Electronic Engine ECU and Sensor Operation

ECU basics

The Engine Control Unit is used to control the operation of internal

combustion engines. Typically this involves the control of fuel quantity and

spark timing as well as other ancillary controls. The ECU is a microprocessor

based electronic circuit that is capable of executing its code at very high

speeds and thus able to monitor and control the engine to crank angle

resolution.

The ECU operates off look-up tables to determine the appropriate value of

fuel quantity and spark timing. The look-up tables would usually be

determined through experiment on the same engine.

ECU, Sensing

The ECU requires knowledge on the engine status in regards to its

Crank angle

Engine rpm

Engine load (determined through Manifold Absolute Pressure or Throttle Position Sensor),

Coolant temperature

Air temperature

Exhaust Oxygen (Lambda) sensor

The sensors used are not unique as there are many manufacturers of sensors.

Crank and Cam Sensors

The function of the crank and cam sensors is to provide knowledge of angular

position and speed of the engine to the ECU. The ECU requires knowledge

of angular position of the engine crank so that spark and fuel are generated at

the desired crank angle.

Usually these sensors are inductive type, two wire (or three wire) and operate

on the principle that a voltage is generated in a coil when iron (a tooth) goes

past the sensor at some speed. Other types of position sensing is sometimes

used such as optical triggering or hall effect (hall effect requires use of

magnets).

Manifold Absolute Pressure (MAP)

The MAP sensor is used to provide intake manifold pressure measurement

which can be used as an engine load indicator. Sometimes this is also

referred to as Manifold Air Pressure, however the use of the word Absolute is

more descriptive as it has to be appreciated that the pressure being measured

is not gauge but absolute. Note that gauge pressure refers to pressure

quantity above atmospheric pressure. Ambient pressure is 100kPa (14.7 psi)

in an absolute scale and not zero. MAP sensors are typically three wire

(ground, signal and supply) and vary in their pressure measuring range

depending on application. Naturally aspirated engines typically utilise 100kPa

sensors while turbocharged (or supercharged) engines utilize 200kPa or

300kPa sensors.

Throttle Position Sensor (TPS)

Usually a potentiometer directly connected to throttle body’s butterfly shaft.

The overall electrical resistance of the potentiometer can vary from one

sensor to another. However the overall resistance has practically no effect on

the throttle position measurement. The ECU reads the voltage at the wiper

which is a function of the orientation (angular position) of the shaft.

Coolant and Air temperature

The coolant and air temperature sensors are usually thermistors. Thermistors

are resistors whose resistance changes with temperature.

conjunction with a pull-up resistor, the thermistors and pull-up resistor make a

potential divider whose voltage output depends on temperature. The voltage

is read by the ECU to provide temperature measurement. The thermistor has

two electrical terminals and therefore two connections to the harness,

however sometimes the coolant temperature sensor has one side of the

thermistor grounded to the engine and hence the sensor will have only one

electrical terminal.

Oxygen (Lambda) sensor

This sensor has seen a lot of evolution over the years. The fundamental

principle is based on the production of a voltage by zirconium dioxide element

when exposed to fresh air and exhaust gas. The most basic sensor is the

one-wire sensor. The single wire provides a voltage that changes in relation

to exhaust oxygen. The output signal of the single wire sensor referenced to

chassis ground. The two-wire sensor provides two electrical connections one

for ground and the other for signal. Therefore the two-wire has better signal

quality compared to the one-wire (note that the single wire’s ground

connection to the chassis is through the possibly rusted exhaust system ).

