Poor Idle when warm

Good afternoon I’m new to group. 1999 996 911. Have an idle problem. Cold is great until warm. 650-700 rpm. AC on 950 perfect when warm. AC off back to 650-700. I changed Crank position sensor, Warm up valve on throttle body. Removed and Cleaned intake manifold. No codes. Runs great , 78000 miles. Oil cap and filler tube appear to seal. Any ideas please would help. thanks!

Hi All, An introduction and some questions... I am new to this group with an '04 X50 Cab we bought a couple months ago with 47Kmi. The car is bone stock and runs well and pulls strong at 1Bar, but the idle is high when the car is fully warmed up. So, here is what is happening: The car idles on the tick when cold - just below 800rpm. It is stable (of course a little high right at start-up, but still below 1000rpm) Once the car is fully warm (like when you can check the oil) the idle won't settle down from 1000. Then after several tries, it will throw a P0507 Idle past stop. I have also noticed that if the cooling fans turn on, it will go right down to just below 800 but when they stop, it will climb back up to 1000+ and hunt a little. Here is what I have done so far: I preformed a pressure test (removed the airbox and attached there) and found few small leaks which I was able to fix (a loose vacuum line, loose turbo output, and the pressure sensor). When I got the car, one of the large hose clamps on the air-box was completely loose (I was hoping that was it). All that has been fixed and it will hold pressure from just after the air box just fine (~10PSI). I have also removed and completely cleaned the throttle body. It was pretty dirty. I cleaned the wiper contacts and lightly greased the gears.. I pulled and cleaned the MAF sensor with electronics cleaner. It seemed just fine. I cleaned the pressure sensor on the intake plenum and replaced the "O" ring (it was one of the air leaks). I replaced the air cleaner element. At this point I'm not sure what to do next... If I remove the oil cap at idle the idle will increase a bit and goes back down a bit when I put it back on. I keep wanting to think it is the throttle body just not settling back down, but I have tried lightly tapping on it with the plastic end of a screwdriver when the idle is high and that does nothing. If it holds pressure I'm assuming no vacuum leaks... Weird how it reacts to the fans though... In the service records, the car had been to the dealership a couple times for this issue at about 20K miles. They cleaned the TB and "reprogrammed". The PO said that it has been doing this for a while, but the CEL always re-sets itself when he fills it up with gas so I think he just got used to it. None of this was found during the PPI I guess the two possibilities at this point are the MAF and the throttle body? They are pricey to play parts roulette with though. The symptoms seem pretty specific and repeatable... What are my next steps? In a hopefully unrelated issue, after the car gets hot and sits in the garage or is idling, it seems to overflow windshield wiper fluid... If I pull the cap under the hood you can see the level rise and fall dramatically. The leak is just forward of the driver's door - about three or four ounces. I know this is where the accumulator will spill too, but this is water with a blue tinge and not brake/hydraulic fluid. It evaporates off the driveway. I also have a weeping right front radiator so I will be replacing all three with those super-cool aluminum ones when they get here next week. Thanks so much, Franny

Quotes source post: 4832382

Quote: Also have a 3.4 car. I’m guessing you’re not seeing any present codes to do a basic scan? Rough idle, I always look at the MAF sensor, but usually it should be accompanied by a code, however it could be on its way out. How does the battery voltage look when running? Sometimes the voltage regulator behind these alternators can be funny as well. Do you have any brief smoke on cold start up? If there’s any symptoms there, id look at a possible bad Air oil Seperator replacement. If your oil cap is hard to remove when the engine is running, you may have an AOS problem- that’s an easy way to attempt to test it. AOS is fairly common replacement on these cars, I know someone was looking at creating an AOS solution. Also double check the coil packs, if they are the older style units with a 996 part number vs 997- there could be some stress cracks, possibly causing a weird hunting or low idle if they are arcing. Let me know if any of this helps get you started or pointed in the right direction. Welcome to 6 Speed!

Quotes source post: 4105925

Quote: Sorry, but this is not bone stock. Find out what flash you have and perhaps it needs adjustment.

Good afternoon Well as far as startup warm or cold NO smoke at all. I will try the cap trick. Does the handbrake affect anything as far as startup as was hard to get running yesterday when run after an hour drive and parked 15-20 minutes. Could the Fuel filter be a case? I try to troubleshoot with the easier things first. Also carbon canisters? I appreciate your help. I keep plugging away at this. Stand by. best regards ConnieD

Thanks Wross, It is my understanding that the X50s boost to 1bar from the factory. Is that not true? Franny

Quotes source post: 4832740

Quote: Handbrake should have no affect at all. I would definitely look at the fuel filter, do you know the last time it’s been changed? I do mine every few years regardless as cheap preventative maintenance. The fuel pump could possibly be on its way out too, but I feel most of these cars tend to get more mileage on them before we see many fuel pump issues. Hopefully all injectors are clean and spraying freely too. There are a lot of specific components that can cause the hard start issues, but the more you can rule out without throwing parts at it, the better. If you’re still stuck after troubleshooting all the basic items, it certainly wouldn’t hurt to have a good private mechanic take a look over the car at your local Porsche shop. I’ve not ever had any issues with the carbon canisters on either of my 996 cars, but maybe others could weigh in.

