O2 Sensor Compensation vs Temperature Modifiers?

While doing more fine tuning, i have a question that I just can't seem to word right or get a straight answer about.

Example: My ECU has an Air Temperature Modifier to add/remove fuel based on the density of air. In this case at 36*F its adding 7% fuel. Now, at cruise, my Oxygen Sensor is trying to achieve 15:1 AFR. The result I see is that the Oxygen Sensor and the Air Temperature Modifier are fighting each other. My engine normally operates at a Closed Loop Compensation around 92%, but I can see it removing more when the modifiers are in play.

So in my head, the reason additional fuel is added at all (coolant or air temp modifiers) is because the fuel comes out of suspension or is not atomized enough to be burned completely. I would expect this to mean that the additional fuel is there to MAINTAIN the desired AFR. Meaning that the extra unburnt fuel should not reflected in the oxygen sensors readings.

What I have been told is the opposite; that the additional fuel results in a lower AFR (like during warm up when I run roughly a 11-12 AFR). My head isn't quite wrapping around this. What is the point of having modifiers active when the oxygen sensor is running at all if it's just going to fight them anyway?

OR am I interpreting my Closed Loop Compensation wrong and its NOT removing the fuel that the modifiers are adding?

This question is mostly for Air Temperature modifiers. Unlike engine Coolant Temperature Modifiers, Air Temperature modifiers are based on a very standard engineering curve for Air Density. I already have tuned my ECU to stop adding any Coolant Modifiers once Closed Loop Compensation is engaged (at 145*f Coolant Temp), but Air Temperature Modifiers should be in play all the time.

some stuff in there is accurate, most is in its correct context.

temp modifiers are normally used to account for changes in air density. colder air is more dense, therefore has more oxygen and needs more fuel to maintain whatever AFR is being commanded. you COULD use a temp modifier to add/remove fuel due to mixing/vaporization issues, but if the O2 sensor is active and being used, it will only be a temporary change.

normally, you would change the target AFR to reflect mixing/vaporization issues, that way the temp modifiers are correct even in this scenario.

you may see some benefit to bringing coolant temp compensation back into your calibration, but it would be entirely application specific. some engines don't need much additional fuel because of low temps, others do, the exact combination of components will determine how much is necessary. if your O2 trims right around the O2 activation point differ in an otherwise identical scenario at full temp, then that is a good indication of how much change would be beneficial in the coolant temp compensation.

Stoichiometric ratio, about 15:1, indicates the correct amount of fuel and oxygen for both components to react with each other completely and leaving none of either component left over. Providing fuel and air in this ratio does not guarantee any of it will react in any percentage.

Quote:

---

So in my head, the reason additional fuel is added at all (coolant or air temp modifiers) is because the fuel comes out of suspension or is not atomized enough to be burned completely.

---

Yes, there are factors affecting combustion and preventing all the fuel and all the air from reacting with each other.

Quote:

---

I would expect this to mean that the additional fuel is there to MAINTAIN the desired AFR.

---

Sort of. But the "extra" fuel displaces oxygen which usually pushes the end result rich.

Keep in mind the O2 sensor is providing a voltage signal based on the difference between the amount of oxygen in the exhaust and the amount surrounding the outside part of the sensor, and the ecm has a chart that correlates voltage to AFR. How close the reported AFR comes to actual depends on a number of variables.

Quote:

---

extra unburnt fuel should not reflected in the oxygen sensors readings.

---

Fuel will not be reflected in O2 sensor readings. O2 content will be. In the case of the mixture richened due to cooler temperatures, with excess fuel and reduced oxygen the end products will be O2 deficient and the O2 sensor would indicate this as rich.

Quote:

---

My head isn't quite wrapping around this. What is the point of having modifiers active when the oxygen sensor is running at all if it's just going to fight them anyway?

---

Your question is good, but it can be asked differently. Ask yourself this: If the O2 is providing an accurate estimation of the AFR, why is the ecm attempting to deliver too much fuel before O2 correction? Is the air density modifier actually predicting the effect of air temp change correctly? Is there another factor that is not correctly modeled or accounted for?

Coolant temp modifiers should be considered important even when the engine is warm. At low rpm the temperature of the intake manifold and cylinder head affect the incoming fuel/air charge more than they will at high rpm. Heat energy transfer depends on time and more time in the intake tract means more heat transferred. If your software allows you to blend coolant and air temp modifiers based on RPM then you might make it part of your goal to adjust those modifiers to accurately reflect the engine's characteristics.

And even when you have the modifiers correct you can still be foiled by humidity. Moisture in the air displaces consumable oxygen but the only sensor that will see this is the O2 sensor, and it happens after the combustion event. I've known people who will monitor humidity in the air on each tuning day although I've never felt that I need to log and correlate that variable. But even if I don't log it, I do remain aware that it is part of the picture.

Thanks guys for helping me bounce this theory around!

I would agree that there are many variables to consider, and they can never be perfectly accounted for and overlap will occur. Just good to know that this is an excepted compromise in our systems. Unfortunately with the age of Commander 950 system I do not have as many variables or modifiers as I like. It would be nice to have barometric compensation, AFR vs Temperature tables, Coolant Mod vs RPM, and others that I just cant control right now.

One question I still have: How responsive is your Closed loop Compensation (or step size) in response to oxygen sensor readings on your system? I find that a change in engine conditions may rapidly bounce my O2 readings from rich to lean, but if Closed Loop was removing fuel to compensate the rich condition (Ex. -15%) it is still removing fuel for some time during the lean condition as it "walks" its way up towards adding fuel. My closed loop compensation step sizes (in % change) ranges from 1-3% depending on RPM. Are you guys more aggressive with your step size or do you find it's better to keep it conservatively slow?

Considering this engine is an 800lb hunk of cast iron designed in 1955 for busses, I'd say the performance I'm getting right now is still exceptional :happy: