Topics:
- Determining and installing actuators for the engine management system
- Fuel injectors
- Selecting suitable injectors
- Mounting the injectors in the intake manifold
- Ignition
- Preparing with the conventional ignition
- Ignition coil for the engine management system
- Current build-up in the primary winding
- Ignition advance
- Throttle body
- Test setup of the stepper motor with simulator
- Settings for the stepper motor
- Fuel pump circuit
- Completion of the mechanical work
Determining and installing actuators for the engine management system:
The actuators that will be controlled by the MegaSquirt are the injectors, ignition coil, fuel pump and the stepper motor for idle speed control. This chapter describes the process in which the actuators were tested and mounted on the engine block, and which choices were made for this.
Fuel injectors:
The MegaSquirt controls the injectors. The injectors are low-side switched. With a low-side switched component, supply voltage is present, but current only flows when the ground is switched on. In this case, the injector will only start injecting when the MegaSquirt ECU switches the ground. As soon as the energizing is stopped, the injector stops injecting. The quantity of fuel to be injected is determined using the VE table and AFR table.
A MOS-FET switches the injector on and off, thereby injecting the fuel. The amount of fuel determined by the MegaSquirt depends on several factors:
- The ideal gas law, which relates the amount of air to its pressure, volume and temperature;
- Measured values from the sensors in the engine block: pressure in the intake manifold (MAP sensor), coolant and intake air temperature, crankshaft speed and the data from the throttle position sensor;
• Adjustment parameters: required fuel quantity, volumetric efficiency (VE), injector opening time and enrichment under certain conditions.
The injection time must be as long as possible during engine idling in order to obtain good fuel metering. Therefore, not just any random injector can be used on the engine. The properties of different types of injectors must be compared and calculations must provide insight into the required fuel quantity for the specific engine. There was also the option to choose between high- and low-impedance injectors. Low-impedance injectors are suitable for engines where a very fast opening of the injector needle is required. The typical resistance is 4 Ohms. The disadvantage of these injectors is the high current. The heat development that this causes in the MegaSquirt is undesirable. It is possible to use low-impedance injectors by mounting special IGBTs on a heat-conducting plate on the MegaSquirt housing. The decision was made to use high-impedance injectors. In this case there is less heat development and these IGBTs are not used.
The flow rate is very important to determine the correct injected quantity and thus the control. If injectors that are too large are chosen, the injection time at idle speed will be so short that the engine may begin to run irregularly. The injection quantity must be sufficient to inject all the fuel within the available time. The injection quantity is expressed as injection time in milliseconds. A high load at high engine speed is assumed. This corresponds to a MAP of 100 kPa. Based on the engine characteristics, the required injector flow can be calculated. The injector flow indicates how many milliliters of fuel are injected per minute.
Selecting the suitable injectors:
For the project, three different types of injectors were made available. Research revealed which type of injector was most suitable to use in this project.
Each injector type has a different flow; the output after one minute of injection differs per type. Before the injectors were tested, they underwent a cleaning cycle in an ultrasonic bath. With this cleaning method, the injector is cleaned inside and out using ultrasonic vibrations and a special test fluid, so that any old dirt residues cannot influence the flow measurement or the spray pattern. During ultrasonic cleaning the injectors were continuously opened and closed and the spray pattern of each injector was observed; it was a fine mist. When closing, no deviations such as droplet formation or an abnormal jet were visible. After ultrasonic cleaning and testing, the O-rings were replaced to guarantee a good seal when mounted in the intake manifold.
Using a test setup (see image above), the injectors can spray into multiple measuring beakers so that the amount of fuel injected after a certain time can be read. By controlling the injectors at an operating pressure of 3 bar, the quantity of injected fuel can be checked. The fuel pressure on the supply line (the rail) must be 3 bar and the injector needle must be actuated for 30 or 60 seconds with a duty cycle of 100%. After the injectors had been actuated for 30 seconds, the following data could be recorded:
Type 1: 120 ml
Type 2: 200 ml
Type 3: 250 ml
Only one type of injector will be used. The injector size is determined using the formula below:
The injector size is determined on the basis of the effective power (Pe) delivered at a given speed, the Brake Specific Fuel Consumption (BSFC), the number of injectors (n injectors) and the maximum duty cycle with which the injectors are controlled. The result is multiplied by 10.5 to convert from pound per hour (lb/hr) to ml/min.
