Motorcycle fuel systems have switched from analog to digital, from carburetors to electronic fuel-injection. This may put the noses of some old-timers out of joint, but it's just another way to perform the same old fuel-metering task. Every function in fuel-injection has its analog in carburetion.
Why noses out of joint? The issue is transparency. A carburetor is something you can take apart, find its problem, fix it and reassemble – all at the roadside and without a notebook computer. Attractive! But the black box aspect of fuel-injection works so well, cold or hot weather, high mountains or low laying valley's. The first thing every fuel system needs is a reference fuel pressure. It won't do to have our system run rich when the fuel tank is full and lean when only a quart remains. We can't have our mixture lean out because a low battery spins the fuel pump slower than usual.
So, in a carburetor we have a float bowl to maintain a constant fuel level, so that gasoline is always ready to flow from idle or main systems. In fuel-injection this function is performed by a pump and fuel-pressure regulator.
The carburetor meters fuel through calibrated restrictions called jets, traditionally made of brass. I have boxes of them, left over from the Paleolithic. The force that pumps the fuel is the pressure of the atmosphere pushing down on the surface of the fuel in the float bowl. No fuel flows when the engine is stopped because pressure in the carburetor's air passage is the same as in the float bowl. But when the engine starts, intake air rushes through that air passage, and its pressure fails as some of that pressure is transformed into velocity (conservation of energy). As that pressure falls, atmospheric pressure in the float bowl can push fuel through the jets and out into the air stream. The more air the engine pumps, the faster it moves and the more pressure it loses. And so the constant pressure in the float bowl pushes out more fuel. The details are complicated but you get the idea. In the digital fuel-injection system, fuel is metered only by how long each cylinder's injector – a little solenoid-operated fuel valve – is turned on. As a cylinder's intake stroke begins, its injection clicks on and sprays fuel at the regulated fuel pressure into the intake-air passage. A digital computer uses both stored information and updates from sensors to decide how long to spray fuel. Then the injector clicks off.
When the rider suddenly opens the throttle to accelerate up and on-ramp, more air rushes into the engine's cylinders, but fuel, being 600 times heavier than air, accelerates less quickly. The result-unless something prevents it-is a momentary lean mixture and the engine stumbles or quits firing. In carburetors, this is handled in one of a variety of ways. One simple device is the accelerator pump – a little diaphragm or piston pump normally kept full of fuel. Throttle movement operates this pump, spraying extra fuel in proportion to the amount of throttle opening to prevent the lean stumble. Or one of the jets is made as a U-tube, full of fuel at low throttle. One end of this U-tube is at atmospheric pressure, and the other end opens into the low pressure of the rushing intake air. Throttle opening reduces pressure in the airstream, so atmospheric pressure pushes the U-tube full of fuel into it. No stumble. In so-called CV or constant-vacuum carburetors, there is a second throttle, lifted only by vacuum but slowed by a damper. Its slower opening lets the fuel keep up.
In electronic fuel-injection, the throttle shaft has a throttle-position sensor (TPS), reporting its rotation to the control computer. Software is written that commands injectors to stay 'on' for extra milliseconds when that sensor detects rapid throttle opening. How does it 'know' how much extra to inject? It doesn't, any more than an electric motor knows to start running when you switch it on. The computer has been given that information, derived from experimental dyno tests on a similar engine.
How does a carburetor know what mixture to supply? That is built into fuel jets, air bleeds and devices that affect the partial vacuum produced by intake airflow as it passes over fuel orifices. All this was worked out over 100 years of trial and error. In digital fuel-injection, injector on-time is planned for standard reference conditions in dynamometer testing and is modified for temperature, pressure and other changes by using sensor data. When a carburetion guru uses an air-density meter to choose jet sizes, it is the same math.
When an engine is cold, only the most volatile fraction of the fuel can evaporate, leaving the mixture too lean to fire. It needs enrichment. With a carburetor, we have either a choke or a tiny starting carburetor built onto the side, jetted very rich. In either case, we move the choke lever to the 'start' position.
Injection has an engine-temperature sensor and a stored internal schedule of how much to extend injection events to achieve the necessary starting mixture at various temperatures. With a carburetor, we move the choke lever to keep the engine running as its warms up. Injection does this invisibly and perfectly.
When a carbureted engine is cold-started, its idle is too low to keep running, so we fuss with the throttle to make it run. But the injection engine has either a motor-driven throttle positioner or a controllable idle air leak. This lets the computer smoothly maintain idle rpm for us.
As temperature and barometer values change, fuel-injection compensates, just as if a veteran race tuner were on the job with an air-density meter and a jet wrench. Carburetors just passively run richer or leaner. While not accurate or 'environmentally friendly,' carburetors did the job for a century.
I've never had a fuel-injection system quit on me, but I know what I'll do if one does. I'll walk.