The Austin Seven from the Special-Builder's Angle

The Austin Seven is one of the world’s great motor cars; some people go so far as to say the greatest. It was in production as an exceedingly successful economy car, unchanged in the main essentials, from 1923 to 1938, and some 400,000 were built during that time It was entered successfully by the makers and by private owners for races and other forms of competition in virtually standard, side-valve form. It has formed the basis of more amateur-built “specials” than any other make and continues to do so.
 
With the advent of the “Unblown 750 Formula” by the 750 Club, a formula introduced to foster inexpensive racing of Austin Sevens by a club catering solely for these cars, interest in the potentialities of the little Austin from a competition aspect has become greater than ever.
 
Consequently, we asked Holland Birkett, Chairman of the 750 Club and himself an inveterate Austin Seven special-builder, to write this article on the subject. The first part deals with the characteristics of different Austin Seven models as they affect the enthusiast and the latter part describes how more power can be extracted from the engine and increased stability and better braking be obtained from the chassis.—Ed.
———
The problem of constructing a sports car from an Austin Seven is one which has been tackled by hundreds of young enthusiasts, who fall approximately into three categories. The first attempt to put a smart-looking coachwork on a standard early chassis, and mostly abandon the project on finding nothing but fresh air on which to support the back end of the body. The second are the power-to-weight fiends who throw away the body and devote all their energies to the engine, finishing with a monstrosity that neither steers nor stops, but which has good acceleration at certain points in its performance range. The third group are more reflective, and aim to retain the many pronounced virtues inherent in the car, while taking steps to counteract the equally marked faults. This article is addressed to these people.
 
It is assumed that the reader has seen Austin Sevens in pieces, and it would be as well briefly to review the significant changes that have been made to the car during its production life. Taking the 1928 Chummy (open four-seater) as the basic type, we note that it has the classic A-shaped frame with top-hat section side-members, quarter-elliptic rear springs beautifully mounted in the ends, a semi-floating back axle with a flange-fitting torque tube connected to the rear cross-member by a ball-jointed link, and to the gearbox by a solid open prop.-shaft with a fabric universal joint forward and pot-and-trunnion-blocks aft. The final drive ratio is 9:44, or 4.9 to 1. The front axle is solid and mounted on the apex of the frame by a transverse spring and two shackles. The engine is solid-mounted and has a 1 1/8-in, two-bearing crankshaft mounted in a 1 1/8-in. bore roller race at the rear, and one roller race and a ball race at the front. Bore and stroke are 2.2 by 3 in. (56 by 76 mm.), the valves are side-by-side and there are four exhaust ports and two siamesed inlet ports. Cylinder head is cast iron giving a compression ratio of approximately 4.8 to 1 with a .075-in. (thick) gasket. The carburetter is up-draught Zenith with a 15-mm. choke diameter on a cast aluminium inlet manifold. The gearbox is three-speed, second being of the order of 1.8 to 1. Ignition is by magneto driven from a train of gears in the timing case and the dynamo is set transversely and driven from the camshaft timing wheel. The starter drives the Bendix pinion through a pair of spur gears.
 
The brakes are 7-in. diameter, foot rear and hand front, cable operated.
 
The body is aluminium, and examples in standard form which still have hood and upholstery can even to-day give remarkably fine service as general utility vehicles. Weight is 8¼ cwt.
 
In 1929 a number of improvements were made, but not simultaneously. The radiator became taller and the windscreen correspondingly shorter. The engine was altered to coil ignition, an entirely new crankcase eliminating the magneto gear train, the distributor being driven by small skew gears from the end of an entirely redesigned dynamo. Later the big-end size was increased to 1 5/15 in., and the rear main bearing increased to 1¼ in. bore. [This crankshaft can be built into the early crankcase. using the later type rods, flywheel and rear main bearing assembly.] The gearbox lid lost its gate. The differential cage became stronger and its securing bolts increased to 3/8 in.
 
In 1930 the rear axle was provided with a screw-in torque tube and screw adjustments for the crown-wheel bearings, so that all meshing adjustments could be made externally.
 
