Civic Woes

[Bigbadspoon]

I've had this Civic for a few years now, and lately, it's started to do something really weird with the engine speed. While it's not *killing* my fuel economy, it's not helping it. I've only been averaging about 36 mpg lately and this time last year I was around 40 or so. Anyway, any input would be appreciated because I'm at wit's end diagnosing this thing. I refuse to throw parts at it, since I'm in college and pretty much broke (not to mention that's bad practice).

Anyway, every time the O2 sensor signals the engine to switch from rich to lean, the car stumbles a bit and has a small random misfire, until the point comes where it signals back from lean to rich and it runs perfect for the next second or so until the cycle starts again. When the engine is in open loop, this doesn't seem to happen and it runs pretty well, which I guess is because the car just goes off of the manufacturer defaults. I've observed it doing this from idle all the way up to 4k rpm. Occasionally, and probably related to this, it will die on very light acceleration (throttle under 15%) but turn back on when I let off or step on it.

Here is what I have done so far. Please let me know if I've missed something obvious, because I feel like I have...

-Gave it a tune-up. Plugs, wires, cap, rotor, fuel filter, pcv. It was due anyway, so no big loss there, but it didn't fix this problem.

-Run a generic Obd2 scanner on it to see some live data. I can't see O2 cross counts, but the primary sensor appears to be going from 0.1 to 0.9 V pretty regularly. Long term fuel trim is usually around +10-12% depending on the day. It seems to vary a bit. Short term trim cycles at idle based on O2 voltage, but tends to go from about -0.75% to +4%. At higher rpm it changes too quickly for me to see what it's really doing.

-It throws no DTCs

-Checked for vacuum leaks by using the old hand-over-throttle-plate trick. It dies instantly.

-Tested the coil. As far as resistance goes, it's primary is on the high end of spec and secondary is right in the middle. I can't directly check the output voltage, but it can make a spark almost an inch long, so it seems pretty good.

-Bled out any coolant air bubbles. I hear that if bubbles hit the IAC, rpm will fluctuate. I got a few bubbles out, but the problem didn't change at all.

-Checked the coolant temp sensor since false high readings would create a lean condition that the computer would try to compensate for with those rich fuel trims. It was fine.

-Checked the IAT sensor and MAP sensor. Both were good. Engine vacuum at idle is around 19 mmHg.

-Pulled the injectors and cleaned them with gasoline. Checked injector resistance; they were all in spec and all within a few tenths of an ohm of one another. When I listen with a stethoscope, they all pulse regularly; there's no extra clicking or anything I'd expect from a bad spring. They held fluid when I had them out, but it was not pressurized.

-There's no EGR on this car, so that's not a problem.

-I recently pulled the plugs to see if there was anything noticeable. They have a small carbon buildup right along the bottom of the thread, but it's not much; one didn't even have it. They're mostly brown, which I believe means relatively normal operating temps. Interestingly, on most of them, the side facing the arc was a little cleaner-looking, I guess because of how the flame front travels. The electrodes all look pretty much new.

-I cleaned the throttle plate, but it wasn't really dirty to begin with.

-Replaced the IAC with a spare I had lying around, but nothing changed. Interestingly, if I unplug the IAC with the car running the problem gets 100 times worse and can barely even get over 2k rpm because it's cycling so fast. It basically sounds like a racecar hitting a rev limiter, but at speeds barely over idle. When plugged back in it goes back to normal.

So, there it is. Thank you for reading all the way to the end. Hopefully, you still have some suggestions.