Idle fueling targets

something which has bugged for a while is the idle fueling targets on our cars. I had initially guessed that they were rich of stoichiometric and this was recently confirmed when the idle fuel target table was added to RomRaider. our cars have to pass emissions at idle (and they do stock) but I don't see how this is possible with a target slightly rich of stoich (and no air pump on the early cars). if the mixture is always rich of stoich, at some point the cat's stored oxygen will be expended and unburnt fuel will make its way out the tail pipe.

 

is the target slightly rich because a quantity of unburnt mixture will make it out of the engine at idle anyway and the target is a fudge factor to compensate for the sensor responding to fuel in the exhaust stream (ie: burnt gasses + oxygen + fuel)? I'm guessing that the AFR sensor will read rich if presented with a stochiometric mixture that is only partially burnt. are there any automotive or chemical engineers who can shed some light on this?

I don't guess I'm following. On my 05LGT rom with the latest RR ecu defs, there is no 'idle fueling targets' unless its under some directory it shouldn't be.

I'm pretty sure it targets a rich mixture around idle as well, just from watching the AFRs.

 

I've been meaning to log 'final fueling base' from the latest logger definitions to see what it says.

I know its not as 'lean' as the 2.0L(many target ~15.3:1) but its normally ~14.5-14.9 or so at idle on mine.

I don't guess I'm following. On my 05LGT rom with the latest RR ecu defs, there is no 'idle fueling targets' unless its under some directory it shouldn't be.

 

it is in the close loop section as a compensation based on load and RPM. at idle loads and revs, the target appears to be 14.2

 

what I don't get is how emissions can pass this way. eventually whatever stored oxygen there is in the cat will run out and there will be no way to burn off the excess fuel and CO. does the car occasionally switch to a really lean mix to feed oxygen to the cat or is the target a fudge factor to compensate for the o2 sensor reading some unburnt mix?

it is in the close loop section as a compensation based on load and RPM. at idle loads and revs, the target appears to be 14.2

 

I'm still not following. My Load comp table for CL fueling is all negative. Assuming a 'normal CL target' would be 14.7, my load table is taking an additional fueling off of that, making it leaner.

does the car occasionally switch to a really lean mix to feed oxygen to the cat or is the target a fudge factor to compensate for the o2 sensor reading some unburnt mix?

 

Although its not always the case, a cat will generally make a post-cat O2 sensor read a bit leaner than the AFRs are pre-cat.

I'm still not following. My Load comp table for CL fueling is all negative. Assuming a 'normal CL target' would be 14.7, my load table is taking an additional fueling off of that, making it leaner.

 

 

I'm pretty sure the table is an additive compensation to AFR. negative values will drive the AFR lower (richer). there's a big dip at the idle cell in the maps I've looked at.

Although its not always the case, a cat will generally make a post-cat O2 sensor read a bit leaner than the AFRs are pre-cat.

 

so then the compensation is a fudge factor. if the pre cat reading is richer than post cat, that indicates to me that there is residual oxygen in the mix available to the cat. if this is so, then the pre-cat sensor is just responding to the unburnt fuel and not truly measuring the mixture ratio.

 

when I get some free time I'll shove a wideband up the tailpipe to get a post cat reading to confirm what's going on. or better yet, I'll mix up some gas and air in a bag and feed that to a WBO2. I'll make sure the mix is well lean of stoich and see if the sensor generates a rich reading due to the unburnt fuel.

it is in the close loop section as a compensation based on load and RPM. at idle loads and revs, the target appears to be 14.2

 

what I don't get is how emissions can pass this way. eventually whatever stored oxygen there is in the cat will run out and there will be no way to burn off the excess fuel and CO. does the car occasionally switch to a really lean mix to feed oxygen to the cat or is the target a fudge factor to compensate for the o2 sensor reading some unburnt mix?

 

I added "final fueling base" to my CarPC display while I ran some errands, and it looks like it mostly targets around 14.25. The only time I saw anything leaner, I was decelerating or driving downhill.

I'm pretty sure the table is an additive compensation to AFR. negative values will drive the AFR lower (richer). there's a big dip at the idle cell in the maps I've looked at.

 

Regardless of whether its an additive table or not, if you 'add' a positive number to a negative it still subtracts.

 

If it is the case in which you are describing, its backwards to every other additive table in map. I'm not saying you are wrong because I've not studied it, but its seems counter-intuitive, at least to me.

Regardless of whether its an additive table or not, if you 'add' a positive number to a negative it still subtracts.

 

If it is the case in which you are describing, its backwards to every other additive table in map. I'm not saying you are wrong because I've not studied it, but its seems counter-intuitive, at least to me.

 

exactly. that's what I've been saying except that all the compensation is backwards. for example, my 800rpm/0.2 load cell is -0.440. 14.7 + (-0.440) = 14.26. There's another comp that kicks in for temp but it is small. for the sake of argument, lets ignore it for now. 14.26 is richer than 14.7, not leaner. the table is a little weird as negative values drive the mixture richer, not leaner.