In 1931 the frame was made much stronger, fitted with a light, drilled longitudinal member between the cross-members, which also served a vital purpose in providing a centre bearing for the brake cross-shaft, which was thus able to pull on the brakes instead of bending its outer bearing brackets. The radiator became taller still and in mid-1931 the body was panelled in steel with higher sides. The starter was direct drive. The rear axle assumed two dual-purpose thrust races immediately in front of the main pinion bearing, secured by a nut on the pinion shaft. It reverted to flange-fitting for the torque tube but retained the crown-wheel adjusters.
 
In 1932 the same model was continued, there also being it large production in saloons. Concurrently the “de luxe” saloon was introduced, with the wheelbase increased from 6 ft. 3 in. to 6 ft. 9 in. and 2¼ in. greater rear track. This rear axle had longer half-shafts and casings and a final ratio of 8:42 (5.25 to 1). [The two rear axles of this year offer the only direct interchangeability whereby the special-builder can make up a wide 4.9 or a narrow 5.25 axle, although a 1931/2 4.9 or a narrow pinion only needs a sleeve and a packing plate made up to fit it into the post-1932 caing.]
 
In 1933 the crankcase was again radically altered so that the same engine mounting holes in the frame in effect brought the engine ¾ in. forward to make room for a four-speed gearbox (third about 1.6) and it had a starter mounted under the bonnet and a mechanical fuel pump. The clutch thrust race was now held away from the toggles by a big spring. The fuel tank was mounted on the body under the floor behind the back axle. [It is an easy matter to mount both a front and rear tank on the “de luxe” bodies, thus giving a 9 to 10 gallons capacity.] The two manifolds were replaced by a single combined affair in cast iron, and a horizontal Zenith carburetter with a 17-mm. choke diameter was fitted. [The 1-in. S.U. from a Morris Minor fits straight on, and many people prefer this to the Zenith. An M.G. needle provides a good compromise between economy and performance.] There is a constriction in the eye of the inlet port in the combined manifold, probably to aid in heating the incoming charge for rapid warming up. This constriction may usefully be relieved. The final drive casing became D-shaped and in one piece with the off-side half-shaft casing.
 
In 1934 the gearbox was synchromesh on third and top, and a minor alteration to the crankcase was made to provide for rubber mounting the engine. The pot-joint at the rear of the prop.-shaft was replaced by a plain bearing “Hook” joint in a hemispherical casing, calling for a large driving flange with six ¼-in. diameter holes on the pinion shaft. Because in earlier models the brake cross-shaft had tended to twist under heavy pedal pressure, so giving a direct pull to the off-side rear brake and a spongy pull to the near side, the makers fitted a concentric tube over the off-side half of the cross-shaft taking the pedal torque to the middle of the shaft. Many 1933 cars had this feature. Brake shoes and drums were increased in width from 1 in. to 1¼ in.
 
In 1935 the “Ruby” range of cars was introduced, with the larger and better-looking body, radiator cowl, synchromesh second gear, 17-in. wheels, full needle-roller prop.-shaft and, most important to the special-builder, a chassis extension cranked up, over and beyond the rear axle. The rear end of the chassis was also lowered by fitting virtually uncambered rear springs, and the rear dampers were moved back to shorten the arms, making them effective for the first time. The pinion was supported on a generous roller race and the shaft diameter increased at that point. These cars make very reliable family saloons, but the enthusiast may find the power to weight ratio a little tedious with four large adults up!
 
The 1936 car was substantially the same, but the fulcrum pins for the brake shoes were moved outwards and the operating levers increased in length.
 
Big changes occurred in 1937. The crankcase was again modified to provide a centre plain crankshaft bearing, but the block casting and therefore the overall length remained the same. Thus the extra room had to be found by severely skimping the thickness of the centre-main crank webs and off-setting the big-ends. (It is found in practice that the inevitable wear in the rear main roller race occurs—largely due to condensation-corrosion—and then the rear centre main crank web breaks across. The engine is pleasantly smooth but is not recommended for high-performance purposes.) The cylinder head was altered to give a compression ration of 6.2 to 1 and to take 14 mm. plugs. The camshaft centre bearing became plain. The clutch had its linings moved on to a sprung centre plate. The brakes were changed to so-called semi-Girling, i.e., with Austin actuation through Girling enclosed cam and plungers, with the separate stop-adjustment for each shoe by a square-headed cam. The drums were made much stiffer. To cope with the greater brake-torque the radius arms were increased in strength (and weight) and attached to the axle at two points thus firmly fixing the castor angle. The pendant lever in the centre of the brake cross-shaft which actuated the front brakes incorporated an ingenious device for compensating the pull to front and rear. The steering-box worm was made “hour glass” shape to reduce wear.
 
