onagao

I thought a while before posting this, unsure of what sort of reception it would get. Part of me is worried that I'll get a lot of half-thought ideas that don't really point me in the right direction. This is a forum, after all, and I realize that comes with the territory. Nonetheless, I'm going to describe this issue because I really don't have a very good idea of what is causing it.

My mods are in my signature, with the Stillen Headers and Motordyne ART Pipes coming after my dyno tune. I'm currently scheduled to have my car re-tuned at Uprev's shop in Austin on Saturday, to make the most of my latest mods and get everything running the way it needs to be. The problem described below was much more pronounced before my eTune (yesterday), and now only shows up under certain conditions. I checked, and the car is currently not throwing any error codes.

The Problem:
Bank 1 (sensor 1) is reading approximately 1-2 AFR below Bank 2 (sensor 1). Below is a series of charts taken from my Cipher datalog that I captured on the highway.

I reproduced it twice while driving steadily at about 70mph around 2700rpm or so with very little be given in terms of throttle. After exiting the highway and coming to a stop, the discrepancy normalizes and the AFR settles to an idle value of approximately 13.8 or so. Under WOT conditions, the AFR's from both banks are within the expected margin of acceptable differences, hovering around 12.9 or 13.

The problem at work:

Idling shortly after:

Full throttle shortly after:


Really, this issue only comes up under certain conditions, but it makes no sense to me. I got the car up on a lift and we inspected all the pipes, welds, and connections. From the looks of it, there are no cracks, leaks, or visible damage of any kind.

After a quick conversation with Jared from Uprev, he suggested that it probably wasn't a tune issue and that if we thought it was possibly a sensor going bad, a good start would be to swap the sensors and see if the issue persists. However, due to time and energy constraints, I don't want to do that just yet. It's not like swapping the downstream sensors, and is a lot more work-intensive than I had the time for.

Here's the two things that are throwing me off, though. First, when O2 sensors go bad, my understanding is that they typically get really funky with their outputs. This is only intermittent and under certain circumstances. Again, the issue was worse (15.4AFR in Bank 2, 12.2 AFR in Bank 1 while idling) prior to the eTune reflash administered yesterday. Further, the readings look fine under WOT. The second thing is that the correction factor shows a 10-15% correction increase for Bank 1, which suggests to me that it's running rich but reading fine - like something is telling it to correct for no reason.

I can't make heads or tails of this, and if any of you guys have some worthwhile suggestions or have seen anything like this in your personal experience, I will welcome any help I can get. I intend to contact Jared (Uprev) tomorrow sometime and send him the datalog of me catching this issue at work. There's no telling whether or not that will yield a solution, and I'm getting concerned that I'm going to travel all the way down to Uprev in Austin and get my car tuned with something wrong on the car. That's pretty much the last thing I want to do with my time, money, and beloved car.

Thanks in advance, guys.

onagao

Forgot to mention this, but I'm not sure if it's even relevant:

Before the eTune, when it was much more pronounced problem, my ECU threw the following SES code:
- P2A00 A/F Sensor 1 Bank 1 Circuit Range/Performance

After reflashing with the new eTune map, that code went away, but I was shortly greeted by the following codes:
- P0139,02 Sensor Circuit Slow Response (Bank 1 Sensor 2)
- P0137,02 Sensor Circuit Low Voltage (Bank I Sensor 2)

The latter two didn't really bother me much, seeing how I don't really care what Sensor 2 says for either bank (downstream in the ART Pipes). I'm probably just going to disable any codes they can throw at me. I cleared those two, and they haven't come back.

Just for what it's worth, in case it gives any more insight into the nature of the problem.

maXmood

iTrader: (16)

contributing to this thread, might give u a better overview.

from my datalogging, i got the following:

(before flashing with basic tune) *i had the following on the car: stillen intake, plenum spacer, borla TD exhaust, berk HFC*

1- at idle:

Bank 1: bouncing between 14.75 to 14.81
Bank 2: bouncing between 14.81 to 14.89

2- at WOT:

Bank 1: starts @ 14.78 and going rich till 11.84 @ 62mph
Bank 2: starts @ 14.70 and going rich till 12.58 @ 62mph

@ WOT, the margin is always 0.50 to 0.90 difference across the log. sometimes the value is the same on both banks at a given time.

------------------
Basic etune (the one i received when i bought Osiris) *not an etune*

1- at idle:

Bank 1: bouncing between 14.63 to 14.81
Bank 2: bouncing between 14.46 to 15.48

2- at WOT:

Bank 1: starts @ 14.12 and going rich till 11.46 @ 62mph
Bank 2: starts @ 14.28 and going rich till 11.72 @ 62mph

@ WOT, the margin is 0.20 to 0.90 difference across the log. and only sometime it's the same on both banks at a given time.

