this is a curiosity to me.. I'm going to toss it at the experts here and see what y'all think..

background:

I have a cali emission, '02 5.9L.. meaning, it has four o2's.. one on each bank upwind, one on each bank downwind.. the upwinds are located less than 16" from the head ports, and actually into the football sized and somewhat shaped pre-cats, and within six or so inches of the collector pipes..

the post sensors in a stock config are just behind the pre-cats..

I pulled headers sometime back, and dropped one (whoops) of course it landed on the WIDEBAND (an extra sensor all together) sensor, and broke it.. I replaced the headers with like headers sans the boss for the wideband..

current:

I pulled the downwind o2 on the drivers side bank (bank one), and inserted a wideband (five wire) sensor for my a/f gauge.. I figured I'd get some readings to see if my fuel is trimmed correctly, and it was just going to be for a little while..

I zip tied the dangling sensor to the 5wire sensor in the meantime.. I set a code, but turned the narrow band sensor around to where it's hidden behind air flow while driving down the road, and wha-lah- no more codes..


here is the curious part:

-My mileage SHOT up.. like, around 1.5~2MPG's..

-after slaving the computer to it, it seems the cross counts on that bank match the cross counts on the other.. (this should be an upwind sensor thing I understand, and technically, they are independent of downwind in these counts- or, at least I thought they were)

-there are no codes, pending codes, or any differences in either banks fuel trim according to the PCM data from the narrow-band OE (upwind) o2's..

-the only physical difference between the banks is one side is monitored downwind, the other is not (it's zip-tied to the wide-band sensor in the boss intended for it)..


why the huge bump in mileage?
do downwind sensors on these rigs monitor and alter a/f too? If so, this flies in the face of everything I thought I understood about these things.. but it still doesn't explain the matching cross counts from both banks..

I'm pretty dang confused..

thoughts?

 

Maybe the downwind you took off was REALLY the upwind one. Someone who replaced your O2s might have gotten that mixed up.

Or it could be your O2s are not working properly.

 

I thought about the two getting mixed up... and that is the first thing I looked at.. the harness has the downwind sensors terminating the harness over the top of the tranny, whereas the upwind ones terminates the harness under the firewall above the bell housing, if you would..

the sensors themselves could be confused, as in, the wrong sensor connected to the wrong terminal.. but, they couplers are color coded on my truck.. I've never bothered to look to see if the coupler itself would allow this kind of mistake..

all four sensors are less than a year/12k miles old.. they are NGK's, and I went through hell to locate the proper ones due to it being a left coast rig, and me on the right coast.. It could be the sensors.. but the cross-counts is what kinda tells me the sensors are doing what they are 'sposed to be doing.. as far as upwind is concerned..

I didn't know downwind sensors had any bearing whatsoever on fuel trim.. I'm still not certain it does.. the cross-counts are within 1% across the banks.. I just can't figure this spike in fuel mileage when nothing else has been altered except that.. taken by itself, it would seem certain that the down-wind sensors are in the fuel trim game too- but this seems bas-ackwards to everything I've ever been told about how this thing works..

 

I think there was another member that had some fun with post-cat O2 sensors, and fuel economy. It appears that they do indeed have some control over mixture. Although, why having one of yours hanging in space would INCREASE fuel economy strikes me as kinda odd. One would think that if it was seeing a large amount of O2... it would want to richen the mixture? Which should make gas mileage go down....

This might be a question for Hemifever, or, an experiment for someone with data logging software to explore. I always thought the post-cat sensor was for the sole purpose of ensuring the cat was doing its job..... Guess there is more to it than meets the eye huh?

 

I've been trying to do some research about this, on the 'count I can't wrap my brain around it either..