The same car continued production into 1938, a few being fitted with full Girling brakes on the back only. The Seven was continued with the Big Seven until early 1939, when both were dropped in favour of the Eight.
 
From 1929 to 1932 the “Ulster” was produced, in supercharged form, some of which were T.T. Replicas, and also in unblown form. The chassis was lowered about 4 in. by fitting flat springs and a longer torque anchor at the back (like a Ruby) and by bowing the front, axle beam downwards about 2½ in. in the centre to give clearance for a stiff spring with reversed camber. The commonly heard terms “underslung” and “upside-down springs” are, of course, not applicable. The frame was as 1931 with coupled brakes, and the narrow aluminium body had the spare wheel set transversely in the pointed tail effectively blocking access to the luggage space. The steering column rake was increased. The gearbox is very interesting, being three-speed, but with a very high second gear of about 1.4 to 1 giving something in the order of 50 m.p.h. at 5,000 r.p.m., depending on the axle-ratio, and the bottom gear high in proportion. The clutch had double springs and cast-iron liners. The rear axles varied, being mostly 1931-type, either 4.9 to 1 (for the blown car) or 5.25 to 1, or else the special “Army” axle, narrow track, of course, with a screw-in torque tube and a larger casing to accommodate a 9:51 crown and pinion (5.07 to 1).
 
The unblown car had an engine basically resembling the 1930 touring unit, except that the crank (1 5/16 in.) was machined all over, and the very satisfactory jet-lubrication system replaced by a somewhat unreliable feed to the nose of the crank by means of a leather gland. The rods were machined all over, had parallel sterns, and unbushed fully-floating small-ends. The cams were of a somewhat fierce profile, this model being the only Austin Seven whose inlet valves open other than at t.d.c., in this case five degrees before. The lift is 3/8 in. against the touring ¼ in. To cope with this the heavy adjustment for the tappet was replaced by a light hardened steel thimble. The valves and double springs, which have a real job of work to do at high r.p.m., were accordingly much increased in length. Inlet and exhaust were of the same tulip form in K.E. 965. The cylinder head was cast iron, and had a compression ratio or something over 7 to 1. There was an updraught 30 mm. Solex carburetter on a sheet steel, buffer-ended manifold and the famous three-branch taper tube exhaust manifold common to all “Ulsters.”
 
The blown engine differed in having a magneto drive with a water pump on the front of it and the train of gears extended across to the near side for the Cozette vane-type compressor. The valve gear was similar, and with different thimbles could accommodate either the mild “blown” camshaft or the high-lift one supplied for conversion to unblown. The crank had 1½-in. crankpins, pressure-fed from the nose by a system differing in detail and involving a felt seal. The cylinder block was held down by 10 3/8-in. studs instead of the usual eight 5/16-in. ones, so was not interchangeable. The T.T. Replica cars are those whose petrol tanks form the scuttle, as opposed to being revealed by lifting the bonnet, which was shorter than on the production blown cars.
 
