7

This is a follow-up from my initial question which can be found here : LTFT and STFT values are off and don't make sense.

Car : BMW 2015 F10 M5 Modifications : Eisenmann Race Catback Exhaust, MSR Cold Air Intake. Everything else stock.

The summary of my problem is, LTFT values between Bank 1 and 2 are inconsistent where Bank 2 is 10% or more lower (-4% vs -14%) throughout the lower RPM band. At WOT, LTFTs line up and stay equal and behave normally.

Zaid has been incredibly helpful and we narrowed it down to the aftermarket exhaust I have.

Basically there is one big difference between the stock exhaust and the aftermarket one and that is the use of flaps. Each muffler (there are two) has 2 pipes (so 4 pipes total) where 1 pipe on each side has a flap connected to it that is operated by a valve. The valve is connected to a vacuum line. In the stock exhaust, at low RPMs, to reduce noise as well as reduce turbo lag and increase low and torque, the flap is closed increasing the backpressure. Over 3,500 rpm (or around that), the flaps open up. The aftermarket exhausts have no flaps (not just mine but a lot of others) and the standard installation procedure is that they zip tie the vacuum lines and plug them as well (mine has bolts plugged). So in an aftermarket scenario, car operates as if the flaps were always open.

Here is a picture of the eisenmann muffler right next to the stock muffler : https://eisenmannblog.files.wordpress.com/2012/11/8182708317_10988d68b2_h.jpg

Here is a picture of the eisenmann muffler on the car : http://farm9.staticflickr.com/8490/8206133927_6b3bec97bb_b.jpg

The theory we have at the moment is the lack of flaps is causing this rich condition I am seeing. What I don't understand is why this is happening on one side of the exhaust not the other. So far I have done:

  • Check both exhausts and where they are clamped, clamps are extremely tight, and there is a good maybe 4-5 inches of stock exhaust inside the aftermarket pipe (after market slips on), so there is no way a leak can happen at the connection points.
  • Check the vacuum lines on both sides, they are both zip tied and plugged with a bolt.
  • Check the intake installation many times, no air leaks, all clamps are securely fastened.
  • Check turbo inlets and charge pipes, no cracks or air leaks.

Returning the exhaust back to stock is not something I want to do. I spent several hours adjusting the tips of the aftermarket exhaust so that all 4 looked identical. Without knowing for sure, I dont want to touch it.

Any help is appreciated.

Thanks.

Pᴀᴜʟsᴛᴇʀ2
  • 156,290
  • 29
  • 252
  • 491
DerStig
  • 421
  • 2
  • 5
  • 7
  • Back pressure and turbo spin-up both affect total airflow through the engine. Airflow affects the amount of fuel needed to keep things at stoichiometric. Have you reprogrammed the fuel response curves to compensate? I don't understand your resistance to the physical reality here. It's not a Small Block Chevy. You've messed with a carefully balanced system, and you're willing to sacrifice performance to keep your tips lined-up. It's your car, it's your call. – kmarsh Jan 13 '16 at 20:01
  • Again, why this is an issue on one side? If both banks were 10% rich, I would understand. Also please see my comment below regarding the LTFT readings at idle and how bank 2 is stuck at 10% whereas bank 1 moves freely up/down depending on throttle input. I am not disputing that this exhaust might throw off the pressure balance but there is 10% or more variation between each side and thats a lot. You could argue the exhausts are not made equal, but that would contribute to a slight difference, not 10%. – DerStig Jan 13 '16 at 20:07
  • See my comment below. It gets down to, does BMW actuate those flaps independently? – kmarsh Jan 13 '16 at 20:08
  • Yes they do, are you saying if the two sides are off, it is for a reason? I dont think 10% difference is intentional. – DerStig Jan 13 '16 at 23:58

2 Answers2

4

Possible Culprits

It is normal for fuel trims to be slightly different, but a 10% bank imbalance on a car that still smells like Munich seems odd.

There are two things which affect fuel trims:

  • the physical system

    If you are confident that both intake and exhaust are air-tight then we will have to cross them off the list of potential causes.

  • the integrity of the sensors/injectors

    This comment of yours about a slow response from Bank 2 has me wondering whether you have a lazy O2 sensor. It's the only thing I can think of right now, since that is the only bank-specific sensor which provides feedback for fuel trim correction.

How to Test

The idea here is to confirm the slow response of Bank 2 O2 sensors compared to Bank 1.

You'll ideally need some graphing capability. A scope or an OBD-II app with graphing functionality. Pull up Bank 1 and Bank 2 STFT's

Introduce some propane gas into the intake of each bank to induce a negative STFT. Look for differences in response time and STFT magnitude. Based on your description, I'm expecting the STFT for Bank 2 to change more slowly and/or register a bigger % difference.

Community
  • 1
Zaid
  • 39,111
  • 48
  • 142
  • 290
  • Thanks Zaid. I heard anout this test but never tried it obviously. Where can I get the proper propane for this? I have a propane tank at home I use for grilling but its one of the big 15 lb ones. Also, is there anything I should watch out for doing this test? I dont want to blow myself up:) – DerStig Jan 13 '16 at 21:56
  • @DerStig I wish I had the time yo do a Q&A on that. You can watch this video by Schrodinger's Box (about halfway into the video) – Zaid Jan 14 '16 at 05:57
  • I have contacted Eisenmann and asked to have one of their engineers comment on this particular problem. Namely, I wanted to find out if they anticipate car running richer because of the reduced backpressure, furthermore, I asked them to confirm whether or not a 10% discrepancy is expected or plausible because of the exhaust. Their response was the car can run up to 5% richer, which makes sense considering my Bank 1 is around -4% vs -14% Bank 2. But they said an additional 10% richness is not normal and cannot be because of the exhaust. – DerStig Jan 15 '16 at 14:05
  • Could this be a fuel injector leak? How would one know if it were? I dont think its a slow responding O2 sensor because when I let go of the throttle or go WOT, I can see O2 voltages dipping and spiking pretty simultaneously. Could it be that there is a fuel injector leak, a small one, and under WOT, it becomes invisible because the fuel demand exceeds the leak amount? – DerStig Jan 15 '16 at 14:08
  • @DerStig look at the STFT's, not O2 voltages. You want to see the amount of correction, not the end result – Zaid Jan 15 '16 at 14:40
  • And yes, a leaky fuel injector could be to blame – Zaid Jan 15 '16 at 14:40
  • Thanks, what exactly am I looking for at STFTs? Usually bank 2 runs richer, maybe 80% of the time and the values are -3 vs -1 or so. – DerStig Jan 15 '16 at 14:51
  • @DerStig just report the maximum % values you're seeing for both banks when you perform the propane test. If you can't then we can sort out an alternative in chat – Zaid Jan 15 '16 at 15:27
  • Zaid, I watched the video you linked and I have also watched all other videos made by the same author. They are very informative but they havent explained how to do the propane test. It would be great if you can give me high level direction if not an alternative way to test this. I think we should focus on fuel injectors as the culprit, i read a number of times last year stock M5s had their injectors replaced. These cars were very new with 5000-10000 miles. My car could be similar – DerStig Jan 15 '16 at 15:39
  • Let us continue this discussion in chat. – Zaid Jan 15 '16 at 15:40
  • 1
    So today I was fortunate enough to drive a stock M5 with no mods whatsoever and I found out something very interesting. Basically the fuel trims are 10% off between both banks! Except they are in opposite direction. While my car reads -2% vs -11% idle this car was 9% vs 0%. When my car reads -4% vs -14% at 1800 rpm, this car was 12% vs 2%! Unbelievable. So I guess the 10% discrepancy is normal in this car or this code reader doesnt know how to read this car? I have also noticed that between this car and my car, catalyst temperatures were off by as much as 100C. Both idle and while driving. – DerStig Jan 16 '16 at 00:06
  • So now the question is can an intake and an exhaust I have cause such a massive difference making the car run so rich? Or is this maybe a fuel pump failure? So strange. And the catalyst temperatures were very interesting to note, my car shows much higher temperatures (480 vs 570 idling and 600 vs 750 while driving). – DerStig Jan 16 '16 at 00:08
2

Gut reaction?

