Oxygen sensors going bad?

[Pistol_91]

Yeah emissions would definitely put a stop to offroad headers. I have no issues with tuning with no cats. Im not sure what people mean by that. Ive read it alot lately.

Sponsored

 

[engineermike]

I’ve learned lately that cat deletes and anti-foulers cause tuning issues. The rear o2 sensors do more than most realize and deleting the cat affects this. The anti-fouler can only fool the test but it can’t fix the other problems.

Just one example…the rear o2’s are used to calibrate the fronts. Cat deletes can throw this off by up to 20%, with or without anti-foulers. The bad thing is you’ll never know it unless you install an independent wideband. You might think you’re running 12/1 a/f because that’s what your widebands say, but it’s actually 14/1.

 

[K4fxd]

You might think you’re running 12/1 a/f because that’s what your widebands say, but it’s actually 14/1.

Is this an issue if the faosc is disabled?

 

[engineermike]

Is this an issue if the faosc is disabled?

That particular issue isn’t but there are others. And with FAOSC disabled, you’ll perpetually get a “not ready” cat test indicator and your widebands could develop error since they aren’t being calibrated.

 

[HKusp]

I’ve learned lately that cat deletes and anti-foulers cause tuning issues. The rear o2 sensors do more than most realize and deleting the cat affects this. The anti-fouler can only fool the test but it can’t fix the other problems.

Just one example…the rear o2’s are used to calibrate the fronts. Cat deletes can throw this off by up to 20%, with or without anti-foulers. The bad thing is you’ll never know it unless you install an independent wideband. You might think you’re running 12/1 a/f because that’s what your widebands say, but it’s actually 14/1.

Well, that's not good, not good at all.....dammit...

 

[Grimreaper]

Some am widebands do free air calibrations to calibrate. Couldn't we do the same with KOEO? Offset tables are present etc. would have to make changes to entire table as a whole.

 

[junits15]

That particular issue isn’t but there are others. And with FAOSC disabled, you’ll perpetually get a “not ready” cat test indicator and your widebands could develop error since they aren’t being calibrated.

Now you got me worried lol, what else are the downstream used for besides biasing the widebands and checking cat health?

Right now I have the less than perfect GESi cats, but FAOSC enabled and no non-foulers. I didn’t even know there was risk, I hadn’t got around to disabling FAOSC because I need to use HPT to do it, but I didn’t think that it was an issue. I assumed that the logic would theoretically be able to handle a “no-cats” condition without damage as that could happen on an otherwise stock car. Maybe I was wrong...

Hmmm

 

[junits15]

I'm not so sure FAOSC is the direct culprit of biasing issues with no cats, though I could be wrong.

There's a good paper in IEEE on this subject, written in 2002 but that would be when this tech was being first deployed. https://ieeexplore.ieee.org/document/1023217 if you work at any company that does engineering you likely have access to this for free.

Its doing 2 key things:

1. Biasing the upstream sensor
2. Estimating the oxygen storage capacity of the cat

The first item is what we're concerned with, biasing the upstream sensor. However, if this paper is to be believed and is still in use today, biasing activity only occurs when the cat is completely full or empty of oxygen. More simply, if I'm understanding this correctly, the biasing activity only occurs at steady-state rich or lean conditions. When the cat is completely empty or completely full of oxygen. In that scenario, the sensors should read the same mixture.

Estimating oxygen storage capacity would be how we get our cat life estimation, which would explain why disabling this system causes the monitor to go non-ready as we are essentially disabling the mechanism by which the ECM checks the cats. It looks like its using a complex filter to achieve this.

IDK I'm not sure, and its entirely possible that the tech is different today and is using protected IP to accomplish more, but it seems like we would know this by now if upstream O2's were that badly out of bias that often. hmmm....

 

[engineermike]

What else are the downstream used for besides biasing the widebands and checking cat health?

Slight correction to my earlier post…they are there for cat testing and FAOSC, minimum, and it performs FAOSC in multiple ways. I’m not sure if it does other cat testing other than looking for switching. I do know that ford had toyed with ideas like using the o2’s for humidity and other air property sensing during dfso.

 

[junits15]

Slight correction to my earlier post…they are there for cat testing and FAOSC, minimum, and it performs FAOSC in multiple ways. I’m not sure if it does other cat testing other than looking for switching. I do know that ford had toyed with ideas like using the o2’s for humidity and other air property sensing during dfso.

Interesting, what tipped you off that there was an error? I assume the only way to know is with a second wideband?

 

[engineermike]

Interesting, what tipped you off that there was an error? I assume the only way to know is with a second wideband?

Two local cars.

One car gutted cats and did anti-foulers. He later noticed that his fuel trims were suddenly pretty far off when he had previously thoroughly dialed in the maf. He did a lot of testing, but noticed that turning FAOSC off resulted in fuel trims returning to normal.

The second car was a flex tune. It wouldn’t learn below 12/1 a/f stoichiometry even using all stock maf and injectors, which equates to about 50% ethanol learned. Turning off FAOSC resulted in a learned a/f of 10.2/1 which is the lowest it’s allowed to learn using stock ford parameters. This lack of learning can only be caused by poor/inaccurate fuel trims.

 

[junits15]

Two local cars.

One car gutted cats and did anti-foulers. He later noticed that his fuel trims were suddenly pretty far off when he had previously thoroughly dialed in the maf. He did a lot of testing, but noticed that turning FAOSC off resulted in fuel trims returning to normal.

The second car was a flex tune. It wouldn’t learn below 12/1 a/f stoichiometry even using all stock maf and injectors, which equates to about 50% ethanol learned. Turning off FAOSC resulted in a learned a/f of 10.2/1 which is the lowest it’s allowed to learn using stock ford parameters. This lack of learning can only be caused by poor/inaccurate fuel trims.

Do you believe it’s biasing base fuel or the upstream hego directly? Wack trims make me think it’s biasing the fuel, my uneducated guess would be that if we were biasing the sensor directly there would be no way to detect an issue unless you had another wideband.

 

[stannypack]

did the cats go on the roush tune or custom?

 

[engineermike]

Do you believe it’s biasing base fuel or the upstream hego directly? Wack trims make me think it’s biasing the fuel, my uneducated guess would be that if we were biasing the sensor directly there would be no way to detect an issue unless you had another wideband.

I think you’re probably right. The only way I was clued into an issue was knowing both cars had good fuel trims before removing cats and bad fuel trims after.

 

[K4fxd]

The only way I was clued into an issue was knowing both cars had good fuel trims before removing cats and bad fuel trims after.

And turning off FAOSC fixed it?

Sponsored