You love late-model muscle cars, but do you understand what makes them tick? Do you know why your '10 Camaro SS not only has more power than equivalent cars of the '60s, but also with vastly better drivability and fuel economy? Why the 1980 Turbo Trans Am was a lackluster machine, but the 1989 model was such a killer? There are a lot of reasons that engines experienced such improvement, and so many of them boil down to Electronic Fuel Injection. EFI so revolutionized hot-rodding that the original High-Tech Performance magazine was born (and became GM-only in 1998). Back then, GMHTP readers were a distinct minority compared to those who thought carburetors would forever reign supreme, but today even the most diehard old-school aficionados have been converted after seeing the incredible combination of power levels and drivability EFI makes possible. These days, you probably take EFI as a given, but it wasn't like this 15 years ago. Back then, many saw it as a hindrance to backyard mechanics fixing or modifying their own vehicles. To love EFI requires knowing the ins-and-outs of how it functions, so we figured a "get to know your EFI" article was in order. After all, there's a lot of unseen business that goes on between the time gas is pumped into your tank and it ends up as motion down the road; this article should be a good first step at figuring it all out.
What Came BeforeCarburetors, that's what. Beautifully simple in their basic operation, carburetors took advantage of the basic premise of physics that a fluid in motion (in this case, air flowing through the venturis) is at a lower pressure than stagnant air, thereby sucking fuel into the flowing air stream and letting the mix find its way through the intake manifold to the intake ports, and past the intake valves into the cylinders. Carburetors fell by the wayside when emissions standards grew more stringent during the 1970s and 1980s; emissions-compliant carburetors proved too complex and expensive to be worthwhile. A better way to more precisely meter fuel was needed in order to achieve more consistent, predictable compositions of gases and particulates in the exhaust, thereby allowing catalytic converters to do their job reliably over the life of the vehicle. That better way was EFI.
(Note: Many muscle car aficionados are familiar with the mechanical fuel injection systems seen on Corvettes of circa-1960 vintage; while very progressive in design, their lineage did not lead directly to EFI. A study of these systems is left to the Google skills of the reader.)
BackgroundThe first production cars with EFI systems date back to the late 1960s (with some notable aborted attempts a decade earlier), but it didn't start showing up under the hoods of GM muscle machines until the mid-1980s. Since then, EFI systems have evolved tremendously, advancing in step with increases in computing power, tightening of emissions limits, and desires to improve fuel economy. The reader's basic knowledge of the four-cycle spark-ignition engine is assumed for the purposes of this article, and we'll leave the components of the spark side largely out of the equation. Still, it's worth noting that while spark plugs still ignite the mixture in the cylinder, distributors and the snake's mess of spark plug wires have become a thing of the past as computer-controlled individual coils now send the energy to the plugs. Also, we're largely ignoring small details like Idle Air Control (IAC) valves and like tangential equipment that has come and gone as EFI has evolved, though some may be mentioned where they happen to come up.
Types of Electronic Fuel InjectionWhat follow are the three main styles of EFI, their differences having mainly to do with the placement of their fuel injector(s). In increasing order of sophistication and cost, they are:
Throttle Body Injection (TBI): In operation, the earliest and most crude form of electronic fuel injection wasn't that different from a carburetor, employing one or more large, low-pressure fuel injectors attached to a throttle body assembly. Like a carb, this setup relied on the intake manifold to mix the fuel with the air and keep it mixed on the way to the intake ports. GM muscle car enthusiasts first saw TBI show up as the Cross-Fire Injection systems of late C3 and early C4 Corvettes, and they could be found on certain F-body engines through the end of the third-gen era. By the mid-1990s, GM finally removed these low-cost systems (and the not-much-more-advanced "Central Port Injection" setups) from its trucks and said bye bye to TBI.
Port Fuel Injection (PFI): Also known as Multi-Port Fuel Injection (MPFI or MPI). Still the most common form of electronic fuel injection on production vehicles, the GM V-8s we know and love have had PFI ever since Tuned Port Injection (TPI) appeared for 1985. By using individual fuel injectors for each cylinder and placing them just upstream of the intake port, PFI systems do not rely on the intake manifold for air/fuel mixing purposes, instead precisely metering a measured amount of fuel to the intake port of each cylinder. PFI systems generally have the injectors mounted as close as possible to, and aimed directly at, the intake valves. (While the cathedral-port design of early LS engines certainly look to be designed largely for injector aiming, in reality the unusually tall port shape had to do more with maximization of port cross-sectional area between the pushrods.)
