Substitution of Lambda probes

In the situation, when we have to replace Lambda probe (hopefully, it’s truly damaged, not some error message, which could be connected with Lambda probe, appears in DME error message memory), quite often (and it’s not wrong) we prefer to spend less money. Accordingly, the question – are there any alternatives, which are a substitute to OEM Lambda probes – appear.


Wideband probes

The situation with wideband probes is quite simple. Supplier of these probes: Bosch. Two variants are possible:

  • Bosch LSU 4.2
  • Bosch LSU 4.9

These probes differ with a nominal dynamic (Nernst) resistance of cells in working mode (80 Ohm vs 300 Ohm), heating strategy (LSU 4.9 requires less heating power), with the curve “current of pump cell vs Lambda”. The difference, why the substitution is practically impossible – connector.

They are strongly different.

Additionally, in the connector, the calibration resistor is mounted (its nominal is unique for each probe) – this obstacle complicates the substitution of probes even more. And finally – taking into account, that prices of LSU 4.2 and LSU 4.9 are quite similar, there is no logical explanation to try to substitute one with other. Max cheap wideband probe could be found by your local Bosch distributor. You just have to know some basic “parameters”:

  • type of probe: LSU 4.2 or LSU 4.9;
  • length of wire (Bosch offers several lengths of “universal” connector wire, till even 100 cm).


For example, for N52, N43, and N53 series engines, LSU 4.9 wideband probes are used. Length of wires depends on the type of the engine, and also for which bank the probe is intended.


Narrowband probes

Narrowband probes are relatively more simple products, you can think – situation should be even more simple. But – it completely opposite! Narrowband probes are manufactured by several large brands, there is a wide choice with “universal” probes. Which ones to choose?


First parameter: number of outputs. For modern cars (with this we understand even 25 years old BMW’s), narrowband probes with heating and the separated sensor element are used. The probe has 4 outputs: two for heating, two for sensor element.

Second parameter: the strategy of heating of the probe. At this moment the situation becomes much more “interesting”. In older BMW’s (in older Siemens DME, for example, MS41/42/43) and even in relatively newer Bosch DME (for example, intended for N42/45/46 engines) narrowband probes, which need 12V for heating are used. Accordingly, DME in right moment connects voltage of 12 V “directly” to probe. Or, when checking live data of the probe, we see PWM = 99 %. In newer DME (starting with MSV70, also MSV80, MSD80, MSD87 etc., also Bosch MEVD17), control probes of newer generation are used – the heating power for them is significantly reduced. The average heating voltage from 12 V is reduced to 5 .. 7 V. Even more – to ensure an optimal (stationary) temperature of the sensor element, its heating is continuously (changing PWM) changed – depending from a load of the engine, RPM and other obstacles. These narrowband probes of “older” and “younger” generation CAN NOT be substituted! What happens, if the type of probe is chosen wrongly?

  • if in the older engine (for example, M54) the probe, which is intended for N52/N43/N53 series engines, is installed, it’s damaged immediately! Supplying it with 12 V for heating (instead of required 5 .. 7 V), the heating element of probe ”receives” 3 .. 4 times larger heat power – they are overheated in one moment;
  • if in the newer engine the narrowband probe, intended for “older” engine, is used, it’s not heated up properly. Probe (its heating is managed according to appropriate map) is supplied with 3 .. 4 times less heating power than necessary, its chemical efficiency is insufficient.

With first case situation is quite simple – the probe will not work (its voltage, checking live data, will be 0.4 V and stationary for connected probe and 0.0 V if the sensor is disconnected from DME), when checking resistance of its heating element, it won’t be possible to measure it (for a healthy probe, resistance of its heating element is several Ohm).


In the second case, the situation will be more interesting:

  • in heated-up conditions (its heating PWM in range of 30 .. 70 %), the probe has reduced chemical efficiency – Nernst resistance of cell is not correct: 0/256/512 Ohm, but – strongly increased (reaching several tenths of K Ohm);
  • the range of voltage, generated by a probe, is narrowed, it means:
    • the minimal value of voltage (in overrun mode) is not below 0.1 V, but around 0.2 .. 0.3 V;
    • the maximal value of voltage (in kick-down mode) is not at least 0.85 .. 0.9 V but don’t reach over 0.5 .. 0.6 V.

Actually, the probe works correctly, not even taking into account decreased temperature – the cause of narrowed voltage range is peculiarities of construction of Siemens DME. Inputs of wideband probes are connected to a voltage reference source (voltage source: 0.42 V) via 47 .. 220 K Ohm resistor. This voltage source is used to identify short circuit of probe’s connection (if such appeared by some reason). This resistor, mentioned before, instead is the cause, why the range of probe’s voltage is narrowing (the probe is being “loaded”, its increased internal resistance works as a voltage divider).

Third obstacle – both for older and newer narrowband probes, the plugs can be identical! Thanks to this nuance, the situation can appear, that as if appropriate (not – universal) probe can work improperly.

And finally – the fourth nuance. Yes, BMW has thought about, that you won’t accidentally confuse the connection of narrowband probes: each bank (if more than one narrowband probe is used), has its own, different plug. Unfortunately, even in this situation, customers manage to mix up the probes! How? If probes of both banks are being changed at the same time (or even just screwed out on time) and they are screwed back in the inappropriate drill hole! Yes, the wires will be connected in the correct way, but probe itself will be in an inappropriate bank!


How to identify the inappropriate type of control probe?

1. for the started engine, wait till the heating of control probe(s) is started. If necessary, maintain increased RPM (2000 .. 3000 RPM) for a moment. Open ../F5/F6

In the image: heating of N53 series control probes. For N43 series engine: only left-side reading (1-st bank) is actual.

At the moment, when PWM increases from 2 .. 10 % (de-frost: pre-heating) to 30 .. 70 %: the probe is turned on (heated) for working mode.

Open ../F5/F2/F6

In the picture: dynamic resistance (Nernst) values of cells of control probes are marked. For N53 series engines – values of both banks are actual, for N43 series engines – pay attention only to indications of the 1-st bank. Not later than 5 .. 10 seconds after starting the heating, the  Nernst resistance has drop from max: 65500 Ohm to 0/256/512 Ohm. If resistance stays strongly higher (in the image: above 40000 Ohm) – chemical efficiency of the probe is incorrect (decreased) and it (probe) has to be replaced!


At the end. Unfortunately, for none of “universal” probes the strategy of heating is indicated, no exact datasheet is available. And, even if the datasheet would be available, there is no warranty, that as if even with similar heating strategy, the probe actually will not sometimes get overheated or sometimes will not cool down. Heating maps, created by MSV/MSD are calculated for one exact model of the probe – nobody knows, how this algorithm will perform with “similar” probe. Due to these causes, I don’t suggest to even experiment with any alternatives or “universal” probes! I’ve come across many situations when new probes have to be replaced by OEM products. Of course – consumption of time, money and nerves – at least double.

MSV/MSD don’t controls (just measures) internal (Nernst) resistance of cells (chemical efficiency of the probe). In case of an inappropriate type of probe, the error messages regarding it will be recorded late or will not be recorded at all, but trim or wideband probes and control of CO catalytic converters will be performed incorrectly. Typically – error messages regarding wideband probes and, after a certain time, also CO catalytic converters – will be recorded. The engine will work unevenly, the composition of exhaust gasses will be incorrect. Causes: incorrect trim of wideband probe; individual adaptations of cylinders turned off; later – control of fuel mixture control after Lambda probes (the engine will work in open loop mode) is turned off.