Tuesday 31 May 2011

WS7 EXHAUST GAS ANALYSIS


The reading bellow illustrates current content of ambient air. This content will be a starting point for the emission analysis. The machine will examine the proportion of gases measured either in ppm (parts per million) or % ( percentage).



After the engine was started and was running cold we memorised this data:


The concentration of two gases such as carbon monoxide (CO) and hydrocarbon (HC) indicates that the air/fuel mixture is quite rich immediately after cold start. Not all CO molecules can combine with oxygen and fuel (hydrocarbons) is not burnt completely. Other figures are just normal for this stage of engine running.
                                                                                                  
The readings have changed with engine reached warm condition:






CO has fallen dramatically. This means that mixture is not so rich anymore. Lambda tell us that fuel mixture is just slightly rich. CO2 and O2 are normal. But look at hydrocarbon. 253 ppm is relatively high for this car been warmed up. To my understanding of what is going on here are some problems with ignition. Fuel could be burnt a bit more efficiently. Putting in account that lambda  is 5 % richer than stoichiometry and proper figures for CO2 and O2 I suspect ignition system, probably just spark plugs.


As it is seen on the screen at 2500 RPM carbon monoxide increased in 16 times compared to the idle RPM. At the same time there are no changes in hydrocarbon and slight changes in O2 and CO2. It's reasonable because the air fuel mixture is richer at higher RPM.

 Experimented with MAP sensor we simulate the situation when engine runs very rich. Have a look at the figures below:
HC shows that fuel could not be burnt, it was too much fuel. CO indicates uncompleted chemical reaction, hard to reach free atom of oxygen to get them combined. CO2 and O2 underlined oxygen deficiency.

All the sudden acceleration always means instantaneous fuel injection and rich mixture in the combustion chamber:


CO jumped up to the 3.467%. This is sign of too rich mixture. HC increased while CO2 and O2 decreased. By blipping the throttle we do not allow fuel to be burnt completely for such short period of time.


Simple experiment with unplugged spark plug allows watching unburned fuel which is represented in high volume of hydrocarbon:


With disconnected injector we have lack of fuel. This imbalance was read by the machine immediately:




The last experiment we did was turned on the air conditioning system:


The engine runs slightly lean with a small load on it by air conditioning compressor. CO2 and O2 are normal. But after all our experiments HC is 25 ppm. It looks like the engine was not warm enough in the beginning of the experiment.
All these tests above lighted up the key points of the modern engine management system. The evolution of engine performance and efficiency from the economical point of view are based on the principle of continuous reading signals from sensors and feedback to the ECU while it is in closed loop mode. Moreover, this is the way how the emission can be reduced considerably what is the major for responsible manufacturer. To get this system working the oxysensors were employed. The faster these sensors come to the operation temperature the less emission we’ll have. That’s why O2 sensors differ in their wiring. 1-wire type is the old one with signal wire and grounding through the body. 2-wire type has signal wire and separate earth wire. 3-wire has signal wire, ground wire, and heating element wire. The heating element is used to reduce time to reach operating temperature and lower emission. 4-wire oxysensor has supply voltage wire and ground for heating element, and also ground for signal wire. All of these mentioned sensors are switching type or they also refer to as HEGO ( heated exhaust gases oxygen sensors).
But 5-wire O2 sensor is wide-band sensor. The wires in this instance are heater voltage supply and ground, reference ground for both senceing cell and pumping cell, signal from the sencing cell to the ECU, and current wire for the pumping cell. The construction of this type wide range air fuel sensors can differ and so far numbers of wires can be higher. The invention of WRAF sensors allows to control how lean and how rich is the mixture. Thus, this sensor is not switching type device. It provides more precise data for the ECU.


WS6 OXYGEN SENSOR ON THE VEHICLE

Before starting the experiment we located the O2 sensor. It’s placed in the exhaust pipe before the catalytic converter. It is three wires device, where two white wires are heater and ground and black wire is O2 sensor signal wire.

This is Zirconia switching type sensor. When the back probe was inserted we got this oscilloscope reading:



From this pattern you can see that voltage goes as high as 0.785 V and as low as 0.05 V. Unfortunately, our oscilloscope could not calculate an average voltage, so that we put on the paper approximate value of 0.4 V. The signal has 8 cross counts.
The signal is not uniform but fluctuates. For the normal condition I’d suspected misfire. However, in our case, it happens due to usage of aftermarkets ECU modules and program. The way how the O2 sensor is wired also plays role. Therefore, the oxysensor itself is working. This is a next in turn example that it would be wrong to say instantly that the device is faulty. The way of thinking must allow you to analyse the whole circuit and find out the real cause. This time this is how the O2 sensor is wired.
Consequences of this sort of wiring can be seen from the next test. The O2 sensor was heated up and RPM came down. It is expected to watch the pattern with cycling signal. But the signal we captured is not a good one:


The highest voltage is 0.016 V, the lowest voltage is 0.01 V. An average voltage is approximately around 0.011 V. We could not register the cross counts. The reason why it is so is described above.
Sudden acceleration brought us much better result:

 
The highest voltage was registered at 0.818V. This sort of O2 sensor sensing underlined that it reacts faster than 100 ms and sends the signal to the ECU that mixture is rich at the moment.
Once we accelerated and then dropped the pedal all the sudden we got this pattern:


The signal trend is normal, time of O2 sensor reaction is within the specification. But once again, bad wiring to the ECU module evoked fluctuations of the signal.
Instant acceleration in order to catch the pattern and count the time of O2 reaction is shot on this image:



Even though it is hard to count in our case, it is clear that from lean to reach sensor switches faster than 100 ms and to switch backwards takes a bit longer.
On the basis of all these tests and the information that O2 sensor is not wired proper way I tend to conclude the Oxygen Sensor is in operation condition. The more accurate conclusion can be made after testing this device or wiring it properly.