Sensitivity is a good quality in a person, or for that matter in an engine, especially in these days of Kyoto Protocol-driven environmentalism. Under the hood, however, controlling emissions is a lot harder than negotiating a treaty. Or is it? In a sense, “negotiation” is what goes on with every power stroke, as the ECU takes information from the sensors and sends commands to actuators. We all know how to use the diagnostic tools needed to make decisions about sensor and actuator performance, but what’s going on inside that little black box? More than you might imagine.
How much air?
Measuring how much air the engine ingests is the entry point for any ECU’s fuel mapping system, since operation at or near the magic 14:1 air-fuel ratio for ‘stoichiometric’ combustion is the key to economy, power and low emissions.
Getting that information to the computer is handled several ways. Older systems use an electromechanical system that uses a spring-loaded flap or vane that rotates in response to the rush of air flowing past. Rotation of the pivoting shaft operates a potentiometer that creates a variable resistance to an applied voltage. That “analog” signal is converted to a digital form usable by the computer inside the ECU. It’s a workable system that has been installed in millions of cars and light trucks, but there are significant disadvantages. The biggest issue with velocity sensing is that the system isn’t solid-state. Not only are there moving parts, but the vane is susceptible to mechanical damage and wear. From a performance perspective, it’s also a relatively crude device, since the mass of the moving parts limits the response time of the ECU to changing airflow in the engine.
A cheaper and better solution is the mass air flow sensor. The MAF operates without moving parts, but uses the same basic idea as the velocity type: convert airflow into a variable resistance to an applied current. The difference in a MAF sensor is that the air flows past a heated wire mesh or film, cooling it in direct proportion to the mass of air flowing past. The film or mesh is made from materials whose resistance changes predictable with temperature, allowing the cooling effect to accurately reflect the air mass going into the engine. MAF sensors have no moving parts, are less restrictive to airflow and react quickly to changing flow in part throttle conditions. They are, however, susceptible to pulsation and backflow of air in the intake. Solving that problem in modern MAF sensors involves conditioning the signal by using a small, smart integrated circuit “chip” directly in the sensor to weed out spurious signals and feed the ECU a signal that’s closer to the actual engine air demands. From a service perspective, the amount of intelligence built into the sensor affects the ability of the technician to measure its performance when diagnosing driveability issues. ‘Scoping the output signal to the ECU is still useful, but with an in-sensor circuit modifying the signal, dedicated test equipment may be the only way to troubleshoot in reasonable time. One sometimes forgotten check is for contamination. A ‘burn-off” cycle of high current to temporarily heat the mesh to very high temperatures is built into many systems (Bosch is an example). A failure or open in the current feed to the burn-off circuit could still leave a functioning sensor, but with poor or inaccurate response.
Temperatures, of coolant, intake air, ambient air and even EGR gases are almost as important to driveability and performance as crucial sensors like O2 and MAF types and are generally easier to diagnose. That’s because internally, the sealed units operate by temperature sensitive resistors called “thermistors” that usually fail as an open circuit. Mechanical damage can create an intermittent, but the simple, sealed nature of temperature sensors and the common two or three wire circuits they feed, makes troubleshooting generally easy. A common test called up by factory manuals involves immersion of suspect sensors in boiling water while attached to an ohmmeter. It’s a workable test, but in many shop environments, it’s a little time consuming compared to in-car testing as the engine warms up. A common off-car shortcut is to play a torch flame across the sensing end while watching the ohmmeter, but this technique should be avoided, since overheating the sensor can cause internal damage to the thermistor, or can crack leads internally, leading to a difficult to diagnose “intermittent”. Understanding how the ECU uses the temperature information can also help pinpoint temperature sensor failures. A typical example is late model Toyota vehicles, whose ECU compares coolant and intake air temperatures at startup. If they’re the same, the system assumes cold start conditions.
Knock sensors are different
It doesn’t take long for a technician to realize that whether it’s MAF, IAT, CT, EGT or TPS, sensing temperature (or position in the case of throttle and EGR valve opening) is about varying a resistance to an applied current. Knock sensors, however, are different. Knock sensors use a piezoelectric ceramic element that acts like the push button igniter in a gas barbecue. In both devices, the piezoelectric material responds to pressure by generating a small current, which in the barbecue is jumped across a spark gap to ignite the flame. In a knock sensor, the flame triggers the current, as rough combustion sets up a 10,000-15,000 Hz vibration, which microscopically compresses the piezo element in the sensor. You can’t boil this one in water, and it’s difficult to simulate engine operating conditions on the bench for this device, so diagnosis depends more on experience, OBD and modern equipment. Failures are surprisingly rare considering the environment knock sensors operate in. Piezoelectric materials can also work in reverse, changing shape when an electric potential is applied. Uses for this property have been limited, but Siemens recently unveiled a common-rail diesel injector that uses piezo-actuation instead of conventional wound coils. Better emissions and lower levels of NVH (noise vibration and harshness) are claimed for the system.
There is far more to sensor technology than can be covered in any one article, and as sensors proliferate in both body and chassis systems, understanding them all will become more difficult. As sensors add signal conditioning electronics in the same package, the multimeter may not be enough for even simple temperature sensor diagnostics. Right now, however, there’s still room for experience and good intuition, and there may be in the future as well. But what about the oxygen sensor? That “mother of all sensors” is a unique and specialized instrument that deserves an article all by itself. Watch for it in the May issue!
Will hydrogen fuel cells mean the end of sensors?
With the recent run-up of world oil prices, the drive to perfect hydrogen fuel cells is picking up steam, with everyone from major auto makers to U.S. President George Bush predicting that fuel cells and hydrogen are the on the horizon. Will this mean the end of sensors? No, but they will change in function and design. NGK Spark Plug has developed three sensor types that may debut on the roughly 50,000 fuel cell vehicles that NGK expects to be on the road by 2010. The hydrogen concentration sensor uses a proton-exchange membrane in a structure similar to current O2 sensors, (and the fuel cell itself) to sense concentrations from 0 to 80 percent. The hydrogen flow sensor and leak sensor use high-tech silicon micro heaters derived from diesel glow plug and O2 sensor technologies, and work a little like a mass airflow sensor. The flow sensor works by measuring the change in the resistive balance of the micro-heater in the circuit as hydrogen flows past. The leak-monitoring sensor uses the micro heater and a special sensing electrode where a measurable chemical reaction with stray hydrogen takes place. It appears that if and when hydrogen powered fuel cell vehicles hit the streets, technicians may troublesh
oot sensors in similar ways to their internal combustion ancestors.