The “65” or “Nippy” and “75” or “Speedy” range was extant from 1933 to early 1937 inclusive. They were lowered in the same way as the “Ulsters” but built on long frames with the short extensions. The neat little two-seater body is well known, its appearance being somewhat spoiled by the strange retention of the 3.50 by 19 wheels and tyres. The “65” had an aluminium body (1935 and subsequent Nippies steel) and jet-lubricated 1½-in. crankpins, which were prone to big-end failure. The crankcase was grooved to provide big-end clearance and the ribbed cast-alloy sump held 3 qt. of oil. The camshaft was like that of the unblown “Ulster,” but with inlet opening at t.d.c. and a larger-diameter front bearing and timing-gear taper. The valves and double springs were intermediate in length between those of the “Ulster” and touring types, with normal screw adjustment for the tappets. The cylinder head was similar to that of the “unblown” “Ulster,” but the manifolding was new. A cast-iron exhaust and a cast-aluminium downdraught inlet manifold nested together joined by a (very) hot spot. The carburetter was a 30 mm. downdraught Zenith with a 21 mm. choke tube. The gearboxes, following the production cars as to synchromesh developments, were close [sic] ratio four-speed, with 5.6 to 1 axle and third gear of 1.49 to 1. The Speedy was a Nippy with a lighter body having a drooping pointed tail and draught deflecting sills round the forward parts of the cockpit sides. It also rejoiced in a 25 mm. choke tube and pressure-lubricated big-ends as on the blown “Ulster.” They are very rare. All Nipples had two-bearing crankshafts, jet-lubricated except for the 1937 models. These had the three-bearing crank as standard, but by paying extra for a “sports” engine one got the Speedy engine, a surplus unit left in stock when the Speedy was dropped in 1936.
 
The vans did not follow the passenger cars in all particulars; they never had the upswept chassis extensions. In 1935 they had 18-in. wheels which, in conjunction with a motor-cycle 3.25 tyre, make a very light front wheel assembly, and certain vans, as well as most Nippies, had crown-wheels and pinions 8:45, or 5.625 to 1 which interchange with the ordinary 5.25 to 1.
 
Let us now consider what faults are common to all Austin Seven models and what steps we may take to overcome them. The most outstanding, of course, are the poor brakes which received periodic attention from the manufacturer, and the monumental oversteer which makes anything approaching rapid cornering a fearful dice. Dealing with the latter point first, let us examine the factors which cause the oversteer. The greatest is in the non-lowered car, where the roll due to centrifugal force flattens and so lengthens the outer rear spring which steers the back axle towards the outside of the corner. A second factor is that, for reasons we need not go into, if the weight transfer from the inner to the outer wheel on a corner is greater at one end of the car than the other, that end will tend to drift outwards. Since there is more “top-hamper” at the back of every Austin Seven and the frame is singularly flexible in torsion, this phenomenon obtains. A further fraction of rear axle steering is caused by lateral movement at the front spring shackles. Finally, the fitting as standard of small-section tyres at the rear is in itself conducive to oversteer. Our remedies are therefore obvious. The shackle roll, though a small matter, also contributes towards wander on bumpy roads, and Messrs. Cambridge Engineering tackle the problem by replacing the front shock-absorber shackles with single rubber bushes, which is very effective except possibly in cases where a large suspension movement is visualised for trials purposes. Another way would be to pack the spaces between the ends of the spring and the upright part of the axle with rubber pads. Some people simply immobilise one shackle. Swing axle suspension disposes of this problem, but will be discussed later. The question of the rear spring camber is easy; never contemplate fast motoring unless you have flat rear springs. The question of weight transfer is more difficult; lowering the car helps, and so would widening the rear track by fitting offset rear wheels. The frame may be stiffened by box-sectioning or by the body structure. The real care, of course, is to have 15 or 16-in. rims built offset outwards on to standard wheel centres, and to run with 5.00 or 5.25 section at the rear and some smaller section, certainly not greater than 4.00, at the front. From this point, the degree of over and understeer (assuming flat rear springs) is simply a question of adjusting tyre pressures to taste. To increase understeer, more pressure at the back and less in the front. The transmission is very strong if the parts are in good order and properly assembled. Elimination of end-float in the pinion shaft is essential and the hubs, which are tapered and keyed, should be carefully ground on to their shafts.
 
