The Perfect Top Gear
These days it is hard to find a car fitted with an overdrive, but for many years they were a welcome option, and any classic car enthusiast has likely had the pleasure of engaging the “added cog”. The function of the “perfect top gear” - taking into account the final drive or axle gearing and the rolling radius of the tyres
– is to allow the engine speed, at which maximum power is developed, to correspond to the road speed at which the tractive effort at the driving wheels is the same as the sum of wind and mechanical resistance to motion. Only then can it be claimed that the car is correctly geared to reach the highest speed of which it is capable.
But as with life, so much is a compromise. And car manufacturers knew the reluctance of the average driver to change down from top gear once they had engaged it. What people wanted was lively top-gear acceleration: the easy way to provide it was to lower the gearing (usually in the final drive) and so increase the torque multiplication. This produced two unwanted side-effects: the car was fussier at all speeds, because the engine would be working at a higher crankshaft rate for a given road speed, and there was some risk of the engine being over-revved and ultimately red-lined, run beyond the speed at which it developed maximum power.
Either way this meant the engine would wear out more quickly if subjected to sustained high-speed cruising, for which top gear ought really to be higher, not lower: a top gear that was theoretically too high would prevent the car from reaching its proper maximum speed unless aided by a downhill gradient or a following wind. In that event, it would still be running within its normal safe working range, while at cruising speeds only slightly lower it would seem more at ease and consume less fuel. For many years, these opposing requirements were compromised by making the third and fourth ratios of the conventional four-speed gearbox very 'close' or numerically similar. The car might be very nearly as fast in third as in fourth; it might even be faster, as in the case of certain Frazer Nash models of about 1930 and certain Citroen models thirty years later.
The snag was that it was a four-speed car with virtually three-speed performance, as it were: a car with two top gears. Restoration of acceleration and hill-climbing prowess called for another gear, and when the need was crucial and money was no object, a five-speed gearbox could be made - as was the case of Delage whose high-revving 1926/7 straight-eight 1½
-litre could peak at about 8000 rpm in fourth gear and then relax for the rest of any long straight at 6500 in fifth. In those days, as still today in many front-engined rear-drive cars, top gear was direct, a 1:1 ratio in which the output and input shafts of the gearbox were coupled together. This eliminated the frictional losses involved in the meshing gears of the lower ratios, and thus made top gear (which was most used) mechanically the most efficient.
In a case such as that of the Delage, fifth gear would not be used much: it would be better to make the other 'top gear', fourth, direct drive, and provide a geared-up fifth - an overdrive gear. The same solution had been found in other, earlier cases such as the Rolls-Royce Silver Ghost
which, for a time, had a four-speed gearbox in which third was direct and fourth was geared up. Rolls-Royce found that customers insisted on going everywhere in overdrive, which could be neither as quiet nor as lively as direct third, and Rolls himself could not educate them to think of third as top when there was another higher ratio available!
By the 1930s, the American manufacturers were occasionally running into the same problems. A three-speed gearbox was the norm there, but some of the fastest cars needed more. The answer was to make two top gears available where they really counted - in the back axle - so that both were direct drive as far as the gearbox was concerned. Unfortunately the gear-change mechanism was clumsy and the extra unsprung weight in the axle undesirable. Two decades later and the problems were only exacerbated - as there were now highways, better fuels, oils, bearings and tyres
- all of which encouraged people to drive for long distances at or near maximum speed. What manufacturers needed was to enable this to be done on the rising portion of the power curve.
The Laycock-de Normanville Overdrive
The answer was a two-speed epicyclic gearbox attached to the rear of a car's normal gearbox, providing a continuation of the direct drive line when all three elements of the gear set (sun gear, planetary pinions, and annulus or ring gear) were locked together by the hydraulically operated cone clutch, or a higher ratio when the clutch moved to lock the sun to the casing. The apparatus was known as the de Normanville (Captain de Normanville was the patentee) or more fully as the Laycock-de Normanville overdrive. It was little less than sensational at the time, featuring in the then new Austin-Healey 100
, the Jensen
and optionally in the popular Standard Vanguard
Many other British cars, mainly of the high-performance type, soon acquired it as an option or even as standard equipment. By the end of the decade, it had even reached foreign manufacturers. Ferrari used it in the 250 GT 2+ 2
model that appeared in 1960
- but as was too often the case with Ferrari engineering at that time, the installation was botched. It was a feature of the Laycock mechanism that the hydraulic operation of its clutch employed oil from a supply shared with the gearbox and pressurised by an integral pump worked by an eccentric on the overdrive input shaft. In the Ferrari, the gearbox had to be filled with an oil that was suitable for the overdrive; alternatively, the latter had to suffer an unsuitable oil to preserve the gearbox. From owners we have spoken with, it seems it didn't work out that well.
