Basic Engineering Principles

From PrecisionCodeWorks
Jump to: navigation, search

Open loop Vs Closed Loop Management systems

Electronic Engine Management Systems utilize two forms of management when it comes to precisely controlling an engines operation. These are known as Open Loop and Closed Loop control systems.


Closed loop.png

In a Closed Loop system the ECU sends and receives a constant stream of data from various sensors throughout the engine. These sensors include but are not limited to

Bosch in out.png

1, Engine management control unit with ambient pressure sensor

2, Battery

3, Ignition switch

4, Engine management system relay

5, Cruise Control lever

6, Linear pressure sensor

7, Brake pedal and clutch pedal switches

8, Lambda sensors upstream and downstream of the converter

9, Timing sensor

10, Engine coolant temperature sensor

11, Detonation sensor

12, Turbo pressure sensor

13, Engine rpm and TDC sensor

14, Intake air temperature and pressure sensor

15, Electric fuel pump relay

  • Idle Up VSV’s and Fuel Pressure Sensors – Fuel Pressure Sensors are almost exclusively fitted to racing engines for fine tuning of the injector maps. The data it receives is constantly monitored, and adjustments are made to the Injector Pulse-Widths and timing along with Spark timing to ensure the engine runs as economically as possible for road car applications.


Open loop.png

In an Open Loop system the ECU ignores the data received from many of these sensors and will simply run the engine according to a base “map” that was programmed into the ECU by the manufacturer. This ECU map is the only reference point the ECU uses when it comes to determining Injector Pulse-Widths and timing and spark timing, rather than receiving and processing inputs from all the engines’ sensors.

The advantage of operating an Open Loop system is that the engine can run on far fewer sensors and inputs to the ECU and still operate to a satisfactory level. These systems are also cheap and easy to program because the ECU doesn’t need the processing power that an ECU operating a Closed Loop system does when it has to process various sensor inputs.

Theoretically and engine running with an Open Loop engine management system can run on only a Crankshaft Position Sensor and Throttle Position Sensor, something that a solely Closed Loop system would be incapable of doing. Where Closed Loop Systems come into their own is when performance and economy are the major concern; because the ECU is receiving and processing so much useful data about exactly how every aspect of the engine is running, it can make very minute and precise adjustments to maximise performance and economy at all times.

One example of why these systems are so useful is the Knock Sensor.

  • Knock Sensors are installed in the engine block close to the top of the combustion chamber, the perfect position to detect a phenomenon we call engine knock or detonation. Detonation occurs when the engine is running too much spark advance or the combustion chamber is so hot that it ignites the Air Fuel Mixture before the piston reaches the top of the cylinder. When detonation occurs the flame front attempts to push the piston back down the cylinder when it is still on the upwards stroke which then causes it to make a knocking noise.

Knock sensors produce a voltage when detonation occurs and tells the ECU. The ECU then proceeds to back the spark advance off a bit until it stops, at which point the ECU then begins to increase spark advance until detonation occurs again. The reason why ECU’s do this is because engines run most economically when they run as lean as possible and with as much spark advance as possible, so knock sensors allow us to run the engine as close to the point of detonation as possible without actually detonating.

When we utilise an Open Loop system the ECU won’t safely be able to run the engine with as much spark advance. To ensure safe engine operation and reliability the ECU will be programmed to run the engine within a safe spark advance limit at all times, rather than running just on the edge of detonation – this system is safe but will not allow the engine to run in its most economical state.

As a brief summary, Open Loop systems are cheap to install but uneconomical whereas Closed Loop systems are more complex and expensive however they offer far superior economy.

Examples of Open and Closed Loop Systems

As was previously explained, a Closed Loop system utilises various sensor inputs from throughout the engine to allow the ECU to make fine adjustments to ignition and fuelling systems during engine operation. One of the most important sensors used in Closed Loop systems is the Lambda/O2 Sensor. This sensor sits in the exhaust system, usually in the collector of the downpipe, where it can monitor the composition of the exhaust gases. The sensor looks for excessive content in the exhaust gases, when the O2 levels are too high the sensor reports this back to the ECU and the ECU then knows that the engine is running too lean. When levels drop too low then the ECU knows that the engine is running too rich and fuel is being wasted, so it then adjusts injector pulse widths accordingly.

When O2 sensors fail we often find that the engine runs far too rich or too lean, resulting in excessive fuel consumption or in the worst case scenario a full engine lean out, which then melts piston tops. To avoid a full lean out however we have an Open Loop failsafe in place; if a sensor or a set of sensors fail, the ECU reverts to a safe mode which is essentially an Open Loop management system that utilises predetermined safe fuel maps to ensure the driver can get the car to a garage safely without causing any lasting damage.

When this fuel and ignition map is implemented we can expect to see a significant rise in fuel consumption, a loss of power and on engine with Variable Valve Timing we also find a drastically lowered rev limit.

