The days of the zirconia upstream oxygen sensor are behind us. Vehicle manufacturers have moved to what is commonly known as the air/fuel ratio sensor, or AFS. With the change in technology has come a steep increase in the price of the replacement part, which means you need know how to diagnose sensor and air/fuel related faults. This month, we will focus on Asian AFS operation, diagnosis and replacement with a focus on what you need to know to fix it right the first time.
Feedback strategies
When a spark ignition engine is in open loop operation, it is using a feed-forward strategy for operation. The Engine Control Module (ECM) is looking at inputs such as engine load and RPM, as well as adjustments for air and coolant temperature and is commanding an injector pulse width to match fuel to the incoming air at a specified ratio. Typically the AF ratio being commanded is 14.7 parts of air to 1 part of fuel by weight. This ratio is commonly known as stoichiometric ratio. While it has been commonly taught that the ECM is primarily commanding stoich, it is not a fixed rule.
In closed loop operation, the AFS is used to provide feedback to the ECM and reports the actual air fuel ratio to the ECM. While the AFS is the primary sensor for feedback, the ECM uses the downstream oxygen sensor to make secondary adjustments for the catalyst.
Why AFS?
For years we have learned about the zirconia oxygen sensor. We were taught that 450 millivolts was stoich, right? Well, it is, but so is anything in the range of 200-800 millivolts. Yes, you read that correctly. The zirconia oxygen sensor’s range is typically in the area of 100-900 millivolts and as such the ECM/PCM doesn’t truly know how rich or how lean the air fuel mixture is. Until the sensor reaches the 200 millivolt range on the lean end and 800 millivolts on the rich end the sensor is assuming stoich. These of course are not hard numbers as the range will vary slightly with sensor temperature and sensor manufacturer. In order to maintain stoich on a system with an upstream zirconia type O2 sensor, the ECM had to toggle the fuel command rich and lean in order to keep the AF ratio within the 200-800mV range. This toggling of AF ratio was an inefficient way of maintaining stoich. The zirconia style oxygen sensor can only accurately see a very narrow range of air/fuel ratios. In fact, it really only sees stoich. Again, it cannot accurately report how rich or how lean the mixture is.
The AF Sensor is considered a wide-range sensor. What does that mean exactly? It means that it can accurately see a wide range of air/fuel ratios. This has become increasingly important to manufacturers as they try to improve fuel economy and tailpipe emissions. The ability to see the exact AF made the switch to the AFS a no brainer. With the accuracy of such a sensor comes the ability to command a variety of fuel ratios for specific engine load and operating conditions. Keep in mind that AF sensors may look similar to their O2 sensor counterparts but they function much differently.
Sensor voltage? Amperage?
The most important difference between the AFS and the O2 are how they operate internally. My advice would be to focus on how to diagnose these sensors and don’t get caught up in their overall operation. While they all achieve a wide-range sensor result they all are designed by different engineers and different manufacturers. Attempting to memorize sensor voltages is a bit overwhelming.
While the zirconia O2 was a voltage generator, the AFS actually outputs a current flow (amperage). When the engine is running at stoich there is no current flow. When the engine is running lean the current flow increases. When the engine is running rich the current flow polarity changes and you get a negative current flow value. The ECM then takes these values and converts them into a voltage. The voltage range will vary from manufacturer to manufacturer with no standard voltage for stoich. For example, Toyota uses 3.3 volts as stoich. Anything below 3.3 is rich and anything above 3.3 is lean. Each manufacturer’s voltage value can be treated the same.
The question I am always asked is which value should be used in diagnosis. The answer is whichever one you are comfortable with. In the Techstream screenshot provided I graphed both AFS voltage and AFS current. You will see the graphs are identical. Why? Because the ECM converted the current into a readable voltage value. Keep in mind the voltage output of an AFS is opposite to that of the zirconia oxygen sensor. Values higher than stoich voltage are lean and values lower than the stoich voltage are rich.
A simple solution
If you put me on the spot and asked me what the stoich voltage value was for each manufacturer I would probably have a hard time remembering them all. I try not to remember anything that I can look up. However, the real reason I have a hard time remembering is because I don’t use voltage or amperage for diagnosis. I use lambda. Lambda is such a perfect solution that manufacturers such as Toyota and Honda provide it as a PID. They even go as far as teaching their techs to use lambda instead of voltage or amperage. It’s really simple once you get the hang of it.