Direct Injection (DI): Also known as Spark Ignition Direct Injection (SIDI) or Gasoline Direct Injection (GDI) to differentiate it from the compression-ignition direct injection scheme of today's diesels, this type of fuel injection can be thought of as essentially the same as PFI, with one key difference: The fuel injectors spray directly into the combustion chambers. This type of EFI has seen increased adoption in production cars in the last few years, but thus far it's seen only limited use in the types of rides that concern GMHTP readers: the fifth-gen Camaro features DI, but only with the V-6, as did turbocharged models of the Solstice/Sky and Cobalt. Though no GM V-8 to date has used direct injection, the wait won't be long, as you can bet on it for the Gen V small-block debuting in the 2014 Corvette.
EFI System InputsThe types of sensors used in EFI vary by system type, age, and its exact execution, but they're very close for both PFI and DI (let's just go ahead and ignore TBI from here on out, shall we?). There are two main types of information the EFI computer needs in order to do its job: characteristics of air entering the engine, and characteristics of exhaust exiting the engine. Both of these, along with other parameters such as engine RPM, coolant temperature, and throttle position, determine how much fuel needs to be injected at any given instant; required spark timing is also a function of many of these inputs, but again, the ignition system is a secondary issue for the purposes of this article (as is the self-diagnostic capability of modern engine computers). Making our way from the air filter to the tailpipe, the sensors and their most common abbreviations are:
Intake Air Temperature (IAT): Most often located between the air filter and throttle body, this sensor measures the temperature of the air being sucked into the engine. Its output is a voltage that varies with temperature. Crucial to the operation of speed density systems, it is also present on MAF-based systems though plays a smaller role.
Mass Air Flow (MAF): Always located between the air filter and throttle body. This sensor generally consists of a heated element that is held at a constant temperature; more air flowing past it has a cooling effect, hence more current is required to hold the element at a given temperature. Its output on most GM EFI systems is a wave signal that varies in frequency and is input to the PCM as Hertz. This frequency, which ranges from 0 to about 12,000 Hz or higher (depending on type of MAF and ECM), is converted by the ECM to g/s or similar measure of air mass per unit time.
Throttle Position Sensor (TPS): Located on the throttle body, it's most often a potentiometer that gives an increasing voltage output as the throttle blade rotates further open. Its output is used by the PCM to determine whether the driver wants the engine to increase its output (accelerate the car), decrease its output (decelerate the car), or remain at constant output (cruise), as well as any rate of change of acceleration or deceleration.
Manifold Absolute Pressure (MAP): Located on or near the intake manifold. Crucial to the operation of speed density systems, this sensor is also present on many MAF-based systems. Its output is a voltage that varies with pressure. Factory units on normally aspirated engines are often referred to as "1-bar MAPs" due to their inability to read anything over atmospheric pressure (occurring during full throttle conditions); supercharged and turbocharged engines generally use 2-bar (up to 15 psi boost) or sometimes even 3-bar (up to 30 psi boost) MAPs.
Oxygen (O2) Sensors: With most other sensors located in or about the intake tract, these are typically the only sensors located in the exhaust stream. Also known as lambda sensors, they output a voltage (typically 0-1 V) that varies with the oxygen content of the exhaust, which is an indication of the air-fuel ratio (AFR) of recent combustion events. Too much oxygen indicates a leaner-than-stoichiometric condition, while too little oxygen indicates a richer-than-stoichiometric condition. These sensors enable "closed loop" feedback control capability where the EFI system can make slight adjustments to fuel delivery to dial in AFRs as close as possible to those desired. Most factory O2 sensors are of the "narrowband" variety that can only indicate AFRs close to stoichiometric (14.7:1) and therefore have no use for anything beyond part-throttle driving conditions. So-called wideband O2 sensors are now used on some newer EFI systems; these can detect a much wider range of AFRs including those present under wide open throttle.
Other Sensors: Various other sensors exist on EFI systems that do not have to do with measurement of the intake or exhaust stream, but have important functions nonetheless. Here are a notable few. The Coolant Temperature Sensor (CTS), as you may guess from its name, gives a voltage to the ECM that varies with coolant temperature; low temperature readings will cause the ECM to take actions such as entering cold-start enrichment mode, while high readings might cause the ECM to activate cooling fan(s) or retard ignition timing to avoid detonation. All modern EFI systems also have one or more knock sensors, microphone-like devices that are tuned to detect a range of frequencies typical of the sound of abnormal combustion; ignition timing is generally retarded in response to such a signal, even though such a signal might be "false" (such as when different internal engine components alter the sound characteristics of the engine, or external parts like exhaust systems physically rub against the vehicle's chassis). Newer setups, especially those incorporating returnless fuel systems (more on that below), typically have a fuel pressure sensor mounted on the fuel pump assembly, along with another one on the fuel rail. Finally, the latest EFI systems also have both a Crankshaft Position Sensor (CKP) and one or more Camshaft Position Sensors (CMP) that output wave signals indicating the rotational positions of the crankshaft and camshaft(s); in addition to being required for sequential fuel injection systems, the information they provide to the ECM allows it to keep tabs on events like misfires and the status of any variable valve timing systems.
