Not a forgotten make, a forgotten model. A bit of a sad Rover story, in fact. There is no doubt that the Rover is, and was, one of Britain’s great cars, although orientated to the ‘Rovonda’ theme in recent times. The Rover Company Ltd got off the motor-car mark in 1904, with production models, some of which were good, some not so successful, and had the distinction of having factories both in the Motor City of Coventry and in Britain’s industrial Midlands, just outside Birmingham.
In fact, Rover mechanically propelled vehicles went back to before 1904, if we include the experimental electric Starley of 1888 and the tricars made by the JK Starley Rover bicycles firm in 1903. But so far as Rover cars are concerned, it began with a sound little 8hp single-cylinder model with tubular backbone chassis, designed by EW Lewis, also made in 6hp form and tough enough to serve in remote corners of the Empire as well as along the lanes and undeveloped main thoroughfares of the Home Counties. Later on Rover adopted the Knight double-sleeve valve engine for its smallest models, but in these small cars it was less successful than in large and stately Daimler, Minerva and similar cars.
The best car made by Rover’s before the Great War was the Owen Clegg-designed four-cylinder Rover Twelve. Clegg had a railway engineering apprenticeship and had joined Rover from Wolseley’s. Not long after he had gone to Rover’s he left to join Darracq, Mark Wild taking his place as Rover’s Works Manager.
The War Office wanted another sound car, the 16/20hp Sunbeam, in such quantities during the war that Rover made it in their Coventry factory, along with 3 ton Maudslay trucks, as well as Rover motorcycles and munitions. After the Armistice the pre-war Twelve was revived, with some notable improvements, but the price of the tourer had risen from £350 to £750, although this was to be progressively reduced. It was renamed the Rover Fourteen, but engine size was unchanged. For a time success was maintained. The Rover name was associated with quality, and there had been pre-war racing victories to embellish its prestige, E Courts had won the 1907 loM TT on a 20hp Rover, at 28.8 mph, beating a Beeston Humber, and this became a production model, priced at £450. The well established Rover Twelve was matched in the early 1920s by the popular little JY Sangster-designed 998cc air-cooled, flat-twin Rover 8. Apparently £400,000 was spent on tooling up to make this economy car. It proved a good move for a time, its successes in trials proving that direct air-cooling was no disadvantage and at first it undercut in price the Austin 7. But the new motorists soon preferred four-cylinder engines and as the price of the A7 came down, bought Austins instead of Rovers, in spite of the latter also having a car-type (as distinct from a cycle-car) transmission. Moreover, by now new models had emerged from other manufacturers and Rover profits turned into substantial losses, in spite of a four-cylinder 9/20hp model having been rushed out, using 8hp chassis parts, in 1924.
It was then that Peter August Poppe saw his opportunity. His name was well known as joint MD of the company which had become Britain’s largest supplier of proprietory engines. William Morris had used T-head White and Poppe power units in his first Morris Oxford in 1912, although after the war he turned to the American Continental and French Hotchkiss engines. In the same way as other engineers, like JG Parry Thomas with his Leyland Eight (Britain’s first production straight-eight), Sidney Guy with the advanced Guy V8, and JA Kemp with his twin-cam 2-litre Maudslay, had planned their ideal cars at roughly the same time, so Peter Poppe set out to design his concept of what the post-war medium sized car should be, starting in 1918.
In 1919 Alfred White had sold his company to Dennis Bros of Guildford. Poppe went to them but by 1924 he had lost patience with their policies. Poppe was a friend of Alex Craig, a Rover director, and no doubt a talk with him set him trotting off to see JK Starley, son of the Rover cycle foundation which had pioneered the ‘safety’ bicycle, who was now in charge of Rover. The Rover board at that time was not possessed of much engineering knowledge and, in a desperate situation, Poppe’s ready-to-produce 14/45 seemed just what it needed.
It was certainly a very advanced design. At a time when hemispherical combustion chambers had scarcely been adopted for racing car power units it is interesting that both Peter Poppe and Parry Thomas incorporated them for their production engines. Poppe did not wish to go as far as using twin overhead camshafts to prod valves with an included angle of 90 deg. Instead he came up with a very unusual valve-gear. He used a single overhead camshaft, driven by a worm-gear from a vertical shaft at the back of the engine, but made it operate the inclined inlet valves through short rockers. To actuate the inclined exhaust valves he used horizontal pushrods, running through tunnels in the head casting, and a further set of rockers. In effect, a combination of overhead camshaft on the inlet side and conventional pushrod operation, but turned through 90-deg, on the inlet valve side of the engine.
