[hotrod182]

Originally Posted by shiv@vishnu

Spot on

Haven't even touched the 02 sensor wiring for almost 2 yrs. Are we thinking the 02 sensor itself could be causing this even with no faults? Or just that there is some kind of misfire that needs to be traced? Also, no popping or backfiring when this happens. Just that strange flutter type noise from the exhaust. So it almost seems as if fuel is being cut to a cylinder.

I will see if I can get it to occur more by lugging the engine in a taller gear, lower speeds. See if I can trace it to a particular bank by 02 sensor monitoring on the BT. I am hopping if I can get it to occur for a longer duration, perhaps it will trigger a fault code.

I just noticed something interesting..... Instead of the normal boost taper, I noticed the boost seems to swell a little right where the misfiring happens.... Meaning, air is restricted going into the cylinder...causing manifold pressure to swell during these episodes? But then again, if air isnt getting in/out of the cylinder, why would it measure lean suddenly. Unless DME is shutting injection?

[hotrod182]

As far as the boost surge during the misfire/lean episodes, I wonder if it is kind of like a two step, where the misfiring actually spins the turbos faster?

[xtyper313]

If it's truly a case of misfire, then at least one of your old plugs should have shown signs of it. Did you check them closely upon removal to help determine the problem cylinder(s)?

[JamesM3M5]

Originally Posted by Joshboody

Hey James
Can you confirm that widebands will read different afr based on spark or not? My understanding that it will reference oxygen between exhaust and air like you said. BUT oxygen total mass will stay constant whether burned or not. Molecular mass remains the same for each element, they just take different forms during/after combustion.

Anyway, I believe 50:1 would allude to no fuel in 1 or more cylinders. And splitting the bank afr would help narrow the issue.

The main portions of the chemical reactions for the pump cell involve O2, not just any form of oxygen.

When the mixture is lean, excess O2 in the exhaust gas flow, Nernst cell voltage > 0.45V, the wideband controller reacts with a current to pump out the free oxygen.

When the mixture is rich, i.e. depleted of oxygen, Nernst cell voltage < 0.45V, the wideband controller reacts with a reversed current to pump oxygen in and consume excess H2 and CO (present in combustion processes with excess fuel) where they react to form water and CO2.

The amount of current is measured and correlates to a lambda value. The goal of the wideband controller is to maintain 0.45V in the reference cell.

[JamesM3M5]

Originally Posted by hotrod182

As far as the boost surge during the misfire/lean episodes, I wonder if it is kind of like a two step, where the misfiring actually spins the turbos faster?

I suppose this is possible. Sounds like a form of anti-lag where the unburned fuel and oxygen combust in the manifold spin up the turbine.

[hotrod182]

Originally Posted by xtyper313

If it's truly a case of misfire, then at least one of your old plugs should have shown signs of it. Did you check them closely upon removal to help determine the problem cylinder(s)?

It is not enough of a misfire to visibly change the plugs. This is only a second or so of misfiring in all the driving that I do daily. I need to get this car running right so we can both go to Sacramento soon!

I spoke with a renown engine tuner today that has built many 1500-2500hp turbocharged engines. We really feel it is going to be an ignition coil, or possibly an injector. Just wish the darn thing would trigger a code. I can try and monitor o2 sensor banks and see if I can determine which bank is misfiring. Then I can swap the coils from the two banks to see if the problem migrates. First of all, I will measure the resistance and spark duration on the BT to see if there are any anomolies. Just by the fact that it is deteriorating, and happens on high torque situations, seems exactly what an ignition coil would do on its way out. The more I look at this, it doesn't seem to be the same symptoms that a bent valve, etc, would cause.

[Joshboody]

Originally Posted by JamesM3M5

The main portions of the chemical reactions for the pump cell involve O2, not just any form of oxygen.

When the mixture is lean, excess O2 in the exhaust gas flow, Nernst cell voltage > 0.45V, the wideband controller reacts with a current to pump out the free oxygen.

When the mixture is rich, i.e. depleted of oxygen, Nernst cell voltage < 0.45V, the wideband controller reacts with a reversed current to pump oxygen in and consume excess H2 and CO (present in combustion processes with excess fuel) where they react to form water and CO2.

The amount of current is measured and correlates to a lambda value. The goal of the wideband controller is to maintain 0.45V in the reference cell.

Thanks for the response... yeah, I looked it up after I posted.