I went straight to fluid dynamics. It is likely that the aftermarket exhaust is made of larger diameter piping. Increasing the cross sectional area of a pipe will, all other things being equal, reduce the flow rate of a fluid through the pipe.

This exhaust is likely tuned for high RPM performance, when the engine needs to move a higher volume of exhaust gasses, and larger piping is needed. At low RPMs however, with a smaller volume of exhaust gas being expelled, the larger piping may have a detrimental effect, slowing the velocity of the gas.

It could feasibly slow enough that your O2 sensors read a higher concentration of gas and presume that the engine mixture is off. It would explain why things level out under WOT.

thunder2709
  • 176
  • 4
  • Thank you and that sounds feasible except one side is severely off. If both sides were doing this, I wouldnt be worried (or worry as much). So I was just driving and the behavior is as follows : low rpm/low throttle input, they are off by 8-10%, high rpm/wot, bank 1 (the good side) reaches to 0.0% immediately whereas bank 2 slowly goes down and the lowest value is -2% or so around 7000 rpm. – DerStig Jan 13 '16 at 19:43
  • Also while parked, idle, I put it in N, and keep it steady at 2, 3, and 4k rpm and bank 1 (good side) reads -6, -7, and -10% respectively whereas bank 2 always reads -10 regardless of rpm. – DerStig Jan 13 '16 at 19:44
  • From your description, there are two vacuum flap actuators, one for each side, I'm assuming they did it that way because they work independently of each other. Are you assuming they work in exact unison? If so, why? – kmarsh Jan 13 '16 at 20:05
  • I am not assuming anything:) Yes there are two independent flaps on each side and most probably the vacuum lines that goes to each directly feed to the turbos on each sidd – DerStig Jan 13 '16 at 20:09
  • So the engineers felt the need to actuate the flaps independently. This means that we can't assume that each side will "do the right thing". One side fixed and one side varying might be some self-diagnostic mode, trying to figure out what is wrong. – kmarsh Jan 13 '16 at 20:12
  • Correct, thats possible. But how can I diagnose this problem? – DerStig Jan 13 '16 at 20:29
  • There are any number of reasons as to why one side is different. Is the exhaust design totally symmetrical? If not gas flow could bottleneck worse on one side. – thunder2709 Jan 13 '16 at 20:58
  • 1
    First, I'd confirm that all relevant sensors are fully operational. Data collected with bad sensors is useless. Check out everything related to fuel and ignition. Verify the turbos are healthy. Do it yourself, or have BMW go over the car with a fine tooth comb. If you check everything and find no red flags, I'd honestly just let it go. The car will let you know if/when something really goes wrong. And frankly, you've modified the airflow pretty drastically from stock, and factory tuning has it's limits. If you want the most from the car and want it to last, you need a custom tune. – thunder2709 Jan 13 '16 at 21:00
  • Thank you, how exactly can one check all the sensors that they are operational? I spoke to 2 different BMW dealerships, and their answer is "until you get a CEL or you get some sort of malfunction codes, there is nothing we can do to diagnose this". They are not even giving me an option. So I need to self diagnose this problem. – DerStig Jan 13 '16 at 22:12
  • If you have no CEL then it's pretty safe to assume that everything is OK, and we circle back to the the point that you have heavily modified the engine's airflow, and factory tuning has limits. You cannot expect the car to properly run after modification to the engine. You might have a lazy O2 sensor as Zaid says, and fixing it might help, but if you really want the car to run perfect and take advantage of your modifications, you have to get a custom tune. – thunder2709 Jan 13 '16 at 22:38
  • I understand what you are saying and I accept that a heavily modified car will not run "smooth". However, I think, also as Zaid also agrees, 10% variation is not normal and makes no sense in this case. By the way, Eisenmann is probably the most reputable exhaust manufacturer in the world, they manufacture Porsche and AMG OEM exhausts. I am tempted to email them and ask them about this, but I hardly think they have designed an exhaust where 1 side is going to run as much as 10% richer than the other side. – DerStig Jan 13 '16 at 22:41