BMW Lambda Sensors (O2 sensors)

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

max lange

Race Dog
Joined
Dec 16, 2010
Messages
628
Reaction score
2
Location
Robertson
Bike
BMW (all models)
For anyone who experiences a bad idle on the BMW R1200, most likely applicable to all bikes, here's my experience with Lambda sensors. Please note that I am by no means an expert and therefore this is my understanding of the function of Lambda sensors and may not be necessary correct. Additionally although this worked for me it may not for all BMW (or any) bikes.

I experienced an idle problem on a 2007 R1200GS, on further inspection and with the use of a GS-911 the issue all pointed to the lambda sensors (O2 sensors located on the exhaust headers). These sensors sample the O2 content in the exhaust gas every second and measure the amount of oxygen that is exiting the cylinder. The oxygen reacts with a chemical in the sensor and generates a voltage which in turn is feed back to the brain (computer). The sensor typically generates voltage between 1 and 9mV and is used by the CPU to determine the time that the injectors energize, to richen the mixture the injector opens longer, i.e. 3.2mS, leaner at 2.2mS.

A rich mixture, a low oxygen content generates a high voltage, typically 9mV; the CPU then leans out the mixture reducing the injector time. Conversely a lean reading produces a lower voltage 1mV resulting in a longer injector time allowing more fuel in the mix.

Therefore these readings bounce between 1mV and 9mV trying to keep a best mixture rate of 4.5mV.

If these are faulty then it could result in poor idle and flat spots when accelerating combined with high fuel consumption. Using a GS-911 its easy to spot, without not so easy other than swapping out the sensors, even then as there can be many factors its not easy to diagnose without some sort of tool.

So back to the GS, both were faulty, one returning 0~1 mV and the other hovering around 2mV and completely confusing the CPU which then defaults to a rich mixture (around 3.2mS injector time) and very poor running at idle. Removing the sensors creates the same result so that's not an option unless you add a power commander or similar.

So replacement became the only (I did test from another bike before continuing), BMW ask R2700 per sensor, ACME predictably R1400 so the search was on. After research we found a Bosch universal sensor @ R200 per sensor.

After I installed the new sensors the bike ran better, the GS-911 reported that the sensor were delivering a voltage between 0 and 0.9V correctly and it plotted a zig-zag graph as expected. However the warm-up time for the universal sensors seems to take longer, so the period where the engine runs a little rougher seems a little longer, this could improve as the electronics adapt.
I also get the impression that the engine is now running leaner, notably more popping.

Oh and remember to reset the adaptation values using the GS-911 after changing the components





 

Attachments

  • DSCF1065.JPG
    DSCF1065.JPG
    170.5 KB
  • DSCF1067.JPG
    DSCF1067.JPG
    343 KB
  • DSCF1069.JPG
    DSCF1069.JPG
    289.9 KB
  • DSCF1070.JPG
    DSCF1070.JPG
    326.3 KB
Top