Lambda is a fantastic tool for diagnosing AFS and related complaints. Lambda is a ratio of available oxygen compared to the combustion processes’ demand for oxygen. Lambda is equal to 1.000 when there is balance between available oxygen and demand for oxygen. In other words, lambda is equal to 1.000 when the current air fuel ratio matches the desired air fuel ratio.
This formula is known as the Brettschneider formula, named for Robert Brettschneider who first proposed it in 1979 in a technical paper published by Bosch. While his original formula is quite complex, it is extremely useful in determining air/fuel ratios and imbalances. iATN.net provides a lambda calculator on its website. You can access it at MotorAge.com/lambda.
The lambda calculator will allow you to enter values from an exhaust gas analyzer to determine lambda. Why is this important? Well, if you had a misreporting sensor, there is really no better way than using the gas analysis and plugging the numbers into the calculator. While this doesn’t happen all that frequently on the Asian products, it is a good idea to keep this method in your back pocket for future reference. By the way, don’t get rid of that old gas analyzer!
Should you suspect that you have a sensor that is not reporting properly the best method is to plug your five gas values into the calculator. Once the numbers are plugged you can use your calculated lambda value to determine the actual running condition of the engine. For example, if your calculated lambda number is .997 It is pretty safe to assume that the engine is running at Stoich. If the calculated lambda value is less than or greater than 1.000 multiply the lambda value by 14.7. The product of your multiplication is the determined AF ratio. For example, if the lambda calculator value is .887, multiply .887 x 14.7. Your resulting AF ratio is just about 13:1 which is slightly on the rich side. For another example, suppose that your lambda calculator returned a value of 1.250. Again, multiply 1.250 by 14.7. The air fuel ratio is roughly 18.4:1 (lean). Do this a few times and you will have the hang of it. Always remember a lambda value of less than one is a rich mixture while a value of greater than one is a lean mixture.
If you are having trouble trusting the calculator you can always use manual testing methods to drive the mixture lean or rich and plug the values into the calculator to see the change in lambda value.
While I prefer to use a scan tool to do so, you can substitute pulling off a vacuum hose to create a lean condition. To drive the mixture rich the use of propane is another option. The bottom line with both methods of testing is that you are looking for a near instantaneous change of the sensor values. While most newer vehicles will code if a sensor is lazy, manual inducing a rich or lean condition can provide some peace of mind that the sensor is reporting rapidly.
As with any diagnostic routine, the best way to learn is by doing. Try these simple formulas and test methods on known good vehicles before trying to fix a problem vehicle.
λ= Current A/F Ratio ÷ Stoich
Current A/F Ratio= λ × Stoich
To use lambda in diagnosis you can simply take the lambda value and multiply it by stoich (14.7).
Stoichiometric Ratio
14.7:1
3.3 Volts
0 mA
1.00λ
Factory tooling
Toyota factory tooling is the least expensive factory tool on the market. A Techstream Lite Mongoose cable from Drew Technologies retails for around $500. Techstream software can be purchased through subscription at www.techinfo.toyota.com.
The Techstream provides a great active test for the AF sensor. This test provides the abilities of manually creating rich or lean conditions by changing the fuel injection volume or pulse width. With the scan tool you can now achieve the same desired conditions from the driver’s seat of the car. The active test allows you to toggle the air fuel ratios instantly rich or lean while graphing sensor response. This is my go-to test when I suspect a lazy sensor. Many aftermarket scan tools provide similar functionality.
To scope or not to scope?
Using a digital storage oscilloscope for AFS is not a realistic or worthwhile venture. Remember that the AF sensor outputs a current flow, and as such we would have to use an inductive current clamp to graph the value on the scope. Being that the sensor generates just milliamps, the use of a highly accurate low current clamp at a cost of over 700.00 needs to be used. My advice: take the $700 and take your significant other on a weekend getaway.
When it comes to diagnosing AF sensors it is important to remember that the manufacturer has gone to great lengths to build diagnostics into their ECM. Diagnosis should be done primarily with a scan tool. In the case of Honda and Toyota, these sensors are relatively reliable. Many of the early problem sensors resulted in published Technical Service Bulletins, and using TSBs as a first step in diagnosis is a best practice when fixing Asian vehicle fuel control faults. Always remember the K.I.S.S. method. Diagnostic routines shouldn’t be complicated. Use the info provided here to work smarter and not harder.
