Advanced air-fuel and oxygen sensor diagnosis
Portillo is the head technician of Car Clinic, a state-of-the-art automotive repair facility in Mahopac, N.Y. He has been trained by Automotive Technician Training Service and is TST certified. Portillo’s real-world, in-depth diagnostic articles appear in Auto Service Professional on a regular basis.
Not long ago, Auto Service Professional covered all of the basics concerning how to diagnose common oxygen and air-fuel sensor codes (see the March/April 2012 issue). This time, we are going to discuss more advanced sensor diagnosis so you can catch those really tough problems that are not simply a matter of checking the heater circuit and having a specification.
Very brief review
Air-fuel and oxygen sensors work in tandem, before and after the catalytic converter. The PCM compares the readings in order to analyze catalytic efficiency, and whether the vehicle is running rich or lean.
When the air fuel or front oxygen sensor senses a rich fuel mixture in the exhaust, the PCM takes that information and then tries to do the opposite to make a fuel mixture that is perfect (called “Lambda”) by sending fuel trims in the opposite direction. Air-fuel sensors reflect a lean condition when their voltage increases and a rich condition when their voltage decreases. Oxygen sensors work the opposite way, with an increase over 450 mV reflecting a rich fuel mixture and a decrease below that number a lean fuel mixture (see Figure 1).
Post-cat oxygen sensors, when good, feature a steady voltage usually between 500 to 700 mV. If it zigzags, the catalytic converter is highly suspect.
Quick air-fuel sensor check. Are you convinced the A/F sensor is stuck lean or rich, but don’t have the right specification? Until advanced milliamp clamps are mainstream where you would be looking for a specification of 0 amps (+ or –0.03 mA), you will have to either put a digital multimeter in series hooked up in the amps port. This is time consuming and putting the meter in series between the wrong wires can fry the PCM. A better method is to stick an emissions analyzer in the tailpipe. If the rear oxygen sensor has elevated mV (something in the 800 mV range) and Lambda is rich, you likely have an A/F sensor stuck lean. An A/F sensor stuck rich is much more rare, but theoretically can be approached in the same way (low rear oxygen sensor voltage and lean Lambda).
Tough P003X oxygen sensor codes. You get an oxygen sensor code, the repair is almost always an oxygen sensor, right? Well, if you have a code in the P003X (X equaling 1 through 9) range, you should expect that the vehicle has the wrong sensor or a module problem. This can be best illustrated in a case study. We received a phone call from a local shop about a 2003 Toyota Camry 2.4L four cylinder 2AZ-FE California emissions vehicle with a strange P0031 Oxygen Sensor Heater Control Circuit Low B1S1 (see Figure 2). This is how the story goes:
They get a “heater circuit DTC,” probably a different code number, checked the sensor and it had an open circuit. A new sensor from the dealer was ordered and installed into the vehicle, but the light stayed on. The resistance of the new sensor’s heater circuit appeared good, and the PCM was duty cycling the heater circuit nicely, which confirms the integrity of the wiring. The shop and all of its mechanics were stumped and decided to put a computer in it.
That’s where we come in. We were paid to reprogram the keys to the new PCM and when we did the check engine light still went on with the same code. Now, we had a diagnosis on our hands (see Figure 3).
On this vehicle, even though this particular code says “oxygen sensor heater circuit,” what they are really talking about is the front A/F sensor. By looking at the wiring schematics, this particular air/fuel ratio sensor is a four wire sensor with two signals wires from the PCM which provide voltage to the sensor. One of these wires is -3.0 (White) and the other +3.3(Orange). The third wire is a heater element circuit (Brown/Red) wire and finally a battery voltage wire (Brown/White). The E.F.I fuse supplies B+ to the E.F.I relay which then supplies battery voltage to the A/F sensor heater element when checking the B+ (Brown/White) wire (see Figure 4).
It is necessary to visually inspect fuses and relays nowadays for voltage drops situations. When we began diagnosing this car, we were getting 12 V key on. We also checked the fuse and relay which all tested good, so obviously that’s why we were getting 12 V.
