In Fuel Injection Simplified, Parts 1, 2, and 3 we explored the benefits of modern fuel injection, became familiar with its parts, and reviewed its operation, respectively. In this final article we look at the history of fuel injection and some ways fuel injection is evolving.

Not long after the diesel engine industry switched from coal dust to light fuel oil, one man supplied all its fuel injection technology. That man was Robert Bosch, a developer of the technology since 1912 and a name now virtually synonymous with it. By the end of WWII, fuel injection's already proven war application inspired a young hot-rodder named Stuart Hilborn to adapt the technology to his Ford racer, and the American fascination with fuel injection began.

The earliest fuel injection systems were entirely mechanical, meaning that timing, pressure and distribution were, like Hilborn's, all handled by mechanical means. German WWII fighter planes for example used fourteen fuel pumps -- one for each cylinder--mounted in a common housing containing a rotating valve. This engine-driven system's output varied with rpm, and a spill valve rerouted fuel back to the tank in inverse proportion to throttle opening; i.e. more throttle, less return to the tank, and visa-versa. Systems managed by throttle opening and rpm became known as "Alpha-N" type, the phrase derived from the engineering symbols for angle (the throttle's) and speed (the engine's). Today they are more commonly called "Speed/Density" (the engine's speed, the intake's density). As for the injectors, as in all mechanical fuel injection systems, the warplane's (and later car systems -- think 1980s VW Rabbit) were spring-loaded and oscillated rapidly to maximize atomization. With parts and controls like these, early fuel injection systems were little more than controlled leaks.

Mechanical fuel injection had one big drawback: with only throttle opening and rpm for controls, output was very linear. This was a problem because the fuel needs of an engine are anything but linear. The resulting mismatch created wide mixture variations. When American carburetor manufacturer Rochester put fuel injection in Chevrolet's Corvette, it addressed linearity by including a venturi linked to a fuel metering valve, thereby adding flow information to the standard Alpha-N control format, resulting in an output more closely matched to the combustion characteristics of the engine. The Bosch company went a step further by making the venturi hourglass-shaped, to more closely approximate most engine's fishhook-shaped rich-lean-rich fuel curves. Each represents the best in mechanical fuel injection systems.

Rochester eventually sold its technology to Bendix, who improved it by replacing many of the mechanical parts with electronic ones, creating the first mass-produced electronic fuel injection system. Detroit failed to sell more than a handful of fuel injected cars however, prompting Bendix to turn its rights over to Bosch, who persistently refined the system until the early 1970s, when emissions and fuel economy concerns coincided with advances in electronics technology to make fuel injection more acceptable and affordable. A few years later, microprocessors turned electronic fuel injection into "smart" systems.

In 1980 Kawasaki introduced motorcycling's first mass-produced electronic fuel injection (EFI) system. This L-Jetronic (Bosch) patterned EFI system relied on a swinging air flow flapper valve plus engine rpm inputs into the ECU to calculate basic fuel discharge. The cylinder head mounted engine temp sensor and the air pressure sensor then added information for final lean/ rich adjustment immediately before injector operation. This 360° system's injectors redundantly squirted with each crankshaft revolution, an inexpensive design that also offered increased fuel vaporization in the intake manifold. Later more sophisticated 720° systems would squirt only when needed.

Most current powersports fuel injection systems are what is known as indirect, meaning the fuel injector sprays fuel into the intake manifold. Systems that spray fuel inside the cylinder directly are called "direct injection" type. Found on both two-stroke and four-stroke engines, one of direct injection's big benefits is injection so vigorous and fast that the engine can run with greatly reduced valve opening durations, potentially limiting exhaust emissions. Other fuel injection system developments include ECUs that contain exhaust catalyst protection programming that shuts off ignition or fuel when abnormal combustion might harm the catalyst, automotive-like "flight recorder" circuitry that stores vehicle use data for law enforcement purposes, the ability to update ECU mapping live during routine dealer visits, On-Star-like satellite linking, more advanced onboard diagnostics, and much more.

And then there's fly-by-wire. On many machines today, particularly sport bikes but also some cruisers and even personal watercraft, an ECU-controlled servomotor is inserted between the throttle grip and actual throttle butterfly, for "ride by wire" operation. The rider operates the throttle grip, which either turns a bellcrank on the throttle body connected to a sensor, or communicates with the sensor direct, and thus requested throttle opening position data is sent to the ECU. The ECU takes that request plus the usual sensor inputs and decides how and when to move the throttle butterfly. This technology interposes one more layer between the operator and the engine, for increased engine response and improved emissions. It also brings the throttle body closer to the point of being cruise control ready, and many ride-by-wire systems incorporate this feature.

There is also a newly-developed type of throttle position sensor (TPS) in use on many new powersports vehicles, one that does not use a wire coil and wiper to communicate throttle angle information to the ECU. A fuel injection system's TPS is one of the most delicate sensors and thus very sensitive to forces that might cause it to fail. Dirt, electrical system voltage spikes, and just plain wear -- all play havoc on TPS. Many new TPS are therefore Hall-effect type. A Hall-effect sensor starts with a printed circuit board that has voltage running through it. When a magnetic wiper nears the circuit board, it modifies the incoming voltage and sends the modified signal out the other end. The Hall-effect TPS is better than the traditional kind because there is no physical contact, thus no wear and higher accuracy. Even some of the old style contact type TPS exhibit some new twists. Many vehicles have redundant technology TPS. Instead of just one winding in the TPS, the redundant TPS boasts two parallel full-time windings that work together. The ECU monitors the signals from both, and when it "sees" a significant difference defaults to the extra winding, "knowing" that the other has failed.