2008 S6 - O2 sensor or something else?

A6 All/ S6 All
#21

All the stuff you posted sounded like my situation. I put brand new injectors in at 91k and it’s been smooth sailing since then (now have 103k). Prior to that it was constantly chasing my tail. I did discover a few issues here and there that needed fixing, but ultimately the injectors were the root cause for me. Seems like that is the case for others too.

I am actually thinking about changing all the injectors at carbon cleans 50k apart just to deal with this issue proactively going forward. I have the old injectors that I pulled. Maybe I’ll send those out to JHM for cleaning and swap them in at 130k or so when I plan to do the next carbon clean.

#22

Its a hard call. You can look at the LTFT but the car is giving a rich code and then a lean code. So that is hard to get to the bottom of. Regardless of what the situation is injectors are a major thing you really want to make sure you know 100% are good. Your in a tough spot not knowing if they are good or not.

You got the code for lean and rich on the same day so its hard to tell. You almost want to clear the code and see how long before it comes back.

#23

I bit the bullet and ordered injectors from JHM. $600 after the core charge isn’t the worst thing in the world. Plus it’s obvious that this is just one of those things that literally HAS TO BE DONE to every C6 S6 on the road. Stupid me for thinking the previous owner actually replaced them.

Justin - as a side note, I did drive around last night with VCDS in hopes of seeing anything out of the ordinary. I did notice at one point the short term fuel trim (ECU 2, group 33, I believe)did jump up to about 25%. I did not think of viewing the long term fuel trim. Also, I did not see consistent misses on any of the cylinders. I think #5 had 2 misses during the 10 or 12 mile run that I did. The misses did not correspond to the lean condition, BTW.

The CEL seems to come back on after 10/20 miles after clearing it.

#24

[quote=“Nebuchadnezzar,post:23,topic:8810”]
And lean is what you’d think if it was bad injectors.
Yeesh. These cars can be frustrating at times.

#25

Ok from what you mentioned thats a VERYVERY big deal that means that your car is maxing out the additve for fuel support.

when it comes to the injectors from JHM its worth the money 100x over. I have been really looking into a S6 and I have been working with a lot of FSI guys over the past 6 years and injectors are the quickest way to wad up a motor and the other injector cleaner service is a half ass way of cleaning injectors.

A quick example. There was a big hp Golf R car that was running a stage 3 turbo kit his motor blew but it was just one cylinder that was damaged. He thought it was a bad injector that did the damage. So he sent his injectors out to the typical injector cleaning service only to have them tell him the injectors were fine. Thinking that the injector was still the case even tho the first typical test said it was fine the typical test only tests at 5bar MAX thats 70psi thats what the car idles at. JHM tests the injectors at over 1400psi… needless to say when the guy with the golf R sent his injectors to JHM… JHM found the bad injector told the customer. The cusomer had numbered the injectors and the number of injector JHM found to be bad was the same cylinder that was bad on his motor.

Moral of the story. you can save a few dollars getting cheap injector testing and cleaning but whats the point on saving a few dollars when it costs over 10 thousand dollars for a new motor. knowiing what I know now and knowing that the standard injector cleaning service tests and flows injectors at 70psi and JHM cleans flows and tests them at 1400psi to me it would be foolish to get your FSI injectors cleaned with the standard service. 70 psi clearly isnt enough

#26

Yea, in hindsight I should have just done the injectors when I had the intake off. Like you said… the cost isn’t necessarily that bad given you send your old ones back and get the core charge back, and the risk for prolonged operation with bad injectors can be well into the thousands.

I got the shipping confirmation from JHM today and they should get here on Thursday. Looks like I’ll break it all down on Friday night after work and then put everything back together on Saturday and cross my fingers.

Ironically I scheduled my car for its annual safety and emissions testing this Wednesday. I thought I had a good chance of fixing this issue for good with the intake but I guess not! I might have to see if I can “sneak” her through emissions testing by clearing the CEL and making sure all monitors are set and PRAY that it doesn’t shit the bed between it being parked and them bringing it in and scanning it.