After these checks we pulled out our scan tool to begin looking at PIDs. We noticed that the signal for the A/F sensor was a steady 3.29 V, which indicates almost a perfect air-fuel mixture as per the Toyota specification. Coolant temperature was 120 degrees F, STFT 0% and LTFT +2%. The fuel trims were practically perfect, which made sense given the A/F sensor data. Everything seemed normal, but our code did not indicate the sensor was not working properly. Instead, we had to dig deeper into why the heater circuit was throwing the code (see Figure 5).
There are only three possible causes. The sensor, the wiring, and the PCM. Being that it had both a new PCM and sensor from the dealership, it was time to whip out the labscope to check the integrity of the wiring and to see if the heater circuit was duty cycling properly.
How do we do this? We check power on the heater circuit from the PCM and at the sensor itself, and look at the difference. A voltage drop between the two can reflect a wiring or connection issue, while a lack of duty cycling at the PCM or the sensor itself can help us condemn an individual component (see Figure 6).
We did this test by hooking up the labscope to the black/red heater element wire to the air/fuel ratio sensor and to the PCM E9 connector pin No. 4. Then we started up the car and did our readings. What we found was a good PCM pulse width modulated control signal duty cycling between 0 to 14 V on our ATS EScope. There was no voltage drop between the two signals, as you can see in the picture, red and yellow are overlapping each other.
After we did this reading, it was no wonder why the other shop thought it was the PCM. Everything appeared to check out, lending plausibility to the theory that the PCM had some sort of internal logic error. As we have just shown, we had checked everything that was related to this heater circuit problem from checking powers and grounds, scan tool PIDs and using a labscope. Yet, all we proved is that we had a functioning A/F sensor.
Then, our patience paid off. We decided to keep the car running and let the labscope do its thing. Then, all of the sudden, the heater circuit was not pulsing between 0 to 14 V! It was just reading a steady 14 V. Then, the check engine light turned on. Can you guess why?
The car warmed up and went from open loop to closed loop! What the other shop was doing during their diagnosis is that they just stopped after taking their readings. Since the car was allowed to run a little longer, we caught this fault.
Let’s take a look at the scan tool when the car goes off of open loop mode to closed loop. It shows that at the fuel system No. 1 PID is “OLFAULT.” That means the engine computer cannot go into closed loop, so the PCM knows that the heater circuit element is not being commanded/grounded; but the PCM is smart enough to compensate for that issue to maintain its fuel strategy even though the A/F sensor is not operating as designed. OLFAULT is another term for “LIMP-MODE,” but for fuel control.
Why would the PCM go into this mode and not closed loop? It is almost impossible that two PCMs would do the same exact thing and a little too coincidental that when the vehicle wants to go to closed loop that the sensor would stop duty cycling. So we figured it had to be the sensor.
A brand new sensor?!? The heater circuit element showed 2.8 ohms of resistance, so the circuit itself was good. But, if you have a code in the P00 as opposed to the P01 range for the oxygen sensor, suspect that something is confusing the computer.
We went onto Worldpac speed dial to look up the part number and company of the OE sensor for a 2003 Toyota Camry with the same engine with CA emissions.
What we found was that the A/F sensor listed was different from what was on the car. Both sensors were Densos, but the OE sensor on Worldpac had four holes on the bottom and a skinnier top when compared to the one on the car. The car’s A/F sensor had only two holes on the bottom and a fat top. Who was wrong, the dealer or Worldpac?
Our advice? Never trust the parts guy.
We talked to the other shop’s foreman to tell him that he had the wrong sensor and he replied, “No way, it’s from the dealer!” We told him to humor us and check to see if he had two Camrys in the shop in his receipts. Less than 15 minutes later he called me just to let me know that it was the wrong sensor that they installed in the car! They had a 2005 Toyota Camry that needed an oxygen sensor on a future date and apparently they grabbed the wrong sensor off the shelf. That’s why the P0031 came back when the car was about to enter closed loop.