The result of the calculation shows which injector is suitable for this engine configuration. It is not a problem if the result deviates by less than 20 ml from the calculated value. By adjusting the software in the MegaSquirt this difference is compensated again. The following table gives an overview of the data used in the formulas:
The first step is to determine the injected fuel at the torque speed. Every two revolutions of the crankshaft a certain amount of air is drawn in. At the torque speed the volumetric efficiency is highest. Because of the engine characteristics (among others valve overlap), the engine fills best at this speed and efficiency is highest. It is estimated that the volumetric efficiency will be around 70%. In formula 4, the air volume that is present in the engine at that moment is calculated.
In formula 5, the quantity of injected fuel is calculated on the basis of the air volume present. The engine power reached at the torque speed is calculated in formula 6. The ratio between the quantity of injected fuel and the power gives the BSFC in formulas 7 and 8.
The actual BSFC in formula 6 is multiplied by 3600 to calculate in kWh. The BSFC of a petrol engine is often between 250 and 345 g/kWh. The lower the value, the more efficient the engine is. Formula 8 gives the ratio between the fuel flow in pound/hour and the effective engine power. This percentage is carried over into formula 9.
From the result of formula 9 it became clear that the injectors with a flow of 200 ml/min are suitable to use on the engine. The difference of 7 ml is negligible because it is compensated for in the software when filling in the VE table.
Mounting the injectors in the intake manifold:
The electronically controlled injection system makes it possible to remove the carburettor that is part of the classic setup. The carburettor is therefore replaced by a throttle body (for air supply) and four separate fuel injectors. The intake manifold was retained and modified to enable the conversion to the engine management system. Fuel injection takes place in the intake manifold. It was decided to mount the injectors as close as possible to the intake valve. Designers of automotive engines mostly choose to mount the injector under an angle in the intake manifold. The fuel is then sprayed onto the intake valve. In the present project, however, a layout was chosen in which the injectors are placed at an angle of 45 degrees relative to the air channels in the manifold.
The intake manifold is made of cast aluminium. It was decided to attach aluminium sleeves to the manifold. Manually machining them to the correct size was not an option because the sleeves had to have different dimensions than the standard size of a drill. This meant that the manufacture of the sleeves had to be outsourced to a company with suitable equipment. The sleeves could then be TIG welded to the manifold. The choice to mount the injectors upright instead of at an angle was made for the following reasons:
- The installation process: it is easier to position the sleeves in a straight, horizontal arrangement. Welding the sleeves onto the manifold is simpler because it is easier to weld all around them than if the sleeve were at an angle.
- The finishing work: During welding the sleeves become slightly oval. The deformation is caused by the heat released during the welding process. This was taken into account by making the internal diameter of the sleeves smaller than the external diameter of the injectors. The finishing (reaming) is less risky: once the inside of the sleeves has been made round again, the diameter is optimal for the injectors and sealing by the O-rings is guaranteed. The height of the sleeves is important; the injector must not be inserted too far into the manifold. The end of the injector must not obstruct the airflow. Based on information from: (Banish, Engine Management, advanced tuning, 2007) it was decided to mount the injectors in the manifold so that the ends are located exactly in the ports of the manifold; the airflow is not hindered.
- Fuel injection: Because the mixing of the fuel mist with the air is optimal before the intake valve opens, it does not matter much whether the injector sprays exactly onto the intake valve or just in front of it in the intake manifold.
With simultaneous injection the injection occurs every crankshaft rotation (360°). All four injectors inject at the same time. This means that fuel is also sprayed into the intake port while the intake valve is not open. Some time later, the intake valve opens and the fuel still enters the cylinder.
The sleeves were specially machined to size on a lathe. The internal diameter is slightly smaller than the external diameter of the injector; because deformation occurs during the welding process, there must still be the possibility to remove material during finishing by reaming. This means the diameter becomes slightly larger because material is removed. The diameter must not become too large because then the rubber O-ring on the injector may no longer be able to provide a proper seal. A good seal is very important; air leakage along the injector results in a lower vacuum in the intake manifold.