The ECM & the two main styles of EFI systemsThe brain of any electronic fuel injection system is the ECM (Electronic or Engine Control Module, sometimes referred to as the Electronic Control Unit or ECU). As these computers have grown in power and scope, they're more modernly known PCMs since they're responsible for control over the entire powertrain. Older applications had an ECM and a separate TCM for the automatic transmission (manual transmissions require few, if any, electronics).
The approaches taken by the ECM in determining how much air is going into the engine at any given instant have changed and evolved over the years. These approaches ("styles," if you will) also differ in some of the components included in the EFI system. The style distinction has by and large dissolved in recent years (really, since the debut of the LT1) as the latest ECMs use a combination of the two approaches to achieve the most accurate fuel delivery possible under all engine operating conditions--hence the existence of both a MAF and a MAP on the latest engines. The following descriptions are basic and simply designed to foster an understanding of the fundamental differences between the approaches.
"MAF-based:" The amount (mass) of air entering the engine is measured by the MAF, with corrections made based on atmospheric conditions. The main issue here is ensuring the MAF is properly calibrated to a given engine's intake tract and other basic characteristics, such that the air flow indicated by the MAF represents reality. If the factory anticipated that higher-than-stock maximum air flow rates (i.e., horsepower levels) might occur out in the field (such that the MAF output doesn't "peg" at some maximum level the sensor is capable of reporting to the ECM), custom tuning is often not even required on MAF-based systems for moderate levels of engine modification. As a further example of the versatility of this approach, the ECM's knowledge of boost level is not even required for forced inducted engines. The best example of "pure" MAF-based systems were those used on the original TPI engines from 1985-1989. "Speed Density:" The amount (mass) of air entering the engine is not directly measured, but rather is inferred based on the temperature of incoming air (IAT), the pressure in the manifold (MAP), and engine RPM. Such systems require a volumetric efficiency (VE) table stored within the ECM to determine how much air (of a certain calculated density in the manifold) actually gets sucked into each cylinder on each intake stroke, and this will vary greatly based on engine RPM and characteristics of engine components (displacement, camshaft profile, intake manifold flow characteristics, etc.). "Pure" speed density systems existed on 1990 and later TPI engines. Advantages of a pure speed density strategy have diminished with the trend toward modified cars retaining today's advanced OEM, reprogrammable ECMs and the availability of larger and higher-flowing MAF sensors. However, the speed density approach is still used with very lumpy cams (allowing, for example, a more stable idle), with intakes where use of a MAF is problematic for packaging concerns (such as with carb-style manifolds), and on all-out race engines where any intake restriction from a MAF is unacceptable.
EFI System OutputsJust as the inputs (sensors) used in EFI have increased in sophistication over the years, so has the delivery of spark and fuel--particularly the latter. The basic EFI system outputs, and a few of the more recent nifty add-ons, are as follows, with fuel injectors and some of their attendant issues treated separately afterward.
Ignition System: One or more coils deliver spark to the engine's spark plugs at the instant commanded by the ECM. Ignition (spark) timing varies by parameters such as engine RPM and load, can be retarded in response to the ECM detecting abnormal combustion events (based on inputs from the knock sensor(s) or crankshaft position sensor).
Fuel System: Getting fuel to the engine starts with an in-tank electric fuel pump that sends pressurized fuel to the fuel rails via fuel lines running underneath the vehicle. Both return-line-equipped and returnless systems exist, with a switch to the latter on newer vehicles being required due to evaporative emissions concerns. Exactly what pressure the fuel rails experience used to be controlled by a vacuum-actuated regulator mounted on the fuel rails, but today is determined by an in-tank regulator that is part of the fuel pump module. PFI engines typically require fuel pressures in the neighborhood of 40-60 psi, though some new setups such as that used with the supercharged LS9 use a supplemental pump to deliver up to 87 psi under hard-throttle conditions (lower fuel pressures at part-throttle allow the sufficiently wide injector pulse widths favorable to emissions and drivability; more on injector pulse widths shortly). On DI systems, extremely high fuel pressure of 2,000 psi or more is required; electric pumps can't efficiently provide this, so (in a twist of irony harking back to the carbureted days) a supplemental cam-driven mechanical fuel pump is utilized to jack the pressure at the fuel rail up to that needed.
