Will stock O2 sensor read down to 11 AFR if moved after turbo? Any negatives?

Heard the stock O2 sensor will read down to 11 AFR if moved after the turbo. True? I'd LOVE to not have to mess with my aftermarket wideband O2 sensor and just be able to log right from the stock sensor (it's a pain for me to hook up the wideband, plug in an extra cord into my computer, yada yada...).

Any drawbacks? I guess if it was it a place that didn't read correctly, you could get lean conditions in Closed Loop which could cause damage. I have 2 bungs in my downpipe so I guess I could hook up both and see what the reading are.

Seems like everyone would do this if it worked so I'm guessing I'm missing something. The info was on NASIOC and might be for STIs or something.

On stock twinscroll cars the front O2 goes to the DP. Our USDM ECUs are not setup for the added delay in the exhaust gas hitting the DP O2 sensor. So closed loop AFRs I have heard can get a little less precise. It shouldn't make a difference in Open Loop.

As you suspect, main drawback is not being able to read below 11.2AFR or whatever the sensor floor is (I do not remember).

Often in the Subaru community (or any community) a flock mentality dominates. I have read isolated incidents of experienced tuners doing it this way, mostly due to running twinscroll headers, but one case on RomRaider of someone doing it for the heck of it. Being able to read AFR to the high 10s is a really nice benefit, since you gain a lot of safety for minimal loss in power.

You might want to research this more before jumping into it.

I'm going to run my car this way soon, will see how it goes...

We had a discussion about this recently, and I've thought of doing the same thing. As ClimberD mentions, the stock sensor is a WBO2 but not scaled to read below 11.2, which isn't that low. The other issue with delay time seemed to stem from earlier ECU iterations... it came up on a 16bit car. We aren't sure if it's still a problem though guys like NSFW would have found it if was causing AFR fluctuations.

A friend of mine put the point to me in the best way. Why do you want to move the O2 sensor back there? To scale your intake? To be sure you have 11.2 or just >11.2 AFR at high revs / load? If this is the answer, and it was for me, it raises another question. How accurate is the stock O2 sensor at richer AFRs? How are you going to calibrate it? NSFW has a an aftermarket WBO2 installed right next to his stocker, so he knows how accurate his is, but how are you going to check?

Just do a permanent install with your aftermarket WBO2. The latest generation of Bosch LSU sensors are much more forgiving in pig-rich environments and last much longer than they used to.

I have a scatter plot somewhere that shows the stock sensor and my PLX wideband reading within about 3% of each other...

... down to 11.2AFR. Right?

Thanks all. Good feedback.

IF moving it didn't affect it's originally designed functions (e.g. didn't cause delays or something due to moving it), it would still be nice to just quickly read down to 11.2 any old time, then hookup my other wideband only when needed.

Anyone every look into using a wideband's 0-5V output and splicing that into the ECU?

Wideband's controller would control the O2 sensor, and you would need to scale the sensor in the ECU, which wouldn't be all that difficult. Unless you have something like AEM's programmable output, in which case no ECU programming required. Even then, is there a programmed AFR floor, regardless of sensor output voltage, and can it be over-ridden? I have not seen that table...

Anyone every look into using a wideband's 0-5V output and splicing that into the ECU?

To what end? If you're going to install an aftermarket WBO2, just leave the stock one in place to carry on controlling CL fueling, and use your WBO2 to scale the intake for OL, plus the usual monitoring / logging duties.

...is there a programmed AFR floor, regardless of sensor output voltage, and can it be over-ridden? I have not seen that table...

There is a table that sets the rich limit for the O2 sensor, and it can be edited, but again... to what end?

Anyone every look into using a wideband's 0-5V output and splicing that into the ECU?

To what end? If you're going to install an aftermarket WBO2, just leave the stock one in place to carry on controlling CL fueling, and use your WBO2 to scale the intake for OL, plus the usual monitoring / logging duties.

 

There is a table that sets the rich limit for the O2 sensor, and it can be edited, but again... to what end?

 

Logging wideband O2 is an inconvenience to me, and this is such an easy fix...

Pros:

-Cheaper Wideband O2 replacement than Front O2.

-Read accurately to 10.8. Maybe stock can do this???

Cons:

-Installation and setup.

 

I'm not saying YOU or ANYONE should do it, I'm simply wondering if it's doable. Sounds like it is.

 

If you're in my situation (tuning via Cobb AP and AccessTuner, JDM twinscroll, want easy logging without laptop) then it makes total sense. I only have one bung in my DP. Why add a second... I would strongly consider it if my Front O2 came up for replacement. I already own the AEM wideband, so it would be a no cost wide-er band replacement.

