Syonyk

I finally got around to reconnecting my O2 sensor and throwing a LED on the O2 sensor signal output from the ECU (up in the diagnostic connector on the S4s).

I observed some behavior that I *think* is normal, but I want to get confirmation on it, and if it is how all the FCs behave, I think there may be some room for significant improvement in city mileage with a heated O2 sensor. My O2 sensor was replaced about 15k miles ago with a generic $20 sensor, which may also explain some of this.

On the highway, after a mile or so, the O2 sensor test lamp was flashing between 4 & 7 hz, with a 50-60% "rich" duty cycle. This seems to be what it should be doing, and I was got 21.8mpg on the tank, with 50 miles of short city driving, and the highway running between 80 & 90 mph with the headlights up. So, I think it's doing it's job. Sadly, this isn't much better than I was getting with the O2 sensor disconnected & the SAFC leaning things out at cruise, but that's a different thread.

The interesting thing I noticed was that in town, the O2 sensor wasn't switching very rapidly at all. During all the city driving I did, when it was in closed loop mode I would only get a 1-2hz cycle rate, with an 90+% "rich" duty cycle. The light would flicker dark, then light up again. However, right after coming off the highway, I would get the 4-7hz flashing for a mile or so when running at city speeds, usually until I hit the first stoplight, then it would revert to the lazy flashing.

Based on my understanding of O2 sensors & exhaust gas temperatures, this indicates to me that the O2 sensor, in city driving, is not getting hot enough. It's cycling, but it's cycling slowly and not providing much useful information to the ECU.

So, my questions:

1. Has anyone else observed this behavior (the lazy flashing in city driving)?

2. Would not having a working air pump reduce EGTs at the O2 sensor location? I have no emissions control junk left, but I don't believe the air pump would be injecting air until somewhere past the O2 sensor (or it would mess up the readings).

3. Is a cheap $20 generic sensor going to respond at low temperatures differently from an OEM sensor? If an OEM sensor is cycling rapidly during city driving, but an aftermarket replacement isn't, this could seriously affect city economy.

4. Has anyone tried a heated O2 sensor for an extended period of time with the stock ECU? If the heated sensor can start cycling rapidly in city driving, it should have a noticeable effect on fuel economy. Many people have noticed that not having the O2 sensor connected doesn't seem to affect city economy much, and a slow cycling O2 sensor isn't much more useful than none at all.

I would try this myself, but I honestly don't do enough city driving to be able to get any decent numbers. My FC is a highway cruiser, and it appears the O2 sensor is hot enough on the highway that a heated sensor wouldn't change anything.

Thanks for sticking with me.

-=Russ=-

SureShot

From my observastions, the ECU is stubborn & difficult to predict.
I have an AFR gauge wired to the narrow band O2 sensor, and an S-AFC the shows throttle position, revs & manifold vac/pressure.

The O2 sensor only runs the fuel show after you have been crusing at light (5-10%) throttle, medium high vacuum, & steady revs for a second.

All other times the ECU follows it's internal (rich) map tables.

The other sensors that affect the fuel feed are:
Air flow meter
Intake air temp sensor
Coolant temp sensor
Throttle position sensor
Manifold boost (or pressure) sensor
Atmospheric air pressure sensor (like a barometer inside the ECU)
Engine speed (RPM) sensor

Common culprits are the intake air temp sensor & water temp sensor.

* The sensors are reliable, but the connectors can get oxydized or dirty.

* Also the sensor values can drift if the engine to body to ECU grounds are not perfect.

Syonyk

Agreed.

Can you tell the difference in the bounce patterns between around town closed loop, and closed loop on the highway after a minute or two? If yours does the same thing as mine, it should be "lazy" around town, even in closed loop mode.

I'm pulling the data from the ECU O2 status light, not the O2 sensor directly, but based on my observations with expected EGTs (higher on the highway), I'd be seeing the same thing off the O2 sensor directly. I may try that here with a spare voltmeter.

-=Russ=-