The measured vacuum will then no longer correspond to the calculated vacuum. This affects the injection, which is determined using the VE table. The vacuum plays a major role in this. The characteristics and settings of the VE table are described in a following chapter.
A bevel was filed on the underside of the sleeves so that the shapes match those of the intake manifold. This ensures the sleeve stands as straight upright as possible. The image below shows the intake manifold with a sleeve during the assembly process. The sleeve is tack welded on one side so that the effect of welding on the material can be carefully inspected. It was unclear whether the aluminium of the manifold contained too many impurities, which would make welding more difficult. This turned out to be fine. To prevent the sleeves from shifting position during welding, holes were drilled in the manifold beforehand and the sleeves were held in the correct position with a specially made jig. In this way the four sleeves were welded all around. A final check revealed that the joints between the sleeves and the manifold were airtight.
The connection between the injectors is normally formed by a solid injector rail. This tube with fittings, often made of an aluminium alloy, is custom made by a designer. On the Land Rover engine used for the project, two injectors are located right next to each other, but the distance between the pairs of injectors is quite large. The dimensions of the fuel rail and the distance between the air channels of the intake manifold did not match. The rail therefore had to be modified.
Shortening some sections and lengthening other parts is very difficult to do by soldering; contamination from old fuel, which is very difficult to remove from the inside of the rail, can lead to poor adhesion. Because this involves fuel, the safest method was chosen: the sections on which the injectors are mounted are connected to each other using high-quality fuel hose. Flaring was applied to all ends and sturdy hose clamps were used to prevent the hoses from slipping over the flares.
The image below shows the intake manifold during machining. The supply line (indicated by number 1) is connected to the outlet of the fuel pump. Under a pressure of 3 bar, fuel is supplied to the inlet of the four injectors. The pressure regulator (3) adjusts the pressure depending on the intake manifold pressure, because the pressure difference between the fuel pressure and the vacuum in the intake manifold must remain 3 bar. Via the return line (2), the fuel flows back to the tank. There is a continuous circulation of fuel. Injection only takes place when the injectors are actuated by the MegaSquirt ECU.
- Supply line
- Return line
- Pressure regulator
- Pressure check
- Heat shield
- Throttle connection
- Vacuum connection
- Injector cylinder 1
- Injector bracket A
- Injector bracket B
- Intake port cylinder 1
In production passenger cars, the injector rail is attached to the intake manifold using clamps or lugs. The injector rail thereby clamps the injectors into the manifold. Because a flexible fuel hose was chosen as injector rail for this project, the above method is not possible. Therefore, it was decided to clamp the injectors into the intake manifold with a custom-made bracket. The brackets consist of two parts: the upper part (bracket A) and the lower part (bracket B).
Bracket A has two notches that can be slid over the injectors. This allows the injectors to be pressed into the manifold by means of the flat sides. Both brackets A have slotted holes so that the distance at which the injectors are located in the slots can be adjusted. Brackets A and B are screwed together: bracket B is attached to the same stud with which the manifold is mounted to the engine. A slotted hole makes it possible to adjust the bracket vertically. The further the bracket is moved downwards, the more firmly the injector is clamped.
Ignition:
The conventional ignition has been replaced by an electronically controlled ignition system with an ignition coil that is controlled by the MegaSquirt. To first have the engine run completely on the original technology, the conventional system with contact points must initially be connected. Only after a number of operating hours can it be established that the engine functions correctly, after which the installation and adjustment of the electronically controlled ignition, among other things, can begin.
Preparing with the conventional ignition:
The Land Rover engine was originally equipped with an ignition system with contact points, which is nowadays also called a conventional ignition system. The image shows this type of ignition system.
With closed contact points, the build-up of the primary current begins. The current is limited to 3 to 4 Amps by the resistance of the primary winding. When a current flows through the primary winding of the ignition coil, a magnetic field is built up. Both the primary (3) and the secondary winding (4) are located in this magnetic field. When the current through the contact points (10) is interrupted by the cam on the distributor shaft (9), a voltage is induced in both windings. About 250 volts is generated in the primary winding. The difference in number of turns causes an induction voltage of 10 to 15 kV to arise in the secondary winding. At the moment the points open, the spark plug spark is generated.
The induction voltage can be limited by allowing the primary current to continue flowing for a short period after the contact points have opened. This is achieved by a capacitor that is connected in parallel with the contact points. The capacitor is a time-determining element which, depending on its capacity, effectively regulates the height of the induction voltage. It also prevents the contact points from burning.
Ignition coil for the engine management system:
The engine management system will provide the control of the ignition coil. The classic ignition coil with distributor will remain present on the engine to serve as a test setup, but will no longer form part of the functioning of the combustion engine. A Distributorless Ignition System (DIS coil) was chosen, literally: “distributorless ignition system”. This type of ignition system does not use a distributor. Another option would have been to choose a Coil On Plug (COP) coil. In that case a separate coil is connected to each spark plug. A COP coil is also called a pencil coil. The disadvantage of a COP coil is that heat dissipation is not as good as with a DIS coil. Also, when using COP coils, a signal from a camshaft sensor is required, which is not present on the current engine.
The missing tooth on the crankshaft pulley serves as the reference point with which the ignition timing is determined. With the DIS coil, two spark plugs are actuated at the same time at an ignition event. The DIS coil is in fact a unit containing two coils. When the pistons of cylinders 1 and 4 move upwards, one is in the compression stroke and the other in the exhaust stroke. Nevertheless, both spark plugs will produce a spark. The spark that occurs in the cylinder that is in the compression stroke will ignite the mixture. The other spark, the so-called “wasted spark”, occurs while the exhaust gases are leaving the combustion chamber. The wasted spark is a spark that is generated when no mixture is ignited. The ignition energy is low; despite the spark there is little energy loss. It is also not harmful.
The figure shows the operating diagram of a four-cylinder petrol engine with a DIS coil. In this operating diagram there are two ignition markings per ignition event; one of them generates the spark to ignite the mixture, the other is the wasted spark. A DIS coil can be controlled by the MegaSquirt with only two pulses.
When the compression stroke takes place in cylinder 1 and the exhaust stroke in cylinder 4, the MegaSquirt controls primary coil A via pin 36 on DB37 (see figure below). This control is based on the crankshaft reference point (between 90 and 120 degrees before TDC). The MegaSquirt then controls primary coil B, which is responsible for spark generation in cylinders 2 and 3, 180 degrees after coil A. There is no reference point for coil B, but the ignition timing can simply be determined by counting the teeth on the 36-1 trigger wheel.
Between coil A of the ignition coil and pin 7 of the processor a 330 Ohm resistor is shown. This resistor limits the current and induction voltage of the control pulse. Because this resistor is not present as standard on the MegaSquirt PCB, it must be mounted afterwards. To the left of the vertical dashed line in the figure below, the internal circuit of the MegaSquirt is shown. The components shown (the two 330 Ohm resistors and the LEDs) had to be soldered onto the PCB afterwards.
Current build-up in the primary winding:
It is important to gain insight into the current build-up in the primary winding. Not only the current strength, but also the charging time of the ignition coil can be determined with this. The charging time depends on a number of factors which the MegaSquirt must take into account.
The self-inductance coefficient (L value) of the selected ignition coil is 3.7 mH. Together with the ohmic resistance R, this determines the maximum primary current and the rise time of the curve. A small L value and resistance ensures that the current rises quickly after switching on. Using the known data of the ignition coil, the build-up of the primary current can be calculated.
The following formula shows the general solution of the first-order differential equation used to calculate current levels, charging and discharging times in order to represent the switching phenomenon with a curve.
The equation is:
where the time constant (Tau) is calculated as follows:
The maximum current, according to Ohm’s Law, would be 28 amperes:
In reality this current will not be reached.
The coil is switched off earlier. The reason for this is explained further on. Filling in these values in the general formula gives:
The figure shows the charging curve of the primary coil. From time T0 to 1 Tau, the coil is charged to 63.2%. This is a fixed percentage for the charging time of a coil. The result of formula 13 shows that at 1 Tau the coil is charged to 17.7 amperes. At t = 5 Tau the final value has practically been reached.