I learned something I didn't know:

o2 sensors sample BOTH interior and exterior (to the exhaust stream) air.. Meaning, the interior probe sniffs the air as it passes, and at the same time sniffs exterior air- the comparison happens literally in between the inlets.. So, if you have any type of material fouling the sensor, you've got a problem you didn't know you had... and it's doubtful the computer will catch it.. I knew (it's pretty obvious) the interior of the exhaust pipe was sniffed, i didn't know the exterior of the sensor sniffed air too..

apparently, you can close the vents (little holes) on the exterior part of the sensor, and create a situation where the interior is 'passing' more air past it than the exterior, which is (in the sensors eyes) going to register less voltage- hence, it thinks it's leaner than it is... it dumps fuel.. likewise, if you foul or somehow obstruct the interior passages of the sensor, you will alter the voltage (spikes) and make the computer process that there is a rich condition, and it will pull fuel..

that was a pretty enlightening little bit of info I stumbled across.. but it kinda begs the question: why two sensors (four, on a cali rig)?.. If the upwind can do the same as downwind and compare interior and exterior air, why does it need the down-wind sensor at all if not just to tattle about whether you have a cat or not?

I don't understand the purpose of a downwind o2 other than for that reason alone, especially in light of that little discovery..

the best I can tell, the downwind sensor- if in fact it also samples air on both sides of the pipe- and if in fact it does have bearing on fuel trim- is reporting perfect running conditions as far as a/f??? dunno.. I'm at a loss here..

 

"the best I can tell, the downwind sensor- if in fact it also samples air on both sides of the pipe- and if in fact it does have bearing on fuel trim- is reporting perfect running conditions as far as a/f??? dunno.. I'm at a loss here.."


I might have to start agreeing with this theory. A couple months ago, I got a p0138 IIRC, for the post cat sensor. The motor was not running as good as it was. I replaced it. And it went back to running as good as it was before the code came up.

 

i can second that Zman. a few months ago i had the same thing happen

 

that's the thing.. not only has my mileage shot up, the truck seems to be running a lot stronger..

a few months ago, I blew out a header gasket on a trip, and the o2's reported lean conditions and started dumping fuel like it was cool, on an engine it thought was starving but was in fact drowning.. I may as well had been pumping gas straight out onto the ground it was burning it so quickly.. I'm talking 5~6MPG's.. I later fixed the gasket (and dropped the header breaking the wide band sensor), and still had a leak at my collector flange..

I fixed that leak, and ran the software on the truck while in operation.. I watched it close loop, open loop when I stood on it, and close again at cruise.. That's how I know my sensors are working right at a quick glance.. I watched the fuel plane out.. I watched the cross counts (number of times it crossed 24.5v) plane out- and match the other bank (which wasn't leaking before) for the first time in weeks.. I knew I had sealed the exhaust on the drivers side at that point..

So, I get into a discussion here, and get all confused about what a/f readings I was getting before I broke the wide band sensor.. I get to worrying that I'm running really rich (my software can't count Lam's with the narrow band sensors).. so- I pulled the downwind o2 sensor on bank 1.. (the pre-cats on a cali rig are monitored, the big one isn't- I know this isn't optimum to monitor behind a cat for a/f, but I just wanted an idea).. I stuck the wide band a/f o2 sensor in the downwind bank 1's boss..

I zip tied the narrow band sensor to the wide band.. I had the inlet facing forward and to the side of the exhaust pipe.. I set a code.. a p0138.. I moved the zip tied sensor to be parallel to the pipe and hidden behind it.. 600 miles later, no code.. So, I left the wide band sensor in that boss..

that's where the fun starts..

The truck is running stronger and almost at perfect a/f.. a touch rich actually.. right at 14.5:1..

the only thing that has changed is that sensor is dangling..

I've got another ace up my sleeve...

I'm going to pull bank2's post o2 sensor and put a slug in the boss, and attach it to the exhaust pipe just like I did bank1's, and see what happens to counts and trims..

this could be interesting..

 

I understand the need to figure out why it's happening but if having a dangling O2 sensor has caused the truck to run stronger and upped the MPG a little I'd be inclined to hang my hat and call it a day.

 

Three little letters. O C D.

I look forward to seeing the results of this little test though..... I may yank my postcat O2, and see what it does on a non-cali rig.

I wonder if the non-fouler trick would be just as effective? I do believe my cat is rather hollow.