The question of brakes is a vexed one. There is a tendency to dismiss the whole matter as a question of brake-drum size, and while these are admittedly small, they are also light, which is one of the virtues we are seeking to retain. It must be remembered that the effective size of the brake drum is a proportion of overall tyre diameter, so small wheels and tyres help in this way as well as by saving unsprung weight. The chief trouble lies in the actuating mechanism. Referring to earlier models, the brake cross-shaft twists and softens the pull on the near-side brake. The rather heavy device introduced to overcome this tends to stop working either by seizure or by the off-side brake lever bending and bottoming on the pedal lever. (One wonders if it would not be simpler to fit a rubber pad in the off-side cable ball joint to equalise the sponginess.) And there tends to be a lot of friction in the cross-shaft bearings. The front brakes, which are the more important, suffer from a marked negative servo, because brake torque tends to rotate the axle as far as the very springy radius arms allow, and because the cables are attached at the bottom of the brake assembly this immediately takes the brake off again. The rational treatment, of course, is to drive the front brakes through Bowdenex cable, incidentally eliminating the oblique pull of the standard layout. Cambridge Engineering have for years marketed a conversion set. As regards adjustment, attention should be paid principally to shimming the shoes in the brakes themselves, so that the angle between cable and lever never becomes greater then 90 degrees. If the normal coupled layout is retained it is usually found necessary to give the front brakes a considerable lead over the rear because. of the negative servo aforesaid, and then it is found that the front brakes come on when the steering is turned. A simple cure for this is to fit direct external pull-off springs to the front levers. The long levers of 1936 are interchangeable with earlier models. All types of brakes are better using high-friction linings, and oil from the back axle can be prevented from entering the brakes by drilling small holes in the underneath of the axle casings about between differential casing and hub.
 
Some people have made hydraulic conversions. One has fitted two modified Morris cylinders into each Austin front brake, giving two leading shoes, and another has grafted small Wolseley brakes on complete, necessitating the use of Big Seven wheels. This has the advantage of eliminating the pendant lever, which is the lowest part of the car and a big nuisance in cross-country going, and also obviating the need for the heavy 1937 radius arms. But there can be no doubt that semi-Girlings on the front, even with normal actuation, can give very good braking indeed if properly maintained. An important maintenance point incidentally is the radius arm ball joint, where there must be no free play.
 
The swing axle i.f.s. system was popularised originally as a conversion to Fords and Austin Sevens by L. M. Bellamy, and conversion parts are now supplied by North Downs Engineering at Caterham. It has a great many advantages and some pronounced snags. Briefly, the front axle beam is cut in the middle and two short tubes are welded horizontally at right-angles to the cut ends. These have Silentbloc rubber bushes pressed into them and the half axles are pivoted to an inverted U-bracket fixed, with the spring, to the chassis nosepiece. At the back of the bracket there is a swinging arm on a vertical pivot, and the two half track rods are universally jointed (usually by two rubber bushes each) to this, these joints being approximately in line with the main axle pivots and with the radius arm ball joint. It will be seen that this at once eliminates the front shackle sidesway, permits lowering the front end by eliminating the clearance between axle and spring, and affords a simple way of widening the front track—a very desirable measure—and it retains the high roll centre which most i.f.s. systems lower to the ground. It further obviates the need for limiting suspension movement except as regard mudguards, etc., so that a soft front spring may be used which will give a safe, comfortable and pitch-free ride on rough surfaces and greatly facilitate steering and braking in trials conditions. Another seldom appreciated virtue of the suspension is its behaviour in really extreme cornering conditions, when the outer wheel assumes an outward camber due to the side thrust from the axle pivot, and this gives a “final understeer” which tends to prevent the car from spinning round. This means that steering control is retained to the last, so there is a net increase in safe cornering speed.
 
The disadvantages, apart from the question of cost and complication (for example the way the centre track rod arm fouls the sump), chiefly arise from a phenomenon known as gyroscopic precession. When a disc is rotating at speed and a force applied to tilt it, it not only resists but tries to tilt also in a prime at right angles. It will be seen then that when the front wheel rises over a bump, it travels through an arc of radius of the half axle beam and there is a force of some violence tending to rotate the wheel about the steering pivot. The residence to tilt imposes a big strain on the king-pins, which will either eventually break or make big oval holes in the end of the axle. (It is possible to fit kingpins up to 9/16 in. diameter—and desirable.) The precession imposes shock loadings on the steering parts, and the four rubber bushes in the centre of the track rod, though seemingly a sponge where none should be, probably serve to protect the track rod steering arms from fracture. It might be unwise to have ball joints in this position.
 
