Random misfire, possibly injectors? (warning, long post, I’m desperate)

[GHuggins]

Yes it does. It'll either be one of two. Either set your min airflow from 1200 up to a really low gain or climb or increase it to where your just below self cruise settings. It opened the throttle plate up a touch more is why the problem went less noticed. The same principle applies to retarding the timing to cause more engine load. So you might be able to combine both now?

[Michael_D]

Update, in the name of science and lunacy......

My new ID 1000 injectors, new larger 108mm CIA bridge, and new 6:1 ratio honeycomb arrived via pony express last week. Got them all installed and spent three days re-calibrating MAF and VE. Yah I know, I should have done the injectors first then the new bridge/screen; or vise versa, to test each independently, but my time not earning a paycheck is limited. I wanted to just get the damn thing recalibrated and be done with it.

I was pretty surprised at just how much I had to work both air modes. Roughly 20% increase in both VE and MAF, everywhere...... Then I had to re-cal idle and spark. All on the street, dealing with this stinking rotten rich after flash BS.

The engine seems to like these changes, a lot. I can put along in second or third gear, under 1800 rpm and not feel like a bobble head doll. Fifth gear, 2000 rpm and up, no stumble/miss/cough/whatever.... Smooth. Still some slight bumps under 2000, but not all that noticeable. Fifth gear, 1600rpm is doable. My bank/bank trim imbalance is improved as well.

I think at this point, about the only other things I might try, is a newer ECM, or possibly fit a second honeycomb in the CIA bridge. Not sure if the second screen would be worth the effort though, cause it would have to be right up next to the throttle body. I do have one to try, if I get so inclined......

This log is after a drive, nothing disabled. I tried to hit all the gears. My MAF low table LTFT+STFT histo, is pretty much <2% error in each cell.

trim histo.png

8-04-15_FINAL DRIVE TEST.hpl

[schpenxel]

Glad to hear the injectors helped!

[barum]

I stumbled on this thread after searching tips while struggling with same issues with one car I've been working with. The cam is quite similar 238/250 LSA113+3 .652 .630 with true 16.9 deg overlap @ .05".
Other mods include ported LS7 heads w/ 11:1 compression, OEM intake valves, Ferrea hollow exhaust valves, PSI LS1511 springs, Comp 3/8" 0.135 pushrods, ARH 1 7/8" headers + x-pipe w/ cats and MSD intake. Stock air filter + tubing + throttle body. The car did have GT500 2014 injectors at first, but even though I've made my own calculator (yeah, I'm awere DSteck has a working one) to give me full control of regression functions (not a big fan of high degree polynoms which are easily overfitted like neural networks), SPA and support for higher pressure datapoints, I decided to get things sorted with OEM injectors to limit the number of unknown variables.

The car is '06 Z06, so the ecm is even older version. It has ~50 000 miles on the clock so the couplers might flex a bit more and these random misfires agitate all the slacks and frequencies in powertrain very easily making it even more annoying to the owner.

Similar checkings plugs, wires, resistances, swapping around parts, etc. About same amount of misfires in each cylinder.

I've played a lot to smooth out spark and air flow controls by reducing different gains at smallish rpm errors to make sure they are not amplifying anything unwanted. I too have tried different spark advances from low (20 deg) to high (over 40 deg) having learned that something between 23 - 28 deg seems to work slightly better in the problem area. I even did case learn procedure which I think (read: imagine) helped a bit but not much. Also cylinder balance showed to be ok, although I did run that test at idle which led me to believe that it's idle speed controller related why results looked like this
#1 96%
#2 97%
#3 97%
#4 98%
#5 98%
#6 98%
#7 99%
#8 99%

I'll rerun it just out of curiosity with idle set at 1200 rpm to see if anything changes.

I've compared a lot of logs and tunes from this forum and it seems rpm/MAP/MAF/MAF Hz/throttle pos look similar to anyone else's logs in terms of smootness, there was only one guy (with MAST 305 heads + cam) who had a bit smoother running engine than the others. I've compared older logs from this car with the stock cam and yet no significant difference can be seen.

I have a feeling that in the beginning these random misfires appeared only under 1400 rpm but now they seem to extend upto 1700 rpm or so. One reason might be that they were masked by some bucking caused by spark controls and spark smoothing. The car has been run in MAF only mode. While calibrating MAF I noticed LS7 has seen three sets of injector data over the years, none of which seem to be just spot on judging from the MAF scaling results so I left '06 injector data untouched.

The interesting part is that whether run in CL or OL, the random misfires feel about the same. AFR in OL was 13.8-13.9 to be sligthly on the rich side, but no help. This would have me thinking it's not lambda related, not solely at least, but more to do with MAF fluctuation however similar it seems with the stock cam.