Congrats on the car. Sorry to hear of the problems. What does the car do when you put the A/C on? There is a vacuum line that goes up to the charcoal canister/ gas tank, one of the valves up by the gas tank may have failed? You can disconnect and cap in the engine bay easily. I will get a pic up of it later today.

Quotes source post: 4832740

Quote: Have you had any luck with your troubleshooting?

If you catch all the leaks in the intake air system it can be related to the fuel tank ventilation system according to the OBDII manual… If you got access to Durametric there are a couple of valves you can check in the system for proper operation...

Ya know, you guys might be on to something... What would you expect to happen if you remove the gas cap while the engine is running (warmed up)? I tried that and got nothing. No rush of air, no change on idle and no difference if I put my palm over the opening. I expected some change. Tim, if you can find that photo that would be awesome. Thanks guys, Franny

May be this will help you located the pipes going to the tank and the valves. They are under the right front wheel well liner. One is located besides the filling pipe and the two others are a bit in front of it….I remember not long time ago helping someone with one of those valve found at fault. But cannot found the thread ;-(

Thank again, I'll pull the liner and have a look. Looks like I may have to invest in the Durametric software. Franny

Ok here is some pics. The line has a check valve on it so a boost leak test will not show anything even if the line can by wide open on the other end, under vacuum it is a leak tho. The line, press in the two grey tabs to release. Released. Cap off the engine side. See if that drops your idle.

Ok, I have some results... Tim: If I remove and plug that line (thanks so much for the photos), it makes no difference Now, I do feel a nice strong vacuum on that line so that is a plus, but in no way could I notice a difference (revving a few time and holding at 3K for a few seconds and then to idle) I tried the AC and, boom - right to the correct idle and holds through several revving (my neighbors must love me...). I turn the AC off, it goes right back up to ~1000K and hunts a little. If the rear deck lid fan fires up, no change, but if the two front radiator fans turn on it settles right down to the correct idle and stays put. Does any of that help? Sounds a little sensor-y... I would think a leak would mess with the idle across all different scenarios, no? Maybe with the AC or the front fans on it bypasses some system/sensor? I haven't done the bit under the right wheel because I don't have the Durametric kit. Can I check them without it? Maybe just 12V to the appropriate pins? Looks like one is in my future. Seems like it is a must with this car (or a strong want). I actually feel a little better eliminating things and seeing it actually work properly in certain conditions... Thanks for your suggestions! Any more ideas from here? Franny

Ok. It is a vacuum leak. Once you have a load on the engine the Leak is providing enough flow that the T/B at its stop is enough to control the idle. Try the brake booster line next. It is right behind the line you just opened. You will need to remove the grey clip (on a white coupling) in order to separate the coupling. All these lines have check valves the boost leak testing will not show the leaks.

Ok, so I found what your were describing, removed the gray clip and there was even a little wush of air (engine off...) when I pulled the tube out. I replaced the tube, but not the clip and went out to warm it up again. Got back and pulled that tube - lots of vacuum! The car almost stalled and I sealed it with my rubber gloved finger. No real change. I had someone run the rpm up and down a couple times with it connected and with my finger on it and no real difference. Tried the AC thing again and the idle went down. Put the tube back on and reinstalled the clip. I'm guessing when I find it it should be pretty obvious, right? Thanks again for all your help! Franny

Tim, Do you think there is any possibility that the TB is just inaccurate at the very extent of idle and with the AC on, it has to open a little more getting it out of the bad spot? Not sure why it would be bad just at idle though... Any ides on what to check next? Maybe some of the valves under the fender well as JP suggested? Can I check them the same way as the other two? Thanks, Franny