Thanks everyone thus far on your input through this repair. I’ll be sure to keep this updated as I make progress.

#27

So today I installed the new injectors from JHM and got everything put back together and still have the same code as referenced earlier in this thread (p2089 too lean, bank 2) after driving for about 15 miles.

While in for PA state inspection, I had the shop do a smoke test and they found two leaks which I addressed (One at the oil separator and the other at the suction jet pump for the brake system). Why I couldn’t find them with starter fluid is beyond me when I sprayed everything down a few weeks ago.

I realize this doesn’t mean that there are no additional vacuum leaks present since typically only the more prominent ones are spotted and then it’s just a process of elimination afterwards as you track down the smaller ones. I decided to buy a cheapo smoke machine on eBay for like 75 dollars. I’ll hook this up to my compressor and see if anything pops up later on in the week and report back.

In the interim, does anyone else have any suggestions? I am leaning towards swapping the MAFs around and seeing if the lean/rich issues follow the sensor. Could one of the high pressure fuel pumps be going bad?

As a side note, I will say that I do notice the car has picked up some more power, most noticeable at the top end so the injectors weren’t by any means a waste of time and money.

#28

I tested my original manifold (it was the updated version) and found that it was leaking at the seams at the corners.
Mechanic pressurized it at low pressure and put soapy solution around the perimeter. It bubbled up quite a bit. I’ve got a video of it saved somewhere. We replaced the manifold, which was unfortunate because it appeared ok otherwise. But, in my case we ultimately replaced the injectors to eliminate the codes.

#29

No matter how you slice it. the JHM injectors are a good idea. I have heard of a few guys that say they can feel a stronger upper rpm with the injector cleaning.

Tracking down the vac leak is a big hill in some cases but with the injectors I would wait to see if you get a new code for the car being lean.

#30

Man, I’d be PISSED if my new intake was leaking like you described!

Cleared the lean code and it came back within another 15/20 miles or so.

So I smoke tested the intake/crank case both cold and hot and came up with nothing. I will say, if you have a compressor buying a pre-made paint can smoke tester from eBay is a pretty good investment. (I know, I know… you can make them yourself for like 20/25 dollars but I was feeling lazy!)

Upon further inspection I noticed on cylinder 5 the coil pack connector is missing its tab. Has anyone had experience of one of these backing off a little intermittently and causing issues? Could my issue very well be due to a coil pack connector backing off, causing a cylinder to essentially misfire which in turn makes the o2 sensor detect a lean condition and then the fuel system tries to compensate by adding more fuel in and then it starts firing correctly again causing it to run rich briefly?

#31

If the connector is not making a good connection you generally get a different code. So its possible but I would think that might not be a big issue. If you can obviously make sure you do what you can to get the connection to the coil pack secure

#32

Good point on it throwing a different code if that was the case. It’s probably not an issue because of the fact that the loom for the wiring is so stiff it will hold it in place. In theory!

I’ve been driving around with VCDS for the last day or two and logging output from bank 2 (where all my issues are coming from). I noticed that after I got a p2098 code (bank 2, post cat o2 sensor too lean) the bank 2 long term fuel trim at idle and long term fuel trim partial values BOTH went to 0% and they are stuck there! Meanwhile the short term fuel fuel trim for bank 2 continues to fluctuate, but is biased towards rich.

Anyone got any ideas?

If the pre-cat o2 reads rich and the post cat o2 reads lean, could this be as simple as someone previously servicing this car have reversed the wires?

#33

http://zenacad.com/chris/media/o2graph.jpg

Logged rear o2 data from banks 1 and 2 in vagcom, viewed in excel and made this graph to better visualize the data.