Toyota’s 2ZZ-GE engine drops the rev limit from 8200RPM to 6200RPM to ensure that the VVTL-i system isn’t activated when the ECU is unable to precisely monitor and adjust fuelling and ignition requirements. From this we can gather that modern cars actually use both Open and Closed loop systems, reverting to Open Loop when sensor failures occur. A fully Open Loop system will not rely an O2 sensor to send back fuelling information, instead it will rely on a far more basic system such as a crank position sensor or in the 4A-GE’s case, a MAP sensor, TPS, Coolant Temperature sensor and AIT sensor. The 4A-GE ECU operates on a basic predetermined fuel and ignition map that was programmed in by Toyota on the EPROM chip. This fuel map will not change and the ECU only adjusts the fuelling according to the data received from these sensors. For example the ECU will increase injector pulse widths when it notices an increase in throttle angle and a drop in manifold pressure when the butterfly valve is opened right up. Similarly, when the engine begins to run very hot due to coolant passage blockages or a broken radiator fan, the ECU then starts to enrich the fuel mixture in an attempt to reduce combustion chamber temperatures. One problem with this system is if the coolant temperature sender fails – when a wire breaks the sensor reads a very high resistance and the ECU then reads this as extremely high engine temperatures. To combat this it then massively over-fuels and the engine then cuts out due to flooding. The 4A-GE ECU does however operate a failsafe system and I have personally seen it run and drive without a MAP sensor – the most important sensor for the ECU to read from is the TPS sensor; with that disconnected the engine simply will not start. This is not to say that it ran well without a MAP sensor though, quite the opposite – the engine over-fuelled wildly and “coughed” during test drives, making power delivery extremely erratic and low.

Interaction Between Engine Management Systems

The interaction between all engine management systems in Closed Loop ECU’s is critical if we are to obtain maximum performance and economy. If we could look at an ECU’s programming easily, we would see that the ECU relies on more than just one system for best fuelling and ignition control. For example, the ECU will not be programmed so that when the TPS sensor reads 25 Degrees of Throttle Angle the injector pulse is 0.8m/s no matter what other sensors read. Instead you will find that it is reliant upon other factors such as the oxygen density flowing into the engine and the manifold pressure. If the Throttle was open by 25 degrees with 12% oxygen content to offer a 0.8m/s pulse width, you would find that the pulse width then increases when the oxygen content rises to ensure maximum power and a clean burn of fuel without leaning out from the extra oxygen content.

Another example of systems that are reliant upon each other is indeed one that I mentioned earlier – the knock sensor and the ECU’s control over the ignition timing depending upon the knock sensors’ readings. The ECU knows that it must keep the engine as close to detonation as it can without actually causing detonation, so the ECU continuously increases spark advance until the engine knocks at which point it then backs it off before it starts pushing closer towards detonation again. In conjunction with the knock sensor the ECU will also continue to monitor Coolant Temperature – remember that higher combustion chamber temperatures increase the likelihood of detonation, thus spark advance will need to be backed off to compensate for that as a more economical solution to simply dumping extra fuel in the combustion chamber.

More modern engine management systems are now ‘learning’ the drivers style of driving – for the first time the driver takes a new car out of the show room the ECU will monitor and store information regarding the RPM’s the driver likes the change gear, the RPM range the driver uses most, the throttle openings the driver uses most and generally how the driver uses the engine. The ECU then optimises itself from the data gathered to ensure maximum performance and economy to match the driver. For example, when the ECU recognises the driver using the bottom of the rev range more often with small throttle openings, suggesting low speed town driving it will adjust the Throttle Actuator.

Most modern cars use a fly-by-wire system whereby the throttle pedal isn’t directly connected to the throttle lever; instead an actuator opens and closes the throttle body butterfly valve according to the voltage produced by the potentiometer on the throttle pedal. Adaptive ECU’s that note low RPM averages will then ensure that the butterfly opens by very small amounts during town driving to make it easier for the driver to manoeuvre and drive at low speeds. This benefits cars with very large throttle bodies where it is quite difficult to move by the pedal by such small amounts so as to prevent the throttle from snatching and over revving. This is why some cars have a sport mode.

Sport mode generally ignores all data gathered about the drivers’ style and maximises engine performance at the expense of economy – the throttle response will be drastically sharpened making the car tougher to drive at low speed as the throttle is sharper and more likely to snatch away from the driver at low speeds. The suspension is also firmed up in cars with active suspension, gear changes are sped up with automatic vehicles and cars with EPAS also find that the steering is sharpened up to offer more response to the driver.

Hall Effect Sensors – Wheel Speed/Crank Position Sensors

Hall Effect sensors operate on the principles suggested by Lorentzes’ law. When we pass a current through a piece of semi-conductor and then apply a magnetic field to it at a right angle to the current, a voltage is produced by the separation of charged particles within the semi-conductor. This voltage is tiny and must be amplified, so either the sensor or ECU is fitted with an amplifier to ensure the signal can be read by the ECU which operates on signal voltages between 0.1V - 4.9V. We apply the Hall Effect in vehicles in the form of a crank position sensor; there is a small current applied to the semi-conductor within the sensor and the magnetic field comes from the teeth of the flywheel. As the flywheel rotates the teeth pass the sensor and trigger it because of the magnetic field they produce. There is a single tooth missing on the flywheel, so when the flywheel rotates there is one gap in the signal produced by the sensor which the ECU notes as Top Dead Centre of Cylinders 1 and 4. The problem with this system is that Hall-Effect sensors are very expensive to produce and they can only tell the ECU that both cylinders 1 and 4 are currently at TDC, however they cannot tell the ECU which one is on the compression stroke. To get around this problem we fit the camshaft with sensors to inform the ECU of the camshaft position and thus of which cylinder is due to spark.