Another practical snag is that it is almost impossible initially to set the front spring exactly right. This set is critical if the front wheels are to be somewhere near upright, and it is almost always found that some trial and error is necessary. This suspension is also conducive to front wheel patter, aggravated by loose chassis rivets, flexible engine mountings and heavy front wheels.
 
In connection with widening the track, it should be noted that the axle pivots should be as close together as possible; 2½ in. centres is the usual and this gives almost exactly 4 in. track increase if a single beam is divided in the centre. Further increases call for two beams. It is simpler to use two ball joints on the front cross-member than to alter the radius arms. For a wider track it is necessary slightly to twist the steering drop arm to avoid its ball being cut through the neck by the drag link—this may be done cold—and, in the interest of steering geometry, it is desirable to retain a right-angle between the drag link (viewed in plan) and its steering arm. This may necessitate making a longer drag link (out of two standard ones) and getting the steering arm bent forwards. This should be done hot by someone who understands the appropriate heat treatment. A periodic inspection of this arm is needed, for they sometimes break, but usually show a visible crack for some time beforehand.
 
If the front end is lowered the drag link will lie at an angle (viewed from the side) which any student of steering geometry will recognise as a source of “wheel fight.” A rigid bracket for raising the steering box should be fabricated, and the appropriate rake for the column can easily be incorporated in it. Fore-and-aft rigidity is the chief criterion in designing the bracket. At. the same time the brake pedal can be raised into a position more favourable for the application of powerful pressures from a lowered seating position. As an alternative to a divided axle, the front of the car can be lowered as on the “Ulster” and Nippy models.
 
It must be emphasised that whatever the suspension arrangements, the damping is critical and the standard shook-absorbers are not adequate for speed competition purposes. The pre-Ruby rear ones are too long in the stems: these should be halved in length (requiring long shackles) and the number of friction discs doubled by the addition of one static and one moving arm on the same centre bolt may be used, and P.R. tubing makes excellent bushing. A large friction area is necessary for both accuracy and stability of adjustment. Hydraulic dampers are preferable to the friction type, but call for a special installation.
 
If it is intended to use the car for both domestic and competition use, it is good policy to have a “cooking” and an “eating” engine. The former should be of the same or equivalent type as the latter; for instance a 1928 magneto engine to partner a blown “Ulster,” a 1929-32 engine for a production unblown “Ulster,” a forward starter unit for a Nippy and so on, so that the change-over involves no infuriating changes in the installation. As a matter of fact it is possible and in many ways preferable to make your “eating” engine out of an ordinary touring unit, for you can then financially survive a blow-up, as parts for the sports units are very difficult to obtain.
 
The writer is further of the opinion that the unblown “Ulster” and “Nippy” camshafts are somewhat overrated in that while they increase maximum b.h.p. by means or quick lift, high lift, long dwell, quick drop and large overlap, this is done at the expense of high valve gear stress and to the detriment of power at low and middle revs. If power is only available over a small range of r.p.m., it is obviously necessary to be well provided with close-ratio gears, and these are sadly lacking in the Austin range. If you are using a 4.9 to 1 axle (on which high-lift camshaft. best revs. are unobtainable in top gear) and one of the wide-ratio gearboxes, you will get better performance with a touring camshaft.
 
If you are not the proud possessor of a sports engine with pressure-fed, fully machined crank you will be able to get admirable service from the standard touring 1 5/16-in. shaft. If there is more than 1½ to 2 thou. ovality the journals should be ground and the rods re-metalled. Then the standard jet lubrication will be perfectly adequate for sustained use at 4,000 and 5,000 r.p.m. in an unblown engine, and as the crank has no serious vibration period in that region it is exceedingly unlikely to break. With a 1 1/8-in. crank, on the other hand, fracture is inevitable. The standard rods have no great margin and the writer has had one pull in halves at about 5,500 r.p.m. The remedy is to use rods that have no file marks or other indentation on the surface of the stem, and to have them shot-peened before remetalling as this simple measure strengthens the rods to the extent of another 500 or so safe revs. Shot-peening should not be confused with shot-blasting.
 