According to the coil specifications, the primary current of the coil after charging is 7.5 A. The current will not rise any higher than this. The time needed to reach 7.5 A is called the dwell time. The dwell time depends on the battery voltage, which in this case is 14 volts. If the charging process is not regulated, the current through the coil will be a maximum of 28 amperes according to formula 12.
The coil according to formula 14 is charged to 17.7 A at t = 7.4 ms. The actual charging time is shorter, because the coil is charged to a maximum of 7.5 A. The time required can be calculated by entering the known values into formula 15.
The primary current build-up is stopped at 7.5 A. This prevents the coil from becoming excessively and unnecessarily hot. The most important thing is that the coil is charged as optimally as possible in the shortest possible time. The figure shows the charging curve up to t = 2.3 ms.
When the battery voltage drops, for example during engine starting, this affects the dwell time. It then takes longer than 2.3 ms before 7.5 A is reached. Using the now familiar formula, the new charging time is determined. The maximum current is determined on the basis of the battery voltage:
The charging time to 7.5 A with a maximum of 20 A is calculated in formula 17:
In the figure, the charging time at 14 volts is shown with the black line, and the charging time at 10 volts in green. The lines drop to 0 at the same moment; this is the ignition timing point. Because a lower battery voltage requires more time to charge the primary coil, the MegaSquirt must switch on the primary current earlier.
The black lines (rising and falling) indicate the dwell time at a battery voltage of 14 volts. The green line indicates the advanced charging time at a lower voltage: this gives Δt. The actual charging time in that case is therefore Δt + 100%.
Later in this paragraph this is clarified with an example and figure 36. The charging time is extended while the ignition timing remains the same. If this would not happen or not sufficiently, it would affect the energy released at ignition. The primary current would then be switched off too early, so that the current of 7.5 A is not reached. The extension of the charging time of the primary coil (dwell time) is, in formula form, a function of the battery voltage. Calculating the dwell time at different voltages gives a different maximum current in the coil.
By assuming that the battery voltage can drop to 6 volts during starting and rise to 14.7 volts while charging, a curve can be drawn by calculating a number of intermediate values. The figure below shows the dwell time correction for the DIS coil used. At every increase of 2 volts, a (red) dot is placed in the graph. Because a previously entered dwell time of 2.3 ms at a voltage of 14 volts was entered in the TunerStudio program, a correction factor is formed from this voltage. A voltage of 14 volts therefore equals 100% (no correction).
It has now been made clear that the charging time increases by as much as 315% at a battery voltage of 6 volts.
Under unfavorable conditions, the battery voltage can drop to as low as 6 volts. This means a weakening of the ignition spark. Extending the dwell time (the time that the primary current flows) compensates for this, so that sufficient ignition energy is obtained even at this low voltage. This means that Δt from figure 36 is tripled (2.3 ms * 315% = 7.26 ms) compared to the dwell time indicated in black of 100% (2.3 ms).
The coefficients indicated in red in the figure above can be copied directly into the TunerStudio program.
Some time after the primary coil has been discharged, the build-up for the next ignition begins. The higher the engine speed, the faster the coil is recharged. In figure 37 two curves are shown in which the primary current rises to 8.85 A. The ignition timing point is at the point where the line drops to 0 A.
Determining the ignition timing:
The ignition signal is determined on the basis of the crankshaft reference point.
On the toothed ring of the crankshaft pulley, 1 of the 36 teeth is milled away at 100 degrees before top dead center of the piston of cylinder 1. Between 100 and 0 degrees, i.e. during the compression stroke, the microprocessor of the MegaSquirt can determine the ignition timing. It takes into account the advance.
The figure shows the two-channel oscilloscope image in which the upper image shows the crankshaft reference point and the lower image shows the control signal from the MegaSquirt to the DIS coil. The control signal has a voltage of 5 volts (a logical 1) and lasts approximately 1.5 ms.
