The car tipped to retire in the opening laps crossed the finish line first. Behind it lies a story of engineering stubbornness, false diagnoses, and one last-minute idea — on the back of the steering wheel.
When John Barnard left McLaren at the end of 1986 and moved to Ferrari — more precisely, set up his own technical center in Britain and worked for Maranello remotely — he brought with him a habit of seeing a race car as a whole organism, not a collection of components. The carbon-fiber monocoque Barnard wanted to make as narrow as possible for aerodynamic reasons ran into a physical obstacle: the conventional mechanical gear lever passed through the cockpit, taking up precious space and dictating the design’s minimum width. The lever — a seemingly secondary detail — was setting the geometry of the entire car. Barnard decided to get rid of it.
The solution looked radical: replace manual shifting with an electrohydraulic system, with commands sent through paddles behind the steering wheel and the clutch engaging automatically on the move. At first, Barnard envisioned something simpler: two buttons on the wheel — up and down. It was Piero Lardi Ferrari, the founder’s son, who took the idea as far as the classic paddles on the back of the wheel: the driver would not have to take his hands off it at all. Seven gears, sequential logic. There was no clutch pedal whatsoever — and it was precisely its absence that freed up space in the narrow monocoque. For starts and pit stops, the driver controlled the clutch manually with a steering-wheel-mounted lever — the only moment when the car demanded old-school technique. After that, at speed, the hydraulics took over.
Funny thing is, Barnard was not the first person in Maranello to have such an idea. Back in 1979, Mauro Forghieri tested a semi-automatic transmission on the Ferrari 312 T3: Gilles Villeneuve drove a hundred laps of Fiorano in it — and rejected it. The exposed wires on the steering wheel unnerved him, and Enzo offered no support for the innovation. Ten years later, the technology returned, but in a different form and with a different motivation: this was about monocoque aerodynamics, while faster shifts were a welcome bonus.
Prototype 639 with the new transmission went testing in late 1987, and the tests continued throughout 1988. The car was not sent onto the track in race trim — turbo-equivalency rules were in force that season, and Ferrari was fighting with different weapons. But even in the hothouse conditions of testing, the gearbox behaved nervously. Gears refused to engage, the system lost contact with the electronics, and time after time the mechanics stripped the unit down, trying to find the source of its tantrums.
What gets me is how long they looked in the wrong place. The engineers were convinced the problem lay in the gearbox itself: the hydraulic valves, the electronic control unit, the cam-position sensors. They rebuilt the gearbox, modified it, installed it again — and got random failures all over again. Barnard pressed Magneti Marelli, demanding that it “beef up” the alternator: everyone thought it was overheating and failing to supply enough current to the car’s electrical system. The real culprit emerged later — and it had nothing to do with the alternator.
The new naturally aspirated V12, designed to reduce friction losses, had a crankshaft with only four main bearings. That layout left the crankshaft unbalanced: at certain revs, vibrations arose that made the alternator belt slip on its pulley and stop charging the battery properly. The alternator’s output dropped, the onboard electronics went dark, and the electrohydraulic gearbox, deprived of control signals, froze. From the outside, it looked like a transmission failure. But the transmission was not to blame — it had simply been left without electricity, like an elevator during a power cut. The issue was resolved together with Magneti Marelli, by reworking the alternator drive and the electrical supply. The false lead consumed months of work, but in the end gave the engineers something smooth development rarely does: an understanding of exactly where each component’s limits of reliability lay.
Enzo Ferrari died in August 1988, and an anxious uncertainty descended on Maranello — the new management questioned the team’s entire technical direction. The electrohydraulic gearbox was both the centerpiece of that direction and its main target: testing had been uneven, the season was approaching, and betting on unproven technology in the first race was not an engineering risk. It was a management nightmare. The answer was a backup plan: build a version of the car with a conventional manual gearbox. A safety net, Plan B, call it what you like — the point was to ensure the team would not find itself on the starting grid with a car incapable of reaching the finish.
