Fuel Injection
LAMBDA CONTROL
For proper operation of the catalytic converter, the fuel-air mixture must be stoichiometric.
This means that the mixture must be neither rich nor lean but consist of exactly 14.7 kg of air to 1 kg of fuel (Lambda = 1).
Regardless of how the air mass flowing into the engine is measured, and regardless of how accurately injection time is calculated, the fuel-air mixture will inevitably deviate from Lambda = 1.
For this reason the system is equipped with an oxygen sensor for each cylinder bank, mounted just after each exhaust manifold.
The front sensor is connected to pin 47 and the rear sensor to pin 28 of the electronic control module. From pin 10 of the control module the oxygen sensors are grounded via the exhaust pipe and grounding point G7L on the engine.
The engine exhaust gases pass the oxygen sensors. A chemical reaction enables the oxygen content of the exhaust gases to be measured. The output voltage of the oxygen sensors is proportional to the current oxygen content.
IMPORTANT: The oxygen sensor receives reference oxygen from the ambient air via the connecting cables. For this reason, contact cleaning sprays and grease must not be used on the sensor connectors.
If the engine runs rich (Lambda less than 1), the output voltage of the sensor will be about 0.9 V. If the engine runs lean (Lambda greater than 1), the output voltage of the sensor will be about 0.1 V. Sensor voltage changes very rapidly when Lambda passes 1.
When sensor voltage is low the control module adjusts the injection time slightly so that a richer mixture is obtained. The oxygen content of the exhaust gases drops and sensor voltage rises rapidly to about 0.9 V as Lambda drops below 1 (rich) and the control module then shortens the injection time slightly. This sequence of events is repeated continuously.
To measure the oxygen content of the exhaust gases, the oxygen sensor has to be at a certain temperature. A preheater element is built into the sensor so that lambda control can be engaged as soon as possible after starting and to ensure that the temperature of the sensor at idling speed is sufficiently high.
The preheater element consists of a PTC resistor, the output power of which diminishes with rising temperature. The preheater is supplied with power via the fuel pump relay and is accordingly activated as soon as the engine starts.
Lambda control is engaged as soon as engine temperature rises above 32°C (90°F) under partial load or 38°C (100°F) at idling speed.
On the engine the two oxygen sensors each control their own bank of cylinders.
Lambda control is disengaged under high engine load conditions since a rich mixture is required for optimum performance.
In the event of a fault occurring in one of the oxygen sensors or its electrical circuit, Lambda control of both cylinder banks takes place through the good oxygen sensor. The MIL (CHECK ENGINE) lamp will light up.
Adaptation
The control module first calculates the injection time using information from the mass air flow sensor and engine rpm. On the basis of information from the oxygen sensors, the injection time is then corrected so that Lambda = 1 is obtained.
The Lambda control function can adjust the calculated injection time by ±25%.
If the control module has calculated that the injection time ought to be 8 ms, the lambda control function can adjust the times down to 6 ms or up to 10 ms if necessary in order to obtain Lambda = 1.
Multiplicative Adaptation
For instance, if the control module calculates that the injection time should be 8 ms but the lambda control function adjusts injection time to 9 ms because the car's fuel pressure is somewhat low, then the control module will "learn" the new injection time. This is done by correcting the basic calculation on the basis of engine rpm and information from the mass air flow sensor so that the result is an injection time of 9 ms.
The correction factor in this example is: 9/8 = 1.125 = +12.5%.
The correction factor, 12.5%, is stored in the control module's memory so that it can be used for calculating the injection time regardless of engine load and rpm. The correction factor may be +20% to -24%.
That is called adaptation.
Conditional for the implementation of adaptation is activated lambda control, a coolant temperature above 70°C (158°F) and non-operation of the EVAP canister purge valve.
The adaptation described takes a fairly long time because of slow age-related changes in fuel pressure, the flow of fuel through the injectors, etc.
Using an ISAT Scan Tool, a multiplicative adaptation readout can be obtained for each bank of cylinders. Differences between the cylinder banks, positive or negative values, are perfectly normal as long as the value does not exceed +20% or fall below -24%.
Additive Adaptation
Further adaptation takes place at idling speed. This differs from multiplicative adaptation in that it takes place very rapidly (although the same conditions must be fulfilled as for multiplicative adaptation). Another difference is that the control module does not store the value as a percentage but as injection time in ms.
For instance, if the calculated injection time is 3.05 ms and the Lambda control function corrects the time to 3.10 ms the control module will adapt the value +0.05 ms. This value is then added on top of the injection time, regardless of engine load and rpm conditions.
The adapted value may be 0.38 ms.
The purpose of additive adaptation is to compensate for air leakage, the effect of which is most noticeable at idling speed.
Using an ISAT Scan Tool, an additive adaptation readout can be obtained for each bank of cylinders.
Differences between the cylinder banks, positive or negative values, are perfectly normal as long as the value does not exceed +0.38 ms or fall below -0.38 ms.