Ignition advance:
No knock sensors are used in this project. It is possible to process information from knock sensors, but simply mounting a knock sensor is not sufficient. Processing the signals is complex. The knock signal must first be converted into a yes/no signal or into an analog signal that indicates the intensity of detonation.
The conversion of engine vibrations into a knock signal is done by an interface circuit. This circuit is not present in the MegaSquirt II. Therefore it was decided to set the full-load and part-load advance safely, so that the engine does not enter the knock region. The full-load advance curve to be set must be determined within the knock limits. The centrifugal and vacuum advance data of the conventional ignition is determined from the factory data in the engine manual. The points can be plotted in a graph (example in the figure below).
The pink line indicates the original mechanical advance. This is partially linear due to the mechanical construction of the centrifugal weights. The black line shows the map-based control in the MegaSquirt; this line follows a curve. It is important to stay out of the part-load and full-load knock regions; therefore the map-based control in part-load is limited (red line) and the advance in full-load does not increase further than in the situation with mechanical advance (red line). The actual map-based control follows the blue line.
First, the full-load advance curve had to be entered in the spark advance table. At higher engine speed and lower load, more advance is required. In part-load, the advance is added to the full-load advance. On page 7 the completed ignition advance table and the advance settings for a cold engine are shown.
Throttle body:
In the original condition, the air/fuel supply was controlled by the carburetor. For the engine management system, the carburetor is replaced by a throttle body and four injectors that are mounted in the intake manifold. This provides a more accurate and controlled injection than with the carburetor, where an air/fuel mixture is formed centrally in the manifold and distributed into four channels. The throttle valve is opened by a Bowden cable that is manually operated from the instrument panel.
The MegaSquirt II does not support an electronically controlled throttle body. Therefore, Bowden cable operation is the only option that can be used.
The throttle position is passed on to the MegaSquirt by means of a voltage. The magnitude of the voltage depends on the opening angle of the throttle valve. The throttle position sensor is a potentiometer with a supply voltage of 5 volts (see figure). Connection 3 and a ground connection 1 are necessary. The wiper (pin 2) takes a position on the resistor which depends on the throttle position. The wiper is therefore connected to the throttle valve. When the wiper has to bridge only a small distance over the resistor (the wiper points to the left) the resistance is low. In the figure, the wiper points to the right (the ground side), which means that there is a high resistance and thus a low signal voltage.
With the throttle body used, there is a voltage of 600 mV on the wiper when the throttle valve is closed and a voltage of 3.9 V when the valve is fully open. The ECU receives the voltage and uses it to calculate the opening angle of the throttle valve. A rapid increase in the opening angle means that the engine is being accelerated; the ECU responds to this by briefly enriching. This is called acceleration enrichment. The throttle position sensor is not used to determine mixture enrichment under different operating conditions; the MAP sensor is used for this.
Test setup of the stepper motor with simulator:
After the MegaSquirt had been modified in hardware, the breakout box could be used to check whether the control signals for the stepper motor were being received. The illumination of two-color LEDs shows that a control signal is present. The steps in which the stepper motor is driven can be followed by looking at the change of the colors. The colors alternate between red and yellow. In the “Idle control” menu in the TunerStudio program, the stepper motor data can be entered. Besides the type (4 wire), the number of steps can be set. This also includes the initial position in which the stepper motor must be when the engine is started. Furthermore, the time can be set for how long it takes to adjust one step.
The number of steps depends, among other things, on the coolant temperature; a lower temperature requires a larger opening of the stepper motor. The steps can be set relative to the temperature in a graph. With the simulator, it can be checked whether the stepper motor is actually being controlled correctly. Because it is first checked on the simulator instead of on the engine, problems during starting or running of the engine due to a possible hardware or software problem can be prevented. Because mainly the coolant temperature and engine speed influence the opening angle of the stepper motor, by turning these potentiometers it can be checked whether the control is correct. The gauge on the dashboard in TunerStudio will display the adjustment in terms of the number of steps moved.
Settings for the stepper motor:
The figure shows the settings screen for the stepper motor that is used for idle speed (idle control).
The steps in which the motor is adjusted were determined in advance using an Arduino. The number of steps required to move to its home position (homing steps) also has to be entered. The stepper motor is active during the warm-up phase (algorithm) and energizes the coils at standstill (hold current between steps).
