Vehicle was brought in with a concern of rough running. Step one on any diagnosis is to confirm the complaint during a complete test drive. During the test drive, the engine feels like one or more cylinders is misfiring. There is also a steady check engine light.
The first thing we need to do is to hook up our DrivePro and check for codes
We can see the codes 29CD-COMBUSTION MISFIRING CYLINDER 1, 29D1-COMBUSTION MISFIRING CYLINDER 5, and 29CC-COMBUSTION MISFIRING SEVERAL CYLINDERS.
The customer explained the other codes are pre-existing and will be diagnosed later.
Opening the hood, we are confronted with this mess of covers and plastic…
Once the engine covers are removed, swapping plugs and coils on this engine is not difficult. However, to simply replace swap a coil to diagnose, we miss out on understanding the failure and having complete confidence in our diagnosis.
An automotive scope will allow us to monitor and review the voltage and amperage of this ignition system over time. We will be able to see issues and trends before the engine computer can pick them up.
For this vehicle, I want to monitor the amperage of the ignition coils while triggering off a known cylinder. We can accomplish this by finding a central voltage or ground point for the ignition coils. We will also connect to the DME control to the cylinder 1 ignition coil. By triggering off cylinder 1, we can identify remaining cylinders in our scope pattern using the vehicle firing order.
Using a wiring diagram, identify cylinder 1 primary circuit, indicated by the blue arrow. Paired with the engine firing order, we can tell when each cylinder’s coil get fired.
The orange arrow points to the main power feed for all the coils (spliced). If we hook our amp clamp here, we will be able to see power flowing through all 6 coils.
The amp clamp is connected as shown and the setting used is . We inserted a backrpobe into cylinder 1 for out trigger. With our Pico Scope set up, channel A is set for 100v dc, and channel B is set for -5 to 20amps dc. The timebase is set to 1 second per division and we are sampling at 4 million samples per screen. Setting the scope up this way will allow us to collect a lot of data on a screen and also allow us to see any trends. We will also be able to zoom in and review each coil firing event.
We only captured about 5 ½ seconds of data, and that’s all we need. If I select an entire 720 degrees of engine rotation, we will be able to see all of the coils firing…
All 6 coil firing events are now on screen in order. We are using the number 1 ignition coil. This paired with a good know firing order we can make sense of this mess. One thing to note, on BMW vehicles we expect to see a multistrike ignition event during engine starting and below 1500 RPM. This capture taken at idle shows one ignition event and an additional strike. Normally, with no misfire there would be 5 events after the initial ignition event.
This is the engine firing order: 1-5-3-6-2-4. Let’s apply this to the scope capture…
Look at the number 1 firing event. It looks different than all the others. Let’s take a closer look at this event…
Look at the steep amperage rise at the beginning of the coil firing. You can see that it goes from 0 to 2.8 amps immediately. If you are a veteran scope user, you would identify this as a shorted ignition coil and end your diagnosis.
What if you were not?
This is a 6-cylinder engine with codes for 2 of the cylinders. We now know which cylinders the computer thinks are the problem. This also means we have 4 known goods to compare. Let’s take a look at a good one from this engine…
Look at the arrow here. What you see is a nice gradual climb in amperage. This is not going to be a lesson on the inner workings of ignition coil primary and secondary windings, just a quick lesson on comparison. We now know we have a bad number one ignition coil
We had a code for cylinder 5 as well. Let’s take a closer look at the coil capture…
Cylinder 5 looks like all the other good ones. That cannot be, our scan tool told us it was misfiring. It must be a problem. Let’s take a closer look…
This waveform looks beautiful. It has a nice gradual amperage climb and all of the good indicators of a healthy ignition coil.
We know we have a bad number 1 coil. Let’s change it and recapture the coil pattern…
Not only is the vehicle running smooth, the coils are now all firing multiple times at idle, as expected. This is how this system operates when it is working correctly. This is another indicator that we fixed the issue.
After a code clear and test drive, the misfire codes are gone. What about cylinder 5? Why did the DME set a code for cylinder 5 misfire?
This would take some deductive reasoning. This engine computer does not know that a cylinder is not firing. It detects misfires by reading the speeding up and slowing down of the crankshaft through the crank sensor. A cylinder that is not contributing will allow the crankshaft to slow slightly and the next cylinder will speed it back up. Since cylinder 5 was next in the firing order, it was flagging a misfire as well. Once we repaired the coil, it is no longer seen as misfiring.
An engine computer also assumes everything in the engine is in peak condition. Imagine if you have a little slop in the chain, the computer may not be able to correctly pick up which cylinder is not fully contributing to the party.
The moral to this story is you always need to test. We as technicians have become too comfortable with letting our scan tools to the work for us. This can be understood because scan tools have become far better then what they used to be. A lot of us forget one of the most important lessons when diagnosing vehicle systems: read the data, then verify the data.
A computer is a device that only knows what it is told. It can only make decisions on the inputs going to it. We always need to be verifying the inputs to make sure the computer isn’t lying to us. Taking these steps will aid in making your diagnosis solid.