges Roesch joined Clement Talbot Ltd. he found that drawings and specifications were crudely made, leaving vital design requirements to individual foremen to decide without an organised co-ordination. He immediately introduced a new system whereby everything was accounted for and strictly subject to his control as chief engineer. A laboratory was established to check supplies of materials. Parts which could safely be standardised were listed and so were materials with their chemical analysis, heat treatment, mechanical properties and the use they were suited for. This system, devoid of red tape, was so foolproof that a saving of time and material was very soon apparent, to the satisfaction of everybody.

As soon as an opportunity occurred, central planning of manufacture was introduced. A jig and tool designer and his staff together with rate fixers were engaged and trained by Roesch alongside his design engineers. In this way the same spirit and methods were introduced on the manufacturing side and very soon a perfect understanding was established in all engineering activities. It was this spirit of co-operation, individual initiative and enthusiasm based on mutual trust which Roesch had inspired, which yielded the results.

The production engineer (one of the finest) and all the foremen and workers were trained to the highest standard of workmanship. They took great pride in what they did and derived immense thrills from the performances which Talbot cars gave.

The manufacturing skill became such that nearly everything from radiators to double helical gears could be made cheaper than they could be bought for outside.

Amongst the many methods which were developed was the grinding of the faces of cast-iron block and heads. At that time it was not possible to avoid hard spots on these castings, with consequent damage to milling cutters. A perfect joint surface was obtained. Main and big-end crankshaft bearings were burnished between rollers and thus acquired a harder and highly polished surface. Tile aluminium piston skirts were diamond turned. The cylinder bores were honed in a home-built machine. The white metalling of connecting-rods was achieved with perfect control of the result.

Later the manufacture of the Talbot Wilson gearbox was a distinct achievement. This was before the Wilson box was adopted on the London ‘buses.

Another magnificent production was the Sunbeam 4.5-litre straight-eight engine, designed in 1935 and shown at Olympia in 1936.

The crankshaft was supported by ten bearings. This presented a difficult problem to line them all correctly. The guided reamer which had long replaced scraping at Talbots was not to he relied upon any more. The factory did not possess a decent boring machine so Roesch decided that a special and extremely rigid jig should be designed to fine bore the crankshaft bearings. At that time Mr. Vandervell (who had raced Talbots) had started to make thin-wall bearings and this was just what Roesch wanted to insure the necessary interchangeability. Four engines were built in which all the parts were found to interchange without any selective assembly.

It is interesting to add that this engine was over 6 in. shorter than another unit of V-construction of the same capacity shown at the same time. Roesch had designed it so that a four or a six-cylinder engine could also be made with common parts and the same tools.

The Firm’s Destiny

It might have been thought that the prosperity and fame which Clement Talbot had attained would have enabled the concern to grow in strength. Its destiny, however, was to be the reverse.

No sooner had the firm turned losses into profit with the 14/45 model than, instead of being able to re-equip its factory and consolidate its position, the Sunbeam and Darracq companies fell into difficulties. Changes in administration did not improve matters and a situation arose in 1934 when assets had to be sold and Clement Talbot, Ltd., thus passed under the control of Rootes Securities Ltd.

It is sad to conclude this brief, but we trust accurate, story by relating that four years later this small but highly efficient and brilliant engineering team had ceased to exist.

A few months before the war Georges Roesch had left, later to join Power Jets and devote himself to gas turbine research, and most of his design staff joined the aero-engine industry.

The Firm’s Achievement

Since this occurrence many years have passed and the pioneer work which Clement Talbot did may now be assessed in the light of present-day practice.

To do so it is permissible to take one example amongst many and to compare the neat, high-efficiency, 3-litre 105 engine which Georges Roesch designed 22 years ago with America’s latest creation, the new Ford six-cylinder power unit of 3.5 litres, described and also compared with another engine in the last July issue of the “Transactions of the Society of Automotive Engineers.” The two constructions follow similar lines, both units have push-rod-operated overhead valves of large area, use fully-filtered oil for the bearings and have crankcase ventilation. Yet the power given by the new engine at 100 b.h.p. does not exceed that of the old one with 500 c.c. less capacity, which was thus lighter, cheaper to build, and took less space in the car.

The shorter Talbot engine had a higher speed but there does not appear to be anything in the American design to lead one to think that its life is likely to exceed that of the British unit although it has a lower piston speed. The Roesch valve gear without mushroom tappets and adjustment on its moving parts was lighter and more efficient. The staggering of the valves (used on the winning 1952 Le Mans Mercedes) also resulted in better ports, improved breathing and more effective cooling all round the exhaust valve seats to lower their temperature to a minimum and ensure an endurance which has been consistently proved in long-distance races.

Apart from the fact that seven bearings were provided in the old engine and only four in the new one, it can be seen that the difference in the two performances is not explained by variations in design.

Engineering Outlook

In spite of the immense resources available and the large expenditure of time and money bestowed on this American development, it is seen that it has hardly caught up with the progress realised long ago at a fraction of the cost by the British engine.

This fact emphasises the danger, which has been growing, in thinking that large allocations of money for design developments may yield proportionate results.

The quality of a design and the rapidity of its evolution are not necessarily improved by increasing the number of draughtsmen and engineers working on it. To get this improvement there must be an increase of skill available.

Since 1938 up to now the incentive to create something new has been lacking, but should a period of intense competition lie ahead then the importance of design will again become vital.

After all, the art of designing a new car is to arrive at such a harmony between all the factors at play as to be able to use the minimum of material and labour in its construction. If these two requirements are not satisfied, all things being equal, the car is either too heavy or too expensive or both. It is therefore evident that the larger the production of a car is planned to be, the better and the more advanced the design should also be.

In this way production can really come into its own, sales are ensured, early obsolescence of plant is eliminated and there remains plenty of time to improve and plan ahead.

The Future

The passage of time has shown, as no other means could, that Clement Talbot was not only ahead in ideas but, what is important economically, in the way they were applied, which in many cases has not been improved upon up to date.

This due tribute to the magnificent spirit of endeavour and achievement of the British firm and its designer gives a measure of the advance which, with new possibilities and knowledge available, should be attained and surpassed in the effort to meet the changing conditions now facing the industry. 

Apart from the suggested merger of owners of Roesch Talbots with the Sunbeam Register, referred to elsewhere, John Wrigley has formed a Talbot Light Car Register, for 8/18s and 10/23s. He has 14 members and will be glad to enroll others, who will receive the breezy newssheet and details of spares sources. Wrigley’s address is Wellington Lodge, Station Road, Wokingham, Berkshire.