Diagnosing misfires and cat codes on today's cars requires information specific to the vehicle you are testing.
When I first started in this business, you could diagnose most running problems based on a generalized set of rules. Of course, this was in the day of carbureted engines and mechanical ignition systems. Today, the Engine Control Module (ECM) does a pretty good job of keeping an eye on things. While there are general methods used by the ECM for testing, the specific testing procedures programmed by an individual manufacturer can have an impact on how you proceed with your diagnosis and repair.
After all, if the ECM considers a test as "failed" and sets a code, your repair must meet the requirements for that particular system. Otherwise, the MIL light will come back on.My 'Check Engine' Light Is 'On'
Today's subject is a 2002 Mitsubishi Lancer that came in with a "Check Engine" light complaint. There were two codes stored in the ECM. The first listed was a P0421, a catalyst efficiency code, and the second was a P0304, a cylinder No. 4 misfire. Considering the mileage on the car, it certainly was possible that the catalytic converter was bad. And we all know that misfires can quickly overheat the cat and damage the substrate. But is that what the ECM is testing for? Time for a little research.The Cat Efficiency Monitor
Generally, the ECM tests the efficiency of the catalytic converter by gauging its ability to store oxygen, which the cat needs to burn off the emissions left over after the combustion process. This test typically relies on input from the pre-cat oxygen sensor and the post-cat sensor. Right off the bat, we know that these two sensors have to be in good shape. After all, the ECM's decision is only as good as the information it receives. I've seen several "false" cat codes caused by the recent replacement of a post-cat oxygen sensor combined with an aged, lazy pre-cat sensor.On this vehicle, all of the methods used are an example of a functional test. The test(s) can be intrusive or passive and can occur at idle or at cruise. I would like to know how this ECM runs its tests so I can try and duplicate it in the bay or on the road. Looking up the information on the Diagnostic Trouble Code (DTC) Set conditions, I learn that this monitor has to have the following conditions met before it will run its test:
- engine rpm less than 3,000
- MAF frequency between 63-169Hz
- intake air temperature over 14 degrees Fahrenheit
- barometric pressure over 11 psi
- throttle open
- ECM in closed loop
- vehicle speed over 0.93 mph.
Wait A Minute!
Because this monitor relies on the sensors' input, I also have to consider any other factor that would affect the sensor switch rate. Catalytic converters need the correct feed gasses coming in to work properly, so any factor that would cause these feed gasses to be abnormally rich or lean could impact the ECM's test.
Let's see, there was a misfire code stored as well. Could that have an impact on this monitor? Sure it could. Every time that cylinder misfires, a short burst of unused oxygen, raw HC or both (depending on the cause of the misfire) is going to get to the converter and likely cause a switch in one or both sensor(s) in response. I think we better fix that first before condemning the converter.
It was obvious from the visual inspection that this car was not maintained, so I decided to pull the plugs first and take a look. The plug gap was so huge I could drive a truck through it. But that wasn't all. Leaking plug tube well seals were allowing oil into the plug wells, and there were definite signs of arcing down the body of the plugs themselves. OK, first order of business is to stop the leaks and install new plugs and plug boots/wires.Looking a little more, I noticed a gray dusting on the side of the ignition coil for cylinders 1 and 4. This engine uses a DIS ignition system, with the two individual coils mounted over one pair of cylinders and a spark plug wire leading to the mating cylinders. I know that worn secondary components can stress a coil and lead to its failure, so I thought I'd better test it to be sure. Using a spark tester, I installed it in the wire going to No. 4 while leaving the No. 1 plug in place and started it up.
While the spark did jump the maximum gap I had set, it didn't last long. The car died within seconds of starting, and that wasn't typical for this test. I swapped over to the No. 1 plug wire and ran it again with the same results. I was really beginning to believe that this coil was weak. A quick test with the scope and a current probe confirmed it. The coil was not producing all the energy it should. Let's add a coil to the parts list. I also recommended the second coil for preventive maintenance, figuring that if one was damaged by the weak secondary, the second was right behind it.
To be on the safe side, I checked the manifold vacuum to get an idea of the mechanical integrity of the engine. There was no abnormal response either at idle or while performing a "snap throttle" test. Looks like the engine itself is OK. I had no reason at this time to check the fuel system, so I left that alone for the time being.Now that I had an idea of what was causing the misfire, I recommended that this fault be fixed before looking at the converter code. It was likely that the code was real, considering the misfire issue and the mileage, and I found it a little odd that the customer hadn't complained about the obvious drivability problems.
After The Miss
With the oil leaks fixed, fresh coils and secondary ignition in place, it was time to check that cat code. Since we now know how the ECM tests the cat, we can design a few tests of our own to check it out. One way is to clear the codes, and by doing so, reset the monitors. We could then drive the car until the monitor is complete and recheck the results.
It just so happens that I took a look at those test results while retrieving the codes. The cat monitor is a non-continuous monitor, and the individual test results for it and all the non-continuous monitors are found in Mode $06. Mode $06 has come a long way in a short time, with test definitions and even conversion factors listed in both major aftermarket service information systems. If you can't find it there, you can always go to the manufacturers' Web site. You can find a listing of them at www.nastf.org and www.oem1stop.com.
Another test option would be to use a scope or scan tool to graph the action of the sensors while driving the car in accordance with the manufacturer's drive cycle. But this would mean I either had to run extended leads into the car or make sure my scan tool PIDs were set to read only what I wanted in order to get a good refresh rate on the data.
Since the ECM is the customer I have to satisfy in repairing the cat code, I figured I'd let it do the testing. I reset the monitors and took the car for a drive. Keep in mind that each monitor has its own defined drive cycle: the operating conditions that must occur in order for the tests to be completed. Some monitors can complete a short while after engine-start, with just the engine idling. Others, like many EVAP monitors, require a very specific "cold soak" followed by specific driving parameters to complete. This is why it is possible to repair a code, only to have the car return weeks or even months later with the same code reset. It's not enough to just drive around the block and see if the MIL comes back on.
Unfortunately, age and lack of maintenance resulted in this cat's early demise. That wasn't too much of a surprise. Typically, the cat efficiency monitor is a good judge of cat health, but you still should take the time to verify that there are no other contributing factors before condemning the cat. I always recommend a converter equivalent to the OEM. Many aftermarket "universal" cats do not have the same oxygen storage capacity as the original and can fail the monitor's tests after only a few months of use. They're just not worth the savings. The new cat was installed, the repair verified and the customer sent on their way, satisfied.
