The '86-'87 Mustang 5.0 port...
The '86-'87 Mustang 5.0 port fuel-injected engines used SD fuel control, with a return-style fuel flow system. Note the lack of a MAF sensor on the air inlet.
On the earlier EEC cars, the fuel system was a return-style system, where the electric fuel pump in the tank simply pumped fuel at full flow all the time. The (manifold vacuum modulated) fuel pressure regulator then bypassed the excess fuel back to the tank. With the newer returnless fuel systems, it gets a bit more complicated. Now the ECU senses fuel pressure in the fuel rails relative to the manifold vacuum, and controls the fuel pressure across the injectors by pulsing the pump voltage. With this system, fuel is not needlessly pumped around and around, heating up in the process. But it requires a special pump to work with pulsed voltage.
Now that you understand the basics of fuel delivery, you need to understand how the ECU decides the proper injector pulse width or duty cycle for the required fuel flow, to achieve the desired A/F ratio.
There are basically three strategies used for EEC fuel control: alpha-N, Speed Density, and Mass Air. The alpha-N is the simplest system, but the least precise for part throttle and engine load transitions. Hence it is used only in racing applications or for coping with a failure mode on passenger vehicles. Speed Density and Mass Air can be much more precise, over a wider range of engine operating conditions, but their precision is only as good as the programmed volumetric efficiency calibration for SD, or air flow sensor accuracy for a Mass Air system.
Alpha-NFor alpha-N, the "alpha" refers to the angle of opening of the throttle blade, and "N" refers to the engine rpm. It's that simple. For the "black box" to know what's going on, a Throttle Position Sensor (TPS) will be needed, as well as some method of measuring the engine rpm (like a crank position sensor, or output from an electronic ignition distributor). Fuel control assumes a simplistic relationship between airflow, alpha, and N, i.e., at larger throttle openings, and/or higher rpm, more air flows into the engine, so it needs more fuel to maintain the proper A/F ratio. Therefore, EEC fuel control can simply be a programmed table of injector PW's for each value of alpha and N. This works well for naturally aspirated race engines (where radical cams leave little to no manifold vacuum), and the engine is either idling or at Wide Open Throttle (WOT), with not much in between. The in between and transition parts are where Alpha-N is not so precise, since the airflow varies with more than just rpm and throttle opening. For example, this system will not work well with a boosted application, since the ECU has no way of knowing if the airflow is being pushed into the engine, or with what pressure.
Here's an example of an Alpha-N...
Here's an example of an Alpha-N fuel control table. Cell entries are actual fuel injector pulse widths for various throttle blade angles and engine rpm.
Another downside to Alpha-N is the painstaking process to determine the correct PW for each rpm, and at each throttle position. Then, of course, as soon as something in the engine is changed, like say a new intake manifold is bolted on, you have to go back and recreate the PW table from scratch. Even if you change to larger (or smaller) injectors, you need to rebuild the entire PW table.
Speed DensitySD systems take fuel control a step further. Actual intake manifold pressure is now measured, using a Manifold Absolute Pressure (MAP) sensor, as well as Inlet Air Temperature (IAT, previously known as Air Charge Temperature, ACT, in older EEC systems), in addition to the previously sensed TPS and engine rpm. Now the ECU fuel control programming includes a desired A/F ratio table, the injector flow rates, engine cubic inch displacement (CID), a volumetric efficiency (VE) table, and the programs necessary to instantaneously calculate inlet airflow, required fuel flow (for the desired A/F ratio found in the A/F ratio table), and finally the correct injector PW.
Here's how SD works: The ECU will first sense MAP and IAT. Using the ideal gas law, it can calculate the instantaneous inlet air density (hence the "density" in SD). From the engine rpm and MAP, the ECU will look up (interpolate, if necessary) the programmed VE. Using the VE, engine displacement, and rpm, the ECU can calculate the instantaneous mass airflow into the engine. Looking up the desired A/F ratio from the base fuel table, the ECU will calculate the required mass fuel flow to achieve the desired A/F ratio. Knowing the number of injectors, and flow rating of the injectors, the ECU can calculate the required DC for the injectors. Finally, knowing the DC and rpm, the ECU can calculate the required PW for the injectors. Whew!
Let's illustrate with an example. Let's say we have a 5.0 engine operating at a manifold vacuum of 5 in Hg., rpm of 3,000, IAT of 80 degrees Fahrenheit, desired A/F ratio of 14.6, and it has eight 19-lb/hr injectors. For these conditions of rpm and MAP, the VE table calls for 85 percent.
After 1999, Ford went to a...
After 1999, Ford went to a returnless fuel system. Said system is found in any stock '03-'04 Cobra. A fuel rail pressure sensor (found on the driver-side fuel rail), monitors the fuel pressure, so the ECU can regulate the fuel pumps as necessary to maintain a constant pressure drop across the injectors.
From the ideal gas law, D = p/(RT), where D is density, p is absolute pressure, T is absolute temperature, and R is the gas constant for air. Before we can calculate the density, we need to fix the units so it all works. To save our sanity, we'll do the calculations in metric units. For the air temperature of 80 degrees F, the absolute temperature would be 300 degrees Kelvin. For an intake vacuum of 5 in. Hg, the absolute pressure would be 29.92-5 = 24.93 in. Hg. = 84.42 kPa. In metric units, R = 0.286 kJ/kg-K, the injectors would flow 8.36 kg/hr and the CID would be 0.005 m3.
So our inlet air density would be: D = 84.42/(0.286*300) = 0.984 kg/m2. Mass air flow is then: Ma = 11/42 D (VE)(CID)(RPM) = 11/42 (0.984)(0.85)(0.005)(3,000)= 6.72 kg/min. For an A/F ratio of 14.6, the fuel flow must then be: Mf = Ma/14.6 = 6.72/14.6 = 0.43 kg/min.
For eight injectors, each would need to flow 0.43/8 = 0.054 kg/min = 3.2 kg/hr. Since the injectors can flow 8.36 kg/hr when wide open, we only need an injector duty cycle DC = 3.2/8.36 = 0.385 or 38.5 percent. For a four-stroke engine (two revolutions per intake event), the time interval between intake events is t=2/rpm (in minutes) or 120,000/rpm (in milliseconds). At 3,000 rpm, the interval is then t=120,000/3,000 = 40 ms.
Finally, the required injector pulse width, PW = DC(t) = (.385)(40) = 15.4 ms. Isn't math fun?
 For SD and MAF fuel control,...  For SD and MAF fuel control, a Base Fuel table is used, which gives Open Loop A/F values for various rpm and Load values. For early EEC-IV, the Base Fuel table actually targeted A/F from load and Engine Coolant Temperature, as shown here. |  For SD fuel control systems,...  For SD fuel control systems, a Volumetric Efficiency Table is required, as shown here for an '87 Mustang application. Load is calculated from MAP. |  All '89-and-later Ford EFI...  All '89-and-later Ford EFI systems use MAF fuel control. The large 90mm '03-'04 Cobra MAF shown here is an excellent retrofit for earlier performance applications. |