Hi All, I spent a bunch of time yesterday going back over everything and I just can't seem to find the smoking gun... I disconnected the two vacuum lines detailed above (Break Booster and not sure what the other is for) and capped them off really well. I went for a drive (no brake assist, so I was really careful) and got the car good and warm and no change in behavior. Still idling at 1000 and sometimes drops down - but only when hot (cold the car idles just fine). Re connected the lines. I pulled the intake plenum and stared at the TB hoping I would get some inspiration. I had someone operate the throttle a bit and it moves smoothly, but at idle it hums for about 10 sec and then stops with a slight movement closed and then is quiet. It does it every time... Then I pulled the TB again just to double check. It seemed just fine, but I lubed the big drive gear this time. It is a compound piece with and internal coil spring in addition to the main return spring all functioning to get the rest butterfly position to be open just a smidge. In the resting position, the butterfly is open a bit (you can see in the above photos); maybe 1mm or so. You can push the valve completely shut, but you are pushing against the secondary spring. The movement there is not very smooth... Like two pieces of plastic rubbing against one other. I got a little white lithium grease in there and it felt slightly better. I cleaned the gasket seal and pulled the gasket out and turned it around. I'll replace it next time I'm in there, but it seemed and felt just fine. I put it back on the car and checked it's movement again with my helper. No change there... ...still had that 10sec pause with the hum at idle and then slight close and silent. I re-installed the intake plenum. I did another pressure test at the junction of the airbox. At 15psi it would hold pressure only loosing about 2psi a minute. I think that is pretty good. The problem is that it doesn't test vacuum though, but with the large two vacuum lines ruled out that leaves only one line I'm wondering about. Number 19 in the diagram. Where does that go and can I isolate it to be sure it isn't an issue? So, I got the car all back together and went out for my normal drives. The idle was just fine even once it got warm. We drove it to work today (a better test with a bit of stop and go) and it was idling at 1000 for a few seconds and then dropped down to normal (700-ish?). When it got to 700-ish it would stay there. Can I rule out a leak and focus on the TP or am I still hunting down some leak somewhere? I'm not sure what to check next. I guess this is a dumb question, but what is "normal"? Should it always return to the 700-ish immediately or does it step down sometimes over a few (3-8) second time period? I can see it doing that to allow the turbos to spool down a tad. Also, in the service records the dealership did "re-program" the TB so I'm wondering if they added this pause... Thanks again, Franny

this may be the dumbest suggestion given the expertise the experts are sharing with you but.. as near as i remember, every time you disconnect the maf, you should allow the ecu to reset the TB. turn ignition to on position approx 30 seconds, until you hear the tb reset. than start as per usual. it can't hurt though any idle hunting ( to 1k .. then back down to 750/800 or a/c "on" to normalize ) i've ever had was a vac leak for sure. add: no "step down" on idle. should be 750/800 always.

Can you log iat temp water temp and oil temp? I am thinking you may have a bad temp sensor that is putting the ecu back into cold start mode?

Thanks Tim, I did clean the MAF, but did not perform any diagnostics on it. I found a thread that explains how to check the voltage output. I'll do that as a start. I don't have a logger... I'm thinking a Durametric is in my future That is an interesting theory and I'm up for anything at this point... It does come off the cold start cycle and the idle will go down from stone cold. In fact, that is the envelope that the idle behaves normally. Then as it gets hot enough to check the idle, it sits high, then falls and may creep up just a tad. I actually noticed that it was maybe 50rpm high and when I let out of the clutch in neutral, it slowed down ever so much. I guess that load was enough. Thanks '02996ttx50 for the suggestions. I have re-set the TB as you described after pulling the battery neg connector. Nice to know there is no step down. So, when you found a leak, was it obvious enough? Like a tube that had pulled loose? Do you think you would have held pressure for a while? I'm just trying to actually eliminate any areas I can. Thanks guys... This is super frustrating! The first note of it in the service records is about two years ago and it managed to allude the dealer then and also the PPI. They didn't get the car warm enough. I have fixed a few small issues so that is a plus! Franny

Franny, The part numbered 9 in the vacuum line diagram may have something to do with it. Try pinching hose numbered 12 and see if that helps. Part # 9 is a venturi and a check valve. It may be damaged??

Thanks Tim, Isn't that the break-boost line? If so, I disconnected it at the end (4) and sealed it off (first photo). No change The other end of it goes right back to the intake plenum... I would love to get in and crimp tube 18 and eliminate that whole mess. Hey, how about that bit that goes off to the right (2nd photo). Looks like it goes to the oil tank? I can't seem to get a diagram for that run. Oil gets hot... Is there any check valves there that could fail? I'm making the wild assumption that since I can hold pressure that it must be a check or solenoid valve. Does that sound reasonable? These things could be a little easier to get to... Maybe I'll get some vice-grips from HF and grind the teeth smooth so it won't harm the tubes and crimp a few sections off. Keep the suggestions coming. It is going to be epic when I finally find it! Thanks again, Franny

Quotes source post: 4122754

Quote: Have you thought about just taking the car to a competent shop in the Denver area? Might be easier.