The position of the stepper motor depends on the coolant temperature. When starting a cold engine, the valve must be slightly more open than with a warmed-up engine. The figure below shows the settings screen for setting the steps (Steps) in relation to the coolant temperature (Coolant). When the engine is cold, the stepper motor is fully open during idling. During the warm-up phase, the stepper motor closes slightly.
It is also possible, when starting the engine, to set the position of the stepper motor on the basis of the coolant temperature. This is called the “Idle Cranking Duty/Steps”. The figure below shows the settings screen.
Fuel pump circuit:
The MegaSquirt controls switching the fuel pump on and off. Transistor Q19 in the figure below protects transistor Q2 against excessively high currents. At too high a current, the transistor may burn out. When the current through the collector-emitter section of Q2 and R40 increases, the saturation voltage at the base of Q19 is reached. Transistor Q19 starts conducting, causing the base-emitter voltage at Q2 to decrease.
Connection FP-1 PTA0 is controlled internally by the MegaSquirt. An input signal from the crankshaft position sensor (a Hall sensor or inductive sensor) is required to drive the transistor circuit. If the signal is lost, for example when the engine unintentionally stalls, the power supply to the fuel pump is immediately cut off.
The output of the transistor circuit (FP1 OUT) is connected to the fuel pump relay. Pin 85 of the relay is the output of the control circuit. When the relay is energized, the main current section (pin 30 and 87) is switched so that the fuel pump receives a supply voltage to operate.
An electronic fuel pump with an operating pressure of 3 bar is used. The fuel is routed through the fuel filter to the fuel rail, where the pressure is present at the injector inlets. The injector will inject a pre‑calculated amount of fuel into the intake manifold at the moment a signal is received from the MegaSquirt. Not only the control by the MegaSquirt determines the amount of injected fuel, but also the fuel pressure in the rail.
At a higher rail pressure, a larger quantity of fuel will be injected with the same control signal. The rail pressure must therefore be adjusted based on the vacuum in the intake manifold. The pressure difference (∆P) must remain 3 bar at all times. The figure shows the diagram of the fuel system. The pink, yellow, orange and black lines represent the electrical connections. The red line indicates the fuel supply and the blue line the fuel return.
Completion of the mechanical work:
The following three photos show the engine in the final stage of the mechanical modifications.
Photo 1:
This is the side where most of the installed components are visible. The dashboard for operation and the MegaSquirt ECU are also located here. Below the photo is a legend with a description of the numbers next to the components. You can open the photos in a larger size by clicking on them.
- Throttle valve;
- Fuel line for the injectors;
- Connecting pipe for throttle body to intake manifold;
- Fuel pressure gauge;
- Intake and exhaust manifold;
- Dashboard with fan switch, indicator lights for alternator and oil pressure, ignition switch and master switch;
- Vacuum hose for MAP sensor;
- Lambda sensor;
- Fuel hoses (supply and return) together in a heat‑shrink sleeve;
- Fuel pump / tank unit;
- Fuel pump relay;
- MegaSquirt;
- Rear silencer.
Photo 2:
This photo shows the other side of the engine. Here you can see the carburetor (15) and the conventional ignition system (17). The purpose of this classic ignition is to make the spark plugs in the test setup (14) spark. It naturally has no function for the engine itself, but it does provide insight into how the ignition system in classic cars worked.
Number 20 indicates the mechanism of the transmission brake. By means of a Bowden cable, the rod of the brake drum can be pulled so that the output shaft of the gearbox is braked. The transmission brake is used to load the engine briefly when a gear is engaged.
14. Test setup of mechanical distributor ignition;
15. Carburetor;
16. DIS coil;
17. Mechanical distributor ignition with vacuum advance;
18. Rear side of dashboard;
19. Mechanical fuel pump;
20. Mechanism for transmission brake;
21. Classic ignition coil.
Photo 3:
The top view of the engine with the test setup for the ignition and the fuel rail is clearly visible here.
The mechanical modifications have been completed. The engine cannot be started yet, because some data first need to be entered into the MegaSquirt.