------------------
e-Tune
1- at idle:

Bank 1: bouncing between 14.41 to 14.65
Bank 2: bouncing between 14.81 to 15.03

2- at WOT:

Bank 1: starts @ 13.69 and going rich till 12.33 @ 62mph
Bank 2: starts @ 12.81 and going rich till 12.72 @ 62mph

@ WOT, the margin is 0.40 to 1 difference across the log. and never the same on a given time..


hope that helps a little..

rcdash

iTrader: (18)

The driver's side (bank 2) reads leaner at low loads (vacuum) because of plenum design. The plenum neck curvature enhances airflow towards the passenger side intake runners which empty into the driver's side cylinders. If you have bank to bank closed loop correction (Haltech, FCON, etc), it's not a problem. At load you will find afr equalizes. Nothing to worry about really since under high vacuum, a lean running cylinder isn't a big deal.

If you think the AFR discrepancy is excessive, I would look for an exhaust manifold leak that is drawing in free air and fooling the O2 sensor.

Vince@R/TTuning

iTrader: (7)

Pretty much what RCDASH said...your still within a 10 percent cross value so i wouldnt worry about it. Also your technically not running lean...your still running at the target air fuel, your ECU just needs to add 10% more fuel to get it there...

onagao

Thanks, guys for the insightful replies. Hopefully, I'll have this sorted out this afternoon when I get my car tuned... but there's no guarantee.

Here's what's really bugging me about this: It's not that it's adding fuel to make it reach the target AFR in Bank 1. Rather, it's adding fuel and causing it to run rich. As though it's reading things the way it should, but reacting completely irrationally. And the worst part about all of this is that it just gets worse and worse with time. Last night, after feeling like my car was running a bit weak, I sat down in my car with my computer and plugged in Cipher to check on it. (It also had just thrown a new SES code). The error code was P2A00, and sure enough, the car was idling unnaturally - 14.7 in Bank 2, 13.3 or so in Bank 1. Correction factor for Bank 1 was over 15% higher than that of Bank 2.

It just doesn't make sense to me.

And if this were something that would only affect low-load applications, I might not care so much. However, the ECU is clearly doing its trimming, learning, and adapting under these circumstances, and it's going to negatively affect my car. That's definitely unsettling.

Any more help?

mzdfc3s

iTrader: (2)

i thought i would hit on your "sensors going very quarky" prior to or during failure. i have had P2A03 sensor 1 bank 2 ses repeatedly for several thousand miles. dealer replaced that sensor twice, and also the ecm. still ses comes back on. i just got the car back again interested to find out that they replaced the rear bank 2 sensor based on some tsb(to which he couldnt tell me a reference number?) so i left skeptical like i did the last 4 weeks in a row. but 200 miles and no ses? maybe this is it but i still have suspicions. data logger should be here this week so I can see what is really happening. i dont know if that is relevant but owell..


oh yeah, you can tell there is something cause the ecu to dump massive amount of fuel. it stinks PIG rich.

Chris@FsP

iTrader: (4)

That's the thing though...he's not running at the target afr on one bank. It's over correcting for no known reason. We all know that the two banks aren't always identical, but his runs really close 90% of the time, then for no reason at all, bank 1 starts over correcting, causing that side to be richer, and not hitting its target afr.

If it were a true airflow issue, the ecu correction would be bringing it back to target, not throwing it rich.

Chris@FsP

iTrader: (4)

Matt, give me a call or email to let me know what they find out.

Vince@R/TTuning

iTrader: (7)

Your right... I just glanced over it earlier and didnt notice what his true air fuels were...hrrmmm Also if the ECU is compensating off of the wideband readings it should be pulling fuel...

If your using and Etune i assume you have a cipher cable...you might wanna reflash the ROM again and see if it helps. I have had cars on the dyno where i had funky issues after i flashed it(I.E. missing target idles and funky trim issues, and i would reflash it again and the would seem to clear up. These ECU's are pretty amazing but if you have a quick disconnect during the reflash process, you might not even see it happen but the flash will still go through.

Keep us posted

maXmood

iTrader: (16)

well, IMO, i think you should flash back to ur stock values. do another datalog and see if it has anything to do with the tune/etune...

starting from the very basic would give u an idea.

onagao

Again, thanks for the suggestions.

Well, the tuning session is over. I'm actually quite upset now. Not at Jared or Uprev. They're great guys with a great product and great service - more than I'm about to say about another company.

