[jfitzpat]

Originally Posted by WolfPlayer

Fact is - and I don't understand why you don't see this - all that I am saying is that if you take the stock rear O2 sensor and make it see lean conditions then the ECU responds and richens up the mixture. Simple as that.

I know that the ECU *can* make some adjustments to AFR based on the rear O2 sensor.

The reason I don't 'see it' is it is actually very, very hard for an ECU to make meaningful fuel adjustments from a post cat sensor. There are tons of SAE, etc. papers on the subject. I've also done work on measuring cat efficiency on the fly myself.

So, we are looking at an ECU doing something very hard, and not terribly useful, or some other explanation. Since there are plenty of simple explanations (ex. the front and rear sensors are both just grounded to the exhaust system, when you remove the rear sensor the removal of the amp or so or heater current could cause a ground shift in the front sensor) to test and even more logical complex explanations (ex. sensor removal is detected as an error and the mix is automatically enrichened to O2 starve the cat to stop it from overheating), I'm not willing to jump several steps ahead.

Originally Posted by WolfPlayer

That is good enough for me.

Obviously, but look at where we are - in a forum devoted to tuning product. And look what we are discussing, AFR. If the rear sensor truly has an open loop effect, and if it was designed to be post cat - then just removing the cat should cause fuel maps to be executed with an unexpected offset.

In other words, if I don't understand the full impact of the rear sensor, then I don't really understand how to properly tune without a cat. Leaving the sensor in may make an obvious symptom go away, but it still leaves a phenomena that potentially impedes performance tuning. You might not care, but I would.

Once something is understood, effective workarounds can be devised. Ideally, the whole narrow band learning thing might be hacked out of the ROM and wideband closed loop control put in its place. But loading the heater circuit with a 12 ohm, 20 watt resistor or, say, programming the analog output on an LC-1 might be more practical to a typical user here. But the validity of either would require a true understanding of underlying cause(s).

-jjf

[lackskill]

Originally Posted by drees

The proper way to test the role of the rear O2 sensor is to substitute the appropriate resistance to simulate a lean and a rich signal and see if the ECU leans or richens up the mixture accordingly.

jjf - I know you're on the Enginuity forums, so if you can read ROMs, there are STi ECU dumps there for you to dig into if you want to verify the role of the rear O2 sensor there.

I know its OT, but it would seem that his time would be better spent looking for a MAF delay table for tuning out the rich conditions associated with turbo spool on a FMIC. It seems that running a FMIC is more useful than removing the rear O2 sensor.

[STLSTI]

Originally Posted by WolfPlayer

No cat.
Rear O2 installed.
Wideband installed well upstream of 2nd O2.
Car targets 14.7 on the LC1 at idle.
Remove rear O2.
Plug port.
Car now targets lower 14's on the LC1.
Reinstall Rear O2.
Car now targets 14.7 on the LC1 at idle.

Test on somebody else's STI. Same thing.

t

I started playing with my fuel and Wolf is 100% correct. My car shows the same behaviour(sp). Catless DP second O2 plug-in but no installed. LC-1 shows idle of 14.4 . Installed second O2 car shows LC-1 shows 14.7.

[overdose]

yep mine does the same thing

[drees]

OK guys, I had an epiphany after reading the wikipedia article and think I've figured out why the STi runs rich with the rear O2 sensor sitting free air.

The wikipedia article mentions that often engineers will program the ECU to fatten up the mixture after reading lean since running the cat slightly lean of stoich will quickly lead to it's untimely death.

I notice a similar correction on my WRX after coasting to a stop in gear, the car will idle at .95 lambda for a little bit before pulling fuel and settling in on stoich again. In fact, looking at a recent log, the ECU runs around .92-.93 lamba any time I get back on the gas, even with a tiny 5% TPS after coasting in gear.

I assume that the logic in the STi is slightly different and perhaps relies more on the rear O2 sensor for determining that the cats are running dangerously lean, so fattens up the mixture to compensate.

I would think that simulating stoich at the rear O2 sensor would avoid this behavior.

Here's the relevant portion of the wikipedia article:

Overly rich fuel mixtures are not usually a problem - there is too little unused oxygen for the exotherm to be large enough to cause damage. A slightly lean of stoichiometric mix is far more dangerous, as the oxygen level is elevated, allowing a very large exotherm, and many engine manufacturers design "rich excursions" as a catalyst protection measure in the engine control software.

So, if you're not running any cats in your system, you would obviously benefit from removing any cat saving logic and you probably completely rethink the low load fueling of the engine to save fuel.

[jfitzpat]

Originally Posted by drees

OK guys, I had an epiphany after reading the wikipedia article and think I've figured out why the STi runs rich with the rear O2 sensor sitting free air.

