Oxygen sensor voltages

[thewifes325]

The car in question is a 96 328i. I need some help understanding my O2 sensors.

I replaced all 4 this spring with Bosch sensors as preventative maintenance. Car ran fine (and is running fine). Then the week before I was to bring it in for inspection, the check engine light came on. According to my Snap-on scanner, the codes were for cat inefficiency on both banks. Not being too excited about having to dump a grand into new cats, I did some investigating. Which leads to my confusion.

With the car cold and running, I see 4.7 volts from my O2 sensors. Once I start driving I see the voltage fluctuate between 1.7 and 0.7. This is on all 4 sensors. They all bounce around with no specific pattern.

Correct me if this is where I am going wrong, but from the reading/searching I have done, the 1.7 to 0.7 fluctuation is a result of the heating circuit. I have tested all four sensors’ heater resistance and they were with 1% of each other. I tested from ECU connection to check the wiring also.

Looking at the wiring diagram, I don’t see an O2 sensor heater relay for a 96, although others mention it.

Suggestions? Am I wrong in zeroing in on the O2 heater circuit?

[Joylove]

I think the ecu is a bit shifty about the heater coil impedance. I think (and a google will confirm) that they need to be about 4 ohms, and 15 ohms will throw a fault and cause the fuelling to go open loop.

Also how are you measuring the sensor voltages? A digital multimeter won't be quick enough to catch the fluctuations, you need a scope or an analogue meter.

[randomy]

I've been trying to get a good idea of how my O2 sensor operate myself lately. I have been using a version of the dealer software that has a test plan, and reads the data from the ECU.
I had the same results you are seeing. Nice and steady at 4.7 when the car is first turned on, but sporadic readings after it had warmed up (though I was seeing readings as high as 3.8).
The test plan says something to the effect that the measured voltage should be between 0.7 and 4.7 volts with wild fluctuations. Further, it discounts the O2 heater circuit for the fact that the CEL is not illuminated.
What are you using to monitor the voltage while you are operating the vehicle?

[Joylove]

OBD2 communication rate is also not fast enough to see the toggling between 0.7V and 4.7V. You must use an Oscope or an analogue multimeter.

It's always 4.7V when cold because you run rich mixtures untill the car warms up.

[thewifes325]

I've been using my Snap On scanner to monitor. I will borrow a scope from work and check things out this weekend. I'll also double check the ohmic values of the heaters. I thought they were all a bit above 4k, but I'll restest to be sure.

Not sure what it means, but the Snap On has entries for O2 Heater % for all for sensors. The fronts stay in the 60-70% range even when warmed up. The rears stay right around 80%.

@randomy - Didn't want to hear that sporadic readings is expected. Trying to find a way not to spring for a new catalytic converter. :-)

[xdgt03]

Yes on the initial reading. It is running in open loop at max rich. The fluxuating reading while driving seem right unless the are fluxuating while maintaining a constant speed. As you accelerate and deccelerate the mixture will fluxuate based upon the volume of air coming in. This should mean a lower voltage when you lift off the throttle and a higher voltage when you press the throttle.

[Eric93se]

I've been using my Snap On scanner to monitor. I will borrow a scope from work and check things out this weekend. I'll also double check the ohmic values of the heaters. I thought they were all a bit above 4k, but I'll restest to be sure.

Not sure what it means, but the Snap On has entries for O2 Heater % for all for sensors. The fronts stay in the 60-70% range even when warmed up. The rears stay right around 80%.

@randomy - Didn't want to hear that sporadic readings is expected. Trying to find a way not to spring for a new catalytic converter. :-)


If your cats are in fact beginning to clog/fail you might want to consider replacing the ignition coils if they are over 100k.

If you want to try and rejuvenate your cats before they're fully clogged try some additives like:
http://www.amazon.com/Cataclean-Engine-Catalytic-Converter-Treatment-/dp/B002BVXM92

http://www.amazon.com/CRC-05063-Guaranteed-Emissions-Formula/dp/B000CIPUR8

[Joylove]

I've been using my Snap On scanner to monitor. I will borrow a scope from work and check things out this weekend. I'll also double check the ohmic values of the heaters. I thought they were all a bit above 4k, but I'll restest to be sure. 4R not 4KR!


Not sure what it means, but the Snap On has entries for O2 Heater % for all for sensors. The fronts stay in the 60-70% range even when warmed up. The rears stay right around 80%.

@randomy - Didn't want to hear that sporadic readings is expected. Trying to find a way not to spring for a new catalytic converter. :-)

Sporadic readings are correct. The engine ECU when running closed loop (off 'choke' - very warmed up) will attempt to position the 02 sensor so that it toggles from high to low (rich to lean) continuously. This is (almost) the optimum fuel/air ratio.

In theory the O2 sensor could toggle from low to high on almost every spark in that cylinder bank, if the ECU was capable of making things that perfect that quickly.

In practice it might do 5 sparks lean, try a bit richer, 5 more sparks, try a bit more richer etc until the 02 sensor toggles to indicate rich running. Depends on how Bosch decided to implement their software control loop.

[flyfishvt]

Do you have a copy of the SnapOn BMW manual? If you don't then send me your email and I'll get you a copy of the PDF file. It goes into great detail about all the test parameters on those sensors.

I just took a quick look and it seems you readings are in line but you should double check just to be safe.

My question is if the readings are all in line then why are you getting a CEL. One thing you can try is to use the scan tool to reset the O2 sensor "adaptations". It should clear the CEL and let the DME re-learn the sensor readings. This will probably require you to complete an approved drive cycle to get the o2 sensor OBDII test to read as "COMPLETED".

Clearing the adaptations and completeing a drive cycle may clear the CEL long enough for you to get it inspected so you'll have more time to get it fixed. If you need to find out what an approved drive cycle includes just do a google search for BMW drive cycle. You'll get plenty of hits. A complete drive cycle takes about an hour. maybe a little more.

[thewifes325]

Do you have a copy of the SnapOn BMW manual? If you don't then send me your email and I'll get you a copy of the PDF file. It goes into great detail about all the test parameters on those sensors.
Thanks for the offer, PM sent.

I will update this weekend when I have some time to some more testing.

[Eric93se]

My question is if the readings are all in line then why are you getting a CEL.

The CEL is for bad cats not bad 02 sensor. Assuming the brand new OEM sensors are bad is a tough route to follow.

We don't know how many miles is on the car or how well its up kept.

[thewifes325]

The CEL is for bad cats not bad 02 sensor. Assuming the brand new OEM sensors are bad is a tough route to follow.

We don't know how many miles is on the car or how well its up kept.

See my first post . . . I was trying to make sure before I spent the money on a cat, that the 02 reading I was seeing was correct. The car is well maintained, just over 150k. Not burning, leaking and fluids. Mileage wise it definitely could be that cats, my wallet wants to be sure first.

[flyfishvt]

Just one other thing comes to mind. What is the code or codes you are getting? I'm thinking it's odd that both cats would fail at exactly the same time. Have you considered it might just be a fuel trim issue and not a cat issue?