It is difficult to see what Poppe hoped to achieve with this complex layout. In later times cross-pushrods were used to actuate fully inclined overhead valves, but the camshaft has been retained in the basechamber. Certainly the 14/45 Rover had the sparking plug in the desirable central position in the head, which a central overhead camshaft precludes. Poppe may have sought to reduce reciprocating weight by using shorter rockers (and lighter pushrods) than a central camshaft would have permitted. He also claimed that heat expansion could be controlled to an extent that tappet clearance would never vary by more than 1 1/2 to 2000th of an inch from this cause. But the fact is that the layout still required two rocker shafts and a full set of rockers as well as the cross pushrods, nor did it reduce the height appreciably of this long-stroke engine. Incidentally, the valve cases were duplicated, giving the uninitiated the impression that the 14/45 Rover had a twin-cam engine anyway.
Having introduced this complexity, Poppe adopted others. Instead of the steering box being mounted on a chassis sidemember, it was at the top of a vertical shaft in a casing forming part of the crankcase, the “drop-arm” actuated by the worm and sector actually being horizontal, but connected to a normal draglink; Poppe’s idea may have been to improve engine accessibility and give a better rake to the shortened column, apart from obtaining a more rigid mounting. The carburettor, on the offside, took in warmed air from a passage cast across the top of the crankcase. The crankshaft ran in four plain bearings, of which the rear one was double, to accommodate the worm-drive for the vertical shaft driving the camshaft. In an age when bearing trouble was a problem, a very thorough lubrication system was provided. A sump holding about three gallons of oil had two large filters removable downwards. The submerged oil pump was driven from an extension of the camshaft-drive shaft and fed the drive gear and main bearings at some 70Ib sq/in. The big ends were pressure-fed and oil ran up the hollow tubular conrods to the gudgeon pins of the aluminium three-ring pistons. Oil under pressure also ran up the hollow camshaft-drive shaft to feed the top worm and by oilways both rocker shafts. From the camshaft casing lubricant returned by gravity to the sump via the steering box, so that the steering was automatically lubricated.
In addition to this thorough engine lubricating, the four-speed gearbox was also pressure lubricated from the main system, via the clutch and the driving and driven gear shafts, after which it fed the box itself, surplus oil flowing into the clutch pit, from whence it was flung by the flywheel back into the engine crankcase. The designer of this ingenious engine had also taken steps against overheating. The belt-driven fan, with an oil chamber to lubricate it, also drove a water impeller. A passage within the water jacket fed water from the base of the radiator up to the impeller, which sent it to cool the cylinder head, the block being thermo-syphon cooled so that previously warmed coolant did not directly reach the head. From the back of the head the water was returned to the radiator by a single large pipe. The radiator had the headlamps mounted on it in previous Rover fashion. The ML magneto was driven from the top of the camshaft-drive shaft and was mounted vertically for accessibility. The starter was mounted on the clutch housing and the dynamo, on the gearbox housing, was chain-driven from the clutch-shaft.
The gearbox cluster could be withdrawn from the rear of the box without taking it from the car, and gearlever rake was adjustable. Poppe also had ideas about equalised braking; the Rover’s four-wheel brakes were operated by pedal and lever, with a failsafe linkage, but because, he said, the rods were usually coupled to the crossshaft on the off or driver’s side, the shaft twisted and so the o/s brakes came on before the others. In his car the various brake rods went into a spur gearbox, acting as an equalizer — more complication! The brake lever was at first a pull-up cranked affair, a reminder that postwar “Auntie” Rovers had a shepherd’s crook lever. Those are only some of the unusual features of this ingenious 1924 car. The engine had a bore and stroke of 75 x 120mm (2120cc) and was claimed to run up to 4000 rpm and give a bmep of 105 lb/in. The cast-iron cylinder block was separate from the alloy crankcase, the single-plate clutch ran in oil, the wheelbase was ten foot and the tyres 32″ x 5 1/4″ . The gear ratios were 20.3, 13.75, 9.3 and 5.0 to 1. Care had been taken to reduce the chore of chassis lubrication. Final drive was via an enclosed propshaft to a spiral-bevel back axle, unusual on a Rover. The speedometer was driven from the gearbox. Suspension was by half elliptic springs, with a 5 inch travel at the rear, but no front shock absorbers. The exhaust manifold was four-branched and an SU carburettor was used.
That was the car which the 54 year old Norwegian Peter Poppe presented to Managing Director JK Starley in 1924. He specified it as able to do 65 mph and 25 mpg, sell in open form for £450 and weigh 25 cwt. At that time Rover employed 1500 workers at the main Meteor factory in Coventry, a like number of hands had been taken on to produce the Rover 8 at Tyseley near Birmingham, bodies were made at Parkside, Coventry, the quality motorcycles at St James’s, and there was a nearby repair depot at Fleet Works, Coventry, showrooms in London’s Bond Street, and the Seagrave Road repair depot in Fulham. But things were desperate and the 14/45 design was readily accepted.