Blue line is bank 1, sensor 2 (good - no known issues). Orange line is bank 2, sensor 2 (bank that gives problems). Red arrow points out where bank 2, sensor 2 is slow to respond.

Looks like the initial diagnosis of a bad o2 was correct.

#34

I remember reading this and not being able to find the post. When the 02s go to 0% that is not good and depending on where is happening that can create a point where you can do damage to the motor. So just made sure to check the actual AFR data to make sure your still good.

#35

Very possible but keep in mind that bank 1 and bank 2 are both on the same side of the motor (passenger side) I can’t tell how your driving when your logging but its possible that the data plot could have been light throttle off and on where the rpms were low enough that there was a time gap in combustion cycle. Not saying that’s the case just mentioning it as it might be something to think about when looking over the data.

#36

Hey guys… sorry to bring an old thread up but I’ve got some more information about this topic.

Quick recap: Was getting random rich/lean codes from bank 2 - both upstream & downstream sensors (o2 sensors assigned to US passenger side, cylinders 4 & 5). Went through and did carbon cleaning, along with a new intake, new injectors from JHM, new coils and plugs. After doing this unfortunately the codes still come up, and are in fact preventing the car from passing emissions because the drive monitors cannot be set.

With that being said, the feedback I got from the indy shop was that the sensors appear to be working correctly as they can get data from them and they respond accordingly, but they believe something weird might be going on within the ECU that is preventing the drive monitors to be set. I took the car to Audi to verify whether or not the updated software had been installed that addressed some o2 issues on our cars. Unfortunately the update had already been installed and they too said they can see data coming from the sensors, however recommend swapping out the o2s.

Anyone have any suggestions as to what direction to go next? I’d just hate to go through the trouble of dropping the engine, swap sensors only to have the same issue again.

Other things I’ve tried:

  • Swapped MAFs from side to side - combined output at WOT seems to be around 310g/s, which is maybe 10% lower than it should be (could this be causing issues on just one single bank of o2s?)

  • Swapped coils and plugs from 4 and 5 and moved them to 3 and 4 - no change

  • Smoke tested crank case and addressed all vacuum leaks

Latest code I’ve pulled using vagcom:


008600 - Oxygen (Lambda) Sensor B2 S1 
               P2198 - 002 - Signal too High (Rich) - MIL ON
             Freeze Frame:
                    Fault Status: 11100010
                    Fault Priority: 2
                    Fault Frequency: 1
                    Reset counter: 255
                    Mileage: 181092 km
                    Time Indication: 0
                    Date: 2018.06.19
                    Time: 18:46:23

             Freeze Frame:
                    RPM: 2012 /min
                    Load: 83.1 %
                    Speed: 73.0 km/h
                    Temperature: 64.0°C
                    Temperature: 33.0°C
                    Absolute Pres.: 1000.0 mbar
                    Voltage: 13.335 V


Readiness: 0100 0101
#37

Rich codes can be a small part of the 02s starting to fail. Check the LTFT data for that bank and then check the other banks that might start to tell the story. There might be an issue that you just haven’t caught yet.

#38

Thanks Justin… I’m just hoping it’s something over-looked as opposed to dropping the engine to swap o2s.

If you were diagnosing this, what specifically would you be looking for with respect to the LTFT data from the other banks by using vagcom? I did a few runs this past weekend logging data from different banks - if you think it makes sense, I can share the entire log for review.

In one of the runs, I logged LTFTs from bank 1 and bank 2. At the end of the run Bank 1 LTFT was at -0.4 and Bank 2 LTFT was at -3.5 (Bank 2 always throws an error code). Right after that run I logged o2 data from the second ECU. Bank 3 LTFT at the end of the run was at 0, and the LTFT for bank 4 was -.8.

I guess this would be a good visual representation of what the LTFTs looked like after the runs:


    Front of car
bank 3     bank 1
0 LTFT    -0.4 LTFT

bank 4     bank 2
-.8 LTFT   -3.5 LTFT