EFI Tech, Part 2

December 9th, 2011

Text and Photos by Dave Emanuel

Photo by Doug Adams

Acronyms in Action

EFI systems don’t respond directly to changes in rpm or engine load—they do it through sensors, device controllers and a decision maker known as a control module.

The alphabet soup of acronyms that is part of any electronic fuel injection system is often a source of confusion. The definitions of those acronyms were addressed last issue. This issue, those acronyms spring into action.

If you have years of experience working with carburetors, you may wonder why electronic fuel injection requires so many sensors while a carburetor doesn’t require any. Simply stated, a carburetor is a direct-response mechanical device. Rather than sensing conditions, processing the information and responding with commands, a carburetor is designed to respond directly to changing air flow conditions and throttle position settings. Air flow acts directly upon the fuel discharge circuit, so as it changes, so does the amount of fuel that enters the air stream. While the accelerator pump circuit isn’t referenced to air flow, it is controlled by throttle movement and position.

In contrast, electronic fuel injection has no means to directly respond to either throttle position or air flow, so sensors are necessary to communicate information to a controller, which in turn uses electronic “commands” to alter fuel flow as required. However, as their names imply, an engine control module or ECM (also called an engine control unit—ECU, or powertrain control module—PCM, if it also controls the transmission) typically controls all aspects of engine operation, including ignition timing.

A supercharger and a few engine modifications put this street-driven Corvette into the low 11s, under the control of the original powertrain control module, which was reprogrammed with EFILive's FlashScan scanning (data monitoring and logging) and tuning software. This software allows the tuner to access and modify the tables inside original equipment GM ECMs and PCMs for maximum performance and fuel economy.

As soon as electrical power is switched on, an engine controller, (subsequently referred to as an ECM) begins collecting sensor input and processing information. In a typical scenario, an ECM’s first actions following “power on” are to monitor sensor input, activate the fuel pump and set the idle air controller (or electronically controlled throttle) to the prescribed cold stat idle position. If the ECM doesn’t receive an rpm signal within two to four seconds, it turns the fuel pump off, and waits until it receives an rpm signal before switching the pump on again.

With most systems, the ECM is live whenever power is applied and if scanning software is connected, you can monitor sensor and device controller status. But the ECM doesn’t get down to serious business until the starter is engaged. That’s when it starts translating the whole range of sensor input data into device controller output commands. That translation is performed by using a set of inputs to point to a cell location within a specific table. Within that cell is the data that will be communicated to the relevant device controller.

ECMs and PCMs come in many different flavors. These controllers were used on pre-1993 vehicles, and although they use different styles of removable PROMs (in which control data is stored), they operate in similar fashion. Over time, the capabilities of factory control systems have been expanded, which necessitated the use of larger capacity PROMs. (More control functions require more tables.)

While the starter is cranking the engine over, the relevant tables (which vary depending on the system) typically relate to throttle or IAC position, ignition timing and fuel flow. The sensor inputs that are used most commonly to determine the cell locations to be used for device controller command data are coolant temperature, inlet air temperature, manifold absolute pressure, engine rpm and throttle position.

Essentially, the system operates like a multiplication table on steroids. If you didn’t know the answer to how much 8 multiplied by 7 is, but had a multiplication table, you would find the answer by locating the row for number 7 and following it across to the point where it intersects with the column for number 8. The cell at that intersection would contain the answer, in this case 56. If an ECM were looking for idle speed, it would find the row for current coolant temperature and follow it across to the column for desired idle speed. Most systems have at least two columns, idle speed in park/neutral and idle speed in gear, so the ECM would move across to the value in the appropriate column.

Many tables are far more complex, but the process is the same. In the case of ignition timing, one axis is associated with rpm and the other with engine load. When the ECM searches for an ignition timing value, it finds relevant engine speed, than travels to the cell at which that rpm intersects with the load value that’s being reported by the manifold absolute pressure (MAP) sensor. (Some systems may use computed engine load based on mass air flow sensor input rather than MAP values.)

Unlike a multiplication table, the reference tables within an ECM do not contain an intersection point for every possible combination of row and column data. Consequently, an ECM has to incorporate a means of being able to read between the lines. That capability varies from one system to another, but regardless of the protocol, it’s likely that when monitoring ECM operation, you may see values that appear to be off the mark. As an example, if the rows within a table are arranged in increments of 400 rpm, the system has to determine appropriate values for engine speed inputs that are between the rows.

Another factor that often biases controller output is a separate “modifier” table. Some engine controllers (such as those used in late-model GM vehicles) contain tables that alter ignition timing based on coolant temperature, inlet air temperature, air/fuel ratio and EGR operation. These controllers may also contain high and low octane, base spark in gear and base spark in park/neutral tables.

That’s one of the reasons that accurate monitoring is necessary when tuning. If a system contains both high- and low-octane tables, it probably incorporates a sliding scale between the values in those tables to determine actual advance settings. The scale slides according to the amount of spark knock detected by the system. After the sliding scale arrives at a figure, that value may then be modified according to the data in the air and coolant temp tables. So it’s entirely possible that the actual timing commanded by the ECM will differ from the value in the primary spark table by 5 or 10 degrees.

Irrespective of the number of tables within a system, or the manner in which the values in those tables interact, the process of determining final values is always the same—sensor input data points to a specific cell within a table. Depending on the system, the value in that cell may or may not be modified by the values in related tables. That’s relevant to tuning, because reprogramming an ECM involves nothing more than changing the values within tables. (As opposed to changing the way an ECM processes information, “programming” changes the result of those processes. Going back to the multiplication table analogy, reprogramming an ECM is the equivalent to changing the data for 8 times 7 from 56 to 54, or 58.)

Throughout the years, ECMs have become considerably more sophisticated, but they still rely on much of the same sensors and device controllers. Basic system components include:

 

Sensor Inputs                                   Device Controllers

 

RPM IAC or Throttle Position
MAP Fuel Injector Duty Cycle
Throttle Position Ignition Timing
Coolant Temperature Fuel Pump
Inlet Air Temperature
O2
Camshaft Position
Knock

 

Some systems may include a mass air flow sensor instead of, or in addition to, a MAP sensor. If a system incorporates sequential, rather than batch fire of the injectors, or if it operates without a distributor, a crankshaft position sensor is also a necessity (so the system knows when cylinder number 1 is at TDC).

Many aftermarket EFI systems are designed to control high performance and race engines, and consequently may appear to be relatively unsophisticated. In fact, they simply eliminate many of the tables and protocols that address emissions, fuel economy and engine protection issues. Consider the need, or lack thereof, for multiple spark tables. In stock vehicles, high and low octane tables are included to accommodate the driver who will purchase a vehicle with a high compression engine and fill the tank with regular fuel. Rather than relying on the knock sensor to retard timing after knock occurs, a system can bias spark settings far enough towards low octane values to eliminate knock before it occurs. Hopefully, the owner of a race engine is sufficiently aware of octane requirements that he/she selects an appropriate fuel, and spark knock is never a concern.

Another consideration regarding race engines is that they may be equipped with camshafts that create highly erratic MAP sensor or MAF sensor readings. In such instances, effective fuel and spark control may be achieved only by referencing rpm and throttle position data.

Whether you’re working with a race car or tow vehicle, the only difference is the number and types of sensors and device controllers. The tuning process is the same; find the values required to make your engine behave and plug them into the ECM.

 

Tags:

Advertisement

Leave a Comment

You must be logged in to post a comment.