When an engine with coil-on-plug (COP) ignition starts to misfire, there are two challenges: finding out which cylinder is misfiring and finding out why. Even if you find a bad coil, simply replacing it is not the whole repair, because like so many other parts of a vehicle, COP ignition coils don’t really die, they’re murdered. We’ll discuss how and why later; first let’s focus on finding the misfire.
Ignition misfire diagnosis can be relatively quick and simple if you have the right tools. Fortunately, there are a lot of different “right tools” available, so it all comes down to the tools you know how to use. Most techs are comfortable using a scan tool, and if the malfunction indicator light (MIL) is on, a scan tool might be enough. But if the diagnostic trouble code (DTC) indicates a random misfire (P0300), the only thing you’ve learned is that the powertrain control module (PCM) “thinks” there’s an ignition-related misfire but it doesn’t have enough information to pinpoint the cylinder.
Remember, the PCM detects misfire with the crankshaft sensor: the crankshaft decelerates as a piston comes up on the compression stroke, then accelerates again on the power stroke after the cylinder fires. If it doesn’t accelerate as expected (per a very complex calculation), the PCM interprets that as a misfire and looks at all the input signals and output devices trying to determine the cause.
Trouble codes with a number lower than P0299 indicate the PCM has found a problem with something in the fuel control or air metering system. The codes from P0300 to P0399 indicate a malfunction in the ignition system, and while these codes can be very specific, any code can be misleading. For instance, P0316 indicates the PCM has detected an ignition misfire within the first 1,000 engine revolutions after start-up. This code doesn’t tell you the real problem... it merely indicates the result of the malfunction. So even though it’s a P03XX code, does the PCM really know enough to be certain it’s an ignition malfunction?
We found a 2006 Dodge Dakota with a 3.7L V-6 engine that consistently misfires on cylinder No. 1, but only for a few seconds immediately after a cold-start. The PCM stored ignition misfire codes for cylinder No. 1, but a detailed and thorough diagnosis shows nothing wrong with any of the COP ignition coils. As it turned out, the misfire is a result of the conditions under which it occurs: only at cold-start and only on that cylinder, plus the position of the engine in the vehicle.
When the engine cools off overnight, the fuel in the rail cools and condenses, leaving a small pocket of vapor at the highest point in the rail, right next to injector No. 1.
Naturally the misfire clears up quickly as the rail fills with fuel, and there’s no reason to chase this problem any further (how could you possibly fix it?). But the question remains: Why did the PCM report this as an ignition misfire?
The answer is in the rules that regulate how OBD II works. According to those rules, all the sensors and output devices in the engine management system must be monitored to make sure they operate as expected. A section of software in the PCM called the comprehensive component monitor (CCM) checks most items only after specific conditions are met (coolant temperature, drive time, etc.), but some items are monitored continuously any time the engine is running. This includes fuel injectors and ignition coils.
The PCM monitors the fuel injectors by looking for open or short circuits. It monitors the ignition coils by monitoring the current in each primary circuit to see if it rises to the correct level and then falls again in response to the firing command.
In our misfiring Dodge, the injectors and ignition coils both pass the monitor tests, but the PCM has detected a misfire, and it must be recorded. Since the injectors pass the monitor test, the misfire is reported as a malfunction in the ignition system. This case may be unusual, but it shows why ignition misfire codes don’t always mean there’s something wrong with the ignition system or that there’s nothing wrong with the fuel system.
Here’s a little-known fact that can help you isolate the problem. You already know that when a cylinder misfires continuously, the PCM will turn off the dead cylinder’s injector to avoid sending gasoline straight into the catalytic converter.
However, it doesn’t remember the misfire, as it discovers the misfire anew each time the engine is started. So the PCM will operate the injector during cranking and for a short period right after start-up until it knows for sure the cylinder isn’t firing. If you’re watching the injectors on a ’scope and they all operate during the first 200 engine revolutions at start-up, the injectors are OK.
Let’s assume the scan tool displays a more useful code, like P0351 (Ignition Coil A Primary/Secondary Circuit Malfunction). This code is usually set because the ignition coil doesn’t pass the CCM. Remember, the CCM monitors the rise and fall of current in the primary circuit, and remember that power is supplied to all of the ignition coils through the same circuit. That means the CCM only needs to monitor one circuit to look for current rise/fall in each ignition coil. (By the way, ignition coil A is the first coil in the firing order, B is the second, etc.).
The monitor will see no change in primary current if the winding inside the coil is open or shorted. However, a broken power or ground wire or a failed coil driver can all cause the same symptom and the same trouble code. Here’s where you need to know if the COP assembly has its own driver (transistor) that controls the ground side of the primary circuit. If there are just two wires connected to the coil, the driver is not in the COP assembly; it’s in a separate ignition control module or (most likely) inside the PCM.
If the coil has more than two wires, you’ll need a wiring diagram to determine how to test it. A coil with three wires has one for power, one for ground and one that carries the command signal from the PCM that operates the coil’s internal switching transistor. If there’s a fourth wire, that one sends a firing confirmation signal to the PCM.
