ENGINE MANAGEMENT – SPARK IGNITION
2.3 COMPUTER CONTROLLED IGNITION SYSTEMS
Figure 2.26 Engine speed related timing advance curve and mechanical advance mechanism
a Ignition advance curves
b Mechanical advance mechanism using unequal length springs
then, as the bob weights are flung out under centrifugal force, the stronger spring then also acts against the bob weights.
The difference between the two advance lines shown in Figure 2.26a illustrates the inability of a mechanical advance system to provide the correct timing at all engine speeds. It was therefore necessary to try to achieve the best compromise for timing advance, which meant that the timing was not correct for much of the engine speed range.
As well as not being able to provide the correct timing at all times when the system was new, when the timing mechanisms began to wear matters became worse and timing was often much too far away from the desired value to enable the engine to operate efficiently.
Mechanical and mechanical/vacuum systems could not respond quickly enough to the required changes in timing and the accuracy deteriorated over time due to wear. It was therefore necessary to eliminate mechanical systems and to use electronic or computer control for the ignition timing functions.
Load related vacuum retard
Figure 2.27 shows a simple vacuum operated timing retard system. The illustration shows the mechanism acting on a base plate onto which the contact breaker assembly were mounted.
When the throttle was closed at idle speed or during deceleration (upper left diagram), the vacuum was blocked from reaching the diaphragm and therefore any timing changes were dependent purely on the mechanical bob weight system (engine speed related).
When the throttle was then partially opened (light load conditions), high manifold depression acts on the diaphragm, which in turn pulls the diaphragm
against the spring. The movement of the diaphragm pulls the linkage, which rotates the base plate in an anticlockwise direction, against the direction of rotation of the distributor shaft. This would have the effect of advancing the trigger signal and timing.
When however the throttle was then opened further (high load conditions), the intake depression would initially reduce (higher pressure), the spring would then force back the diaphragm and in turn, this would allow the base plate to move back in a clockwise direction (the same direction as the distributor shaft rotation). The opening of the contact breaker and timing would therefore retard.
Whilst the vacuum advance/retard system was reasonably effective, it could not accurately control timing for all the variations of load that occur and again a compromise was inevitable. Adding the compromise of a vacuum system to the compromise of a mechanical advance system, resulted in an inaccurate timing control system.
For electronic systems, near the trigger mechanism located in the distributor (e.g. inductive, Hall effect or optical). Refer to Figures 2.11, 2.12 and 2.19. The movement of the base plate and distributor shaft altered the timing in the same way.
Changing operating conditions
Apart from the lack of accuracy of mechanical and vacuum based timing systems, there are other reasons why greater flexibility of timing control is needed.
Air/fuel mixtures need to be altered more rapidly, so the ignition timing must be altered accordingly; in addition, ignition timing requirements differ with temperature. In effect, anything that affects the speed of combustion, i.e. the flame speed through the combustion chamber
Figure 2.27 Vacuum advance/retard mechanism
and the burn time, will require a different timing advance value.
Although many adaptations to basic mechanical and vacuum systems were introduced, the accuracy and speed of change of timing remained less than satisfactory for modern requirements.
With computer control, and the use of a number of sensors, it became possible to obtain much more accurate control of ignition timing. Rapidly changing operating conditions could therefore be sensed (including rapid changes in engine speed), and the computer or ECU could then alter the timing as necessary. Systems were therefore introduced where the switching action of the ignition module was controlled by the ECU. A trigger signal is still produced by an inductive or Hall effect trigger (or other type of pulse generator), but the signal is passed to the ECU, which modifies or ‘phases’ the signal to achieve timing control.
Figure 2.28 shows the basic layout of such a system.
Note that the ignition module can be separate from the ECU or, in many cases, the module is integrated within the ECU.
The simple example in Figure 2.28 shows an ECU receiving a trigger signal provided by an inductive sensor. A vacuum pipe connects the intake manifold vacuum (depression) to a pressure sensitive component in the ECU thus providing the ECU with engine speed and load information. See section 1.5.4 for information on electronic type pressure sensors.
Computer controlled ignition systems 63
2.3.2 Digital timing control
If we assume that an ECU receives a trigger signal from some form of engine speed sensor, and load information via the vacuum sensor, the ECU can then calculate the required ignition timing. Within the ECU is a ‘look up table’ or memory, which contains all the relevant timing data applicable to the different load conditions; by comparing the conditions with the data in the look up table, the ECU can determine the timing advance required.
