AFR Riddle

So I was doing some light datalogging on the way home from work and thought I'd see if I could get a few more cruise data points at various RPMs by maintaining highway speed in different gears 5,4,3 when I decided to go back to 5th the car was bucking a little and I noticed the AFR was running 17+. I thought maybe the ecu got goofed up so I turned the engine off and restarted but no change. Great, must have lost a hose and be sucking air after the maf. But then I noticed as I got off the highway that it was running pretty rough and was down on power, but occasionally would get a little spurt of power. AFRs still running way to lean. "I know, an injector must have gone bad." I get home and start unplugging injectors while idling and sure enough cyl #1 isn't firing. Ha, found it! I swap in a spare injector and start the engine, no change. #1 still not running, great, now what? bad coil pack. I go to swap the coil pack with another cylinder and find the#1 coil is unplugged. Huh? plug it back in and all is good. But now I'm puzzled, injector squirting, but plug not firing and the wideband shows the engine is running super lean. Shouldn't it be showing rich? (engine SR20DET)

In reply to bigben:

maybe the ECU is cutting fuel to the cylinder with no spark?

If the cylinder didn't fire, the oxygen (O2) that ~should~ have combined with the fuel (HC) in combustion to produce H20 and C02 didn't happen, so all that Oxygen just headed out the exhaust like nothing happened (because nothing did happen).

O2 sensor only sees a ton of O2 in the exhaust, and goes "OMG! Clearly we are WAY too lean!" It doesn't see the raw unburned fuel there as well. If you had an HC sensor, then we'd be laughing, because then we'd know exactly what's going on.

SkinnyG wrote: If the cylinder didn't fire, the oxygen (O2) that ~should~ have combined with the fuel (HC) in combustion to produce H20 and C02 didn't happen, so all that Oxygen just headed out the exhaust like nothing happened (because nothing did happen). O2 sensor only sees a ton of O2 in the exhaust, and goes "OMG! Clearly we are WAY too lean!" It doesn't see the raw unburned fuel there as well. If you had an HC sensor, then we'd be laughing, because then we'd know exactly what's going on.

This. You have to remember that it's an oxygen sensor and not a mixture sensor, which can lead to all kinds of situations where extra fuel is being dumped through the engine and yet it suggests that your "mixture" is "lean."

The guys above are correct. Misfires will show as lean due to extra oxygen in the exhaust.

In reply to rslifkin:

You will also show lean if the engine is running so rich that it is misfiring... and you will show lean if there is an exhaust leak upstream of the O2 sensor or if the O2 is too close to the tailpipe, since fresh air will get pulled into the leak by high velocity exhaust gases.

In reply to Knurled:

It can even read lean if there's a leak right AFTER the O2 sensor, or the exhaust pipe is very short after the O2 sensor. The acoustics of the exhaust will suck fresh air back in after the exhaust pulse, causing a lean reading. This is why most O2 sensors recommend at least ~10 pipe diameters of tubing after the O2 sensor.

In reply to mck1117:

Yep, that too although that is more a case of "induced problem". That happened on the LS3 Solstice, car was wired up and ready to go for everything except the exhaust, which ended 2" after the front O2s because the exhaust was going to be subbed out to another shop. (The engine bay looks like it was designed specifically for the LS/T56 combo. The floorpan, however, does not have any room for a left side exhaust) Start the engine up to verify everything and it just kept running richer and richer. Well, yeah, it'd do that.

Makes me wonder about those little devices that clamp a wideband into the end of the tailpipe. That can't be accurate at all, for all sorts of reasons.

Thanks for the responses, I guess I need to go back and read up on how an oxygen sensor works. I get that a mis-fire or non-firing cylinder adds extra oxygen and therefore the sensor reads this as a lean condition; however what I don't get is if it is only reading oxygen content then how does it read conditions richer than stoich? At stoich, theoretically, all oxygen is consumed and oxygen content should be zero. How can the sensor read negative O2 content?

There's still a little bit of oxygen left at stoich, it's just not very much. The amount of remaining oxygen declines further once you get richer than stoich.

Stoichiometric combustion means that all the fuel is consumed, not all the Oxygen. If all the Oxygen were consumed, there would be no need for catalytic converters, because there wouldn't be any Oxygen left over to combine with Sulfer or Nitrogen.

So what we have here is theory vs. reality. [An "ideal" fuel/air mixture in which both the fuel and the oxygen in the air are completely consumed is called the "stoichiometric"]

"Ideal" doesn't actually exist. There is no such thing as the perfect chemical reaction, but sometimes it can get close.

Some notes;

Most engines I've worked on has about 0.8% O2 left in the exhaust before the catalyst system. Only some special engines are lower than that.

And for the O2 sensor, it's actually measuring the balance of reductants to oxidants. The sensor is a mini catalyst, and the reaction is what generates the voltage. It will give good and accurate readings even when what is left is in the single didgit of ppm.

For a misfire, even fueled, it's not mixed well enough for the balance to burn properly. It will ignite, but catalysts won't be happy. And remember, the O2 sensor is a catalyst.

bigben wrote: Thanks for the responses, I guess I need to go back and read up on how an oxygen sensor works. I get that a mis-fire or non-firing cylinder adds extra oxygen and therefore the sensor reads this as a lean condition; however what I don't get is if it is only reading oxygen content then how does it read conditions richer than stoich? At stoich, theoretically, all oxygen is consumed and oxygen content should be zero. How can the sensor read negative O2 content?

Magic!

Actually what the O2 sensor reads is not "oxygen", but rather "lack of oxygen relative to the outside world". (Thus why an exhaust leak blowing onto an O2 sensor can cause it to read wrong, too) Running rich makes for less oxygen in the exhaust than the cell is calibrated for, which causes ions to flow across a junction, which is where the current coming up the O2 signal wire is coming from.

Widebands simply alter the switching point. (Well, that's grossly simplified, what a wideband does is have two cells, and separate chamber between the cell that the exhaust gas element is exposed to, and the one that is exposed to the outside world. Then you have a controller that pumping a small amount of current in one direction or the other so that the "text mixture" in that chamber makes the exhaust element read stoich. Air fuel ratio measured is a function of how much current it takes, and in which direction. This is why widebands need controllers, while narrowbands just need a voltmeter)

I may have had a class on the subject recently.

But you are right, incomplete combustion can throw things off.... and since the whole point of closed loop control is to keep the catalyst happy, more and more manufacturers are putting heavy weighting to what the REAR oxygen sensor is doing when factoring fuel trim. Some manufacturers have three O2s per bank, or wideband rear sensors, too.

SkinnyG wrote: If the cylinder didn't fire, the oxygen (O2) that ~should~ have combined with the fuel (HC) in combustion to produce H20 and C02 didn't happen, so all that Oxygen just headed out the exhaust like nothing happened (because nothing did happen). O2 sensor only sees a ton of O2 in the exhaust, and goes "OMG! Clearly we are WAY too lean!" It doesn't see the raw unburned fuel there as well. If you had an HC sensor, then we'd be laughing, because then we'd know exactly what's going on.

That is how my Espirt (RIP) died. I think. Ignition problem, raw fuel dumped out the cylinders, catalytic converter, FIRE.