Rough running when engine warm

[Stefan]

My car is a 1999 Boxster with a 1999 3.4L 996 engine. The car has 140k miles. The engine has 84k miles.

The has a problem for several months now. When I start the car cold, it always runs perfectly but as the car warms up the idle gets progressively rougher. When the engine is running rough, the entire engine shudders and it sounds like misfires but there are no codes. The problem is intermittent. When warm, the car sometimes idles perfectly. The car goes from idling smoothly to idling rough randomly. I have never gotten any CEL codes since this problem started happening.

Using a Durametric, I confirmed that the "rough running" parameter fluctuates when the problem is happening.

Things I have eliminated:

(1) MAF - problem happens identically with the MAF unplugged

(2) Ignition coils - all checked, no cracks or other visible problems, plugs snapped in properly

(3) Pre-cat O2 sensors - I just replaced them and there is no change in idle behavior (although car gets significantly better MPG)

(4) Intake leak - I checked the entire intake and vacuum system for leaks - nothing found (and idle doesn't go up when the problem occurs)

(5) Resetting mixture parameters - I have tried this a few times after various experiments - no change

(6) Lower fuel injector o-rings leaking - I replaced these - no change

(7) Clogged fuel injectors - ran Techron fuel system cleaner through - no change

(8) Idle control valve - I have cleaned this and the flap moves fine. Also, idle is rock solid even when engine is running rough.

Suspects:

(1) Spark plugs. The spark plugs are relatively new (about 10k miles). I took them all out and they all look fine (and identical to each other). The plugs I'm using are the ones from AutoHausAz (Bosch 4-electrode).

(2) Post-cat O2 sensors - They are 140k old but there are no codes and I didn't think these sensors affect mixture (right?)

(3) AOS bellows - I haven't checked this yet but I replaced the AOS and bellows about 25k miles ago and the problem started happening well after that.

(4) Upper fuel injection o-rings - I didn't replace these because it requires removal of fuel injectors (read: fuel everywhere)

---

Any comments or help on what else I should check or how to further diagnose would be greatly appreciated.

[Stefan]

Replaced the AOS bellows since I happened to have an extra lying around. The problem persists. :(

[Loren]

This has to be something on the intake side - either an air leak or exhaust leak.

Odd that there are no fault codes. What does your MAF read at idle and at 2500 RPM (no load)?

Are your FRA and TRA steady at idle and what are the O2 sensors reading at idle?

[Stefan]

I really wish I was getting a code because that would make diagnosis much easier. I'm thinking that it can't really be a misfire because the engine management system is good at detecting those.

At idle...

MAF

17.45 kg/hour (idle).

I haven't checked the reading at 2500 but I know that the value does go up (i.e. the MAF isn't totally broken).

Note that I am using a 560 ohm resistor on the signal wire. Using a 330 ohm resistor, it was 19.05 kg/hour.

I doubt that the MAF is the cause since when the MAF is unplugged the rough running behavior is identical.

FRA and TRA

0.73 (FRA and FRA2)

0.24 and 0.2 (TRA and TRA2, respectively)

They seem steady at idle but the Durametric rev 1 refresh rate is quite slow IMO.

Pre-cat O2 sensors

0.91 - 1.04 (cycling up and down, same range for both sides)

Do you happen to know if using Propane around a cracked oil filler tube bellows will cause idle to go up? I haven't eliminated a crack in that as a possibility. I haven't been able to see a crack and wiggling the hose doesn't affect the rough running.

[Stefan]

I finally got two codes,

P1127 Oxygen Sensing Adaptation Range 2 (Cylinder 1 - 3) Lean Limit.

P1129 Oxygen Sensing Adaptation Range 2 (Cylinder 4 - 6) Lean Limit.

So now I'm going to recheck the fuel pressure regulator, its vacuum lines, and the fuel return line.

Another possible listed cause of these codes is leakage of the fuel injectors. I have to assume they're referring to air leakage because I can't see why the car would be running that rich unless air is being lost from the system.

--

This is also particularly interesting since I was initially trying to chase this down by looking for a leak in the intake system. But that would cause a lean condition, not a rich one.