Truth about O2 sensors...

[Hawkster]

Ok, so as I'm readin thru some of the posts here about O2 sensor problems, I've determined that there's some confusion on how they work and what they do.

An O2 sensor measures the amount of oxygen going thru your exhaust system. In an OBD-II vehicle, there are two oxygen sensors, one in front of the cat and one after the cat.

An O2 sensor in our car works like this: First of all, the sensor must reach a temperature of 600F in order to give a proper reading. So lets assume your car has been running for a bit so the O2S is up to proper temperature. The PCM sends a 1 volt reference signal to the sensor. As the exhaust air passes over the sensor, it is compared to the oxygen content of ambient air (outside the exhaust). If there's very little unburt oxygen in the exhaust (rich mixture) it will send a modified signal of about .9 to 1.0 volt back to the computer. If the oxygen content in the exhaust is the same as ambient air, it's a very lean mixture, and will in turn send a modified voltage of about 0.1v back to the computer. The PCM will now try to compensate for whatever mixture your engine is currently receiving by dumping either more fuel if you're running lean, or less fuel if you're running rich. The "perfect" stoichometric ratio of 14.7 parts of air to 1 part of fuel will read at about .45v at the O2 sensor.

The Pre-Cat O2 is the one that's primary purpose is to see how much Oxygen is coming out of your engine. Generally in newer cars they also have a built in heater circuit. This is used to get the sensor's temperature up to 600F without having to wait for the car to get to full operating temperature. On a diagnostic scan tool, parts catalog, service manual, etc, this may be refered to as the "HO2S" which stands for Heated Oxygen Sensor.

Now, the Post-Cat O2 has a totally different purpose. This sensor generally does not have a built-in heater circuit because it's readings are much less critical to engine performance than the Pre-Cat Sensor. It still compares the amount of oxygen in the exhaust to the oxygen in the ambient air, but it's not used for mixture-controlling purposes. It is used soley to tell the computer if the catalyst is working efficiently. So now not only is it comparing Exhaust to Ambient, it's also comparing to the Pre-Cat HO2S. As exhaust passes thru the catalyst, remaining oxygen is used in a chemical process which I won't go into detail on because it's rather irrelevent to this subject. Anyway... the Post-Cat O2 should read at .45v pulsing up and down in voltage by about .10 - .15v. If it does not pulse, or if it just completely mimics the Pre-Cat O2, the computer will know there is something wrong with either the Post-Cat O2 and throw a CEL light and a code for "Low Catalyst Efficiency" or one for a bad O2 sensor.


So to sum it up, I'm a student at the automotive school Universal Technical Institute and I have entirely too much time on my hands to sit here and type this all out. If anyone needs clarification, let me know. If anyone disagrees with this, lets me know... just because I was taught it doesn't mean it's 100% correct.



Hawk

[gooten1]

Awesome, thank you for the explanation. Makes perfect sense now. So a post-cat o2 simulator would have to fluxuate? That means I couldnt' just put a resistor on the wires and make the computer happy...

[Hawkster]

Originally Posted by gooten1

Awesome, thank you for the explanation. Makes perfect sense now. So a post-cat o2 simulator would have to fluxuate? That means I couldnt' just put a resistor on the wires and make the computer happy...

That's correct. Also, I just read back and realized I didn't say this. BOTH Pre-Cat and Post-Cat O2 sensors fluctuate back and forth. If you were to view the voltages of both sensors next to each other on an ociliscope, they would both be going up and down. The difference would be that the Pre-Cat would pulse from .1 to .9v, and the Post-Cat would show the same pulsations, just in a much narrower wavelength, from like .3v to .6v.

I'm trying to find a picture of a graph online to demonstrate this, but I'm failing miserably. I'll try to do it with just text.

Pre Cat
|
|
|..........___...........___
|........./....\........./.....\
|......../......\......./.......\
|......./........\...../.........\
|....../..........\__/...........\ etc
|
|________________________________

Post cat
|
|
|
|.........____.........____
|......../......\......./......\
|......./........\___/.........\ etc
|
|
|________________________________



EDIT: I'm sorry for all the " . "s but the thing didn't like all the spaces and made my 'graph' a big jumbled mess. If I can get access to a scanner I'll scan one of my print outs from work or school

[DustenT]

This is from my datalogger. It's connected to a WideBand O2 sensor. Illustrates the constant change nicely. You can see how it bounces between rich and lean as it tries to keep the engine at 14.7 AFR. The Wideband has an output of 0-5v, versus the stock 02 sensor with a 0-1v output. Both sensors BASICALLY read the same way.



This is how it look on the display:

[Blue318ti]

Whoa!! now I understand O2 sensors. Thanks guys!!

[wolfgang20878]

if/when they go bad. I'd rather not spend 4x$100 to replace based on milage. My M44 has 90K miles.
-John

[aceyx]

There are only two for the M44.

[GDB]

Originally Posted by wolfgang20878

if/when they go bad. I'd rather not spend 4x$100 to replace based on milage. My M44 has 90K miles.
-John

Your car will throw a CEL (check engine light will come on) when the O2 sensor goes out. Don't replace them before then, it's pointless.

[VMGordon]

I Have a 96 and it just tripped 129000 miles, On the way home today the check engine light came on. From what I have been reading, it seems like the O2 Sensor is the first place I need to look.
Just bought the car and it is my first BMW, so I am new to all things BMW.

I'll try to update my profile as soon as possible,

Thanks.
VM Gordon

[robcarync]

o2 sensors arent really cheap, but they are really easy to replace on your own. it would be best to get your car scanned to make sure it is the o2 sensor code...youd hate to replace the sensor and still have the CEL on

[Bill Strong]

so knowing this info, how could one make a after-cat O2 simulator based on the front O2 readings, so as to not trip the CEL if a cat was misteriously missing?

[Hawkster]

Originally Posted by Bill Strong

so knowing this info, how could one make a after-cat O2 simulator based on the front O2 readings, so as to not trip the CEL if a cat was misteriously missing?

I'm working on it

[J!m]

I would think a small string including an appropriately sized resister (to drop the voltage) and a diode (to prevent back-flow) attached between the sensor's should be do-able. Just need to establish a solid trend between the two and experiment... I don't think the ECU is smart enough to tell the signal is exactly the same, but perhaps it is.

The second option is to remove the CEL bulb...

[Bill Strong]

I thought of that... but it would be kinda nice to know if I had just dumped 5.1 quarts of Mobil1 on the highway...



Hawkster - do you have an ETA?
I am getting ready to build a new manifold back exhaust system. I want to use a resonator in place of the cat. But if I need to I will just flange a high flow 3-way cat and install that.

[jekaio]

the PCM sends no signal to the O2 sensors.

o2 sensors generate their own voltage.

either O2 sensor can be heated, upstrem or downstream.

the heaters on heated o2 sensors run when the car is below 3000rpm.

When a non-heated O2 sensor is below 600F it will run on preset parameters.