Many techs will test a coil primary winding with an ohmmeter, but that only checks the coil itself when it’s cold. If you back-probe the (two-wire) coil’s ground circuit at the PCM connector and look for battery voltage with the engine warmed up but not running (KOEO), that tells you something about the entire circuit. But even these tests have limited value: If the primary and secondary resistance are both correct and the whole circuit is complete, that still doesn’t prove the circuit works properly when the engine is running.
By now you get the point that a scan tool and a digital volt ohm meter (DVOM) don’t always provide enough information for an accurate misfire diagnosis. Any ignition system can be affected by heat and engine load, so testing a coil with the engine running gives you a more complete picture.
There is a new generation of easy-to-use tools that will show without a doubt whether or not the coil is firing. They’re based on magnetic induction, the same thing that makes a coil work in the first place.
By simply touching the inductive pick-up to the top of the coil, the tool picks up the rise and collapse of the magnetic field in the primary circuit. Some tools will simply flash a light to indicate the coil is operating, some display data on a small screen, and some connect directly to an oscilloscope to display a waveform of current flow in that coil’s primary circuit.
Of course the very best way to test any ignition system is with an oscilloscope that can display data from all the cylinders at once. You don’t even have to know that much about what actually appears in the waveform; if one doesn’t look like the others, you’re that much closer to finding the problem. The problem is, if you don’t use a ’scope regularly, you’re likely to forget how to set it up when you really need it.
Remember, the best tools are those you’re comfortable using. It only takes a few minutes to find the fuse that supplies power to all the coils and tap a ’scope into the circuit. Do it often enough to recognize known-good and you’ll quickly recognize something that doesn’t look right on any engine.
Before connecting an oscilloscope directly to the primary voltage circuit, check the ’scope’s maximum allowable input voltage. The primary circuit can create spikes of more than of 400 volts under normal conditions. Most voltmeters can handle this, but most oscilloscopes can’t and you’ll need to connect an attenuator to protect the ’scope. That’s one reason we prefer to look at primary current with an amp probe. An oscilloscope also lets you look at the control signal on three- and four-wire coils. This is typically a square wave signal of about 4 volts that matches the timing of the current ramp in the primary circuit. If you see the command signal but no current ramp, you know the PCM is good but the coil driver is not responding.
What killed the coil?
As noted earlier, even though COP coils are known to fail with some frequency, that failure is usually caused by something outside the coil. The most common causes are worn or incorrect spark plugs, excessively lean air/fuel mixture and liquid getting into the spark plug tubes.
If primary current level and dwell time are correct, a coil will generate enough voltage to meet almost any demand. A healthy ignition coil with old worn out spark plugs might develop an initial firing voltage of 80 kV or more.
That’s the coil doing its best to keep up with the demand, and for a while it will. But when a coil works that hard, the secondary winding overheats from generating that much voltage, and eventually the heat will damage the winding or the driver transistor. The coil will either begin to misfire when hot, or it will fail completely, or it will damage the driver.
To understand how hot a coil can get, look up Ford TSB 13-4-17 or 11-8-2. They include pictures of ignition coils that melted because the coil driver inside the PCM shorted to ground and kept the primary current turned on continuously (primary current normally lasts about 5 milliseconds). In case you’re not familiar with this (infamous) issue, the fix is to replace all the ignition coils and the PCM, which requires reprogramming the PCM, which in turn requires two original ignition keys to reprogram/reboot the anti-theft system.
Combine a worn out spark plug with lean air/fuel mixture, and that 80kV is going to find an easier path to ground. Even the best insulator boot can’t contain it indefinitely; at first the spark will leak through the boot to the valve cover only during acceleration, but eventually it will happen all the time. That’s why boots and connector springs are available separately on some models, so they can be replaced when installing new spark plugs. If the old spark plug has a carbon track on the ceramic, that boot should be replaced because there’s a matching track in the boot that offers an easy path to ground.
If there’s liquid in the spark plug well, even a new insulator boot might not be able to contain the secondary voltage. Chrysler issued a recall for the 2004-06 Dodge Durango (18-024-06) to replace ignition coils that were shorting through the (non replaceable) boot to the valve cover. The repair included installing a redesigned windshield cowl to keep rain water out of the engine compartment.
Other vehicles have had similar problems, but not just because of weather. Leaky spark plug tube seals, coolant leaks and even water or mud splashed up from below have all been known to cause this problem.
So now we’ve seen how the PCM detects and reports misfire, and we’ve shown that COP misfire diagnosis can be relatively quick and simple if you have the right tools. We’ve also discussed the importance of finding out what damaged the coil, because COP ignition really is simple and reliable as any other type of ignition system. ■
Jacques Gordon has worked in the automotive industry for 40 years as a service technician, lab technician, trainer and technical writer. He began his writing career writing service manuals at Chilton Book Co. He currently holds ASE Master Technician and L1 certifications and has participated in ASE test writing workshops.
To read more of Jacques Gordon's articles, click: