Using fuel trims and downstream O2s to analyze intermittent misfire concerns

Editor's note: This article was originally published May 28, 2013. Some of the information may no longer be relevant, so please use it at your discretion.

Let me share with you a technique that will give you a quick and effective way to capture the concern and determine, with a high level of certainty, an analysis path. It uses your scan tool to watch the fuel trims and the downstream O2 sensors. I began teaching this technique to Ford technicians in the Advanced Drivability courses starting in 2000 as a Service Training Instructor with Ford Motor Company.

To understand why this method works, let’s lay the foundation by reviewing Powertrain Control Module (PCM) feedback fuel control. First, the conventional zirconia dioxide sensor is the feedback to the PCM reporting the exhaust stream as rich or lean with a stoichiometric switch point of 450mV. In operation, it is more accurate to say that the sensor is a switch and that stoichiometric (the “ideal” air/fuel mix) is a range that goes from about 350mV to 650mV. A voltage above 650mV then represents a rich (low oxygen (O2) and/or high hydrocarbon (HC) content) air/fuel mixture in the exhaust. A voltage below 350mV represents a lean (high oxygen and/or low hydrocarbon content) air/fuel mixture in the exhaust. A voltage reading between 350 and 650mV is a stoichiometric (14.7:1) air/fuel mixture in the exhaust.

When we monitor the conventional oxygen sensors’ signals, we see that at a steady state operation the typical vehicle’s upstream O2 will be constantly switching rich-lean. With a working cat during steady state operation, the downstream oxygen sensors will be holding steady, usually somewhere in between 350 and 650mV.

Testing has also shown that as a rich-lean or stoichiometric switch, the operation of the sensor does not change based on combustion. In other words, if a stoichiometric air/fuel mix entered the combustion chamber, a stoichiometric air/fuel mix comes out. If a cylinder does not fire due to an ignition concern, the effect on the upstream sensor will be minimal. If the cylinder does not fire due to a lack of fuel, the effect will be pronounced. Note: There is what is called the Characteristic Shift Down, which comes from a sensor being contaminated by an excess amount of HC in the exhaust stream if the over rich or misfire condition lasts long enough.

Getting the switch to switch

To get a conventional oxygen sensor to switch, the PCM drives the mixture rich to lean to rich. This can be seen with a scan-tool that will graph the Short Term Fuel Trims (STFTs). The pattern will be a saw tooth that is just opposite of the sensor voltage swing. By watching the STFT values, you also will see that they will be switching in a ±4 percent range (on a known good car, of course). If the fuel trims spend about an equal amount of time adding and subtracting, then the overall fuel mix is stoichiometric.

Long Term Fuel Trim (LTFT) is the adaptive strategy side of the system and those values will be used for baseline purposes in this analysis. Under normal operation, we expect these values to be within 10 percent of zero. While it takes LTFT in excess of plus or minus 25 percent to set a rich or lean Diagnostic Trouble Code (DTC) and illuminate the Malfunction Indicator Lamp (MIL), I consider anything in excess of ±10 percent to be a concern that requires investigation. Possibilities include contaminated oil, slight vacuum leaks, or a contaminated Mass Airflow sensor (MAF) and are all things that should be looked into.

Add in the converter

To finish the foundation, let’s review engine and catalytic converter operation. When STFT is switching in a ±4 percent range, a stoichiometric air/fuel mix is entering and leaving the cylinder. That is critical to understanding operation and using this technique to analyze intermittent misfire concerns. It is the combination of STFT and LTFT that tells us what entered the cylinder.

If the air/fuel ratio coming out of the cylinder is lean, the sensor will stay at a low voltage. Keep in mind that lean can be too much air or not enough fuel; rich is just the opposite. Conversely, the voltage will stay high if the air/fuel mix is rich. The PCM will decrease or increase fuel pulse width until it gets the sensor to once again switch. Once the sensor is switching, the STFT goes back to switching in a ±4 percent range again. Adaptive strategy will start and the PCM will learn what the new normal is and LTFT will change so that the STFT switch range is above and below zero.

Last part of the system to consider is the operation of the catalytic converter and the downstream oxygen sensors. A normally functioning cat will take excess O2 that is in the exhaust stream and store it. It will also capture the HC and CO and store it. (NOx emissions and cat operation does not effect this diagnostic.) During catalytic operation; the O2, HC, and CO are used and we have H2O (water) and CO2 (carbon dioxide) coming out. There is a balance during steady state operation and the downstream sensor will also indicate whether or not a stoichiometric A/F left the cylinder(s).

It is typical to see the downstream sensor voltage floating in the 350 to 650mV range when the cat is lit off, the system is in closed loop and there are no misfires or other concerns. In fact, most manufacturers monitor the downstream oxygen sensors and will tweak LTFT to improve fuel

economy and reduce emissions. That is one reason your car’s fuel economy improves during steady-state highway driving.

Using this info to fix a car

Now that we have set the foundation, let’s use this information to figure out what is causing an intermittent misfire.