Sure... if the stock front O2 were to go out and I couldn't get one as cheap as a Bosch replacement... I'd think about this approach.

I think there is a reason they limit the rich reading to 11.2 AFR. Perhaps NSFW tried lowering that and checking it against his PLX? I'm just doubtful the reading is useful below that point.

Sure... if the stock front O2 were to go out and I couldn't get one as cheap as a Bosch replacement... I'd think about this approach.

I think there is a reason they limit the rich reading to 11.2 AFR. Perhaps NSFW tried lowering that and checking it against his PLX? I'm just doubtful the reading is useful below that point.

Conversely, 11.2 is TOO Low for the sensor's intended function, which is to do low backpressure closed loop near-stochiometric AFR sensing. It may have the capacity to go well below 11.2, accurately, but what good would that do in a up pipe? None, because it would be pointlessly incorrect in an up pipe. But not in a down pipe. If it's a linear sensor with room to spare down low on its sensing range and linear voltage output, would be a great software solution to the 11.2 AFR, which is a touch too rich for where I want to be.

NSFW did you ever mess with the stock sensor's floor? Any desire to?

I'm in for more info. Having 4 subarus (3 forced induction), I hate moving my WB02 from car to car. I have bungs on the turbo models in the DP, immediately after the turbo, so this would literally be a PNP setup for me.

Mr. D,

I may be wrong about this, but the older (pre-06 impreza, pre-05 legacy) front O2 sensors read in mA. I could probably convert a 0-5V signal to a mA signal with the required speed and accuracy. The problem with my LGT is that in 05, they went to a uA signal. Converting that might get tricky.

edit:

Just pulled up ROMS for each.

02 WRX reads -13 to 10.9mA for AFR of 11.36 to 25.34:1. 24mA conversion is simple enough... offset for negative current... not as simple, but possible.

05 LGT reads -1.3 to .74mA for AFR of 11.15 to 20.28:1. 2mA is getting tight to control. It's not the uA I thought, but a very small current change represents a large change in AFR.

Conversely, 11.2 is TOO Low for the sensor's intended function, which is to do low backpressure closed loop near-stochiometric AFR sensing. It may have the capacity to go well below 11.2, accurately, but what good would that do in a up pipe? None, because it would be pointlessly incorrect in an up pipe. But not in a down pipe. If it's a linear sensor with room to spare down low on its sensing range and linear voltage output, would be a great software solution to the 11.2 AFR, which is a touch too rich for where I want to be.

 

NSFW did you ever mess with the stock sensor's floor? Any desire to?

Yes, of course even 11.2 is way beyond what the sensor can read in the stock location because of the EGBP you'd normally see once into boost and OL fueling, where it's reading is ignored. This however makes it less likely that Subaru would spend the extra $ using a true WBO2 in that location where it's capabilities will go to waste.

Of course just dropping the number in the rich limit table (many ROMs are limited at 11.02) doesn't magically scale the sensor past 11.15 , which is where the it stops many stock ROMs. The scale is much a straight line between sensor current and AFR from full lean at 20.28 AFR down to full rich at 11.15 in most of the later ROMs I've looked at in the last ten minutes or so... It could be that it can be scaled further. There are even 'spare' cells available for this.

[Table2D]

-1.30 -0.87 -0.47 -0.14 0.00 0.10 0.24 0.39 0.53 0.64 0.74 0.74 0.74

11.15 12.17 13.18 14.19 14.70 15.21 16.22 17.23 18.25 19.26 20.28 20.28 20.28

We're talking about moving the front O2 sensor from UP to DP. No moving or touching of the 2nd O2, which is in the rear of the DP.

I don't remember what exactly the rear O2 does, but in some thread on here it was revealed and was a bit surprising considering the thoughtless consensus had been / is that it only check cat efficiency and some long term fueling tweeks. How it actively participates in the equation... escapes me. But either way if you leave it and let it do its thing then car will be happy.

Anyone every look into using a wideband's 0-5V output and splicing that into the ECU?

 

Wideband's controller would control the O2 sensor, and you would need to scale the sensor in the ECU, which wouldn't be all that difficult. Unless you have something like AEM's programmable output, in which case no ECU programming required. Even then, is there a programmed AFR floor, regardless of sensor output voltage, and can it be over-ridden? I have not seen that table...