Alas, as with the other “ideal” designs, the Rover 14/45 proved too expensive to make and it had teething troubles. It is said that the cylinder head weighed around one hundredweight and that when Mark Wild, the Works Director, saw the first one he walked out, never to return until the 14/45 was discontinued! For the time being, a brave face was put on things. What looked like a powerful engine developed only about 20 bhp per litre, the weight came out at 27 cwt and the fuel consumption at 16/20 mpg, top speed at 52 mph. The four-seater was priced at £550, or £100 more than the estimate and at the end of 1925 Rover’s were losing £14 on each one sold. Nor was that all. The camshaft drive was noisy, giving the impression of a run big-end, fuel consumption heavy, and a 4.7 to 1 axle ratio was desirable. So cars queued at the Service Departments, adding to warranty losses. Modifications were put in hand and things improved. Some owners and testers approved of the 14/45 and in Dudley Noble Rover’s had a good publicity man.
It was said that to improve mpg the engine had to run hot, which could be why Noble organised 50 ascents of Bwlch-y-Groes, which in 1925 earned the covetted RAC’s Dewar Trophy. It wasn’t until late that year that the car was in full production, but Noble had amassed six 14/45s for a preview at Burford Bridge, Dorking, and a chassis and tourer had been readied for the 1924 Olympia Show. The solution to the sluggish performance was to increase the cylinder bore to 60mm (2413cc) in 1926 and this 16/50 model ran concurrently with the 14/45. It went better, some owners getting 60 mph but Rover was glad to pass it out on test at 54 mph. But it was very expensive, the saloon listed at £775, the tourer at £575. A Lanchester vibration damper smoothed out the 50 bhp.
Poppe refused to give up and introduced two- and four-seater sports models of both the 14/45 and 16/50hp, which in spite of pedestrian-looking bodies, were fast. Indeed, at Brooklands in October 1925, a sports 14/45 was officially timed, sans mudguards, to lap at 83.94 mph and to cover the mile at 85.49 mph. Again, they were expensive, the two-seater at £650, or at £750 with the bigger engine.
But this encouraged Poppe to build a racing Rover. A lightened 14/45 chassis was used and the engine was a 16/50 with larger valves, a compression ratio of 7 to 1, and twin Stromberg carburettors. It gave about 92 bhp.
The electric starter was retained but the dynamo deleted. Reverse gear was also deleted and there was no differential. The unladen weight was 21 cwt. Two Autovacs fed fuel from a large tank in the tail of the well streamlined single-seater body, which had a full length smooth undershield. Called “Odin”, the car was painted orange, probably to attract attention, and even better streamlined towards the end of 1926. It was driven at Brooklands by Ehrling Poppe (Gunner Poppe had been apprenticed to Austin’s and drove their racing cars and Olav Poppe was Chief Planning Officer at Rover’s in the Land Rover era). On its first appearance the Rover racer gave Malcolm Campbell’s 1 1/2-litre Bugatti a 3sec start. It never won a race but in seven starts that season it was second in three, third in one, once from scratch. It lapped at 102.9 mph, so was good for some 110 mph.
Alas, it was too late. Rover had lost £201,395 in two years, no dividends had been paid since 1923, and the value of a 16/50 chassis had dropped to £380. Starley was again desperate, especially as, before Poppe’s arrival, a 3 1/2-litre Rover Six had flopped after only three had been made. This sad situation was resolved when Col Frank Searle took over. He sold off the remaining Poppe cars as a job lot to David Rosenfield in Manchester, a company that survived the war but seems to have vanished (otherwise it would be interesting to know how they disposed of these cars and at what prices!), got rid of many of the staff, and sent Starley off on a sales tour, sacking him when he had reached New Zealand . . . ! But Searle, too, must have been desperate, because he retained Peter Poppe to design another Rover. It was a far more simple 2-litre pushrod ohv Light Six, selling as a Weymann in 1928 saloon for £395.
It was an improvement but trouble still lay ahead; Searle had a row with Poppe, who left in September 1929. It took Spencer King and others to regain Rover’s good name. WB
40/30/20/10 years ago 1996 Tommi Mäkinen wins the Swedish Rally aboard his Mitsubishi Lancer, but Carlos Sainz provides a stern challenge to finish just 23sec adrift in Escort. 1986 It's…
This Month's Continental Races
F.I = Formula I cars, i.e., up to 1½ litres s/c, up to 4½ litres non-s/c. F.II = Formula II cars. i.e., up to 2 litres, non-s/c. C.= Formula Libre,…
Wheel to wheel -- the month in historic motorsport
Action-packed finale -- HSCC: Silverstone October 15 Though all bar one of the HSCC's championships were settled ahead of the final meeting of the year at Silverstone, packed grids and…