[boulderlund]

I'm still working on my misfire problem discussed here http://www.n54tech.com/forums/showthread.php?t=50142

Basically I am getting misfire error codes on cylinders 1-3 only and occasional mixture too rich errors on bank 1 only. I am also getting underboost errors that trigger around 2000 rpm. The fuel trims are strange with the multiplicative control being strongly negative (-20 to -30) on bank 1
and more normal on bank 2 (-1 to 7). The engine is definitely running rich, especially on bank 1. The injector tips, spark plugs, turbos, and downpipes are all covered with soot.

I have done tons of diagnostics and recently several key hardware replacements, none of which have had any effect. To summarize

I replaced
Both boost solenoids
All vacuum lines
All 6 spark plugs
All 4 O2 sensors
Both turbos
All 6 injectors

I performed
A walnut blast
A cleaning of the vanos solenoids

I am absolutely certain that
There are no boost leaks
The vacuum system/boost solenoids are working correctly

Today I read that a clogged pre-cat will increase exhaust back pressure, which will in turn adversely affect the turbo as well as the O2 sensor readings (and therefore the mixture control). I also learned that the cats can be ruined by an excessively rich fuel-air mixture. My misfire problem arose after I corrected a leaky injector on cylinder 3. It took me some time to understand and correct this problem, and thus the car was driven for a few weeks with presumably very rich mixture.

I am wondering if a clogged pre-cat on bank 1 is causing both my misfires and underboost problems. It makes sense to me since the new injectors, O2 sensors, and turbos should all be working correctly. Also before I replaced the injectors I moved the old ones around and the misfires never followed them. I did the same experiment with the coils and plugs with the same result. Seems like it must be bad commands from the ECU that result in the rich mixture and misfires. I understand that the O2 sensor readings are messed up by excessive exhaust back pressure. At least the bank 1 turbo would also not spool up properly (especially at low rpm) with high back pressure.

I found a few posts where similar misfire problems were corrected by replacing the pre-cats. I also read about some tests to verify that the pre-cats are clogged. These involve removing the upstream O2 sensors to see if the car runs better. Seems like this could be counter-productive since other mixture-related problems may occur without good data from the O2 sensors.

Can anyone comment on the likelihood that my problem is due to a clogged pre-cat? Also are is there a better test that can be performed on them?

[RocketSurgeon]

Your situation is really unusual, clogged precat is a possibility. I think the best option is to pull the downpipes and inspect them. Are you tuned? Some guys have really melted their cats at higher performance levels. You should be able to look in the downpipes and see a problem.
Do you have 6 brand new injectors? I saw that you moved injectors in your previous thread. Injector seals need to be replaced every time the injectors are installed, otherwise the head can get damaged.

Your situation is so weird, especially the fuel trims. Clogged cat/exhaust issues are pretty rare, but I would evaluate that first. Otherwise I might look at the DME

[langsbr]

I'm in for following this. My car has a very similar issue but on bank 2 and it adds fuel instead of removes it. I've replaced everything you've done as well. What I notice is that when my fuel trim on bank 2 is adding 30%, the afr for that bank is stuck at 14.93. It never moves, regardless of throttle input. Even lift throttle it stays at 14.93 when it should go to 239:1 or 18x:1, etc - basically free air.

There's been a few posts in the FB group about secondary o2s affecting fuel trims. Multiple very knowledgeable tuners have said it is so, I just wish I knew how much it can affect them. It seems so counterintuitive to have a narrowband sensor affect fueling, especially when its supposed to be just for catalyst efficiency. I checked one of my secondary o2s while under the car for a different reason today and it was loose -I'll probably replace the secondary one in the coming week to see if it makes a difference.

[boulderlund]

Thanks for the input RocketSurgeon and langsbr. I am not running a tune. Currently I have 6 new injectors, complete with new seals and decoupler rings. I had the downpipes out for the turbo replacement but I did not think to inspect them (I really wish I would have!) Even though they were not inspected carefully, I never noticed and rattling inside, which is supposed to be one symptom of trouble.

If a clogged pre-cat is not the problem then I think it would either need to be the high pressure fuel system or the DME. I did have the HPFP and sensor replaced and a firmware update performed by the dealer shortly before the misfires began. I have logged the fuel rail pressure and posted some data here http://www.e90post.com/forums/showth...2#post22440552
I did not get much feedback, but those who responded did not see fault with the data. I would think that a fuel pressure problem would affect both banks equally, so I am currently thinking either the rare pre-cat problem or equally rare? problems with the DME. I have done some research but never saw any threads where the DME was set incorrectly. Can this happen? Could the dealer have screwed it up?

[RocketSurgeon]

See how your fuel trims diverge like that. I think you might have your wideband O2 sensors crossed. When the fuel trims move in opposite directions like that, it makes me think the lambda sensors are plugged into the wrong ports.

Can you post up more logs of fuel trims?

Unplug one of your wideband O2 sensors and read codes to see which sensor you unplugged. Trace the unplugged O2 sensor wire and make sure it's the same bank that the code specifies.

[langsbr]

It might be worth doing, but the DME is pretty smart. I thought mine might be crossed so I swapped the plugs. The DME said in about 10 seconds that the O2 sensor were mixed up.

