Lambda diagnostics: Solve those system lean problems fast
Craig Truglia is an ASE A6 and A8 certified technician who presently works as a service writer for Patterson Autobody, a repair facility in Patterson, N.Y. A former shop owner and editor of several automotive repair magazines, Truglia combines his Columbia University education with the real-world experience he sees daily in the automotive repair field.
Diagnosing lean problems has become increasingly complicated over the years. Back in the day, all one had to do was adjust a carburetor. Now, thanks to the increased computerization and concurrent lack of standardization in the automotive industry, it requires the mastery of several different systems, with many manufacturers having different nuances.
However, one thing has stayed the same over the years: Lambda. Lambda never changes and it always represents fuel system perfection. If we understand Lambda, no matter how much the sensor feedback technologies end up changing, we will be able to adjust and diagnose vehicles.
What is Lambda? Lambda is a perfect air-fuel ratio in which there are practically no unconsumed hydrocarbons (HC) in the fuel. Now, in reality, no engine is going to burn absolutely perfectly, which is a big reason why even in “good” running engines there will be more HC pre-cat than post-cat, but we are talking about a difference of a few parts per million (PPM) of HC.
For all intents and purposes, if you have a Lambda of 1.0 you have a perfect running engine. If you go below 1, you begin running rich. If you go above 1, you run lean. Anything within 0.97 to 1.03 is normal, but if you go above these numbers and the vehicle has a code for fuel trim or a converter issue, it is worth taking a closer look. However, don’t be hyper sensitive. If the vehicle is running fine and has a Lambda 1.08 or 0.95, that could be “good enough.”
Just remember how it works: above 1 is lean and below 1 is rich.
Lambda and oxygen sensors
Most old school technicians are familiar with the old Bosch oxygen sensor method of Lambda feedback. Then, the oxygen sensor before the catalytic converter essentially gave the computer an idea of the engine’s air-fuel ratio. Obviously, the idea is to get it perfect, i.e. 1.00 Lambda.
Perfect Lambda in an oxygen sensor system is 450 mV. However, the PCM is constantly making adjustments in fuel mixture to get it just right. So, good oxygen sensors have even waves in the 150 mV to 850 mV range while ascending or descending within a 100 mS or less when the system is in closed loop. Now, if the average of the waveform is more than 450 mV the vehicle is running rich. If the average is less than 450 mV, this reflects a lean condition.
For example, it is not uncommon that a prolonged vacuum leak or exhaust leak forces a sensor lean so long, that even when these conditions are repaired the sensor will constantly read 0 V, forcing the engine perpetually lean. If adding propane to the fuel mixture or going wide open throttle doesn’t force the sensor rich and it stays where it is, the sensor needs to be replaced.
A lot of technicians find oxygen sensors pretty easy to understand because their voltage goes up as the amount of fuel is up and the voltage goes down when fuel goes down. It is simple conceptually, but be aware that this is the opposite of Lambda.
Lambda and air fuel sensors
Unlike oxygen sensors, air fuel sensors get it right. They go up in voltage when the fuel mixture leans out and they go down when fuel mixture gets richer.
To know exactly what you are looking at, you will need to know the voltage specifications for the air-fuel sensor on your scan tool. On newer vehicles, the O2 Sensor B1 Parameter ID in datastream will give us the accurate voltage. In a previous article is Auto Service Professional, the following specs were given: 3.3 V (Toyota), 2.8 V (Honda), 1.9 V (Hyundai), 2.44 V (Subaru), 1.47 V (Nissan), 1.00 Lambda (all European manufacturers).
However, as we move into the unveiling of 2015 vehicles, some of these specifications are going to become out of date. What can a technician do in this situation?
The answer is a tried and true tool: the emissions analyzer. While the specifications on an air fuel ratio sensor can change over the years, making your job to ascertain if the vehicle is running rich or lean much more difficult, the emissions analyzer always will have the same known good specification: 1.00 Lambda.
Your emissions analyzer won’t lie to you. It can be calibrated unlike on-board air fuel ratio sensors. Further, it doesn’t fail due to temperature related faults. We will see how this is relevant in just a little bit.
Test vehicle: 2007 Hyundai Elantra
2.0L P0170, P0171, P2195 and P2414 DTCs
This vehicle was not easy to diagnose because it “broke all the rules.”