[starmo]

Do you misfire with meth off? I'm wondering if too much meth combined with cold temps could be causing it

I'm also curious if you turned off your misfire detection using a flash just to test and try it would still show the issue...that fixed it on dezenno's car it seems...if this is truly a misfire then it should show even if the detection is off right? I mean you'd still see those AFR chart lean areas

All this makes me not really want to get meth or nitrous, such hairy issues

[hotrod182]

Originally Posted by starmo

Do you misfire with meth off? I'm wondering if too much meth combined with cold temps could be causing it

Haven't specifically tried to troubleshoot this problem with meth off. My guess is the higher boost (which I only use with meth) is necessary to get it to misfire. Haven't had a problem with meth and cold temps. Just think about how cold it is with the cold nitrous hitting the meth too. And it used to work fine using meth/nitrous before this problem started occurring.

[Joshboody]

misfire detection is an obd 2 thing and doesn't trigger a code unless so many are recorded. I wonder if turning off the car after it happens will keep the code. I don't suggest this though.

I don't know how Dzenno recorded a code for each occurrence, unless BMW has there own logic in addition to obd 2, and you're just not tripping it for some reason. You sure all JB code clearing is off.

swapping bank coils and logging each AFRs is a good idea i think.

[stilov]

I have no "real" input...just going to relate a weird problem I had regarding fuel...

Mine was a bit different although it manifested itself THE day I installed meth, half engine light and the car shut down, couldn't get it to start again.

Anyone I talked to said..."oh ya HPFP" I changed it with the low pressure sensor...no result. With the JB4 I could tell I was getting no lpfp fuel pressure, so I swapped out that too...still no improvement until I took the car to an indy shop and they found a small plastic elbow on the fuel line in the level assembly/pressure part in the drivers side of the fuel tank had separated.

I was basically circling fuel in the tank

Any chance you elbow could be starting to separate and losing pressure for an instant?

[cn555ic]

Didn't you make a thread a couple of weeks ago with issues of lean spikes of 20:1. Now you have lean spikes of 50:1?? This is getting progressively worse if that's the case. It is odd that it just spikes at that specific point and then a/f stabilizes back so quickly. Very odd.

[hotrod182]

Originally Posted by stilov

I have no "real" input...just going to relate a weird problem I had regarding fuel...

Mine was a bit different although it manifested itself THE day I installed meth, half engine light and the car shut down, couldn't get it to start again.

Anyone I talked to said..."oh ya HPFP" I changed it with the low pressure sensor...no result. With the JB4 I could tell I was getting no lpfp fuel pressure, so I swapped out that too...still no improvement until I took the car to an indy shop and they found a small plastic elbow on the fuel line in the level assembly/pressure part in the drivers side of the fuel tank had separated.

I was basically circling fuel in the tank

Any chance you elbow could be starting to separate and losing pressure for an instant?

I was very suspicious that something was going on inside the fuel tank. Especially since sometimes it almost sounds like it is cavitating inside there. But I really think this is too quick of disturbance to be a typical fuel pressure problem. And the fuel trim logs just dont show a fuel supply problem.

Originally Posted by cn555ic

Didn't you make a thread a couple of weeks ago with issues of lean spikes of 20:1. Now you have lean spikes of 50:1?? This is getting progressively worse if that's the case. It is odd that it just spikes at that specific point and then a/f stabilizes back so quickly. Very odd.

I corrected my first paragraph. It is the same 20:1 spikes. I could be much higher, but I don't think the log will even measure anything higher than 20:1. However, it has gotten worse, based on the fact that it is happening on non-nitrous runs now.

[GeorgiaTech335coupe]

At the beginning of the second set of injector replacing, the misfires would progressively get worse and easier to create. First they started only in high gears when lugging the engine. Eventually they would happen enough that it made my car almost undriveable. This happened for each injector, one by one, as I replaced them over months. That was a hard couple months, I hated my car.

The problem seems to be different when I have a coil going out (seems to be the issue now - already replaced 2 at different times, may need to replace 2 more). Bad coil misfires seem to occur more when going WOT in low gears, not just lugging.

When the injectors went out, I could do WOT pulls if I went through the gears from first or second, until it got really bad. With the coils, I can't do WOT pulls in any gear, although sometimes it works.

[hotrod182]

Originally Posted by GeorgiaTech335coupe

At the beginning of the second set of injector replacing, the misfires would progressively get worse and easier to create. First they started only in high gears when lugging the engine. Eventually they would happen enough that it made my car almost undriveable. This happened for each injector, one by one, as I replaced them over months. That was a hard couple months, I hated my car.

The problem seems to be different when I have a coil going out (seems to be the issue now - already replaced 2 at different times, may need to replace 2 more). Bad coil misfires seem to occur more when going WOT in low gears, not just lugging.

When the injectors went out, I could do WOT pulls if I went through the gears from first or second, until it got really bad. With the coils, I can't do WOT pulls in any gear, although sometimes it works.

