For the most part, motorcycles and carburetors have come to a parting of the ways. While a few carburetor-equipped bikes are still on the market, the vast majority of street bikes are now equipped with some form of electronic fuel injection, which despite the complication it adds, is both a better way to go, and undeniably here to stay, protests from the Luddites notwithstanding.
As a rule, internal combustion engines achieve their best performance with air-fuel ratios that hover in the 14.7 to 1 range. Although there's always some latitude, Leaner mixtures tend to degrade performance, while richer ones waste fuel and increase emissions, without substantially increasing performance.
Because lean settings create drivability problems, motorcycles in the pre-emission law days were normally set up on the rich side. This made them easy to start, quick to warm up and provided good performance. True, economy wasn't what it might have been and tail pipe emissions were off the charts, but at the time, gas was cheap and emissions unregulated. As long as the bike ran decently and raw fuel wasn't dripping from the pipes, no one cared how rich their jetting was.
All that changed when the first motorcycle pollution laws were enacted in 1979. Initially, most of the OEM's simply leaned out their jetting, but that led to other problems, and it soon became obvious that a better solution was needed. Initially, it looked like improved carburetors might be the answer and for a time it was. In fact, there are still plenty of carbureted bikes on the market that work just fine.
Unfortunately those days are coming to end, and here's why; Because carburetors rely on fixed orifice jets, overlapping fuel circuits, and volumetric pressure to deliver the correct fuel/air mixture there's not a whole lot of adaptability to them, at least not until you break out the screwdrivers and start changing things, so jetting is often a compromise between slightly lean at some throttle openings and slightly rich at others. This leads to things like slow warm ups, surges at small throttle openings and emission outputs that are borderline legal, and may be pushed over the edge by even the slightest adjustment or change in jetting.
On the other hand, electronic fuel injection systems employ a variety of sensors that tell a computer exactly what the engine is doing at any given moment. After comparing that information to a set of known parameters called a map, the computer determines exactly how much fuel is required to maximize power while creating the lowest emissions, then adjusts the air/fuel ratio accordingly.
Provided the map is accurately written, and with some minor exceptions they're usually pretty good, this allows the engine to receive the ideal mixture under every circumstance, neither so lean that it creates problems nor so rich that it exceeds emission standards, and it's for that reason that they've become the fuel delivery system of choice on everything from scooters to superbikes.
How EFI works
When a carbureted engine is running, airflow creates a low-pressure area in the carburetor venturi. Because the fuel in the float bowl is at atmospheric pressure while the pressure in the venturi is something lower, fuel is forced through the metering jets into the venturi, and from there it's carried by the moving airstream into the combustion chamber. In short, fuel metering is dependent on the size of the holes in the jets and the strength of the vacuum signal in the carburetor venturi. Because the vacuum signal is affected by everything from the engine's rpm and throttle position to the condition of the rings and the type of exhaust system you're using, it's a particularly crucial piece of the puzzle and one that can complicate the jetting process to no end. Granted that's a great oversimplification of how a carburetor works but it hits the high notes and hopefully, gives you some inkling of why it takes a very sophisticated carburetor to provide really accurate fuel metering.
In a sense, Electronic Fuel Injection is a lot simpler than a carburetor. Reduced to its essentials an EFI system is nothing more than a nozzle that sprays gasoline into the airstream whenever a computer tells it to. Of course, the devil is always in the details and those details, in particular the processes the computer uses to determine how much fuel to supply can get pretty complicated.
All electronic fuel injection systems work in essentially the same way; information about the engines operating state is conveyed to the computer via sensors. The computer then compares that data to a stored map of fueling requirements. Once it finds what it likes, it tells the injector to remain open and spraying fuel for a given length of time.
Since all of the hardware is roughly the same, the real difference between EFI systems is in how that hardware is driven, or mapped.
Because we're primarily concerned with V-Twin cruisers here, we're going to limit the discussion to the mapping most commonly used on them, the Alpha-N system, which uses throttle position to compute engine load and the Speed-Density system, which determines load by measuring the intake manifold vacuum.
Alpha-N systems are simple, a little crude by the latest standards and very effective. They don't require a whole lot of sensors or a very big ECU to get the job done, which makes them cost effective, and popular. While their simplicity, especially when it comes to remapping for tuning purposes makes them attractive to the go-fast crowd. As a bonus, they work very well when volumetric efficiency exceeds 100% as it does when forced induction is fitted, so if a blower or turbo is your thing, an Alpha-N system, with some upgrades to measure the increased airflow is just the ticket.
At its most basic, the system works by comparing the throttle angle or "alpha" to the engine rpm, or "N." Once it knows what rpm the engine is spinning and where you've got the throttle it makes a few quick, like microsecond quick, calculations and then selects the correct fueling points from the map. Other sensors are incorporated on an as needed basis to provide as much information as the systems designer deems necessary.
