Select a technical image to view it in detail.
On first sight this diagnostic case looks very simple, almost too simple.
Fault codes and symptoms
The engine is an N43. The DME fault memory holds two entries:
- 29CE: misfirings, cylinder 2
- 2A2C: fuel mixture control, bank 2
There is no freeze-frame data, because the simplest “solution” to a problem was applied beforehand: the fault codes were deleted. No fault code, no problem. The symptoms were still there, though: the engine vibrates at idle.
Cylinder performance
Let’s look at the performance of the individual cylinders:

The marked cylinder is No. 2 (by position). Its efficiency is strongly reduced, to only about half of what it should be. This menu lists the cylinders in firing order, 1/3/4/2. The other cylinders all perform identically, so they are evidently in good order.
At first glance the problem seems solved. There are fault codes for cylinder 2, and cylinder 2 really is performing poorly. The mixture on both banks is also correct, as the bottom line of the menu shows: the Lambda probe voltages are very close to the ideal 2.00 V. So what happened to cylinder 2?
If it were that simple, I would not be writing this entry. Let’s look at what is happening with the fuel mixture, since the DME also has fault codes stored for it.
Fuel mixture adaptations

The first screen already shows two serious problems:
- The LTFT multiplicative adaptations are close to their minimum (-30%).
- The STFT (integrator) is also close to its minimum (-30%).
Both are on bank 2, which consists of cylinders 2 and 3.
The third problem is also on bank 2:

The LTFT offset adaptations have reached their minimum (-1.0 km/stk).
What do these numbers mean? One of the injectors on bank 2 is leaking, and leaking strongly. The DME has cut the fuel supply using every LTFT tool available, both offset and multiplicative (low load). Even the integrator is close to its minimum and removes a further 25% of the fuel quantity. Despite these drastic measures, the mixture on both banks is correct, as the first screen shows. Is something wrong with the LTFT? Is the DME lying?
Injector live data
What do the injector live data say?

In this menu the injectors are listed in the order 1/4/3/2.
The injection times on bank 1 (No. 1 and No. 4) are almost twice as long as on bank 2 (No. 2 and No. 3). The live data confirm that fuel injection on bank 2 is indeed strongly reduced. But if the injection times are half as long and the injected quantity is still what the engine needs, that is conclusive proof that one of the bank 2 injectors is leaking.
One possible scenario is that the injector of cylinder 2 is leaking. It sounds logical: there are fault codes for this cylinder, and it really does “drop out” against the others. But this time the situation is different.
Cylinder 2 performs poorly because it does not get enough fuel. Its fuel supply has been cut by almost half, so it is only logical that its output drops by half. The injector of cylinder 2 works correctly. The mixture in this cylinder is lean, so it is no wonder it registers misfires.
So how can cylinder 3 work at full power? There is only one explanation: its injector is leaking. Although its opening time has been halved, this cylinder still has more than enough fuel.
Conclusion
The cause of all the problems is the leaking injector of cylinder 3.
Sometimes a situation turns out to be trickier than it first appeared. It is therefore always worth analyzing all the data and making sure there are no contradictions.