Rough running engine - no codes

[Octane4evr]

I have a 2003 C4S with the X51 package that I purchased last November. I use the car as my daily driver so I've put about 11k on it since purchasing. The engine has always run a bit rough, like an old V8 with a bigger cam than stock. Kind of a smooth - shudder - shudder - smooth cadence. The motor pulls hard but I feel the shudder even free rev-ing the engine and trying to hold a steady RPM. 

 

At idle, when it's doing the shudder thing, the RPM's jump imperceptibly on the tach but I can see them bounce through the OBD reader. The plugs and coils have been replaced for other reasons and I've used Liqui-Moly injector and valve cleaner several times with no effect. There are no codes in the ECU, no misfire or cam or injector codes at all.

 

My local Indy repair shop thinks it may be the vario-cam not fully disengaging. Oil's been changed twice in my ownership but I've not used any lubrication modifier or anything like Sea-Foam to mess with the oil.

 

Any ideas? The vario-cam thing seems likely but before I spend that kind of time or cash I'd like to be sure.

[JFP in PA]

I would check the MAF values.  Often, when a MAF is either very dirty, or on its way out, rough idle is a common complaint.  The VarioCam system can be fully tested with a Porsche scan tool like the Durametric system (which can also check the MAF values); I would test that before even thinking about replacing any components in the system as they are all rather $.

[flatjoe]

MAF or AOS. Try to open oil cap while idle. If goes worst it is broken AOS. 

[Octane4evr]

So, I finally got a Durametric so I can test the MAF. I plotted MAV and Idle speed in the image below. As you can see, the curves lay right over the top of each other, as you'd expect. What I don't know is if this variation is normal. 

 

I cleaned the MAF with CRC's specialty cleaner. No effect. I pulled the oil cap off with engine idling, slight increase in engine RPM, no increase in the shudder feel.

 

Any ideas?

[DBJoe996]

Pull your throttle body and give it a good cleaning with throttle body cleaner.  Remove the idle control valve and clean it out with carb cleaner, Q-tips and cycle it with a 9V battery.  It should be shiny clean inside.  Once it is all back together, do a TB/idle control calibration...turn ignition key on (not start), wait one minute, then turn key off.  The reason I suggest this is your idle is all over the place from the graph.  Have you checked the oil filler tube for cracks?  Vacuum leaks on the intake system?

 

Following this, use the Durametric and check the cam deviation values, the use the Activations and trigger the VarioCam system on both banks.  Let use know what happens.

Edited October 27, 2017 by DBJoe996

[Octane4evr]

It really does feel like a vacuum leak. I'm more familiar with small block V-8's and this is a lot like it feels when the carb isn't fully sealed to the intake manifold or there's some other vacuum leak.

 

I'll clean the TB, I'm not sure the 2003 has an Idle Control Valve, as you suggest and check for disconnected tubes/cracks in the oil filler. Thanks for the help.

 

Edited October 28, 2017 by Octane4evr
clarification

[DBJoe996]

I remember someone finding that the clamps on the rubber pieces for the intake plenum had gotten loose and it was misaligned creating a vacuum leak.  So while you have the TB out, check the clamps and plenum, #12 here http://www.autoatlanta.com/porsche-parts/hardparts.php?dir=996-99-05&section=107-10

[Octane4evr]

Cleaned and did the TB/Idle reset. No effect on the problem. No Idle control valve on the 2003 engine.

 

Checked around for a vacuum leak. Moved around the Oil Tube to see if there was a leak, no change in engine performance. This time when pulling the oil cap off the engine RPM dropped and got very rough, I may not have had it fully warm last time.

 

I've attached the Durametric screen showing cam deviation, two separate tests. They aren't the same on both sides, not sure what they should be though. When I activate the "valve lift control banks 1 and 2" (There are two of them listed with the same name) the engine shuts off immediately.

 

[DBJoe996]

Obviously something is wrong somewhere.  When I have my Durametric hooked up and activate the Vario-cam on either bank, the engine runs rough but does not stall the engine.  It does seem as if one, or both, of the Vario-cam activation solenoids is stuck.  Maybe someone with more experience can weigh in.  I don't know if this works but you might try turning the ignition on, not start the engine, and while listening underneath the engine where the solenoids are, have someone trigger the vario-cam on one of the banks and see/feel if the solenoid clicks.

[JFP in PA]

When the Durametric triggers the VarioCam units, you should see the real time values jump as well.

 

.

[Ahsai]

For a MkII 996, when one activates the "Valve lift" in Durametric during idle, that will kill the engine due to too much air (high lift at idle). Both my '02 and '03 do that so that's normal.

 

You may want to log the misfire counters for each cylinder during idle and see if there are indeed misfire events. How old are the coils and plugs? Also log the FRA and TRA of both banks.

 

Your cam deviation shouldn't differ that much albeit two different tests. Were they both taken after engine has fully warmed up? When you rev the engine, do these cam deviations change at all? They should be rock solid. They will differ a bit every time you test them after engine has warmed up but they shouldn't differ by 0.75 degrees crank angle.

[Octane4evr]

Over the past couple of weeks I've checked the cam deviation values when the engine is fully warm, they change only slightly one varies from 1.63 to 1.75 degrees and the other from 0.00 to 0.03. They are steady at whatever value they have for that trip throughout the RPM range.

 

The plugs and coils are only about 4000 miles old. The symptoms were present before they were changed. 

 

There are no misfires logged on any cylinder during idle or at cruise.

 

I've attached a log file of a recent trip logging everything I could think of and Ahsai's suggestions. The other thing I checked was the crankcase vacuum at idle. It is 2.5 inches of water.

 

I also found another thread where Ahsai and JFP helped discover that the problem that a 986 owner was having was the MAF. JFP has already suggested this might be and MAF problem. I'm willing to change it out depending on your confidence level after seeing the data. Is there any way to check for a stuck lifter?