If your vehicle has a P03xx code, then by all means follow the pinpoint test related to the code. If not, then we will start by getting a baseline on the vehicle. Baseline the vehicle means to note what the LTFT and STFT values are without the misfire present and note where the downstream oxygen sensor voltage is.

The car needs to be running in closed loop as most vehicles limit datastream in open loop operation. Of course, if the car has gone into open loop, there should also be a MIL code or at least a pending MIL code.

Another piece of data to review is the misfire data. Ford does not list misfire counts in their data stream but useful misfire information can be found in Mode $06. This information will give you a sense for which cylinder(s) have a concern. They may not give a direction, such as fuel or ignition but it will at least point you towards a specific cylinder.

Let’s consider what happens in a “good” cylinder in closed loop. A near stoichiometric A/F ratio enters the cylinder. The plug fires, the mixture burns, and the exhaust is expelled out and into the exhaust system on its way to the catalytic converter. The cat absorbs any excess O2, HC and CO and then processes that into H2O and CO2. This exhaust stream then passes by the downstream O2 and out the tailpipe. As this happens, the upstream O2 switches and outputs a varying voltage between 100 and 900mV. The downstream O2 will stay fairly steady and float between 350 to 650mV.

Now let’s take a look at what happens when things don’t go as planned.

Scenario 1 – Ignition misfire

In closed loop operation, the PCM has created a near stoichiometric A/F ratio in the cylinder. When the sparkplug doesn’t fire that A/F mix enters the exhaust stream. Some fuel condenses out in the cylinder and goes into the oil. This is why the O2 sensor gives an initial lean reading. However, most of that stoichiometric air/fuel mix enters the exhaust stream, goes past the upstream sensor and into the cat.

Anytime there is a change in the A/F ratio sensed by the O2s, the STFT will react and begin to add or subtract fuel to get the O2 to switch. With an ignition misfire the STFT typically will only need to add up to 15 percent to get the O2 to return to switching.

The downstream O2 will show an increase in voltage and go into the 900mV to even 1,000mV range. This happens because the cat has to try and clean the exhaust and so it uses all of the stored O2 to attempt to oxidize the unburned HC entering from the misfiring cylinder. It is unable to oxidize all of the HC and uses most to all of the excess O2 that came with the HC from the cylinder that had the misfire.

The cat is trying to clean the exhaust and therefore the downstream O2 sees a rich exhaust stream, low O2. It will output a voltage above 650mV and usually above 900mV during the event. The one above is at 980mV.

This happens so consistently that you can look at the datastream during the misfire event and if STFT has changed less than 15 percent positive and the downstream oxygen sensor has voltage higher than 650mV that you can with confidence look for an ignition related concern to be the cause of the misfire.

To prove this to students and myself I have intentionally created this concern on many cars over the years to assess the reliability of the test. I have yet to find a vehicle in which this was not seen. Even vehicles with upstream wide band air fuel sensors react the same way.

My observations do show, however, that sensor placement in the exhaust system has an effect on this method. When the manufacturer places the upstream oxygen sensor farther down the exhaust system the effect of a misfire tends to be dampened. As the exhaust streams from the other cylinders mixes with the stream from the misfiring cylinder the change in fuel trims tends to be less.

One thing you may have to watch out for is when the OEM has built in some type of “limp in” mode that impacts the datastream. In this case of a Lincoln Towncar, the PCM determines a fault and goes into what Ford calls Failure Mode Effects Management.

Scenario 2 – A lack of fuel misfire

In the case of a misfire caused by an injector that does not spray fuel the reaction of STFT and the downstream O2 is significantly different. If we have a miss caused by no fuel or little fuel due to a mostly plugged injector, then the mix in the cylinder is mostly air. Isn’t that the same as a lean air/fuel mix? This cylinder full of air is then pushed out into the exhaust stream, past the upstream O2s and into the cat.

In this case the reaction of the STFT will go well beyond 15 percent. Because there is an excess of oxygen going into the cat, the cat has more than enough O2 to oxidize what HC and CO enters and have O2 left over. The reaction of the downstream O2 is to produce voltages below 100mV. In this example the voltage produced is 8mV.

The rule of thumb for a fuel-related misfire concern is STFT going more than 15 percent positive, often into the mid 20 percent range or more and the downstream O2 voltage below 350mV.

The downstream oxygen sensors also can be also used in a similar fashion to analyze concerns such as a fuel pump losing pressure or volume. As the volume drops off and the system goes lean the downstream sensors will go lean as well. Because both banks will be similarly affected you can rule out a single cylinder concern as the cause of the lean running condition.

Keep this tool in your memory toolbox for the next intermittent misfire you’re faced with:

• A 15 percent or less change in trim and a downstream O2 above 650mV points to an ignition concern
• A 15 percent or more change in trim and a downstream O2 under 350mV points to a lack of fuel concern

These scenarios are fairly easy for you to recreate in your shop to see the results and prove to yourself that this method is repeatable to the point of being a realistic method for analysis.

Subscribe to Motor Age and receive articles like this every month…absolutely free. Click here