The downside to a pure Alpha-N system is that it's not the most precise form of EFI on the market and because of that, fuel economy and at times power output may fall short when compared to some of the more sophisticated systems. Alpha-N systems can also have some throttle response issues at low and cruising speeds, particularly if the throttle position sensor isn't precisely adjusted or needs cleaning and they are more sensitive to routine maintenance issues than other systems, something as small as a dirty air filter can throw one for a loop.
Lastly, because they are rather crude, straight Alpha-N systems are at the bottom of the food chain when it comes to emissions and mileage. Because of that, they are quickly falling by the wayside, at least as a stand-alone system.
Speed-Density systems determine engine load by measuring the manifold air pressure, that's the density part, and correlating it to the engine rpm, which is the speed portion of the equation. The ECU then uses that calculation to determine how much fuel the engine requires under the current conditions. Speed-Density systems came about because a system was needed that would lean out the mixture at idle and cruising speeds, enhancing fuel economy and lowering emissions under steady state throttle conditions yet still provide good response when the grip was twisted, especially from low speeds.
The answer was to monitor pressure changes in the intake manifold, via a Manifold Air Pressure or MAP sensor, and use that information to help determine fuel needs. Here's the deal; any changes in the engine's air requirements are reflected by near instantaneous changes in the intake manifold. If the ECU monitors the intake manifold pressure it can use the information to determine how much air is entering the combustion chamber and similarly how much fuel that air will need to combust properly. It sounds simple, and in principle, it is.
Although Speed-Density systems are more sophisticated than the Alpha-N types they do have their drawbacks. The most serious is that they don't adapt themselves very well to single and twin cylinder bikes. By nature, singles and twins suffer from constant manifold pressure variations especially at low speeds. Consider what happens at idle. When the intake valve is closed air pressure in the intake manifold is high because atmospheric pressure is pushing air through the air box, past the throttle and into the manifold where it piles up against the intake valve. As the intake valve opens, the air flows into the cylinder causing manifold pressure to drop. As the valve closes, and the incoming charge refills the manifold, pressure again raises.
These huge swings from low pressure to high and back again occur quickly, but because some time is involved, they complicate using a speed density system. However, there are ways to cope with the problem, and it's important to realize that at higher engine speeds the airflow smoothes out and the fluctuations diminish so it becomes much easier to take your readings. I should also mention that as you add cylinders, manifold pressures begin to settle down, simply because when one cylinder is at a low point another is high, so if you tie a pressure sensor into all your cylinders the overall readings will average out, which works just fine as far as an ECU is concerned.
The last thing worth mentioning is that because Speed-Density systems rely on manifold pressure they do their best work at low and moderate speeds where pressure changes are pronounced. At high speeds where manifold pressure changes have less impact, speed-density systems are less accurate.
Now that we've peeked at the philosophies behind the most common EFI systems let's look at how they actually function.
How they work.
When the key is turned on the ECU runs through a quick self-check that tells it everything is ready to go.
As the engine spins, the crankshaft position sensor tells the ECU how fast the engine is turning, where TDC is and what position the crank is in, while the throttle position sensor and MAP sensor indicate load. Note that both Alpha-N and Speed Density systems employ Throttle Position Sensor or TPS and MAP sensors. However, how the information is interpreted depends on the type of mapping being used.
The ECU compares the supplied info to its stored map, and selects a value it likes, that value representing the amount of time it'll hold the injector open, (its pulse width). If the exact value can't be found the computer will estimate what the engine needs by searching the map for higher and lower numbers and then picking something in the middle.
Once it has a value it likes, the computer will check to see if any further adjustment is needed. For example, if the engine temperature sensor tells the ECU the engine is stone cold or if the ambient air temperature indicates that the incoming air is particularly chilly ECU will assume the engine is in a cold start situation and select a fuel point that's rich enough to fire the engine, depending on the systems level of sophistication it may also increase the idle speed by moving the throttle stop via a small electric motor.
As the engine warms up and less fuel is required, the sensors transmit that info to the ECU, which will then lean out the mixture and lower the idle.
Obviously that's a very simple explanation of how an EFI system operates, and glosses over the literally thousands of inputs, calculations and decisions the ECU makes in fractions of second, but essentially that's how they work, everything else is just a detail.
Since both Speed-Density and Alpha-N systems share common sensors, and hold distinct advantages over each other at different throttle positions it's not uncommon for a manufacturer to piggy back the two systems together. Many EFI's systems use a Speed-density map at low and cruising speeds where the vacuum system is strong, then switch to an Alpha-N strategy at large throttle openings. In essence, this provides the best of both worlds, which is always a neat trick.
Open/Closed Loop Operaton
The ability to adjust itself, based on mapping is EFI's biggest advantage over a carburetor so any discussion of it has to include some comment on open and closed loop operation. A system that's operating in open loop is using the information it receives from its sensors to select a predetermined amount of fuel from its stored map. Once it selects the air/fuel ratio it wants it doesn't particularly care what happens in the combustion chamber or what's coming out of the tail pipe. The majority of fuel-injected motorcycles operate with open loop systems.
