EFIE demystified:
A word of warning: these devices are complex electronic products designed by electronics engineers. Users with basic knowledge about signals and electronics may not easily understand the explanations below. If you find you are having difficulty understanding the explanations below, you should ask a specialist for guidance.
Many people have tried to implement EFIEs in their cars, some with success, some with disappointing results. But why is this?
The answer is relatively simple: because the solutions simply overlook some key parameters of the sensors. All the cars today have oxygen sensors that deliver signals to the computer telling it how rich or lean your mixture is. One particular sensor is the wideband sensor, present in about 90% percent of today’s cars, and the most difficult to tune. These sensors are basically switches designed to change output from about 0 to 1Volt when the mixture is at 14.7:1. This ratio was believed for years to be the best mixture (known as stoichiometric mixture) for burning fuel. Today it seems that the car manufactures admit that they were wrong, and the cars are equipped with wideband sensors, able to go much leaner, to save gas.
However, what do we do with the cars with narrowband sensors? The solution in the market is to use a so called EFIE, which is a device inserted after the O2 sensor, which will modify the output to make it look richer to the computer and make the computer adjust the mixture leaner. The approaches are very similar. One of them is to add a voltage to the sensor, which seems to be logical, but it doesn’t work. Another attempt is the so-called digital EFIE, which compares the voltage with a reference signal and generates a digital output. Please look at the picture bellow to understand the common aftermarket “explanation”.
Many people have tried to implement EFIEs in their cars, some with success, some with disappointing results. But why is this?
The answer is relatively simple: because the solutions simply overlook some key parameters of the sensors. All the cars today have oxygen sensors that deliver signals to the computer telling it how rich or lean your mixture is. One particular sensor is the wideband sensor, present in about 90% percent of today’s cars, and the most difficult to tune. These sensors are basically switches designed to change output from about 0 to 1Volt when the mixture is at 14.7:1. This ratio was believed for years to be the best mixture (known as stoichiometric mixture) for burning fuel. Today it seems that the car manufactures admit that they were wrong, and the cars are equipped with wideband sensors, able to go much leaner, to save gas.
However, what do we do with the cars with narrowband sensors? The solution in the market is to use a so called EFIE, which is a device inserted after the O2 sensor, which will modify the output to make it look richer to the computer and make the computer adjust the mixture leaner. The approaches are very similar. One of them is to add a voltage to the sensor, which seems to be logical, but it doesn’t work. Another attempt is the so-called digital EFIE, which compares the voltage with a reference signal and generates a digital output. Please look at the picture bellow to understand the common aftermarket “explanation”.
The assumption is that the O2 sensor generates a sine wave signal and measuring the voltage will give you some results. This assumption is misleading since the O2 sensor is NOT a sine wave generator. These sensors are tuned to change their output very fast at 14.7:1, so the signal looks more like a square wave. In the following document, written by an ASE Certified Master Technician, you can get some more details and find out why the raising and falling times are critical for the sensors:
hhttp://www.batauto.com/articles/catfailure/
Or, if you read this datasheet from the manufacture, you will also find: “When the fuel mixture goes rich, even just a little bit, the O2 sensor's voltage output shoots up quickly to its maximum output of close to 0.9 volts. Conversely, when the fuel mixture goes lean, the sensor's output voltage drops to 0.1 volts.”
http://tayloredge.com/reference/Science/oxygensensor3.pdf
It is now clear that the switching time is really fast (old sensors less than 100ms, and for the newest sensors less than 25ms) and thus the signal is not a sine wave at all. This is a very important detail because all explanations considering the O2 sensor output as a being sine wave are wrong. If we repeat the tests with faster signals (25ms), this is what you get:
hhttp://www.batauto.com/articles/catfailure/
Or, if you read this datasheet from the manufacture, you will also find: “When the fuel mixture goes rich, even just a little bit, the O2 sensor's voltage output shoots up quickly to its maximum output of close to 0.9 volts. Conversely, when the fuel mixture goes lean, the sensor's output voltage drops to 0.1 volts.”
http://tayloredge.com/reference/Science/oxygensensor3.pdf
It is now clear that the switching time is really fast (old sensors less than 100ms, and for the newest sensors less than 25ms) and thus the signal is not a sine wave at all. This is a very important detail because all explanations considering the O2 sensor output as a being sine wave are wrong. If we repeat the tests with faster signals (25ms), this is what you get:
In conclusion, if you have a slow sensor, you might get some improvement with the existing EFIEs. If your sensor is in good shape and has fast transitions, you are probably in trouble and will need a better EFIE, with the adaptive function.