Since the car has stock MAF location + intake tubing I've been wondering if Halltech's "MAF forward" desing could help by introducing more volume between intake valve and MAF hence reducing the reversion seen by MAF sensor (low pass filtering). I've even considered building longer test tubing where MAF would be located several feet away from the intake to see if it makes any difference.

As a side note FAST and MSD intakes differ in that
- MSD has 9.25" runner length, FAST has 11" measured along the centerline
- likely the intake manifold volume is a bit bigger in MSD
- runners in MSD are only sligthly bent, FAST has quite tight radius close to bell mouth
So it's likely they behave differently when it comes to reversion and symptoms can be somewhat different.

Regarding the closed loop feedback signal (O2 sensor) parameters I've experimented with Airflow Mode by reducing the Normal mode values to those of DOD without any significant change.
I even calculated new lower end values assuming 20-25% lower exhaust gas volume/speed based on lower VE due to cam and adjusted accordingly, no change.

Then I tested Closed Loop Proportional/Airflow Mode table values and found out the only thing that changes is the closed loop O2 sensor oscillation frequency, amplitude is not affected (when keeping load & rpm constant). In the end I used values that give similar frequecy what was seen with stock cam + LTs. STFT change points are the best indication to measure oscillation time (but not on idle) if someone wants to test this.

I tested the EOIT as well, had it delayed by upto around 40 degrees to have 5-10 deg safety barrier between exhaust valve closing and start of injection all the way up to 2560 rpm, but it made engine loose torque ~20 Nm and idle/run was "heavier" in the lack of a better word. The car doesn't have raw fuel smell probably due to cats, not even with stock EOIT values.

However, I find it very hard to believe that these older ecm's wouldn't have any delay compensation at all for the O2 sensor location. Sure, it can be hard coded but it must be there for the control circuit to function properly. I've been hacking some hard coded limits and other stuff on Saab Trionic7 in my past which is why I would be very interested digging deeper in E38 binary files. Too bad, binaries aren't "open" anymore.

A newer ecm could be one (partial) solution but probably close to impossible on these older models as so much else would need to be changed, too. I wonder if multispark amplifiers would help or not?

[Sébast19X]

barum, great testing, and Michael D too.

You're aproach are like scientist !!!!

Well don guys

Best regards,

Sebastien

[Bluemamba]

Any updates to this thread?

[Crashgnx]

Sorry to wake the dead thread but after reading through it I had to comment. My opinion on the MAF, those blade style ones are designed and calibrated for a specific diameter tube and need to be in straight section of pipe. I believe they recommend 3 inches before and after but as far away from bends as possible. I think that is your MAF issue and personally in your situation I would go with full SD tune eliminating it or using it only for the IAT. Mainly because it would take less time to build that tune than to hash out a way to either control the airflow across the MAF or tune around the turbulent air issue. My .02. Love the car though, im sure its a blast to drive.

[Michael_D]

I have not done much of anything to the car's tune since last year. It's been in my garage, with the interior ripped out of it..... Another project to make sure I keep throwing money at a car I drive very little. I should have bought a boat.

So with the MAF, this particular CAI is quite popular for the C6 Vettes. Works great. The last experiment with the honeycomb did seem to help a tad. I can log fuel trim error, and it is very low in any of the first, four - foward gears. When I get into overdrive it when fuel trims start to wander, and the left bank goes leaner than the right. Low rpm/low load on this engine is my nemisis.

Sometime this summer I intend to re-visit EOIT. I think I can improve on that some, now that GHuggins has posted up some pretty darn good step by step instructions (that I have copied onto a word doc).....

I also have to tweak transients a bit. 90 octane fuel really sucks.

[arsautorepair1]

Update?

[Michael_D]

I installed the newer 2010 E38 and am actively engaged with O2 table tweaks. No real wins to report yet.

[arsautorepair1]

I installed the newer 2010 E38 and am actively engaged with O2 table tweaks. No real wins to report yet.

Keep me updated I've been following your progress have same problem..

[arsautorepair1]

bump

[Michael_D]

Very small improvements tweaking the tables. I threw in the towel and tore the engine down. Was going to hone it five thou and fit new pistons, but leak down was under 2%, so left the pistons as is. Inspected bearings and the rods. Perfect. Just finished re-assembling the short block.

Carbon layer in the intake ports, combustion chambers and pistons was quite thick, indicating significant reversion. I'm starting over. New cam is on the bech, 5 deg OL at fifty lift. 670 lift. 228/242 duration at fifty, 115 lsa.

Bought a set of new LS7 heads and had MS90 guides fit and seats re cut. Pedestals milled to fit the Crower rockers. Mild port work, retaining the swirl dam. Hand porting and polishing them now. New MSD intake. I'm currently porting it too. It's a friggin mess out of the box.

I'll report back this spring after I get the new build calibrated.