 

Thanks for all the help!!!

 

2017_11_11_15_41_50_UW_TRIP_KEEGAN.xls

[Octane4evr]

Here's an updated file with the data as numbers instead of text so it can be plotted.2017_11_11_15_41_50_UW_TRIP - Data.xls

[Ahsai]

In this excel file, did you start with the engine already fully warmed up? Or the first row begins with a cold start?

[Ahsai]

In your second spreadsheet, the rough running columns are in the order 1, 2, 3... The typical order is 1-6-2-4-3-5 (firing order). Did you reorder the columns or change the title?

Second spreadsheet column G says Rough running cylinder 2. It has higher variations and magnitude than any other cylinders at idle.

[Ahsai]

36s of idling here (time from 15:20:41.27 to 15:21:16.82). Red is cylinder 2 (per your second spreadsheet heading).

You can see clearly all the other cylinders are approx  +/- 0.8 per second^2

 

I wonder if you should do a leak down and compression test on all cylinders to compare them. If the numbers are OK, maybe replace the cylinder 2 injector?

 

plots.pdf

Edited November 14, 2017 by Ahsai

[Octane4evr]

FIRING ORDER!!! Now I understand why the column order was wacky - Thanks!  I did swap the column order, not just the labels.

 

It took me a few minutes to get the Durametric set up so the car was warm when I started logging the data. I'll create a log file on the way home tonight and add coolant temp to the list and post the results.

 

I see the same thing on cylinder 2 but I wasn't sure what the acceptable range actually is. 

 

I'm not sure how a leak down test is accomplished, is that something a shop would need to do?

[Ahsai]

Ideally, it would be good if you can drive the car until the engine has been fully warmed up (about 1/2hr of driving).

Then do a log for steady idle for a couple of minutes (assuming the rough idling occurs).

 

For reference, I have a stock '03 996 and a '02 3.8L LN nickies 996. Both have the rough running values for all cylinders close to 0.2~0.3 most of the time and from time to time jump to +/- 0.6 per second^2.

 

So it seems all your cylinders have numbers close to the numbers above except cylinder 2.

 

Edited November 14, 2017 by Ahsai

[Ahsai]

Leak down and compression tests are best done by pros if you're not familiar with those. Basically the test will reveal if the cylinders are more or less the same in terms of compression and sealing capability. If one cylinder has a very different number than the others, clearly that cylinder has a problem.

 

If those tests check out, the next step may be the fuel injector.

[Octane4evr]

Thanks for the great info and advice. I've attached two log files, one for a full trip home of over an hour and one of a couple of minutes of idle after the drive. The rough running values for several of the cylinders are higher than you're seeing on yours. This time it's not #2 though. 

 

I have a call into my shop, I recall they did a compression check when they did the plugs and coils. Not sure about a leak down. I'll post that info when I get it.

Idle after 1 hr drive.xlsx

Full trip.xlsx

[Ahsai]

Very interesting data and I don't think I have any conclusion though. Now I'm not so sure it's a single cylinder issue.

 

In the 1hr file, you got some misfires on cyl #2 and #3 @ 2k+ rpm. In general cyl #2 rough running has more variation than other cylinders.

In the 1 min file, the idle rpm varies quite a bit and cyl #1 has the highest rough running variation. I assume you feel rough idling during that time?

 

Going back to what John (JFP) was saying about the MAF, did you ever try to clean the MAF? Also, will the rough idle disappear if you disconnect the MAF?

[Ahsai]

Btw, also check the variocam+ and intake valve lift control solenoids. Two on each cam cover. There's a metal bracket with rubber seal on each one. They need to be air tight. Air leak here will cause misfires.

[Porschetech3]

The crankcase vacuum that you listed as 2.5 in/h2o is too low. I normally see twice that much. Recheck this value and if it is still 2.5 in/h2o), check the bellows at the AOS to see if it is collapsed. The bellows is the rubber accordion looking hose that attaches the AOS to the head. I have seen this hose get soft/collapse/split many more times than I have seen the AOS fail, which is also quite common,  except when the AOS fails the vacuum reading will be high (> 7.0 in/h2o). Check this and if it's not the source of the problem, I have many more suggestions/tests that I can give. lol

[Octane4evr]

I did clean the MAF with CRC's product designed for that purpose, no change. I'll pull the plug on the MAF and see what happens. Any issue in driving it for a long distance without the MAF connected or should this just be a short test?

 

I did feel the rough idle when the data shows the variation.

 

I'm due an oil change, planning that for next week. I'll check the Variocam+ and intake valve lift solenoid seals. 

 

In reading on this in other threads there was some success others have had by changing the IMS and cam chain tensioners. Any thoughts on doing this preemptively? 

 

Porschetech3, I tested this with a home made manometer, see the picture attached for the principle I used to create it, mine's just a length of clear tubing on a piece of plywood but inches of water is inches of water. I've heard all kinds of things about the AOS and faults with it but I can't seem to find a diagram showing where it is and where all the various hoses are and how to access. Any help would be appreciated.

 

I'm going to get my local shop to do a leak down test, hopefully next week. 

 

Thanks for all the help!

[DBJoe996]

The AOS is a piece of garbage square plastic thing located on the back of the engine behind the left (drivers side) intake manifold.  Total PITA to get to it, much less change any hoses attached to it, unless the engine is out.  Coolant runs through it and there is a diaphragm inside to separate the oil side from the coolant water side.  A faulty AOS is not repairable.  Just have to replace the whole thing.  Here is a diagram http://www.autoatlanta.com/porsche-parts/hardparts.php?dir=996-99-05&section=104-10

 

Edited November 16, 2017 by DBJoe996