A closed loop system is one that uses an oxygen sensor to measure the amount of unburned oxygen in the exhaust system and adjust the mixture based on its inputs. For street use oxygen sensors are primarily used to limit emissions at idle and cruise speeds, and to protect the catalytic convertor (if one is being used) from overdosing on fuel. Closed loop systems operate only when there's little demand for power and conditions don't change very quickly, I.E. during idling and cruising. Whenever power is called for, the system must revert to an open loop. Because they cause the engine to run lean, closed loop operation can also cause surging problems at light throttle openings, especially on twin cylinder engines. For street use, closed loop systems are primarily driven by emission control requirements, so I'm positive we'll see a lot more of them in the future.
The single most popular modification most owners make involves replacing the exhaust system and air filter with something less restrictive. When that happens airflow through the engine is increased and so a corresponding increase in fuel is required.
Some EFI systems can cope with minor, (and I have to stress the word minor here) changes in airflow. Most can't, so they have to be coerced into providing more fuel. While there are some ECUs with replaceable chips, those are mainly found on older sportbikes, so we'll leave any discussion of them to the guys who own them. Likewise, reflashing a computer, or writing custom programs isn't for the faint hearted so I'd leave that option to the pros, particularly the ones with lots of experience and a dyno.
If you're planning on doing it yourself, the simplest and most cost effective way to modify the fuel map is by installing an add-on fuel module, or "cheater."
There are two types of cheaters on the market. The first intercepts the sensor signals and then modifies them, before passing them onto the ECU. Typically they'll tell the ECU that the coolant temperature is cooler than it is, so the ECU extends the injector pulse width, resulting in a richer mixture. The second, which is far less common, changes the injector pulse width after it leaves the ECU, leaving the sensor signals as they were.
The boxes are normally preprogrammed for a given application, and can be fine-tuned via trimmer screws. A few can even accept custom programs, and there's a couple that can be programmed on the fly using push buttons. Installation is generally straightforward, in most cases installing a box requires a lot less work than rejetting a carburetor and most are plug and play applications so you're good to go out of the box.
Of course, if you plan on going completely nuts, as far as building a high performance bike that is, the aftermarket will be only too happy to supply you with everything from oversize fuel injectors to new ECU's.
Down the Road
Despite their current level of sophistication, the existing state of the art in EFI systems has barely scratched the surface. As systems become cheaper and more technically advanced, and the government becomes more involved expect to see easier to access on board diagnostics, similar to the OBD II systems that all passenger cars now use. I think we'll also see things like launch and traction control become more common, at least on high end/high horsepower machines.
Already some of the more powerful bikes, the Suzuki B-King comes to mind, are equipped with alternate power curve mappings, one setting for rainy days, another for dry so expect to see more of them over time. Down the road, I'd also look for things like variable valve timing, especially on top of the line sport touring bikes, and more mundane technology like acoustical anti-knock sensors that sense detonation as it occurs, and adjust the fuel and ignition timing curves to eliminate it. All of this and more is already available in the automotive world so it's only a matter of time before it trickles down to even the lowest echelon of motorcycling.
So what's in your ECU?
While most EFI systems operate in a similar fashion each one is a dedicated design fitted and mapped for a particular motorcycle. Fortunately, there's plenty of shared hardware to help keep costs manageable.
Yamaha, Suzuki and Kawasaki (with some exceptions) all seem to prefer Denso processors in conjunction with Keihin throttle bodies. Typically, at low speeds these systems use a Speed Density map, and shift over to an Alpha-N map at larger throttle openings. Normally these systems also utilize sequentially fired injectors and incorporate tip-over switches to kill the engine in the event of a get-off.
Kawasaki cruisers use/used a Mitsubishi ECU with a Keihin dual throat throttle body on their Vulcan 1500-1600 cruisers, though whether the latest versions still do I can't say. The Mitsubishi system, a Speed Density- Alpha-N, set up is particularly friendly to large V-Twins, so although I haven't tried to verify how many other bikes employ it, I wouldn't be surprised to find it's being widely used.
H-D uses a Delphi system and ironically, (or not, depending on your point of view,) it's likely the best, most sophisticated system in current use. Like the majority of EFI systems, it's a Speed Density to Alpha-N system, but one that can also operate in closed loop. Atypically it features ion sensing, which means it can use the spark plug as a sensor to tell it what's happening in the combustion chamber. This is a feature that no other EFI system uses-yet, and pretty much guarantees that H-D will be able to get their bikes through emission testing without problems for many more years.
There are also a few oddball systems out there, early Triumphs and the defunct Excelsior-Henderson used Sagem, a French brand now owned by Johnson Controls. New Triumphs, including the twins use Keihin throttle bodies and ECU's. BMW, of course, uses BOSCH and those guys won't tell you jack. Honda uses its own in-house brand called PGM-FI and last, but by no means, least there's the Magneti-Marelli system as used by Ducati.
Interested in more?
Motorcycle Fuel Injection Handbook
- By Adam Wade, a lot to wade through, (pun intended), but extremely useful and more info per chapter than most of us can process at a single sitting.
Four Stroke Performance Tuning
- By A. Graham Bell. This one is car oriented but includes plenty of good solid info, and does have some motorcycle specific info included.