The look up table effectively contains a three- dimensional map of the timing requirements, a simple example of which is shown in Figure 2.29. The spark timing map provides the timing angle (crankshaft angle) related to engine speed (revolutions/second) and engine load (on a scale of 0 to 1). The example in Figure 2.29 shows a relatively small number of speed, load and timing reference points, but many more reference points can be included on a spark map.
Figure 2.30 shows the same spark timing map with the engine speed at 32 rev/s (1920 rev/min) and the engine load at 0.5 (half load). The 32 point mark on the map is therefore followed across until it intersects with the 0.5 load line; the intersect point identifies the required spark advance, which in this case is 52°.
The timing advance characteristics are established during many tests on development engines (before they go into production). Once the timing
Figure 2.28 Layout of a computer (ECU) controlled ignition system
characteristics have been established, production ECUs will then have a memory and look up tables containing the applicable data.
Because accurate timing is so important to engine efficiency, performance, emissions and economy, improvements in all of these factors are achieved with a greater number of timing reference points.
Figure 2.31 shows a more modern and complex spark advance map (compared with the example in Figure 2.29).
It can now be appreciated that considerable accuracy of spark timing can be achieved with the complex spark timing map. However, because the trigger signal from the inductive or Hall effect triggers (and other pulse generators) discussed so far provides only one reference point per cylinder, this is a limiting
factor in the accuracy of ignition/spark timing.
Additionally, the trigger mechanisms so far discussed are usually located within the distributor body, so the rotation of the trigger mechanisms is driven by the distributor shaft which in turn is driven by a timing belt or chain. Any wear or maladjustment of the belt or chain will cause inaccuracies in the timing reference.
A more reliable and accurate means of providing a trigger/reference signal would enable the spark timing also to be more accurate. The following section (2.3.3) covers some examples of more accurate timing trigger/reference signal systems which usually have a sensor and reluctor or trigger disc located on the crankshaft.
Figure 2.29 Simplified three-dimensional spark advance map
Figure 2.30 Simplified three-dimensional spark advance map showing the spark timing at 32 rev/s and half engine load
2.3.3 Crankshaft speed/position sensors
Direct triggering reference from the crankshaft Most crankshaft speed/position sensors are located adjacent to the crankshaft and are usually inductive sensors (Figures 2.32 and 2.33). In some cases Hall effect sensors have been used. Locating a sensor adjacent to a crankshaft allows a reluctor disc (trigger disc with the reference points) to be mounted directly on the crankshaft, with most versions being mounted either at the front pulley or at the rear of the crankshaft adjacent to the flywheel. Sometimes a reluctor disc is mounted at a convenient point on the crankshaft between the crankshaft webs, i.e. the disc is located in the crankcase.
The obvious advantage of locating the reluctor/trigger disc on the crankshaft is that there is no drive linkage (belt, chain or other mechanism); this means that the trigger or reference signal will be
accurately identifying the crankshaft speed or angular position without any losses of accuracy that could be caused by drive mechanisms.
Increased number of reference points
With a crankshaft mounted reluctor or trigger disc, it is possible to use a larger diameter disc, which can more easily contain a larger number of reference points (reluctor teeth). Due to the fact that most of the crankshaft speed position sensors are of the inductive type, the reference points usually take the form of teeth located around the disc.
In most cases the trigger disc is located directly at the front or the rear of the crankshaft and can form part Computer controlled ignition systems 65
Figure 2.31 Typical modern three-dimensional spark advance map
Figure 2.32 Crankshaft speed/position sensor
Figure 2.33 Crankshaft speed/position sensor located adjacent to flywheel
of the front pulley or part of the flywheel assembly.
Because of these locations, especially if the disc has a similar diameter to the flywheel, it is relatively easy to locate a number of reference points around the disc.
Some earlier examples had only a small number of reference points, e.g. two or four, but most modern systems have as many as 60 reference points. One of the reference points or teeth is usually either missing or is a different shape to the rest of the teeth; this enables a master reference signal to be produced.
Figure 2.33 shows a flywheel located inductive crankshaft speed/position sensor. Note the missing tooth, which is adjacent to the sensor in the illustration.
In fact, the example shown in Figure 2.33 has two missing teeth; one missing tooth is a master reference for cylinders 1 and 4, while the other missing tooth is a master reference for cylinders 2 and 3. On this particular system, with a single ignition coil, the high voltage output from the coil is directed via a rotor arm located at the end of the camshaft (OHC type).
