I've seen it with O2 sensor codes, EGR codes, and misfire codes. As you say, they come back again if you don't fix the underlying problem. Just a matter of time and circumstance. My Lumina 3100 throws an EGR code once a year when I take my vacation trip. Usually in mountainous territory, but sometimes when I push the engine at high speed.
It isn't that the OBDII system is flaky, it just saw something that wasn't possda be. Clearing the codes does just that, clears them for the time being. If the sensor wasn't working at all, the code would have came back right away. Since it didn't, then at some point in time, the sensor was "out of it parameters." You can take a scan tool and watch them work. If there is a problem, the code will reset after
10 or so "drive cycles," might not come back for awhile or even at all. It/they are probably getting bad so it wouldn't hurt to replace them.
some engine computers don't seem to be programmed for mountain descents with engine braking. a prolonged engine braked descent will trip a sensor code on an older honda pretty much every time.
Every time your engine coughs up oil or soot it slows the sensor down by covering parts of it. Run the engine hard and it clears up a little.
Changing it now will probably avoid some hassles. A slow sensor can cause the car to cycle between lean and rich every time the gas pedal is moved. It makes the car lurch, lag, and generally feel unresponsive. Those rich to lean cycles will age the catalytic converter more quickly.
The one exception to all of this is a temporary oxygen sensor failure after an oil change. That means the shop made a big miss and oil fouled the sensor's outside vent.
um, it's constantly cycling anyway as part of normal operations - to determine the fuel/air bias to achieve stoichiometry. it's got nothing to do with the pedal being moved.
that's often a sensor temperature problem. cold sensors don't respond quickly so the engine defaults to its "safe" ratio. that gives poor performance.
it's not exactly a "vent". and it's almost impossible to "foul" in practice because it relies on solid state diffusion, not air flow. the diffusion process takes a long time and a thin layer of oil isn't much of a diffusion barrier, especially not at elevated temperatures.
Yes, it's a diffusion through very tiny gaps, filters, or holes in the back side of the case. It might even vent into the cable sheeth. It still temporarily malfunctions when oil burns off of it. Oil smoke and hot soot does a fine job of consuming oxygen.
Also, the ECU has compensation for a certain O2 sensor delay. When it's slower than that due to failure, the negative feedback loop becomes a positive feedback loop for certain frequencies. It causes a bouncing or lurching feel when moving the gas pedal.
unless you have some weird sensor from some fringe manufacturer, it /does/ "vent" into the cable sheath and nowhere else. air "flow" is measured in molecules - it's absolutely tiny.
i think you have bigger problems if you've burned off the sheath.
it doesn't have "negative" or "positive" feedback, it has open and closed loop. if the loop is closed, the computer is able to read a signal from the sensor and uses it to measure stoichiometry by swinging the injection between rich and lean and observing the resultant sensor voltage. it doesn't even try to go into closed loop until the engine has reached a certain temperature. when it reaches that temperature, it will try closing the loop based on it's internal map, plus the last known correction factors. if the sensor isn't responding quickly enough, the computer ignores readings and stays in open loop [and typically sets a trouble code]. some fuel injected cars rely solely on their map and have no feedback loop at all.
bottom line, if you're experiencing lurching, you probably have another issue that's causing it.
Closed loop is another term for negative feedback. Delayed negative feedback is positive feedback for certain frequencies. Stable closed loops are programmed to compensate for a specific delay in the system. A large change in delay (shorter or longer) will cause positive feedback (overshoot, ringing, or full oscillation).
For an example of delayed negative feedback, try balancing a very long stick vertically on the open palm of your hand. Until you learn to compensate for delayed visual feedback, you make rapid back and forth motions that become larger and larger until the stick falls.
ok, but it's not /called/ "negative" or "positive" feedback for car oxygen sensors, it's called open and closed loop. an engine can run just fine in an open loop condition and doesn't become unstable. indeed, default "safe mode" on encountering a problem is to /go/ open loop.
Some OBDII systems need to be driven 50 or 60 freeway miles to actually get all the checks done. City miles don't count. I just went through this with the bug, and a buddy is currently going through it with his son's jap car. Found out the hard way by resetting the computer after a hose detached error, driving it for a month, then getting smog check notification (which I'm sure I didn't get last December when I renewed). Had to drive it home on the freeway, then back the next day to get all checks done - 66 miles each way. Check engine light came on 5 miles before getting to shop. So I had to get a new cat. Funny how that happens a year after ten year warranty expires.
DMV system is even more messed up when the smog rules change. They sent the old letter from last year when the rules were different, the shop I liked in the past wasn't certified under the new rules yet.
MotorsForum website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.