Rough Startups

My car has been plagued with this issue for a while now. Dealer I bought it from looked at it but couldn't find what is wrong with it. They want to wait for the engine light to come on before diagnosing I got a little video that shows what's going on. This is as bad is at get's usually it doesn't hunt that long, but it is what happens every time the car sits for a day or two. Ones it settles at about 1100RPM it will sit there for a minute or two until it warms up, drop down to 750RPM and run like a charm after that. Pulls strong, and starts without any issues while at temperature. If you experienced this before or have seen it somewhere i could use some ideas as to what it might be. Dealer apparently checked for vacuum leaks and compression, it was all good. I did have a topic open a while back about it but without the actual video. Old Thread: https://rennlist.com/forums/996-foru...al-dealer.html Thanks!

Obviously, you're low on gas Sounds like it's running on 4 cylinders. I would connect Durametric and check the misfire counters to see if a particular cylinder is misfiring. In the video, did you hold the gas pedal or the engine was able to hold idle by itself?

Quotes source post: 11670583

I was wondering who was gonna be the first I wasn't touching anything, it was hunting by itself. It is like i said, probably the worst i have seen. Will hook up a durametric in a couple of hours and see what's there. Thanks fro the niput, always helpful! Cheers!

Also note the coolant temp and air intake temp to see if they correspond to the ambient correctly. MAF should read 15kg/hr or a little higher @1k rpm. What are the cam deviation angles, FRA, and RKAT values? This problem shows up only when engine is cold?

Thanks Ahsai, i will make sure i monitor those values. Yes, only when engine is cold. If i go for a ride, shut the car off, and start it up while still warm, no issues, sits at about 750RPM and no hunting or rough idle at all. It doesn't even have to be at full running temperature.

Also search for "Cold start stumble" for similar cases. One particular case was two leaky fuel injectors (not easy to diagnose...pretty much only after eliminated everything else).

Assuming that's a cold engine when you first start it in the video, I'm wondering why your Oil Pressure doesn't shoot up to 5... Just another observation...

...because he's using 0w40 Good eyes.

Nope Motul 5W40. Should it shoot up to 5? I do have the spin on filter adapter with Mobil 1 filter, but i don't remember it being any different with the stock filter.

Quotes source post: 11670743

Mine always shoots up to 5 when cold on both m1 0w40 and Castrol 5w40. LN spin on could have something to do with it but wonder what others with the spin on using same oil are getting...

Ok i did a quick check on the durametric software, no misfires. I tryed to capture some data as Ashai instructed. Unfortunately the software crapped out the first time i started the car when it was completely cold. It ran for about 30 seconds, i reset the software and tryed it again. There was no irratic running the second time around and this is the numbers i got. Let me know what you see. Thanks.

Quotes source post: 11670864

I was kidding. Both 0w40 and 5w40, the oil pressure should still shoot up to 5 when cold. I also don't know the effect of the LN spin on filter but my guess is it should not have any effect because the OE filter normally will also filter 100% and only at the startup instant if the spring is weak some oil will bypass it.

I've definitely never experienced anything like that on my 40th. What muffler do you have? Likewise on mine (with muffler bypass), on a cold startup after sitting a few days, it cranks right up but has a bit of that lope when the secondary air pump is still running. It's nowhere near the severity of yours though; the tach is pretty stable and it just sounds a bit lopey. Like you, I hooked up a durametric, found no misfires or anything suspicious, and just made an effort to drive it more. It also is a bit better when the car has been on the battery tender.

All the numbers look good. I just re-read your old thread now. A lot items have been checked already. One suggestion re the purge valve may be worth trying since it's easy. You just disconnect the fuel vapor line at the quick disconnect then seal both ends before you start your cold engine. Another test is to verify the fuel pressure at start up, which is more involving. Leaky fuel injectors is a strong possibility too but it's difficult to diagnose. I suppose a crude test can be like this. Park the car and leave the engine running, then remove the fuel pump fuse (c4) and let the engine die. In this way, the fuel pressure should be all bled out. Replace the fuse the next time you start the car cold. Leave the key at the last position before crank to leave enough time for the fuel pump to prime before cranking.

Thanks Ahsai i will look into those two things. In the mean time i searched under "Cold Start Stumble" as you suggested, and found this thread. https://rennlist.com/forums/996-foru...old-start.html Doesn't look to promising, he lives not too far from me, and the dealer he speaks about is my local dealer that i took my car to. Too many things check out good, yet the car keeps running bad at every startup.