As you can see in the picture the sensors look totally different from the outside and surely the insides are different too, even though they are Toyota’s Denso sensors.
See Figure 7. The one on the left is the correct A/F sensor and the one on the right is the wrong one. Notice how the A/F sensor has four breather holes and the other only has two? Also, one is thicker than the other.
One last tip. Unless you have an ohms spec, do not be too sure that what you are reading is a “good” heater circuit. The right specification for this 2003 Toyota Camry I4’s A/F sensor is 1.1 ohms while the 2005 Camry’s sensor is 2.8 ohms (see Figures 8 and 9). Figure 8 is the specification on a 2005 Camry A/F heater circuit in ohms and Figure 9 is the specification for a 2003 Camry A/F heater circuit in ohms.
The lower the resistance, the higher the amperage, so this ECM has been programmed to work and command a pulse width on this heater sensor with higher amperage, but keep in mind that this reading was taken when the heater element was cold because that is when the heater element works harder. The hotter the sensor gets, the higher the resistance and the ECM detects this and adjusts the pulse width signal in order to get the amperage lower. The PCM detected something different than what it was designed to work with and set the P0031.
After we installed the correct sensor we had our scan tool connected to prove a point. When the engine was in open loop the A/F sensor voltage read 3.29 Vwith a STFT of 0 percent and a LTFT of 6 percent (see Figure 10).
The vehicle warmed up and the PCM went into closed loop and not OLFAULT! That means the vehicle was 100 percent fixed. We tested the sensor by going WOT.
The A/F sensor signal was at 2.35 V at 3,500 RPM (the engine becomes rich because of fuel being dumped into the cylinders) and then the PCM tried correcting itself and cut off fuel sending the A/F voltage to 4.16 V before settling back to 3.29 V.
Also notice how the short fuel trim was under 10% under load.
This indicates that the PCM is under fuel control in closed loop.
This same test was performed with the wrong sensor, but there wasn’t any change on voltage or fuel trims because the ECM was under OLFAULT (again see Figure 5).
Perhaps if we were smarter diagnosticians, we would have realized that this was not normal operation, but hey, we had to learn the hard way.
Now that we’ve been through it, you don’t have to experience the same grief. ●
Fuel trim control exhaust sensors
The exterior appearance of air/fuel sensors and oxygen sensors may be very similar but that’s where the similarity stops. Using a light bulb analogy, the oxygen sensor sends voltage to the engine control unit and acts like an on/off switch, turning the bulb on and off. The air/fuel sensor receives voltage from the engine control unit and acts like a dimmer switch, making the bulb brighter/dimmer. They differ in output characteristics, so although they may look alike, the two sensor types are not interchangeable.
Vehicles equipped with an air/fuel sensor have approximately 0.4 V constantly applied to the sensor which outputs a current that varies in accordance with the oxygen concentration in the exhaust gas. The ECU converts the difference in output current into voltage, allowing a response that is directly proportional to the present air/fuel ratio in the exhaust system.
The air/fuel sensor operates at a temperature of 1,200 degrees Fahrenheit, which is twice that of an oxygen sensor. Lean mixture indicates higher voltage on the voltmeter. Rich mixture indicates lower voltage on the voltmeter.
In vehicles equipped with an oxygen sensor, the output voltage changes in accordance with the oxygen concentration in the exhaust gas. The ECU uses this output voltage to determine whether the present air/fuel ratio is richer or leaner than the stoichiometric air/fuel ratio (14.7:1).
Light-off temperature between 572 degrees F and 752 degrees F is needed to quickly warm up the oxygen sensor. Three- and four-wire oxygen sensors have built-in heaters.
Lean mixture indicates lower voltage on the volt-meter. Rich mixture indicates higher voltage on the voltmeter.
NOTE: For more sensor diagnostic information and your own copy of DENSO’s sensor poster, go to http://www.densoaftermarket.com/posters/.