Other components: The newest EFI systems have taken on more components in the never-ending quest for reduced emissions, improved efficiency, and increased power output, not to mention the need to incorporate advanced features like traction control and additional self-diagnostic capabilities. Two of the most important are Electronic Throttle Control and Variable Valve Timing. Most engines now have Electronic Throttle Control (ETC), which involves the ECM sending commands to a stepper motor that actuates the throttle blade. This provides myriad advantages, including simplifying cruise control systems, improving traction control and active handling systems, allowing momentary torque reduction during automatic transmission upshifts, and on and on. Variable Valve Timing (VVT) uses ECM-controlled camshaft actuators to alter camshaft phasing relative to the crankshaft; ubiquitous on overhead cam engines, it's now finding its way into pushrod engines as well. Advantages include increased efficiency at low RPM, improved output at high RPM, and a broader torque curve.
Detail on Fuel InjectorsAll other components of the fuel system simply provide pressurized fuel to the inlet of the injector; getting the correct amount of this fuel into the cylinder at the right time is the injector's job, under the command of the ECM. The style of fuel injector used necessarily varies by system type (TBI, PFI, DI), and because of the prevalence and ubiquity of PFI, there is huge variation among types of PFI injectors used and available. We're focusing on injectors and their behavior in a PFI system, but the lessons learned here can be applied to other types of EFI with appropriate modification.
The interior of a fuel injector consists of an electromagnetic coil that, when energized via current commanded by the ECM, opens a check valve and allows fuel to flow through it. When this current is turned off, a spring inside the fuel injector closes the check valve. The theory is simple enough, but when you throw in the fact that not only does this need to happen in a matter of milliseconds, but the injector must exhibit acceptable performance at both docile idle and screaming high-RPM conditions, things get pretty complicated.
The time during which the ECM is commanding the injector to be open is known as the pulse width; pulse widths are at a maximum when the engine is operating at its highest fuel needs per cycle. This happens at peak torque, corresponding to maximal cylinder filling. As RPM climb past peak torque, less fuel needs to get injected per cycle, so pulsewidths fall simultaneously with the need for more pulses per second. You may have heard the term "duty cycle," which has to do with the percent of time an injector stays open. Noting that every injector needs a small but measurable amount of time to go from closed to open (and open to closed), tasking most injectors beyond 80 percent duty cycle is a recipe for unstable fueling, dangerously high AFRs, overheated injectors, and/or a fried ECM.
The cure for a too-high duty cycle is "larger" fuel injectors. This is a misnomer as such injectors are not larger in physical size, but in increased flow capability (lb/hr or cc/min). A 60-lb/hr injector will, in an ideal world, inject double the amount of fuel of a 30-lb/hr injector for a given pulse width. (I say "in an ideal world" because there are many issues surrounding the design and performance of fuel injectors that we are ignoring, such as lag times and voltage offsets that vary by injector.) Can an injector be too large? You bet it can… a very high-flowing fuel injector will have trouble delivering precision amounts of fuel at idle conditions, where pulse widths might need to be so brief that the injector may need to start closing before it fully opens. Helping compensate for this is that most EFI fuel systems vary fuel rail pressure by engine load (some more than others), with fuel pressures at low-load conditions being less than those during large throttle inputs. This means that fuel pressures at idle can stay low enough to keep pulse widths long enough for repeatable, accurate (and emissions-friendly) fuel delivery, so long as you don't select too large of an injector.
Some words of caution on fuel injectors. Most OEM applications use high-impedance injectors (think of impedance as a measure of resistance of the electromagnetic coil). While super-high flow rates are possible with a swap to low-impedance injectors, they are probably only needed if you require more than about 80 lb/hr of injector flow. Regardless, going the low impedance route usually involves swapping to a different injector controller or using an impedance converter to avoid cooking the ECM; low-impedance injectors typically are run off of the more complex peak-and-hold circuitry (with its advantageous faster dynamic response) versus the saturation driver circuitry scheme typical of OEMs. Also important to note is that older-type Minitimer-connector fuel injectors can't be directly swapped with newer-type USCAR-connector injectors; not only are the older-style injectors usually taller, but the connectors require an adapter harness (readily available) to be retrofit.
A few final comments on injector timing are in order. DI offers the ultimate control over injection of fuel, and there can be multiple injections during different strokes based on engine operating characteristics (even during the exhaust stroke to aid catalyst warmup!). PFI doesn't have that luxury, as any fuel injected needs to get past the intake valve before it can get to the combustion chamber. The most modern systems, nearly ubiquitous for OEMs now, are known as Sequential Fuel Injection (SFI) and inject fuel just before or during the period the intake valve is open and air is flowing through the intake port. Less sophisticated systems are commonly known as "batch fire," where injectors fire simultaneously (often once every crank revolution, or twice per cycle), regardless of whether a given cylinder's intake valve is open at that time. One might think of a batch fire system as a "wet port" fuel injection system, where the air/fuel mix sits in the port waiting for the intake valve to open and let it in. It's arguable whether SFI systems allow for more power potential, but they definitely improve emissions.