Quotes source post: 11672500

Yes, same poster (sean) I referred to about leaky injectors. His problem was solved by having two fuel injectors replaced by the dealer. He mentioned it in another thread but he never closed the thread you quoted.

Based on what you wrote the first thing that comes to mind is the coolant temperature sensor is supplying the wrong -- too hot -- temperature and the fueling is lean. I couldn't tell from the audio but if you hear spit back in the intake that's a sign the engine is too lean. (Backfires in the exhaust is a sign the engine is too rich.) If you want to confirm this -- or eliminate this -- you need to capture intake and coolant temperature at cold engine start. I had a coolant temp sensor in my VW Golf TDi act up at cold start. I can't recall the symptoms now -- they would be a different due to the engine differences (compression ignition vs. spark ignition) -- but I used my trusty hand held Actron OBD2 code reader/data viewer to observe coolant temperate and a couple of other things -- only room for 4 items on the display -- at cold start and while the coolant temp reading wasn't that off it was clearly off. I took the car to the dealer and requested the sensor be replaced and it was and the symptoms were no more. Oh, you might try an E-Gas calibration. The behavior from the engine seems a bit much for just this to be the problem, but the calibration is easy enough to do. Keep your foot off the gas pedal during this procedure. Turn the key to the on position. Leave in this on position for 60 seconds. Turn the key off for 10 seconds. Done.

Engine temp looks flat line. My experience is at cold start and idle coolant temperature should increase 1 to 2 degs F. every second.

Quotes source post: 11672531

This seems to make the most sense considering the OP issue is time dependent and based on if he let's it sit for a couple days. FWIW, I had a similiar issue before with the same symptoms on another car.

Quotes source posts: 11672531, 11672571

In the thread i posted, he says that they replaced the two faulty injectors, and even though it seemed like it fixed the issue, it came right back, although not as bad. I am assuming engine and coolant temperature are the same thing? Durametric, i only see the engine temperature. I will repeat the test again tonight and follow the temps a little further and see what happens. Thanks.

Quotes source post: 11672629

Check post #10 here posted on 9/2013 https://rennlist.com/forums/996-foru...l#post10732999 Do you see a later update from Sean?

Quotes source post: 11672629

There are "engine compartment temperature" and "engine temperature" available in Durametric. Engine temp = coolant temp. How much time for the log shown above? I assume only a few seconds worth? was the ambient at around 30C?

Quotes source post: 11672625

Yeah at least quite plausible. Too bad it's a lot of work to remove the injectors for testing. Hope my crude test of releasing the fuel pressure till next start-up works.

Quotes source posts: 11672661, 10732999

Interesting, in his earlier post from 2011 this is what he wrote when he got the car back from the dealer. Post #65. Swapped the injectors from 2 and 4 with 3 and 6. Misfires during stumble moved to 3 and 6. They replaced those injectors and emptied/cleaned the fuel rails. reset everything and the stumble as gone. Engine now runs smoothly. (Awesome, so it was the injectors. Although the work order lists injector valves...not injectors themselves, so maybe he just repaired them? Although he did tell me he replaced them.) Whatever, I got her back and she's fixed. I pay for my new sparkplugs (parts only) and my oil change and head out. I get in the car, which has been parked for two days, and guess what? It does the exact same stumble on start! I took her home as I was late for a shoot already. This morning, she started better, but there was still a hesitation and an rpm dip. Also, as with before, I cannot simply drive away...if I try to drive right away, the engine bogs and I have to over-rev and run the clutch to get it moving. So I don't think it's fixed. Sadly. She does idle more smoothly now (once warm) So when I bring it back for them to fix the leaks, I'll let him have another kick at it I guess.

Quotes source post: 11672664

The log was longer, but i only cut and pasted the first few seconds, once i get home i will check the excel file again and post what is there.

Quotes source post: 11672734

I know. Maybe you can PM him and see what he says.

Just in case you haven't tried to calibrate the e-gas yet, it is probably worth a shot since it is simple to do. I had a similar cold-start problem although not quite as severe. It would stumble and almost stall for a few seconds only when cold. Macster recommended calibrating the e-gas (which I knew nothing about and couldn't find any reference to in my manual). But it worked and the problem has not shown up since. https://rennlist.com/forums/996-foru...0-seconds.html Although for some reason the problem would also only occur when the hvac system was on auto when starting cold. My theory is that the car turns on the A/C briefly on startup (about 7-8 seconds in) which turns on the front fans and pulls a lot of current that somehow makes this problem worse.

