This time I will look at quite a widespread defect of the wideband Lambda probe. As a result of this defect, indications of the Lambda probe are “drifting”, not taking into account that the probe itself is as if able to work (warming up – OK, and the chemical efficiency – correct). The situation is complicated comparing the simple case when there are problems with the warming of the probe, and they are easy to identify with a multimeter. 

A short history of the problems of the car.

A strongly leaking injector was identified for a car with an N43 engine. After replacing the injector, suddenly, the error messages regarding air mass meter; Lambda probe; switching off the cylinders of the 1-st bank appeared. In total – some 10 (!) error messages!

The defect of the air mass meter is apparent rapidly. With a completely open throttle valve, the detected amount of air was strongly reduced. No damages were seen to the inlet manifold. After replacement of the air mass meter, problems with air amount disappeared. I can not prove, but I have suspicions, that the air mass meter was “cleaned” when the first problems with the engine started. It is a common (but completely incorrect) “repair” method, which is popularized by unqualified “repair specialists”.

The situation after replacement of injector and air mass meter: 

a. all injector data – correct;

b. error messages regarding air amount – disappeared. 

The error messages regarding the Lambda probe are still left: 2C39 un 2C3D. The error message regarding control of the Homogeneous mode (2D64) results from Lambda probe problems. Both Lambda probe error messages refer to the wideband probe (before CO catalytic converters) of the 1-st bank.

Unfortunately, the freeze-frame of both error messages (2C39 un 2C3D) is pretty chaotic. DME had recorded the error messages when it was switching from Stratified charge to Homogeneous mode. Accordingly – part of the data corresponds to Stratified charge; part – Homogeneous mode, but they don’t give absolute clarity.

So the diagnostics have to be continued with the analysis of live data. 

At first, let’s check the Nernst resistance (it describes the chemical activity of the probe). The value of suspicious probe is marked with red;

Marked value – correct (correct values: 0/256/512). So – the warming of the probe is fine; the temperature of the ceramic element – correct (around 700 .. 800 oC), and the ceramic element of this probe is as if working!

At the next step, check the indications of the Lambda probe:

Indication of the Lambda probe is quite suspicious (correct value: around 1.00). Such Lambda value could be only if one of two cylinders of this bank is turned off. So at the next step, we check the mechanical efficiency of cylinders:

In the menu, we clearly see that all cylinders work perfectly. At the same time – an indication of the Lambda probe of the 1-st bank is incorrect (and there is no more doubt about it), and instead of 2.00V, it reaches 3.12V (which corresponds to Lambda value around 2.00).

Note: trimming range of the wideband probes is +/-400mV (0.4V). This time the “drift” of the probe is 1.12V (3.12 – 2.00), which exceeds the allowed value around 3 times. Obviously, the probe has serious problems!

For the control, we can check indications of the NOx sensor:

NOx sensor indications confirm correct average Lambda (0.99). So – it proves that those indications of Lambda probe of the 1-st bank really are incorrect (instead of 1.00, Lambda indicates double leftover of air: Lambda around 2.00).

The Lambda probe’s defect corresponds to scenario No.3, described here. Poisoning of the probe (with fuel) is the most popular defect. This defect is quite unpleasant because very often, it is sporadic – the problems appear in some specific modes (typically – in idle, low load mode) and usually tend to disappear without any special warning.

Repair – replacement of the probe.