Therefore the ECU does not need to receive a signal relating to each cylinder, but it does receive signals relating to the TDC position of each pair of cylinders (or other predefined angular position of the crankshaft).
On this system therefore, the sensor would provide the speed signal and the angular position of the crankshaft as each tooth passes the sensor; the ECU can count the number of signals as each tooth passes the sensor thus enabling the ECU to identify the angle of rotation from TDC (or from the master reference position). It is also possible on this type of system for the ECU to assess any changes in crankshaft speed as each tooth passes the sensor.
Using Figure 2.28 as an example, the ECU receives a load signal from the pressure sensor (pressure transducer), which is connected by a vacuum pipe to
TDC No. 1 cylinder Optical pick-up
Rotor plate
Figure 2.34 Optical speed/position sensor located in distributor body
Figure 2.35 Analogue signal produced by an inductive crankshaft speed/position sensor
the intake manifold. This system also uses information from a knock sensor (discussed in section 2.3.4). Also note that a coolant temperature sensor provides information to the ECU to enable changes in timing to occur with changes in temperature (temperature sensors are covered in section 1.5.1).
Note: Some speed/position sensor systems located within a distributor also have many reference points, one type being an optical system with 360 slots located around a disc (Figure 2.34). However, locating a large number of reference points around a relatively small disc requires good manufacturing accuracy and the problem still remains that any wear or maladjustment of the drive linkage to the distributor will result in incorrect timing references.
Crankshaft speed/position sensor: operation and signals
Section 2.2.5 provides an explanation of an inductive pulse generator, and the inductive crankshaft speed/position sensor operates in exactly the same way.
The crankshaft sensors, however, are usually constructed so that the winding or coil is formed around the magnet and this assembly is located in the sensor body, which is then bolted or secured in some way to the engine block or flywheel housing. The sensor is located so that it will be affected by the movement of the reluctor teeth (reference points) whilst the crankshaft is rotating.
Figure 2.35 shows a typical analogue signal produced by an inductive crankshaft speed/position sensor. Note the different shaped pulse produced by the missing tooth.
The analogue signal is passed to the ECU which then converts it to a digital signal, thus enabling the required speed and angular position information to be obtained.
Other types of sensor signal
Almost all crankshaft speed/position sensors are of the inductive type, but where a Hall effect or optical system is used, these types will provide a digital signal in the same way as the older ignition trigger systems discussed previously in sections 2.2.6 and 2.2.7. The only differences compared with the older ignition trigger systems will be the number of signal pulses produced which will depend on the number of reference points.
2.3.4 Knock sensors
A knock sensor (Figure 2.36) is effectively a vibration sensor that responds to those vibrations in the engine that cause pressure waves to occur in the cylinder block or cylinder head. By detecting the vibrations or pressure waves, a knock sensor can detect the vibrations caused by combustion knock.
The knock sensor is an electronic pressure sensor, which with a pressure sensitive crystal that produces a small electrical pulse when it is exposed to pressure waves (such as the engine vibrations). Vibrations caused by combustion knock will result in a slightly different signal (frequency and voltage) being produced by the sensor. When the ECU receives the signal from the sensor, it is able to filter out the normal vibrations and respond to the particular part of the signal that is caused by combustion knock.
Although ECU controlled ignition timing should provide ideal timing for all operating conditions, it is possible that fuel quality could be poor (momentarily or continuously). Other factors such as the temperature in the combustion chamber can also cause short term combustion knock. In most cases, slightly retarding the ignition timing/spark advance will reduce and eliminate combustion knock.
Therefore, when the ECU detects a combustion knock signal, it will respond by retarding the spark timing a predetermined number of degrees. If the combustion knock is no longer detected, the ECU will progressively advance the timing to its correct value (so long as combustion knock does not reoccur).
An ECU can alter the timing for just the affected cylinder. When the knock occurs (when combustion occurs in the affected cylinder), the ECU will then provide the correct timing for the remaining cylinders (for example, the remaining three cylinders on a four- cylinder engine). When the affected cylinder is then due to receive its next spark, the ECU can retard the timing for just the affected cylinder.