Oxygen sensors require an operational temperature above 300°C to function

properly. The three-wire senor has an embedded heater that heats up the

sensor quickly on start-up thus enabling a much faster knowledge of exhaust

oxygen. In a three-wire sensor, usually two wires are for the heater (typically

two white wires) and the third is signal (referenced to chassis ground). A fourwire

sensor has two wires for heater (typically two white wires) and the other

two wires are signal and signal ground. One, two, three and four wire sensors

provide a voltage ranging from zero to 1Volt. A voltage of approximately 0.45

volts indicates stoichiometric condition, voltages lower than 0.45 imply lean

combustion while voltages higher than 0.45 imply rich combustion. The

measured voltage cannot provide knowledge on the Air to Fuel Ratio AFR but

only knowledge whether rich or lean. Five-wire sensors do provide a voltage

that provides knowledge on the AFR. Five-wire sensors are also referred to

as wide- band sensors. Wide band sensors have signal conditioning circuitry

and provide a linearized voltage output with AFR.

ECU, Electronic Control

The ECU controls the engine through fuel injection and spark timing. For

spark ignition engines, the quantity of fuel required is in direct proportion to

the quantity of air inhaled by the engine. The mass of Air to mass of Fuel

ratio (AFR) for ideal operation is stoichiometric. When a three way catalytic

converter is used in production vehicles, the AFR is cycled (through closed

loop control) between rich and lean in order for the catalyst to be able to

perform both oxidizing and reduction reactions. In racing applications the

AFR is typically maintained rich (that is AFR smaller than AFR stoichiometric)

because this produces more power and is safer for the engine.

Fuel Injection

Spark ignition engines operate at AFR close to stoichiometric. The quantity of

fuel required to obtain the required AFR is controlled by the amount of time

the injector is left open, and is referred to here as Duration Of Injection (DOI).

The DOI required at any condition depends mostly on Volumetric Efficiency

which in turn is very dependent on engine rpm. The DOI required is also

dependent on engine load which is determined through the MAP or TPS

sensors. It is noted here that the logical consumption of much more fuel at

higher rpm is due to the fact that the DOI applicable is injected every

revolution (or every other revolution). Fuel injectors are very quick-acting onoff

valves capable of being cycled (that is opened and closed) in the order of a

millisecond. Injectors are available in a variety of flow rates and are also

divided into low impedance and high impedance injectors depending on their

electrical resistance. Peak-and–hold drivers can drive both low impedance

and high impedance injectors while saturation drivers can drive high

impedance injectors only.

Spark Generation

The timing of the spark is critical for optimal engine operation. Typically spark

timing has to be advanced with increasing engine rpm. This is due to the fact

that spark has to be generated in an earlier crank angle if the flame front is to

travel across the combustion chamber at higher rpm while still fully

combusting all gases just several degrees after top dead centre. The optimal

spark timing is also dependent on engine load. Lighter engine loads require

more advanced spark due to a slower moving flame in lower density

combustion gases. In older mechanical systems this spark advance at low

engine loads was achieved by the vacuum advance system. Various types of

spark generation and delivery are available, namely, one coil with distributor,

a coil every two cylinders (wasted spark) and an individual coil for each

cylinder. The spark, as with the older contact breaker setup (make and break)

is generated by the switching-off of current to the coil. This is so because the

coil (inductor) cannot allow the magnetic flux to vanish immediately and

therefore a high voltage is produced which is capable of producing an

electrical discharge across the spark plug gap. The Capacitive Discharge

Ignition (CDI) delivers a quantity of electricity to the coil at a very high voltage

on the primary side of the coil (can be 300V). This high voltage in CDI

systems charges the coil a lot faster and leaves enough time to recharge and

spark the plugs more than once per engine cycle (multi spark).

Using the ECU

The ECU is an electronic circuit using state of the art microprocessor,

memory, signal conditioning and power transistors. The wiring diagram

should be well followed before connecting power to the system. Damage to

the ECU can be done if wiring is not correct or not following the wiring

suggestions. This applies most of all to making sure that ECU pins that are

supposed to be connected to power are correctly connected to the relevant

power, while pins that are not supposed to be supplied with power aren’t

connected to power. It is also worthwhile mentioning that high voltage spikes

(around 350V) are generated by the spark plug coils even on the low voltage

side (that is ECU side). These high voltage spikes are properly handled by

the coil drivers but should not be connected to any other ECU pins other than

the coil drivers.