As it turned out, the problem reared its ugly head while on the dyno, shortly after I mentioned it to Jared. He was somewhat confused by it as well. In his notations, he had it down as "ECU overtrimming. Bank 1 reading rich but still adding fuel." Or something to that effect. His explanation to me was more eloquent than I'm capable of repeating here - mainly due to my ignorance of the topic - but he felt that it is likely the O2 sensor itself. This is supported by the fact that even on my stock ecu map (map 1), it continued the illogical behavior. So it looks like I might need to get a new O2 sensor if it keeps up.

I wasn't exactly in the mood to do any datalogging on the way out thanks to the revelation that I have, in fact, lost power since the last time I was there.

That's correct. I lost power. The last time I was there, before 1) swapping from JWT Popcharger to K&N Drop-in, 2) swapping stock for Stillen Headers, and 3) swapping Stillen HFCs for Motordyne ART pipes, I dyno'd at 270.66whp under very similar conditions. This time, I maxed out at 267whp.

Seeing how I personally tested the ART Pipes to show a gain of 9whp, and the intake swap could, at most, yield a drop in 3whp, this left me only a few potential culprits. Assuming an engine malfunction, I asked Jared to run a compression test. Perfect bill of health - the engine is running exceptionally well. Spark plugs were ragged, so I asked him to replace them. We dyno'd again. Same result: 267whp. I won't waste your time with the math of it all, but somewhere in there, I lost 9+whp, and I have only one modification left unaccounted for. I'll be calling Stillen within the next few days for answers.

At least my car is running reliably right now?

I'll plug in Cipher sometime tomorrow and see if the original AFR imbalance is persisting. I wonder if the spark plug replacement might have fixed something; they did look pretty nasty.

onagao

Update:

Hooked up Cipher on the way back from Texas and conducted some tests. The AFR over-correction issue is, indeed, persisting. I'm pretty sure that it is not tune related, as it continues under both stock and tuned maps when I switch between them. This make me think that it is a faulty O2 sensor that is reading incorrectly under certain circumstances. However, that's really just my thought because I can't think of anything else.

Also, I contacted Stillen and was met with a brick wall of disagreement. I was subsequently transferred to a manager where I was granted the great and wonderful opportunity of leaving a message and hoping I get a call back.

rcdash

iTrader: (18)

???

onagao

I just got back from a trip to Failsafe Performance, where Chris looked around to see if there was any leak. No noticeable amount of carbon deposits (except for one tiny, tiny spot that very well may be the weld itself and not carbon at all), and no audible hissing or sounds of leakage when using his "tool" to listen for leaks. So as far as we can tell, there's no leakage.

If there was, though, wouldn't that lead to the AFR appearing lean? I would think that if it did, it would appear leaner and then correct the value to bring it back to normal. As it is, it's reading normal/rich and adding fuel to make it richer. Am I wrong there?

rcdash

iTrader: (18)

I suppose it can depend on where the leak is in relation to the O2 sensor. I'm not sure if you have aftermarket widebands or not? If you have an exhaust leak near the stock O2 sensor it should read lean and the ECU should compensate by dumping more fuel. An aftermarket wideband can still read rich if it is not seeing the leaked air (but the stock sensor should read lean if the ECU is compensating). If you are reading only the stock O2 widebands and the sensor is reading rich AND the ECU is dumping fuel, then something is wrong with the O2 sensor or an injector on that side.

You could swap sensors side to side and test the sensor that way?

onagao

^^^That's my situation. Stock O2 sensors reading rich and showing a correction factor a full 10% higher than the opposite bank. I'm thinking that your suggestion will be my first step. The only problem is that they're a pain to get to.

Why can't these issues just be straightforward?

rcdash

iTrader: (18)

They are kind of a pain, especially if they are "stuck" on. I'd recommend a special O2 socket and anti-seize on the threads when you put them back on. Good luck getting to them! By the way, can you monitor the downstream sensors with Cipher? That might be a nice double check. I think those are slower sensors but still wideband???

onagao

From what I understand, I cannot. With my current setup, however, I wouldn't really want to. The Motordyne ART Pipes have sensor bungs set up specifically such that I won't get any error codes... which I think means that any readings from them would be meaningless. You're right, though. That would be a good double check.

Right now, I'm torn between swapping the two O2 sensors and just skipping the whole testing part and replacing the suspect O2 sensor altogether.

rcdash

iTrader: (18)

Could just be the clip or harness also - a wire burned through... If you can reach the sensor wiring and unplug the clip, you may be able to do some testing - maybe at least check ground and +12v... I think there are plugs up near the injectors somewhere??? That wouldn't explain the sensor working under load though. Yep, probably just replace that sensor!