Another weakness of the touring rods is the split small ends. It is essential that good quality high-tensile bolts be used in this situation with threads in perfect condition. Star-type shake-proof locking washers are probably slightly more reliable than the standard tab variety, certainly on second application! If the gudgeon-pin hole in the rod is oval the rod must be scrapped. The same applies to stretched big-end bolts.
 
The early or late-type of front main bearing are equally satisfactory except perhaps that the impecunious or parsimonious may defer the replacement of tire later double-purpose thrust type by putting a shim between them. Incidentally, if these races are fitted to a 1929-32 crankcase it is necessary to make a packing ring as they do not stand proud to be held in position by the steel plate. The packing goes between the front race and the plate. It is usual to fit these races with the sides marked “thrust” in contact, in which case the shim goes between the outer tracks. New rear main bearings are delightful and expensive things; they very quickly develop a little play and then last like that for years unless corrosion or over-advance sets in. For this reason the writer usually uses slightly worn ones in the first place, often making them up by selection from the parts of several more worn ones. The initial play is due to crank whip, and some people recommend using a self-aligning ball race, but suggest it will have a short life.
 
The cylinder base joint is a possible source of trouble, and it is advisable to have the upper surface of the crankcase checked for flatness. It is a comparatively simple matter to convert to 3/8-in. diameter studs, which with Whitworth threads in the aluminium will hold any unblown engine together. Stud holes in the crankcase should be chamfered and “Stag” or “Wellseal” jointing compound helps towards is tight joint. It is useful to note that the early cylinder blocks with a sheet cork valve cover washer, as fitted to magneto-type engines, have is very much greater resistance to wear than the later ones. They are also more accurately co-ordinated as between casting and machining, and so may safely be bored out to 60 thou. oversize. Most commercial pistons, particularly Special-loids, are entirely satisfactory.
 
Cylinder head design on side-valve engines is a compromise between leaving plenty of room round the valves and through the passage into the bore for the gas to flow freely, and reducing the head volume in the interests of compression ratio. The early standard head is unnecessarily recessed over the whole of the bore diameter, so is unsuitable for our purpose. The 1937 version on the other hand is excellent in conjunction with touring-type valve gear, and probably the chief advantage gained by the many proprietary aluminium heads, apart from saving some 7 lb. dead weight, is one of improved cooling. Most alloy heads call for periodic resurfacing, to avoid gasket troubles. A head designed to accommodate 3/8-in. valve lift used with a ¼-in. camshaft might be supposed to have some waste volume over the valves and possibly have some constriction into the bore itself if it has a reasonable compression ratio. The edge of the block leading into the bore may be slightly radiused.
 
The questions of carburation and manifolding are, of course, critical to performance. Many people have experimented with two carburetters and are unanimous in reporting no noticeable improvement, and one infers from this that the induction “bottleneck” is in the inlet valves, as in 90 per cent. of car engines, and not in the carburetter. This view is supported by the comparatively high speeds sometimes observed with the old updraught 15-mm. choke instrument. The “Nippy” 30-mm. downdraught inlet manifold, on the other hand, is universally liked, and often used with no hot spot in conjunction with a branched exhaust manifold such as the “Ulster” will take its own 30-mm. Zenith or, with simple adaptors, a downdraught S.U., or a Ford Eight or Ten carburetter, which have choke tubes of 19 and 22 mm. respectively. The writer has found the Ford Eight most excellent with normal valve gear, and the Ten somewhat too large, giving poor opening-up, although others claim to get good results with it, probably at the expense of fuel consumption. There is little to be gained by using a choke larger than 20 mm. Either the 1¼-in. S.U. or the rare 1 1/8-in. will fit on, and although both are too large this is compensated automatically by the slide not opening fully. There is, unhappily, no 1-in. downdraught S.U. made. Whatever carburetter is used, the chief thing is to see that all flanges and ports fit together with a smooth blending of sections leaving no steps or sharp edges for the gas to flow over (the angles in the buffer-end arrangement in the Nippy manifold have to be excused because of their valuable service to distribution). Internal polishing is of secondary importance. It has been noted that home-made downdraught manifolds often give unsatisfactory results; perhaps the internal dimensions of the Nippy one should be followed more closely, because manifold design is not yet an exact science, and maybe Mr. Austin had a lucky accident.
 
