Rough cold idle

So, I've had my 996 for about a little under a month. It's still under the NY used car warrantee for about another week and has an appointment to go in to the dealer for this. Just figured I'd ask here because I'm curious. So, when it's cold out (last few weeks) the car has been idling a little rough. Typically a 100-200 rpm fluctuation. It sorta feels like it's stubling at the lower end, the engine catches it, raises the rpm a bit, then it comes back down, stuble, repeat. I haven't done any major checks other then pull the oil filler cap with the engine running. Wasn't particulary hard to remove, however with it off the engine ran like absolute crap. Now that its warmed up outside the fluctuation seems to be down to about a needle width, maybe a little more. Sometimes it's better then others. Just curious what you guys think?

What year of car? If an early example the idle air valve could need attention. If a later example, with e-Gas, my only experience with a varying idle proved to be a failing AOS. That removing the oil filler tube cap test is ok but even if you can remove the cap with no difficultly does not eliminate the AOS. The AOS may not have deteriorated enough to subject the crankcase to full engine vacuum. The engine's reaction is like I expected, at least at first. If the engine running didn't improve after a few moments with the cap off this suggests the AOS may be at fault as there it is an intake leak that with the cap off is more than the DME can adapt to. The car is new to you so there could possibly be other things going on: Stale gas (though not likely); plugs; air filter; fuel filter. Regardless, I'd not delay but take the car in ASAP. Be sure you know how to demo the behavior.

Could be the MAF

Sounds like it could be the SAI(secondary air injection unit). OP there is a simple test you can do if you unplug your hose from the SAI unit towards the driver's left rear in the engine bay. Looks like a round 4-5" plastic cylinder. This is essentially a fan that pumps additional air into the motor when the it's cold to improve cold running emmisions. You should feel air coming out of it when you first start your car on a cold day. This can only be done when it's cold outside. This unit will not operate when temps are warm outside hence the reason why probably noticed it acting up more when temps were cold outside. That's my guess based on the info given.

Sorry, 2001 Carrera 4. The problem with demoing is, it is a minor problem once things get warm. I'll be dropping the car off next friday, so they wont look at it until monday. At that point it's a crap shoot. We will see. I'll take a look for the SAI when I leave work in another 40 minutes. Though it says it's about 50* out right now so is that cold enough? Macster, the engine should run normally with the filler cap removed? How long should it take for the DME to adjust?

Quotes source post: 11283978

It is your job is to know how to demo, reproduce the behavior. If this means leaving the car sit at the dealer or shop overnight to get nice and cold and then showing up at the shop early in the morning to demo the behavior, well, that is your job. Based on my experience the SAI should come on even at 50F. It has to be pretty warm for it not to come on but I can't give you a number. You should hear the secondary air injection pump running to know it if is on or not. I know I hear the one in my Boxster come on every morning at cold start (approx. 50F) and often in the evening when I leave work for home (>50F). (You can also use an OBD2 code reader/data viewer and observe the SAI setting in real time.) A healthy engine will react to the cap being removed. It will stumble but not die. It will run rough but not real rough and in about the time it takes to read this sentence (at least once and maybe twice) the engine should idle smoother as the DME adjusts/adapts to the air leak. The engine may not run quite as good as it was before you removed the cap, but it will run rather well. 'course, when the cap is put back on it takes a bit of time for the DME to adjust back the other way. The egnine's reaction/behavior is not as dramatic as the mixture goes a bit rich (vs a bit lean with the cap coming off) and the engine, especially a cold or cool engine, can tolerate this a lot better than going leaner when cool or cold. If the behavior gets less severe when the ambient air temperature is higher or if the engine is warmer this can be a sign the coolant temperature sensor is not delivering the right temperature -- it is signaling the DME the engine is warmer than it really is. It can also be a sign the MAF is acting up and not measuring intake air temperature properly. For diagnostic purposes an OBD2 code reader/data viewer can be used to view/observe these sensor values in real time as the engine is started. If the car sits overnight (12 hours) both sensors at least initially should read close to ambient. They may not read ambient. These cars can take a long time to cool down to ambient and may not ever if they sit outside in the sun. This heats the car up during the day and then as the temperature falls it takes the car time to shed heat. The car's "internal" temperature always lags ambient temperature as this can change in a very short time. But given the engine's behavior I think if the coolant temperature sensor or the MAF and its intake air temperature function is bad you'll see this from the readings. (Bear in mind once the engine starts the coolant temperature can increase one sometimes two degrees F per second.)