I have, but the car has already been to both dealerships here and one couldn't fix it under warranty and the other missed it completely at the PPI. We do have some good shops here, but outside of performance shops they mainly focus on older cars. I think the shops are great for large R&R jobs and I wouldn't hesitate to take it in for something like that, but a tricky issue like this I can see a shop chewing through $1500 in shop charges and not getting it. For that money I can replace most of the vacuum system and TB and be pretty assured to actually get it. My experience with shops has been so-so and always expensive. I'm betting this is a $30-$100 part - I just need to find it. A new TB is only $450... Also, this is how I bond with a car... My 3.2 Carrera was a bit of a mess when I got it a year ago and it is a dream to drive now. I also learned quite a bit! That is an enjoyable car to work on. The 356A was a basket case and it is now like butter to drive and both did spend a little time in the shop, but anly about 25% of the work was done there and all of it was big stuff I can't do (like and alignment, AC recharge, etc.). I appreciate your suggestion and it may come down to that in the end, but I'm not ready to throw in the towel just yet Thanks, Franny

Ok, so I have a bit more info... I got a hold of a Actron 9580A diagnostic scanner and the car has no codes active or pending (that it could tell). I plugged it in and had someone drive for over 30min in traffic/highway/city and looked at the active data. The scanner will monitor temp (water) and that stayed right on target the whole time and lined up with the gauge on the dash. It also monitors the O2 senesors and the difference between them. They were pretty much identical. The MAF seemed to be going up and down as you would expect with throttle and seemed to settle in at idle about the same every time. It also monitors the throttle position and the RPMs. That is what I really watched... I was looking for some deviation with the throttle % down and the idle up or some inconsistency. What I saw on the throttle position was exactly what was happening to the idle... Now, the car was actually pretty well behaved today and would settle down to the high 700's pretty quickly (within a couple seconds), but it followed the throttle position - it was the throttle that was not settling right down; not a high idle and the throttle closed. Now I'm starting to think the dealership set up some sort of delay to try and "fix" the issue before. The documentation does say a reprogramming, not just a reset. I don't know... What would cause the TB position to hold for a few seconds at a percent or two high and then settle to the nominal idle percentage (~3%)? I did notice this when I had the intake plenum off and had someone actuate the throttle... But that was with a warm at best engine and certainly not the same as actually driving. More information, but more questions... Thanks, Franny

I set the scanner to record and graph over about a minute or so (78sec). The little blip up to 1270rpm was me to get it to do its little trick before the fans turned back on Graphs are: MAF (lb/M) Calculated Load (%) RPM Absolute Throttle Position (TPS%) Coolant Temp (F) Air Intake Temp (F) Ignition Advance (deg) Short Term Fuel Trim Bank1 (ST FTRM1%) Short Term Fuel Trim Bank2 (ST FTRM2%) Let me know your thoughts... It seemed to oscillate in a minute or so cycle. Pretty consistent. When the cooling fans turn on, it settles down to 760 or so and stays there. These graphs were taken with all accessories off (no AC) and the cooling fans off. I'll post the next graphs in the next post... Thanks! Franny

...and the rest

So, here's a question... What is the correct throttle position percentage for a normal idle with the car well warmed up? Maybe the computer is driving the TB below that to try and get the idle down and then eases up, but the idle then creeps up? Am I still chasing a leak or am I looking at a bad component (sensor, or MAF or TB)? Thanks, Franny

I ran another data log... This one is over about 70 seconds... It was after the ride home so good and hot. I didn't touch the throttle at all, the clutch was in, temp was at 198deg, no AC or fans on. The idle never settled down to 780, but the TB never went below 3.2% either. when it is idling at 780, the TB is at 2.4%... I just don't understand why it won't close the TB to 2.4%. It does when it is cooler. It seems to idle high because the TB is being commanded to be open. Is 2.4% correct for idle or is that too low? Arrrg This is so frustrating... Thanks, Franny

I did not read everything real closely but I would replace the MAF if you have not already. You can get a Bosche for decent price from amazon.

I do have a graph of that too; same run... You think the MAF would cause the computer to command the throttle open a bit? The car did throw another P0507 last night ...or if the computer is commanding a lower throttle and the TB just isn't responding, and the %throttle is just being read off the potentiometer in the TB... That would kind of make sense... Because a P0507 says that the computer is past its ability to lower the idle. Weird if it reads so high... It would seem like it has plenty of room to lower it at 3.4%. Thanks, Franny

I am far from a pro but have found when I have a miss/rough idle etc its the MAF. It can cause some real weird problems and in the scheme of things cheap. Have you cleaned the throttle body? I also like to clean electrical connections with some form of electro clean spray.

it's true. hunting idle is *always*? either a faulty maf, and/or a very minor leak. or lastly the tb sticky/dirty. if the maf is fried? the ecu tries to compensate, and idle is where we see/feel it. also, a tb reset whenever the maf is fooled with, is often overlooked. at the risk of repeating myself lol