Before using the ECU, the wiring strategy must be developed.

Usual Wiring Information and Commonalities

ECU’s are powered from battery voltage, nominally 12V. The battery voltage

is not actually 12V all the time as during cranking voltage will surely drop,

while during charging voltage would be around 13.8V. The spark plug coils,

injectors, oxygen sensor heater, relays, dashboard indicator lights and other

ancillaries will typically run off 12V supply. The ECU internal electronics will

typically run at lower voltage.

This voltage was 5V until recently and now is 3.3V. Sensors will also typically be powered by a lower voltage, typically 5V, however some sensors do get powered by the battery 12V.

Sensor signals are typically between 0 and 5V, one exception is the two wire inductive pickup (used for crank and cam sensors) whose output voltage increases from less

than a volt at low rpm but can reach as high as 20V depending on application.

Due to the fact that ECU electronics and power electronics have a common

ground but a different high side voltage as described above, switching of the

power circuits by the ECU electronics is achieved by closing or opening the

connection of the power circuits to ground. That is, coils and injectors would

have a continuous 12V supply (battery voltage), the ECU would then turn on

the coils and injections by supplying a ground connection to them. Turning-off

of the power is achieved by breaking the connection to ground. Such a

strategy was also used in the past on mechanical contact breakers systems.

At this stage it is appropriate to note that due to the fact that all current from

coils, injectors and other power circuits flows into the ECU through the low

voltage side (ECU side) of these power consumers, the ground current

flowing out of the ECU is very high when compared to the much smaller

current flowing into the ECU from the battery positive supply to power the

ECU electronics. This fact needs to be appreciated to recognize why there

are typically many more ground connections compared to the 12V positive

supply connections. It is advised that all these ground connections are

connected so that there is ample current handling capability.

Another word on grounds, different types of grounds are cited, namely battery

ground and analogue ground. Battery ground is the ground that is directly

connected to battery, its main feature is its huge current carrying capacity, the

current flowing from coils and injectors would be routed to this ground inside

the ECU. The analogue ground is the ground that is used by analogue

sensors, analogue meaning voltage that can vary continuously between

ground and supply voltage. Examples of analogue sensors are TPS, MAP

and temperature sensors. The voltage output of these sensors varies in direct

proportion to the measured parameter. Therefore the ground voltage level of

these sensors has to be very stable otherwise a slight shift in the voltage level

of the ground would be erroneously translated into a change in the measured

parameter value. It should be noted that battery ground would have discrete

shifts in ground voltage level due to the turning on and off of coils and

injectors and turning on and off of other digital electronics. A filter to cancel

these shifts in ground level is typically employed to produce a clean analogue

ground. The supply voltage to the analogue sensors (typically 5V) would also

be a clean voltage, that is it would also be without any voltage shifts due to

switching. Appreciating the differences between these ground and supplies is

important so that connections are made to the appropriate terminals and not

just by whatever happens to seem the easiest physical connection on the

vehicle.

Heat dissipation: Electronic circuits do need to get cooled and cannot operate

at high temperatures. The ECU heats up in part due to the microcontroller

and associated electronics but mostly due to the power transistors associated

with switching on and off of the coils, injectors and other auxiliaries. The

reason behind the heat generated by power transistors is due to the fact that

when switched on, the power transistors would have a voltage drop across

them say of 0.8V. Therefore if a coil draws 5Amps in saturation, it would

translate in 4W (P=IV, P=5*0.8=4) of heat generated in the transistor that has

to be dissipated into the surroundings. Therefore ECU’s typically have there

case that functions as a heat sink for the internal electronics. To make sure

the heat sinking is effective, the ECU should be mounted in a relatively cool

location and if possible have air current or mounted to heat sinking (and cold)

metal parts.