Wow, So its soooo ironic you said this. I just got off the phone with Prosport because I just purchased their Premium Halo gauges along with the wideband O2 sensor. The representative I spoke with suggested that I do this in order to datalog. Basically, he said to use the signal wire from the ecu and splice it to the new wideband and remove the stock sensor. That way the ecu would be reading the from the wideband (0-5v) as oppost to the factory (0-4v). Does this make sense at all? And it thats the case how will that affect my tune? BTW i'm using a V2 accessport as well and just got my tune from Infamous.

We're talking about moving the front O2 sensor from UP to DP. No moving or touching of the 2nd O2, which is in the rear of the DP.

 

I don't remember what exactly the rear O2 does, but in some thread on here it was revealed and was a bit surprising considering the thoughtless consensus had been / is that it only check cat efficiency and some long term fueling tweeks. How it actively participates in the equation... escapes me. But either way if you leave it and let it do its thing then car will be happy.

Interesting. This is a 32bit ecu phenomena, correct? My 02 WRX has a busted old o2 sensor in the rear just to plug the hole and runs... fine? Minus the fact that the TD06-35r maxed out... everything. I haven't tried it on the LGTs.

Wow, So its soooo ironic you said this. I just got off the phone with Prosport because I just purchased their Premium Halo gauges along with the wideband O2 sensor. The representative I spoke with suggested that I do this in order to datalog. Basically, he said to use the signal wire from the ecu and splice it to the new wideband and remove the stock sensor. That way the ecu would be reading the from the wideband (0-5v) as oppost to the factory (0-4v). Does this make sense at all? And it thats the case how will that affect my tune? BTW i'm using a V2 accessport as well and just got my tune from Infamous.

The factory is a small current signal, not voltage. I'm not up to snuff on all WBo2s but I believe there aren't too many capable of outputting a -1.3 to .7mA signal. You 'could' tap it into a TGV sensor feedback if you are TGVless. I believe they are 0-5v. The problem there is that you would have to condition the data afterwards. i.e. x volts on the sensor feedback = y AFR.

You can use the 0-5v output to input to your laptop for logging, worst case. I have used one of the two TGV position sensors, spliced into my wideband's output, so I can log without a laptop. NSFW proposed doing this a while back. Great trick, if you have deleted TGVs.

I do not know what you can do in so far as what ProSport said, but it appears that the stock wideband sensor does not output the linear 0-5V signal you're looking for.

You can use the 0-5v output to input to your laptop for logging, worst case. I have used one of the two TGV position sensors, spliced into my wideband's output, so I can log without a laptop. NSFW proposed doing this a while back. Great trick, if you have deleted TGVs.

I do not know what you can do in so far as what ProSport said, but it appears that the stock wideband sensor does not output the linear 0-5V signal you're looking for.

Yes, I'm TGVless...well I will be as soon as i'm done removing this horrid intake manifold LOL...its taking a while. Anyway, i've now been instructed by a pretty repuitable source to NOT tap into that signal wire considering that it uses impedancane to communicate with the O2 sensor...therefore it the wire is significantly altered in length thats goign to throw off the ecu.

Anyway, i don't know what to do now as far as logging. Appearantly the prosport gauges are analog and not digital.....don't really know waht that means.

The answer is in this thread

Two of us said it.

I tried to get it to read lower than 11.2, but I haven't figured it out yet. I changed the scaling and the 'rich limit' (there's a table for that in RR) and an undocumented limit, but it still won't read lower. The limit might be related to the sensor itself, but there is one more place in the ROM that I want to check - the code that reads the sensor signal. Problem is, I haven't found that code yet.

With 16-bit ECUs, disconnecting the rear O2 has no ill effects, but people with 32-bit ECUs have reported that their cars went rich after disconnecting it. However a couple guys at RR seem to have found the solution to that problem:

http://www.romraider.com/forum/viewtopic.php?f=34&t=8131

Basically it's related to the ECU code that normally causes the AFR to oscillate a few percent rich and lean while you're in cruise (it's an emissions thing). With no rear O2 sensor the ECU gets stuck in the 'rich' mode. But there's a table that can be used to disable the oscillation completely, and one guy has reported that he removed the rear O2 without running rich that way. Before we can be sure it will work for everyone, we need some more people to verify that.

Not that if you have a catted exhaust for emissions reasons, this will reduce the effectiveness of your cat, so if you're a tree-hugger like me you should keep your rear O2 in place and operational. But if you're a catless bastard then that sensor isn't doing any good anyhow so you might as well tune it out.

I can post find the oscillation table for A2WC522N pretty quickly I already have that ROM "under the microscope" so to speak. Other ROMs would take some work but if anyone wants to be a guinea pig let me know. But be prepared to wait a while.