Bernie Thompson of Automotive Test Solutions has made the observation when working on Hyundais that they are the only vehicles on the road whose MAF sensors don’t follow the 1 gram per liter at idle rule. For example, a vehicle with a 2.4L engine and a good MAF sensor should read about 2.4 grams per second (GPS) of air, or a little more. If there is a lower reading, say 2.1 GPS, that might reflect a volumetric efficiency issue such as a dirty MAF sensor not picking up all the air entering the engine.
Now, the point is that on Hyundais, a reading of 2.4 GPS would not cut it. In fact, while unlike every car from Ford to Mercedes such a reading would be good (though readings can often be higher), a Hyundai should have a significantly higher reading, such as 4.0 GPS.
This vehicle came in with very bad fuel mileage. The goal was to fix it for as little money and time as possible.
As with any lean issue, the first thing the technician should look at is fuel trim and compare it to the front air-fuel sensor. However, on this vehicle this actually put us in the wrong direction.
On this Elantra, the vehicle when it started up would have 48.4+ STFT and the air-fuel sensor read 4.9V.
You don’t need to be a rocket scientist to figure out that this vehicle was running really lean. When going wide open throttle, the calculated load when to 100%, which indicated a good MAF sensor. Further, the front air fuel sensor’s voltage would spike downward as the injectors dumped fuel.
Eventually, as the vehicle became hotter, both the STFT and the front air-fuel ratio sensor voltage declined to the point where STFT became 0.
Further, the front air fuel sensor settled at 2.8V. LTFT was 25+, which is lean, but the STFT was good! It appeared to me that the vehicle was really lean around start up and running fine once it got hot.
At this point, this is where my mind started playing tricks on me. I knew the correct specification for Hyundais, but the known good specification was from a 2008 Hyundai Elantra. “Maybe,” I told myself, “the 2007 Hyundai specification is different. The STFT is 0 when the air-fuel read 2.8V!”
A quick look in a parts catalog might have confirmed that both the 2007 and 2008 Elantra AFR sensors had the same part number. However, being that I only had access to a generic OBD II scan tool and not the Carman Scan at the time, which has OE capabilities on Hyundais for model years up into early this decade, how did I know my scan tool was giving me the right reading? Perhaps, like on older models, generic OBD II was giving me an inaccurate front AFR sensor voltage reading.
Scoping the front air-fuel sensor would have given us the right reading either way, but hey, we were being lazy. We figured if STFT is 0%, then the steady air-fuel ratio voltage had to be good.
With the engine cold, the intake was smoked and no vacuum leaks were found anywhere. The sensor would go lean and rich by creating a vacuum leak and adding propane. However, we noticed during such testing that the air-fuel sensor appeared to be working, but the STFT did not correspond with changes in the air-fuel mixture when the vehicle was hot. When the vehicle was cold, the STFT worked normally.
Without an emissions analyzer and a proper respect for the craziness found in Hyundais, the vehicle appeared to be violating a Law of Short Term Fuel Trim nature: STFT always responds to movements in the oxygen/air-fuel sensor.
First, we did a soft reset of the PCM by touching the battery cables to each other with a jumper wire for a few minutes. When this didn’t work, we decided to get a cheap used PCM and plug it in. It didn’t need to be reprogrammed and guess what, the vehicle still did the same exact thing.
Thankfully, we just received an emissions analyzer from ANSED and were able to put the emissions to the test. The emissions were really clean, but that did not concern us. When the vehicle was cold, the Lambda was significantly above 1 reflecting a lean condition. When it warmed up, even though STFT stood stuck at zero, the fuel mixture reported lean, around a 1.235 Lambda. We saw a reading of 100 HC and 1.235 Lambda when the vehicle was warm on the ANSED emissions analyzer. When cold, the converter would clean up less HC and the number would be in the hundreds. Being that the system has no vacuum leaks, exhaust leaks, and a good MAF sensor, our eyes went back to the front air-fuel sensor being stuck lean. The new sensor dropped the AFR voltage down to 2.0 V and a few days later, the vehicle was able to pass a state inspection. Plus, the fuel mileage gauge reported gains!
Conclusion: While an understanding of fuel trim can allow for some quick diagnostics, ultimately unless we know how to read Lambda and have an emissions analyzer, there won’t be any vehicles that we won’t be able to diagnose unless we have specifications we are totally confident in and a factory scan tool. As it often takes years for these specifications to trickle down into our hands, and often it isn’t realistic that we can test every new vehicle that rolls into our shops for them, a knowledge of Lambda and the ability to measure it will be necessary for years to come. ●
Want to read more technical articles? See the entire May/June 2014 issue of Auto Service Professional by clicking on our digital version here.