...

So, if you're not running any cats in your system, you would obviously benefit from removing any cat saving logic and you probably completely rethink the low load fueling of the engine to save fuel.

Yes, this is one of my two main theories above.

To others. Please try to understand - I am saying that I found WolfPlayer's *original* post uncompelling. However, I became convinced of the existance of some phenomena from another poster.

My argument is not that the phenomena does not exist, my argument is that because the sensor is supposed to be post cat, the problem has a high probability of being electrical or *emissions* related code in the ECU. Since many people are removing the cat, exactly what is going on is well worth understanding.

Ground offset problems could be tested by loading the heater circuit on the removed sensor with something like a 12 ohm, 20 watt resistor.

Cat protection/missing sensor error code could tested for by programming an analog output on an LC-1 (or LM-1) to simulate a narrow band sensor.

If WolfPlayer is truly 'right', you probably really need to go farther and simulate some cat efficiency, you wouldn't apply an offset to the stoich point in the simulated narrowband from the LC-1 (like you might if you were trying to trick an ECU to run at best economy instead of peak), you would actually 'shrink' the range. Say instead of 0-1V, you would program .1-.9V. Stoich would still be in the correct place, but open loop readings would look leaner on the rear sensor, which is what the ECU would expect if a cat were still installed. This would presumably be superior to simply leaving the original sensor in place (and reading richer than expected).

So, again, my point isn't that the phenomena doesn't exist. My point is that not fully understanding it means that you can't really compensate for it. And, if something is effecting mixture that is not understood, performance tuning is going to be harder.

-jjf

Edit: Drees, regarding closed loop fueling. Yes, if you have no cat you can simulate a narrowband for the front sensor and 'shift' stoich so that it is actually at about 1.05 lambda instead of 1.0 lambda. This point lean of peak is 'Best Economy'. The engine also runs cooler. The ECU thinks it is holding stoich, so no idle code generally needs to be changed. But you will generally need to enrichen your open loop fuel maps because of the ECU's idea of where stoich is.

If you commute, the savings can be pretty significant. Also, better fuel economy helps offset the environmental impact of cat removal.

[Mr NiceGuy]

With all this talk of rear O2s I am wondering if you guys can help me out. Mine was cross threaded into the pipe by Subaru and the threads were destroyed so I didn't put it into my new exhaust. My buddy threw it out becuase he thought I didn't need it. Now that it seems that I do need it I can't find anyone that sells one for an 06 WRX. Everyone says it shouldn't be bad yet so we don't carry it. I can find them for other years but no one seems to know if it's the same for say a 02-05. Does anyone here know or can someone with an 06 give me a part number. I would really apreciate it so I can get my car running right and get my cruise back. Thanks so much.

[R4ND0M_AX3]

I finally got around to hooking up one of the analog outputs from the LC-1 to the ECUs rear O2 sensor input to simulate the expected signal. Things are already better. I didn't reset the ECU because I don't have my streettuner laptop with me right now and I didn't want to risk the car not working for some reason (I've made realtime changes that aren't in my AP yet).

What used to be low 13's for about 30 seconds after it should have returned to closed loop followed by low 14's is now almost immediatly low 14's. Now I can work on the fuel trims without it feeling like a waste of time.

Thanks Innovate. Now I don't need to add a second O2 bung to my downpipe.


PS wolf.. the link to my post in this threads first post is borked.

[R4ND0M_AX3]

Are you guys getting stoich during cruise? I'm still running around 14.3 (.97λ)

[STLSTI]

R4ndom, How did you connect the O2 to the LC1. Analog one or two?. What color cable did you use?. Thanks

[R4ND0M_AX3]

Originally Posted by STLSTI

R4ndom, How did you connect the O2 to the LC1. Analog one or two?. What color cable did you use?. Thanks

analog1 yellow

[STLSTI]

Originally Posted by R4ND0M_AX3

analog1 yellow

Ok, so analog one the yellow wire to the blue wire on the O2 sensor?.

Thanks

[R4ND0M_AX3]

Originally Posted by STLSTI

Ok, so analog one the yellow wire to the blue wire on the O2 sensor?.

Thanks

Analog1 yellow to D25 on the ECU connector B137. At the ECU the signal wire for the rear O2 sensor is white. This is for a 2005 US STi.

I already had the wires inside the cabin so I decided to hook up at the ECU. I had wanted to get a dead sensor and use its connector at the fatcory location. I don't know what color the wire is down there.

[STLSTI]

Thanks that helps alot.

[kingsdman]

[quote=WolfPlayer]I have my wideband in the bung

dude, whats a wideband? sorry new to sti's. thanks in advance