Overdrive switch from a Triumph Dolomite Sprint - the same unit was used on the Triumph 2000/2500 Series. The early overdrive switches were placed on stalks on the steering column, but the gear knob type became popular.
In Britain, the Laycock-de Normanville overdrive went from strength to strength. It was soon realised that the overdrive could be made operable while one of the indirect ratios of the gearbox was engaged, thus acting as a ratio-splitter: with careful juggling of cog sizes in the gearbox, a fairly regular progression could be achieved in which overdrive third nicely bridged the gap between direct third and top, overdrive second doing the same between direct second and third. Thus a four-speed gearbox might be made part of a seven-speed transmission. It was never thought safe to apply the overdrive to first gear, as the ratio of first was so low as to multiply the torque of the engine three or more times, the resulting output torque from the gearbox being too much for the overdrive to stand.
When allied to large engines, the overdrive might not be able to take the torque in second gear either, and in powerful cars it was thought more prudent to limit its operation to top gear only. All this could be arranged with the aid of simple electrical switches tripped by the gearbox selector rods, for the driver's control over the overdrive was by means of an electrical two-position switch whose circuit could be interrupted by the gearbox switches if the wrong gear were engaged. In the appropriate gear, all the driver had to do was flick the control switch (which might be on the facia or the steering column, and later found favour in the knob on the gearlever) and the job was done.
There was no need to touch the clutch, although in some installations the change up into overdrive would not be made unless the overdrive were temporarily relieved of full-throttle torque. The simple treatment was briefly to release the accelerator - just as a brief blip of the throttle would smooth a downward change from overdrive to direct top.
The electrical circuit was simple in design and execution, it only had to operate a solenoid controlling the valve which admitted high-pressure oil to shift the cone clutch to the overdrive position. It was very easy for a driver to slow down in overdrive top gear, select perhaps first or second gear, and then when accelerating hard again and passing from third to the top gear as expected, they instead found that overdrive was back in use and the engine revs had slumped. Well maybe that criticism should only be levelled at us, but we bet there are others out there that did exactly the same thing - plenty of times.
Go to an Austin-Healey car show or club and some will (quietly) tell you the original overdrive system on the 100 was unnecessarily complicated and the use of the later Borg-Warner overdrive (an unsuccessful attempt to rival Laycock) unpredictable. Of the early overdrive installations, it seems Bristol
had the thing best sorted, as they demonstrated in their 405
model first shown late in 1954
. When overdrive top was engaged and the driver changed down (or simply moved the gearlever into the neutral position) the gearbox switch isolated the driver's relay switch, which snapped back into the normal position; next time top gear was engaged, it would be direct top.
Obviously such a system could only work when overdrive was limited to top gear, as in the Bristol, but the objection was more imaginary than real, as in most other installations overdrive third was so similar to direct top as to constitute a somewhat useless (and mechanically less efficient) substitute for it. That said, there will likely be arguments from car club members of the MG and Triumph persuasion - as with these cars the overdrive compensated for the shortcomings of gearboxes that were never designed for anything but placid family sedans, with ratios much too wide to suit sporting driving.
Over time the cost of manufacturing a five-speed gearbox instead of a 4 + overdrive was negligible, and by the middle 1960s there were many European manufacturers switched to the manufacture of 5 speed units - the Japanese were soon to follow. For cars built in this era, fifth gear was usually an overdrive ratio in the sense of being geared higher than 1I:1. Occasionally, as in the Getrag gearbox adopted for the Jensen-Healey
and optional in the BMW 2002
, fifth was direct and the final-drive gearing (or the tyre
size) may be altered to give an overdrive effect.
It really does not matter whether the highest gear is direct or not, save in the sense that a direct drive is more efficient than an indirect one. If a gearbox has a direct-drive ratio, all the others it offers must lose some power transmission efficiency through at least two pairs of meshing gears; whereas if all the gears are indirect, with the output shaft not coaxial with the input, each ratio involves only one pair of meshing gears, so all are of roughly equal efficiency. All indirect gearboxes were common in front wheel drive
cars, and were the rule in rear-engined cars.
Also see: Laycock Overdrive
| Borg Warner Overdrive
| Overdrive Definition