The main advantage of computer controlled ignition is accurate timing – that stays accurate over the life of the engine
The ideal timing setting is held in the ECU memory in the form of a look up table
Key Points
Computer controlled ignition systems 67
Figure 2.36 Knock sensor located in cylinder block
2.4.1 Limitations of distributor based systems
Restricted dwell time and wasted energy
All of the ignition systems covered so far have two major disadvantages when it comes to providing high energy from the ignition coil to the spark plugs; both disadvantages arise because the ignition systems use a single coil to provide a spark for all of the cylinders.
Time to build up coil energy
The first disadvantage when using a single coil for all cylinders is that it limits the time available to build up coil energy between each of the individual ignition cycles: there is very little ‘dwell’ time available for the current to flow through the primary winding of the coil and build up a strong magnetic field (magnetic flux).
As previously explained, on a multi-cylinder engine there is very little time between one cylinder firing and the next; the faster the engine speed and the greater the number of cylinders, the less time there is available for the ignition coil to build up sufficient energy for the next ignition cycle. On modern high speed engines which operate with relatively weak mixtures, it is essential that the energy available from the coil is sufficient to produce a powerful and long duration spark otherwise emissions and general performance will not be acceptable.
Distributor and rotor arm wasting energy When a single coil is used to produce a spark for a number of cylinders, the energy from the coil (high voltage) is passed via a high tension cable (HT lead) to the distributor cap and rotor arm assembly (see section 2.28 in Hillier’s Fundamentals of Motor Vehicle Technology Book 1). Note that the lead passing the energy from the coil to the distributor cap is often referred to as the ‘king lead’. The distributor cap contains a number of contact points (referred to as electrodes), which in turn are connected to each of the spark plugs via additional HT leads. When the high voltage from the coil passes along the king lead to the centre electrode in the distributor cap, it is then passed to the centre of the rotor arm; because the rotor arm rotates with the distributor shaft, it is then able to pass the energy to the individual HT leads and spark plugs.
Figure 2.37a shows a basic layout of an ignition system with a single coil and Figure 2.37b shows a plan view of the rotor arm and distributor cap.
One problem with the rotor arm system is that voltage is lost or wasted when the current flow flows through all of the HT leads, and especially when the current flows across the rotor arm tip to each of the electrodes. There is a necessary gap between the rotor arm tip and the electrode, and this absorbs or uses some of the energy produced by the coil.
However, although Figure 2.37b shows the rotor arm in alignment with the electrode, in reality, the rotor arm passes through quite a large angle during the period of time that the spark exists (remember that the spark may last 2 ms or more). When an engine is operating at 6000 rev/min, the rotor arm (which rotates at half engine speed) will rotate 50 times in one second or one rotation in 0.02 s (2 hundredths of a second). During the spark duration of 2 ms (2 thousandths of a second), the rotor arm will rotate through one-tenth of a complete rotation i.e. 36° of rotation. There is therefore quite a substantial gap between the rotor arm tip and the electrodes when the energy from the coil is passing to the spark plug. This gap inevitably uses considerable amounts of valuable energy, which reduces the energy available to maintain the spark.
Additionally, the rotor arm tip and electrodes will progressively deteriorate due to the arcing that occurs as the voltage or energy flows across the gaps.
It is also important to note that, when electricity has to jump the gap at the rotor arm tip, this creates electrical interference, which must be suppressed to prevent interference with other electrical and electronic devices.
Using multiple ignition coils (eliminating the rotor arm)
The next progression in ignition system design was therefore based on a desire to eliminate the distributor cap and rotor arm assembly and use more than one ignition coil. Note that there were some engines produced (typically V8 and V12 engines) that did use two coils, each of which provided sparks for half of the cylinders. However, these systems still used one rotor arm for each coil and group of cylinders: in effect there were two ignition systems.
The ultimate ignition system would have one coil for each cylinder, and this is the general rule for modern engines, where an individual ignition coil is either directly connected to the top of the spark plug or there is an HT lead from each coil to the spark plug.
There is however another solution, which is still used on some systems, and this design uses a single coil to provide a spark at two spark plugs. Although these systems are often referred to as ‘distributorless’ ignition systems, the same terminology can be applied to systems that use one coil for each cylinder. For the purposes of differentiation, coils that provide sparks to two cylinders at the same time are referred to within this book as ‘wasted spark’ systems, the reason for which will be made clear in section 2.4.2. Section 2.4.3 covers systems that use a ‘single coil per cylinder’.
For both the wasted spark and the single coil per cylinder systems, the ignition systems do not require a distributor and rotor arm assembly to distribute the spark to the different cylinders: they are both therefore