Carburetter tuning is fairly simple in the case of the Ford and Zenith range. If it is weak at high revs., the main jet (the smaller in o/d of the two in the bottom of the float chamber) should be bigger. If the opening up is weak, the compensating jet should be bigger. The larger the number on the jet the larger the hole in it. You have only to learn the difference between a weak and a rich cut. The absence of a hot spot calls for a larger compensating jet as a rule. In the case of S.U.s it is advisable to enlist the help of someone with a needle dimension book, a supply of needles, and some experience, for this tuning problem is a very complex one. Experts tend to be dogmatic, expensive, and liable to improve one’s petrol consumption from 30 to 15 m.p.g. Perhaps better stick to Ford Eight, which, with 70 compensating, and 75 main or thereabouts, will give entire satisfaction and a fuel consumption of around 35 m.p.g. with hard driving.
 
Reverting to the induction “bottle-neck.” A number of people have tried increasing the size of the inlet throats and valves, and only a few have claimed a noticeable improvement. There are various points. The valve seating should be as narrow as possible (the opposite is true of exhaust valves) and both sides should blend smoothly with the adjacent metal both on the cylinder block and the valve itself. This means that the valve throat must not be reamered out parallel but given a sort of venturi form. The valve should not be of pronounced tulip shape as the thick “neck” tends to fill the orifice when the valve is open. There is only very little room for these modifications and it is possible to make the necessary special valves out of Austin Eight exhaust valves, i.e., about 2 mm. increase in diameter. It must be realised that an improvement in volumetric efficiency means a rise in the peak r.p.m., so that problems of mechanical safety and axle ratios intrude themselves. Also, of course, it is quite possible that many cylinder heads will not be able to stand the necessary removal of metal round the inlet valve area without going through to the water jacket, and it is difficult to see a way of finding out without trying. The writer is using a touring 1931 engine with this valve modification, a Nippy inlet manifold and an “Ulster” exhaust system in an “Ulster” car and believes he gets the same top end performance as with a 3/8-in. lift camshaft, but without the valve gear wear and clatter, and with much less deterioration of middle and bottom range performance. The machining was done by Messrs. F. and F. Knight, Ltd. of’ Farnborough. Hants.
 
Naturally, double valve springs (Terry’s “Aero,” etc.) are needed to obviate valve bounce at high revs. Up to 2 lb. may be readily machined from the flywheel rim without apparent ill-effects.
 
Exhaust systems are probably less critical than many people imagine, except that it seems well enough proved that a high-revving Austin engine prefers a long narrow tail pipe, and the 7/8-in. bore 5-ft. long affair on the “Ulster” is remarkably efficient, as well as making a delightful overrun snarl. The 2-in. dia. copper pipe touch does not pay dividends.
 
Any four-speed Austin Seven gearbox can be fitted to a three-speed car. The magneto ignition unit will need its prop.-shaft shortened by ¾ in., but a 1929-32 coil engine can be pushed forward by this amount to accommodate the longer box. The latter procedure will enable the normal speedo-drive to be used instead of its fouling the front cross-member. The normal three-speed thrust race should be used, with the spring from the four-speed box keeping it against the toggles. In turn, it is most important to keep the pedal against the thrust race; this is easily done by using the spring front the three-speed clutch pedal, which fulfils this very purpose.
 
Austin Seven Gearbox Ratios
Type of Box
Normal 3-speed … – 4th, 1.0 3rd, 1.84 2nd, 3.26 1st.
“Ulster” 3-speed … – 4th, 1.0 3rd, 1.43 2nd, 2.56 1st.
Normal 4-speed … 1.0 4th, 1.66 3rd, 2.64 2nd, 4.37 1st.
“Nippy” 4-speed … 1.0 4th, 1.49 3rd, 2.38 2nd, 3.93 1st.