In simple cases, MSV/MSD engine management units identify damaged Lambda probes, using self-diagnostics, marking the relevant errors. Unfortunately quite often there are no error messages, but still the engine has problems with maintaining the fuel.

In this post – about diagnostics of Lambda probes and finding the hidden defects.


Let’s start with simple part – narrow-band control probes.

First type of test: evaluation of Nernst (internal) resistance. MSV/MSD measures the internal impedance of probes very simply – once in a second (for a few mili-seconds), the signal wire of probe is being connected to +5V power supply thru current source and evaluating changes in voltage. MSV/MSD performs the heating according to heating map, accordingly – it is not possible to evaluate chemical activity using PWM value as reference (unlike the situation, when the internal resistance of Nernst cells – chemical activity of probes – would be stabilized using PWM).


Open INPA ../F5/F2/F6.

For cold probes (for example, after cooling down the engine for 5 minutes), internal resistance of probes should be 65K Ohm.

If Nernst (internal) resistance is lower than value shown in the picture – check for current leakage.

Attention! If only one control probe is used for the engine (4 cylinders versions), ignore the readings of second (bank) control probe – they can be inexact (although the hardware is the same as for 6 cylinder engines, the software can proceed the data of “non-existing” probe inaccurately).

If Nernst resistance is correct, let’s proceed with next test.


Using INPA, select ../F5/F6, scroll down (if INPA loader 2.023 is used), and check voltages of control probes.

For cold probes the voltages should be 0,42V (acceptable deviation from value could be  +/-5 .. 10%). If this voltage is correct, it is worth to continue the performance of heated probe(s), if the voltage differs from required value – below you will find description of possible problem.

Attention! If only one control probe is used for the engine (4 cylinders versions), ignore the readings of second (bank) control probe – they can be inexact (although the hardware is the same as for 6 cylinder engines, the software can proceed the data of “non-existing” probe inaccurately).


In this image you can see the connection of control probes:

As can be seen:

  • DME controls the heating of probes via outputs No.6 and 12 (active signal level: 0V);
  • contacts No.8 and 10 are GND of the signal;
  • contacts No.14 and 16 are probe signal.

Signal pins of lambda probes (No.14 and 16) are connected to internal voltage reference  0,42V via current source.

Taking this in account, in case of suspicion regarding performance of Lambda probes, check the voltage while probes are cold (for example, cooling down the engine for 5 minutes and performing the measurement directly after switching on the ignition – before the probes are heated up), check the voltage of control probes, using INPA (../F5/F6). If the probe or connector has current leakage (very likely scenario, if we take in account, that connectors of control probes for N series engines are placed close to the floor of the vehicle), possible scenarios could be:

  • the voltage is significantly lower than 0,42V, if the power leakage is on probe pin No.3 (or car body);
  • the voltage is significantly higher than 0,42V, the leakage is on probe pins No.1 and/or 2.

Sample: current leakage in probe or connector to probe pin No.1 and/or 2 (menu from INPA loader 1.001):

Problems, what can be caused with such defect: incorrect performance of control probe – incorrect wide-band probe reference value: inappropriate fuel mixture in corresponding bank.

Solution – check the connector for current leakage, check the voltage (as described before) with disconnected probe. If necessary (the leakage is in the probe itself) – the control probe should be replaced.

Unfortunately MSV/MSD:

  • does not control internal resistance of Nernst cells (although it is being measured);
  • does not control voltage for signal inputs for cold probes (although the hardware has specially designed reference supply source of 0,42V).

Defects mentioned before can “cover itself”, because voltage inadequacies for heated probes are manifesting much more less, internal resistances of Nernst cells are not being displayed in common menu of fuel adaptations and/or Lambda probes.

Test for heated probes is described here, but once again shortly:

  • Internal resistance of Nernst cells (according INPA data) should be: 0/256/512 Ohm, allowed value: 768 Ohm; if the internal resistance is 1024 Ohm and more – the probe has aged (it’s chemical reaction is insufficient) and it has to be replaced;
  • maximum voltage, generated by probe, should be not less than 0,80 – 0,85 V (it can be checked in moment of rapid acceleration, minimum voltage: not more than 0,15V (can be checked in overrun mode, braking with the engine)*. If the voltage values are different from required ones – the probe has to be replaced.

*Voltages (min/max), generated by probe, can be evaluated, also performing test block for Lambda probes.


Following – analogous checks for wide-band probes.

Operating principle of wide-band probe:

Despite the obvious complexity, the solution performs simple function: changing the current via probe’s Pump cell, maintain exact voltage (450mV) to Nernst cell pins. Current of Pump cell is a measure of measured Lambda value.

For example: -1mA means Lambda 0,8; +1mA means Lambda 1,7.


Connection of wide-band probes:

Connection of probe:

1. output of reference voltage for both probe cells (Nernst and Pump);

2. voltage of Nernst cell: +0,5V corresponding to the pin No.1;

3. heating, PWM, active: 0V;

4. heating,+12V;

5. pin of probe calibration resistor;

6. pin of Pump cell, 0,5 .. 4,5V (+/-10mA max)*

* current through Pump cell (according to manufacturer’s map from exact type of probe) is a measure of Lambda value. If I=0mA, Lambda = 1,0; Lambda measuring tool measures the drop of voltage between Pin5 and Pin6.

Also PWM of wide-band probe heating is stabilized according to map of heating, chemical activity of probe is not stabilized by PWM value!

In the image: Nernst resistances for wide-band probes:

  • for cold probes: 16K Ohm; if the resistance is displayed lower – check for current leakage (in the connector, wire or probe itself);
  • for heated probes: 0/256/512 Ohm; if INPA shows 768 Ohm or more – the probe has aged and has to be replaced.

Cold wide-band probes has to display their voltage (voltage to output of Nernst cell) 2,0V +/- 10%.


Note: the internal (Nernst) resistance for cold wide-band probes is high, the current don’t flows thru Lambda measuring circuit, therefore INPA/ISTA will display Lambda value 1,0 even in situations, when there is as current leakage in connector, wire or probe itself, and the circuit does not performs correctly. For this reason, exactly the voltage has to be evaluated (can not rely on reference Lambda value for cold probes)!

If the voltage on Nernst cell output is displayed significantly different form 2,0V, the current leakage has to be found and eliminated. If necessary, the measuring has to be performed with disconnected probe – the voltage has to be 2,0V (+/-10%).

Unfortunately MSV/MSD records the error messages regarding wide-band probes only in case of extreme damages. Quite often the inappropriate performance of probes stay’s “unnoticed”. Quite often the existence of problems is not discovered also with test block of Lambda probes. Tests described above helps to perform the evaluation of probe performance in complicated situations.

Test for heated probes is described in details here.


Related entries:

MSD80 diagnostics

N52 and Lambda probes