Quotes source post: 11677581

I did that, and last two startups have been ok, but i am gonna keep watching this. There might be more to it. Thanks.

Here is another cold start with data spanning two minutes from the start. Yelow row is 25 seconds into it which is the first time engine temperature has gone up about 0.7C. Can those RKAT/FRA values really be that dead on? Based on Macasters theory that engine temp should go up 1 deg every second this may indicate that that coolant temp. sensor may indeed be waving bye bye. This chart i made after a long run with engine at temperature in my garage. Any input appreciated, thanks!

The numbers still look fine to me assuming the measured temp was close to the ambient temp at start up. If the temp sensor was bad, it should be significantly off. At warm up it reports 96C, which is right also.

I don't have a way to really format the data nicely and make it available here but I checked a data log file I have from my Boxster and starting about around 73F the coolant temp climbs a deg. or 2 every few seconds while idling. The only time the coolant temperature reads steady is once the engine has reached some level of warm. That is the engine has various levels of warm. Based on its load, small because it is just idling, and the ambient temperature (and starting temperature) the coolant temperature reaches some sort of temperature equilibrium, the coolant temperature flat lines a bit, for 10s of seconds. The unchanging coolant temperature reading your data capture chart has looks wrong to me. Also, that after 2:08 of idling the engine coolant temperature is just 37.5 after starting out at 21.8 looks wrong too. That the temperature reaches 96 is as I would expect. The engine is warmed up. In fact 204F warmed up. The abnormal behavior you are asking about only occurs from a cold start and is not present once the engine is warm nor is it present upon a warm start. Thus if you are going to capture data that helps you ID what is going on you should strive to capture data from the initial cold engine start and keep capturing until the behavior goes away if you can capture that much data.

A couple of new developments: on daily basis, total misfire on all 6, and my fuel consumption has gone up noticeably. I think my MAF may not be working properly. I have a new one on order as well as the eng. temp. switch.

Quotes source post: 11704371

Misfires on all 6 cylinders is certainly an interesting development. Fuel consumption gong up? Not so much. My guess is the fuel consumption is poor because of the way you have been driving the car. I would take care of the misfires then take a close look at fuel consumption (or any abnormal behavior) once misfires are taken care of. 'course the question is what is at the root cause of the misfires? While I have come across at least one report of MAF caused misfires (and I believe the poster) this is a rare cause. But you have ordered the new MAF so I hope it proves to be the cure.

If the misfires were constant on a couple of cylinders then i would think, like suggested before, that it is the leaky injectors, or if it's one bank it could be part of the variocam system, but Total Misfire and increased fuel consumption high probability MAF is either bad, or confused (vacuum leak). We'll see.

Quotes source post: 11704483

For all cylinders to be misfiring doesn't require all 6 injectors be bad. Just two, one per bank. The DME dials back the fuel to address an overly rich mixture due to a leaking injector and the cylinders with good injectors misfire because they are too lean along with the too rich cylinder with the bad injector. There is the possibility that just one bank has a problem injector and the other bank is misfiring due to the other bank's troubles. A misfire just means the DME detected an under (or over) performing cylinder. The cylinder's power stroke does not spin the crank and flywheel hard enough, doesn't impart enough acceleration to the flywheel and if this happens too many times over a span of time the DME flags this as a misfire. If one bank of the engine is running poorly the other healthy bank might develop sympathetic misfires, well, misfires caused by the upset from the other bank. We know at some point cylinders can help one another -- that is the purpose for the resonance flap in the intake -- so there has to be circumstances abnormal to be sure, during which they can negatively affect one another. There is also the possibility the misfires are due to a bad flywheel. This is a rather rare condition (about as rare as all cylinders misfiring due to a bad MAF) but it comes up once in a while. If the dual mass feature has failed enough this can allow the flywheel to have excessive and inappropriate movement. The DME can't distinguish between the lack or or the excess of acceleration imparted to the flywheel by a weak cylinder or by a flywheel that has play. But see what the new MAF does. What you learn after you bolt this on will help you decide what your next move should be, if there is a next move. The new MAF might be the cure. I hope it is. The intake air temperature readings you posted some few posts back sort of makes me like the MAF for this.