In this entry – about one quite typical situation, which is characteristic to OEM NOx sensors.


One of my customers sent me the following INPA screenshot (incorrect values I already marked with red):

Why exactly these values are marked as incorrect?

Lambda binary is different from 1200mV – a proof, that a sensor is on Online mode. Obviously, some moment before all conditions (the exhaust temperature, driving speed, etc.) were appropriate for DME to give a command to the sensor to switch from Offline mode to Online mode. The sensor has followed instructions as if passed all self-tests and started to send the data.

Also, Lambda Lineares (wideband data) are different from 1.00 – there is no slightest doubt, that the sensor sends Lambda data.

Unfortunately, the first problem appears – both data (narrow-band and wide-band) are controversial! By Lambda Linear 0.99 Binare signal would have to be at least 850mV, instead, for 726mV Binare signal corresponds to Lambda Lineare 1.00.

Second fundamental problem: NOx concentration is 0ppm. If the sensor is in Online mode and sends NOx data, INPA will never display 0ppm, because the displayable values are -150..-1; 1..1500ppm.

Obviously, the sensor has not succeeded to complete the self-adjustment procedure of both its cells and its NOx cell is not in working order!

Unfortunately, there are situations, when DME doesn’t record any error messages by these symptoms. Or sometimes the diagnostics “specialist” simply delete these error messages, obviously hoping, that it will fix the situation in some way.

This time DME has recorded the 30DA error message for numerous times, which confirms the damage of the NOx sensor. But, to make it more interesting, let’s assume, that DME hasn’t recorded any error messages. What else could confirm the damage of the sensor?


At first, let’s inspect learnbits: ../F5/Shift+F6/F2. In this case:

First two learnbits of the second column are not colored! If the engine would have run in Stratified charge for at least 10 .. 15 minutes after last deleting of the adaptations, these learnbits (Lambda adaptations for Stratified charge) would be colored.

Any more proofs to the problem? Here: ../F5/Shift+F6/F5:

As you see, DME has not ever performed the injectors tests in Stratified charge after last deleting of the adaptations! Not in idle, not in a mode of an partial load!

It’s 100% clear, that the engine has never worked in Stratified charge – a damaged NOx sensor is a reason for this problem.