The other problem with Hall Effect sensors is that they need to be very close to the flywheel teeth if we are to gain a strong enough voltage within the semi-conductor. If the gap increases by the smallest amount due to the sensor bracket being dislodged, the signal will weaken substantially and will eventually cut out completely leaving the ECU without a way of knowing what the crankshaft position is – for some engines this means they will not start or they will cut out mid operation. Where Hall-Effect sensors come into their own is when we consider that they produce a digital signal that doesn’t need to be converted or conditioned by a conditioner unit within the ECU. This leaves more space within the ECU for a larger processor or allows us to reduce the ECU’s size and complexity. Magnetic inductive sensors require conditioner units because they produce an analogue signal in a rounded wave-form rather than an I/O signal such as that produced by Hall-Effect Sensors.

Magnetic Inductive Sensors – Crank Position Sensors

Magnetic Inductive sensors have been in use for many years now as Crank Position Sensors and wheel speed sensors. They work in a similar way to a coil, wherein a magnetic field is passed through a coil and this induces a current within the coil. Like a Hall Effect Sensor, the Magnetic Inductive sensor relies upon the magnetic field produced by the flywheel teeth cutting through the coil within the sensor to induce a current/voltage. And again like the Hall-Effect sensor, the voltage produced is very small and must be amplified to register within the ECU’s voltage range of 0.1V-4.9V. Again, this amplifier can be built into the sensor or into the ECU, although sensors with the amplifier built in do tend to be more expensive due to the tightly packed electronics.

The real advantage behind putting the amplifier pack within the ECU is to reduce the size of the sensor – sometimes packaging issues are very prominent in small cars, thus making the sensors in various places smaller not only leaves the sensor less prone to being damaged but also means it doesn’t get in the way of other components as much as a larger unit may have done. Another large advantage of magnetic inductive sensors lies in the fact that we have been using them for years – this means that they are considerably cheaper than Hall-Effect sensors. They have been manufactured for long enough and in enough vehicles that economy of scale reduced their price to a point where it makes a lot of sense to use them in road cars. Where they fall down is when we look at the signal they produce – they produce an analogue signal which requires a conditioner unit to convert it into a digital signal. This requires more electronics and more space either in the sensor or the ECU and adds to the complexity of the system. All the conditioner does is cut anything above a given voltage out to flatten the signal curve produced, thus turning it into an I/O digital signal that the ECU can be understood by the ECU.

Thermal Resistors – Engine Coolant Temperature Sensors

Thermal Resistors, commonly known as Thermistors for short, are used to measure the coolant temperature within an engine. These sensors report back the coolant temperature to the ECU in a form of a voltage. From the reading given by the CTS the ECU can determine whether or not the engine is at correct operating temperature. Upon cold start the CTS informs the ECU that the engine is well below operating temperature, at which point the ECU enters Open Loop mode whereby the O2 sensor is ‘ignored’ as is the EGT (Exhaust Gas Temperature) sensor. The ECU activates the Cold Start Injector to add more fuel into the combustion chambers thus warming the engine up faster. The O2 sensor and EGT Sensor are left out of the ECU’s inputs during cold start because the ECU knows they will read very low EGT’s and very high emissions, telling the ECU to reduce fuelling which then results in an engine that takes far too long to warm up. When the CTS informs the ECU that the engine is approaching correct operating temperature the ECU cuts the cold start injector out and allows the engine to run on the standard injectors. Its then enters Closed Loop mode and reintroduces the inputs from the O2 and EGT sensors to relay information regarding engine operation and emissions. At this point the engine runs as efficiently as possible and emissions are as low as possible because the ECU is now making all the minute adjustments to ensure the best economy and performance. CTS’s are a semi-conductor based sensor that changes voltage as the resistance changes within the semi-conductor. As the coolant temperature changes so does the resistance within the semi-conductor, thus altering the voltage output to the ECU. There are two forms of CTS; Positive Temperature Coefficient and Negative Temperature Coefficient. PTC’s increase in resistance as the temperature rises whereas NTC’s decrease resistance as the temperature rises. The most common form of Thermistor is the PTC. Whether a Thermistor is a PTC or and NTC is determined by the semi-conductor material used. Thermistors generally have a temperature range between -50 to 250 degrees Celsius. It is important to note that the changing resistance can only be measured if a reference voltage is supplied to the sensor. This makes the CTS an active sensor rather than a passive sensor like the Hall-Effect and Magnetic Inductive sensors that produce their own voltage. The reference voltage is supplied by the ECU, and the ECU simply compares the input voltage from the sensor with the voltage it supplies the sensor.