The backup car was actually built: a conversion of the “stillborn” 639 chassis, its narrow monocoque cut open to make room for a gear lever. Nigel Mansell, who had come to Ferrari after years at Williams, got in, tried it — and delivered the verdict that became the final nail in its coffin: he preferred the paddles of the production 640. The decision seems irrational: why would a driver who needs championship points choose a car with technology whose reliability is in doubt? But Mansell could feel the speed. The paddles allowed him to shift without taking his hands off the wheel, without losing control in a corner, without that characteristic instant when the right hand yanks the lever while the left alone holds the car on line. For a driver with an aggressive style, it was not convenience — it was a weapon.
26 March 1989, Jacarepaguá circuit, Rio de Janeiro. The first round of the season. The Ferrari 640 — the car the paddock had whispered about all winter. Some called it brilliant, others doomed. A gearbox controlled by electronics and hydraulics, in Brazilian heat, vibration, G-forces — it all added up to an equation with too many variables.
Mansell started, manually releasing the clutch — the only moment in the race when the driver operated the transmission the old-fashioned way. Then the paddles took over. The race was hard, hot, and tense. But the 640 held together. The gearbox shifted, the hydraulics did not leak, the alternator spun, the electronics stayed alive. Mansell finished first.
A win in a debut race is the kind of story that feels scripted, but behind it were months of agonizing refinement, false diagnoses, and the stubbornness of several people unwilling to back down. I like the detail that Mansell — a driver known for physical strength and bluntness — became the first winner in a car whose philosophy was to remove physical effort from the shifting process.
The spectacular triumph in Brazil created the illusion that the problems had been solved. They had not. At subsequent rounds of the 1989 season, the gearbox kept letting down both Mansell and his teammate. Failures came in qualifying, on warm-up laps, in the middle of races. Every finish in the Ferrari 640 remained, in a sense, an event: the car either arrived brilliantly or did not arrive at all. There was no middle ground.
That binary — triumph or retirement — defined Ferrari’s entire championship that year. The team scored points in bursts, without the consistency needed to fight for the title. But every victory proved that the concept worked. Reliability is a matter of time and iteration; shift speed had been solved in principle. The rivals saw it. Williams, McLaren, and other teams began their own semi-automatic gearbox development programs. By the mid-nineties, steering-wheel paddles had become the norm across the grid — without an FIA directive, without mandatory regulations. Put bluntly, the technology won not because of an order from above, but because refusing it meant losing.
In 1997, Ferrari released the F355 F1 — a road-going sports car fitted with an electrohydraulically controlled gearbox and steering-wheel paddles. The direct line of inheritance from the 640 to a production car took eight years. The system had been adapted: the road version operated more smoothly, allowed an automatic mode, accounted for comfort — but the principle remained the same. The driver clicks a paddle, the electronics command the hydraulics, the clutch engages without a pedal.
“F1” gearboxes became a defining feature of road-going Ferraris for a decade and a half. They moved to the 360 Modena, the F430, the Enzo-line supercars. At the same time, the automotive industry was developing its own branches of automated transmissions — Volkswagen’s DSG dual-clutch gearboxes, Porsche’s PDK — but those systems had their own engineering genealogy, and folding them into Barnard’s racing solution would be an oversimplification. The racing paddle-shift layout influenced road-going Ferraris directly and demonstrably; its influence on the rest of the industry was more as proof that drivers were ready to give up the clutch pedal if they got speed and convenience in return.
Ferrari itself eventually moved away from single-clutch electrohydraulic gearboxes to dual-clutch units — starting with the California and continuing from there. But the paddles on the wheel remained. They remained everywhere: in sedans, crossovers, electric cars with simulated shifts. A control element born from the drive to squeeze every last centimeter out of a narrow monocoque outlived the transmission that created it. Today, steering-wheel paddles can be found in cars with no hydraulics, no mechanical clutch, not even a stepped gearbox — yet the gesture remains, because the hand on the wheel wants to shift.