Interesting how your coils were so bad even in daily driving. I think possibly my coil is just bad enough where it is insufficient at these higher boost levels only. What the race engine builder/tuner told me was the the problem will usually manifest itself towards the torque peaks. He was once tuning a 1500+ HP Toyota engine, and he said he had to keep closing the gap on the plugs to help combat the problem. Since I had no problems with stock gap previously, I am going to start focusing on the coils. Just not sure what aspects to check on them. I will try and measure the primary and secondary windings to see if there are any anomolies. Then I will probably buy a few coils and start swapping/testing, etc. Coils are cheap enough and easy enough, for me to start with them. Injectors are not so cheap, or easy with the coding, etc. Remember, if you aren't racing at triple digit speeds, you wouldn't notice my misfire yet. I would assume eventually it will effect everyday driving, idling, etc. At that time I would probably be getting useful fault codes. I'm not really going to wait till it becomes that obvious.

Does anyone know specifically, a good way of testing the coils? Resistance measurements, etc?

[JamesM3M5]

Unfortunately our coils live in a pretty bad environment due to heat. Due to their exceptionally small dimensions, pencil coils are computer modeled to ensure they can generate the voltage necessary to fire a high energy spark. They do not have a lot of extra headroom if they get weak or when we start putting very high loads on them.

[starmo]

Sorry for stating maybe the obvious but Isn't it strange that coils would only now be an issue after all this time running amazingly well? I mean given they're fine to begin with

[Fundahl]

Originally Posted by starmo

Sorry for stating maybe the obvious but Isn't it strange that coils would only now be an issue after all this time running amazingly well? I mean given they're fine to begin with

I've had weak coils only blow out in certain scenarios. They wouldn't always misfire, and only sometimes under load. On a M54 I fixed in a 330i actually. My Z32 coils will consistently misfire when they go and I hit a certain boost pressure.

I would check all your plugs just for fun. Look for spots or a missing/bent tip or something.

[hotrod182]

Plugs were changed with brand new ones, no improvement.
Last night I pulled all the coils out again to inspect them. Also retorqued all the coil ground wires on the ground posts.

Primary Coil Windings:
Cyl#1 1.2 ohm
Cyl#2 1.2 ohm
Cyl#3 1.2 ohm
Cyl#4 1.2 ohm
Cyl#5 1.2 ohm
Cyl#6 1.2 ohm

Secondary Coil Windings:
Cyl#1 2.51 meg ohm
Cyl#2 2.33 meg ohm
Cyl#3 2.24 meg ohm
Cyl#4 2.38 meg ohm
Cyl#5 2.28 meg ohm
Cyl#6 2.45 meg ohm

So unfortunately nothing stands out significantly as far as resistance measurements.

Hard to say if the higher or lower resistance measures are good/bad on the coils.
My next step is to buy 3 coils. Measure them, see what is normal for a brand new coil. Maybe I can see some kind of trend. I will replace one bank, and test. If the problem still persists, I will swap them to the other bank. This should allow me to confirm whether or not it is the coils within two batches of testing. Assuming the problem is being caused by one coil.

I tested a qtr mile run again, to make sure the car is still cutting out before I replace coils. It was fine through the qtr mile, but then got the lean/misfire right after shifting into 5th gear. (130+). This is good. The worst thing that could happen would be if it stopped happening during these preliminary tests, because then I wouldn't easily be able to see the improvement of the new coils.

The more I think about it, I have a real good feeling of the coils being the cause. Kind of bummed out that the measurements don't show anything obvious. The next step would have been to take measurements with the coils hot, but I'm just going to skip that test. If the coils don't fix this, I think I am going to give up till the problem manifests itself in normal driving, and triggers some kind of fault code.

[vasillalov]

If my memory serves me right, Dzenno is running Okada coils with his upgraded turbos and he is seeing good results with them.

Just a suggestion...

[Lulz_M3]

Originally Posted by vasillalov

If my memory serves me right, Dzenno is running Okada coils with his upgraded turbos and he is seeing good results with them.

Just a suggestion...

This is good, may help determine what kind of power the stock ignition system can handle before requiring upgrades. Now if only we had a definitive answer for how much the factory DI can flow before issues come up...

[hotrod182]

Originally Posted by Lulz_M3

This is good, may help determine what kind of power the stock ignition system can handle requiring upgrades. Now if only we had a definitive answer for how much the factory DI can flow before issues come up...

I'm very curious to see how the new coils measure electrically. Perhaps they too will be about the same as my old coils. Or if they are way different than the old ones, I will probably end up replacing all the coils.