[boulderlund]

I posted some fuel trim logs earlier at http://www.e90post.com/forums/showth...2#post22440552
I also show fuel pressure there. The strange thing is the divergence in the STFT when under load. Also strange is that the LTFT (additive control) remains at zero for both banks. I have never seen it move. My understanding is that it should increment once the STFT hits the limit, +/- 34, I think.

I will check to see if the O2 sensor wires could be crossed. I really doubt it, but this is easy to check.

[boulderlund]

I checked the O2 sensor connections. The sensor with the black wire is connected to downpipe coming off the front turbo and the other end of the sensor wire bundle is plugged into the forward connector affixed to the fuel rail. I marked everything carefully when I replaced the sensors and this is how it was set up originally. I also saw two different posts confirming that this is the correct placement for the sensors. On my car, the connector for the downstream O2 sensor with the grey wire (aft turbo) had a grey dot painted on it. This connector is the one closest to the downpipes. I saw a youtube video showing the same arrangement.

I also took some more fuel trim logs and have pasted them below. There are some differences between the lambda integrator and lambda multiplicative control, but both show divergences under load. I also enclosed a log of the measured lambda value. This plot was a bit crowded so I limited the time range compared to the other two plots. I am not sure how much lambda should vary from 1. It seems close while under load, but quickly pegs at 2 when the throttle is closed. Looks strange to me, but perhaps there is a reason for this. Also the bank 1 lambda value has much more noise, which seems to due to the noise in the fuel trim.

The other thing I don't understand is why the additive lambda control never comes off zero. I see the integrator maxing out on bank 1, and I understand this is supposed to trigger an increase in the additive control.

[boulderlund]

I made an exhaust back pressure tester to test the pre-cats for obstructions. It turns out that a standard compression tester will screw into the O2 sensor port. I swapped out the high-pressure compression gauge with a 15 psi gauge that I had laying around. Guess what I found? Bank 1 produces at least 8 psi of back pressure when the engine is revved to 4000 rpm sitting on the driveway! Bank 2, on the other hand, could not open the check valve (at about 2 psi) and thus the gauge stayed at zero no mater how hard I revved it. So it looks like I have the rare pre-cat failure on a car that has not been tuned! I feel pretty stupid for not doing this test earlier since it only took a couple of hours. You guys don't want to know how many hours I spent swapping the turbos!

Well done OP

Looking forward to seeing your results after swapping out the DPs.

[RSL]

+1, I've never even thought to try a compression tester in the O2 bung. Excellent diagnostic work.

[jpsanfor]

I am going to look into the back pressures on my car as soon as I get a chance. If you look at a thread I started, I am having almost identical issues:

http://www.e90post.com/forums/showth...2#post22530322

I studied the service records and noticed that my DP cats have been replaced TWICE before by previous owners... (I've only had the car a short time) I'm assuming the injectors had been leaking for a while and plugging up the cats. So with new injectors and my cats possibly being plugged I will update if this is the solution to my problem. Wouldn't a plugged cat have a specific code though?

Edit: I talked to my mechanic friend and he said cars don't have anything that would sense exhaust back-pressure.

[jpsanfor]

Update on my issue here: http://www.e90post.com/forums/showth...3#post22549283

I found I had a clogged secondary cat vs the primary cat being clogged on OP's car.

[boulderlund]

Based on my exhaust back pressure test I bought a pair of used stock pre-cats on ebay (which were removed for a catless upgrade). When I removed the originals I was astounded to see that the honeycomb from the bank 1 cat was completely missing! Evidently it melted as fuel from the leaky injector burnt there. Guess where all the debris ended up? Yep, plastered all over the secondary cat. I took the car to a local muffler repair shop where they cut out the clogged cat and welded in a new universal unit. Now the car runs perfectly! No more misfires, great acceleration, no more underboost errors, no more CEL, SES, etc!!!

I included a few pictures below. The first is a pile of debris that was loose in the midpipe and the second is a look into the secondary cat. You can't even make out the honeycomb pattern.

Finally, the third picture is my back pressure tester. I will post a DIY for this but, for now, I simply unscrewed the high pressure gauge on my compression tester and replaced it with a 15 psi gauge that I bought some time ago at my local hardware store for about $10. It turns out that the O2 sensors share the same 18 mm X 1.5 thread that is used with some spark plugs. Thus most compression testers will have this thread included, most likely on the step down on the engine end of the hose. Simply remove the metal skid plate (6, 16 mm bolts) under the car in order to access the upstream O2 sensors. Unscrew one at a time and replace it with the tester. I found it helpful to have an assistant spin the hose from the gauge end as I guided the threaded end into the O2 bung. Once the tester was in place I simply ran the car on the driveway, revving to about 4000 rpm. Your compression tester should have a check valve which retains pressure on the gauge. This means you can do the test by yourself and even drive the car around a bit if you think you need to check it under load. I read at least 8 psi on the clogged bank and 0 on the good one.

Nice work OP

[MysticRob]

Wow, good work. I wonder how long and how badly that #3 injector had been leaking. Wouldn't think it would be enough to destroy a cat like that, given this is now the 3rd one installed. Any idea how many miles on the old injectors, the car, etc?

[Chris@VargasTurboTech]

Nice diagnosis work! I wish more people were like you, well done sir.

[korguy]

i just came here to drop a sincere thank you. I was able to resolve my misfires and 30FF after performing the back pressure test you showed. For anyone reading this, do your best to determine which cat is clogged before gutting or deleting them both.