 

the richer CL targets at higher loads/RPMs make sense to me but the idle target still doesn't. I know Subaru wants a smoother idle but they also need to pass emissions at idle. somehow, sufficient oxygen is getting to the cat. from what I can tell looking at the datasheets, the AFR sensor will actually read a little rich if it is exposed to a stoichiometric mixture that hasn't combusted. unfortunately my WBo2 is in a different city so I can run the experiment right away.

If I add up the load and ECT compensations in my tables, I get about -0.5, which would mean a target AFR of 14.2, which is what I see in both my measured AFRs and in the "final fueling base" logs that I took over the weekend.

If I add up the load and ECT compensations in my tables, I get about -0.5, which would mean a target AFR of 14.2, which is what I see in both my measured AFRs and in the "final fueling base" logs that I took over the weekend.

 

This is what I get too.

 

I'm not sure how they do it. It doesn't seem like the catalytic converter is big enough to store that much oxygen. Maybe it doesn't require much? 3 g/s is about idle and it's only 0.5 points off stoichiometric.

I'm wondering what motivated the factory tuners to set it up this way.

 

Is it bad for a cat (not the furry kind) to be exposed to stoich exhaust for long periods?

the load and ECT comp tables are not the only compensations involved. There are more undefined than there are defined. You need to log the CL fueling target to determine the final actual target.

I'm wondering what motivated the factory tuners to set it up this way.

 

Is it bad for a cat (not the furry kind) to be exposed to stoich exhaust for long periods?

 

 

Maybe due to the fact there are three converters? More converters, more oxygen storage?

This is what I get too.

 

I'm not sure how they do it. It doesn't seem like the catalytic converter is big enough to store that much oxygen. Maybe it doesn't require much? 3 g/s is about idle and it's only 0.5 points off stoichiometric.

 

I believe the stock cats hold about 2g of O2. If I remember correctly, when I did the calc to see how long that would last it was only about a min. That's why I suspect the calibration is actually a fudge factor to compensate for the sensor.

Maybe due to the fact there are three converters? More converters, more oxygen storage?

 

Perhaps, but why not just burn at stoich and let the resulting exhaust pass through the cats?

I'm wondering what motivated the factory tuners to set it up this way.

 

Is it bad for a cat (not the furry kind) to be exposed to stoich exhaust for long periods?

 

keeps the cat warm for one thing (assuming there's the occasional burst of oxygen) but I don't think that was their motivation. other possibility is if there's some sort of NOx trap that they need to regenerate but I don't know if our cars have that stock.

there's got to be something else going on inside the ECU that we don't understand yet.

 

http://www.legacygt.com/forums/attachment.php?attachmentid=76454&stc=1&d=1258524974

 

Is it bad for a cat (not the furry kind) to be exposed to stoich exhaust for long periods?

 

http://www.legacygt.com/forums/attachment.php?attachmentid=76453&stc=1&d=1258524974

 

 

but its normally ~14.5-14.9 or so at idle on mine.

 

and that's what I've noticed the stock front O2 sensor reading is at idle

Awesome. Thanks for posting that.

 

All this time, I thought the feedback loop was just oscillating from too much gain, and I wanted to "fix" it.

I laugh when I hear on the news that cars produce too much CO2 and we need to improve emissions... when, given our current technology, the cleanest cars (given their fuel consumption) actually try to maximize C02 production because it's a hell of a lot cleaner than sooty stuff

Revival.

Was playing with some logs last night and noticed my CL idle AFR was ~14.25 with little variance +/- 0.05 or so. Thought my O2 sensor might be getting old ( has 91k ) however Boostin's current map shows a target of 14.26 below 900rpm.

So based on my limited understanding and the info in post #21 are they richening the mixture to lower NOx?

I know "Lean Burn" maps are somewhat popular these days plus the direct-inject diesel can run ungodly lean mixtures at cruise ( 100:1 from what I hear ). Downside is elevated NOx.

So like mentioned in the prior post is the slightly rich idle AFR to help one part of the emission formula? Thanks for any info!

Revival.

Was playing with some logs last night and noticed my CL idle AFR was ~14.25 with little variance +/- 0.05 or so. Thought my O2 sensor might be getting old ( has 91k ) however Boostin's current map shows a target of 14.26 below 900rpm.

So based on my limited understanding and the info in post #21 are they richening the mixture to lower NOx?

I know "Lean Burn" maps are somewhat popular these days plus the direct-inject diesel can run ungodly lean mixtures at cruise ( 100:1 from what I hear ). Downside is elevated NOx.

So like mentioned in the prior post is the slightly rich idle AFR to help one part of the emission formula? Thanks for any info!

 

I'm still convinced there are other fudge factors involved. At idle, the loads are so low I wouldn't really expect much in the way of NOx production. The one thing that does support my theory is that the actual mix coming out the tail pipe is stoichiometric with a slightly rich target. I had to push my targets almost half a AFR point lean of stoich to get the car to burn slightly lean (say 0.2). Does wonders for the smell and the idle roughness isn't bad if you retard the timing a bit.