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

[GHuggins]

Do the leakage test if you've got the tools. Bet those rods are way too long... Your cam has a whole lot of lift and if the lifters are hydraulic they may have caused you to get the wrong pushrod length via plunger being depressed slightly durring measuring process.

[Michael_D]

They are not too long. This I know..... Ain't my first engine building experience. First tuning a modified 7.0, not building.

The lash adjusters needed 1/4 turn to take up free lash, prior to loading the plunger spring.

[GHuggins]

So the rockers are adjustable and you are using hydraulic lifters?

Something I prefer to do after seeing the difference from an older engine builder teaching me is to adjust them while the engine is actually idling. I know, I know sounds horrible, but it's not. With the engine at hot idle and maybe even with the idle rpm turned down more, I set my lash adjustment one at a time by backing them off then reworking. This way the plungers are filled with hot oil, metals are to temp and all you have to do is listen for 0 lash. You find 0 by backing them off till you hear the clickedy clackity then immediately start tightening back down. You'll find your first adjustment to be off anywhere from a little to a lot sometimes. Had an experianced engine builder teach me that years ago. Little bit of a pain as the nuts your turning are moving up and down, but very accurate...

It may just be that your volumetric efficiency is just too low with that cam. Have you noticed any differences at all with any adjustments withen your calibration? SUCH as fueling or air tables primarily?

[Michael_D]

I have full roller rockers with last adjusters and I am using Morel 5294's, high speed hydraulic lifters. The scrub was horrible with the OEM rockers and this much lift, so I went roller tip. M-MOI of all the other bolt on roller tipped options is shit, plus they are aluminum and I do not like aluminum rockers. Jesel wants about 4K for their set up, so I went with the Crower rig. A couple pics are below. The witness marks are after a few complete engine revolutions using a couple modified lifters (made them solid to check rocker/tip geo).

I never use a solid lifter grind, unless the engine needs it. And even under those conditions, I'll use short travel "cheater" lifter prior to full on solid. I hate them, and I grew up with the f'ers. Quite experienced leaning over the fenders of race cars, breathing oil smoke, adjusting lash....

P8318759.jpg

P8308757.jpg

When I calibrated the VE table (with the custom Op System), I was able to get STFT error under 3% in all cells that I could actually grad data in. I was / am pretty happy with VE.

I cannot however, get MAF dialed in under about 2500 rpm. It is all over the place. I can get the curve spot on, looking perfect in second gear, then I check it again in third gear and it's way off. Fifth gear - same thing. Dial it in in fifth gear and it's shit in fourth. So in the end, gave up using both STFT and WB AFR error to dial MAF in in the lower Hz cells and just used LTFT error. I just collected data in these cells, all gears and used a LTFT error histogram for these cells.

I dunno.....maybe with MAF being in the weeds, this ECM is just confused. I don't understand this hybrid / blended system very well. I have been lead to believe that even if you have Dynamic enabled up to 4000 rpm, the ECM still references MAF on some biased percentage under 4000 rpm.

I just flashed the car with a change to the Closed Loop Proportional Airflow Mode as you suggested in the other thread (multiplied the table by *.80). I'll go for a long drive, let the trims settle in, star data logging, and post up the log...

[Michael_D]

Last log and tune file. I started logging right after the flash, so the first 20 minutes can probably be ignored.

<2500 rpm in any gear, any speed, I feel this misfire/surge/stumble/buck - whatever you want to call it. Lower gears are worse. Second gear, around town low speed driving sucks.

5-15-16-01.hpl2015_NW_102_o2 change_02.hpt

[GHuggins]

OK, just for clarification shimming or height adjustment of a shaft style rocker is what more or less sets the ride pattern over the valve stem. Pushrod length is a very small factor to this with shaft style rockers. BUT like you said, you had to turn the lash adjusters in to take up "free play" so even though you could probably get them more accurate by readjusting while it's running, it would not fix your concern and I wasn't even thinking of what your concern actually is until your last post. Sorry about that, didn't have my head on straight apparently... Your fighting the usual lean misfire caused by high lift aggressive camshafts. OTHERWISE as you stated very clearly misfire/surge/stumble/"BUCK"

There are actually a few ways to tackle this and may wind up having to tackle it at multiple angles.

First off.... How much have you retarded your injection timing thus far? HOW are you retarding your injection timing? With your setup I would recommend adding a min of 25 to 30 to your entire injection boundary. If necessary since you can tell if it's gone immediately or not without fuel corrections add more than this until the problem minimizes as much as possible or even goes away entirely. Most people don't add enough and are almost afraid to add this much, but it's OK, just keep adding until there is no longer any gain felt. You can always bring in back down in the higher rpms if you experiance adverse affects.

After getting that as close as possible, go to your timing corrections tables and minimize these to nearly nothing for the first 4 to 5 cells. Do this to the underspeed and overspeed corrections. Then go to your main timing table and while watching where it mostly operates when the problem is present try to smooth these zones out as much as possible.

After all this if your still missing due to a lean condition go to your stoichio table and set the first cell down around 14.1 to in the high 13's then slope or blend the corresponding cells from there.

Report back on how this does. Again, sorry for not putting two and two together to realize what you had going on.

Side note, did adjusting the proportional table to a smaller value help or hurt? If I remember you seemed to think the larger numbers seemed to help. All of the above is to try and get your fueling constant and back to where it needs to be to fix the lean buck issue which the O2 settings can help with...

[Michael_D]

First off.... How much have you retarded your injection timing thus far? HOW are you retarding your injection timing? With your setup I would recommend adding a min of 25 to 30 to your entire injection boundary. If necessary since you can tell if it's gone immediately or not without fuel corrections add more than this until the problem minimizes as much as possible or even goes away entirely. Most people don't add enough and are almost afraid to add this much, but it's OK, just keep adding until there is no longer any gain felt. You can always bring in back down in the higher rpms if you experiance adverse affects. ...

I did screw around with the this last year. Moved it quite a bit, or at least I had thought I did. I did not move the boundary. If I recall, I moved EOIT vrs RPM as far as I could, then moved EOIT Adder vs. ECT to get it to move further. Did not move Boundary though. I do not recall why. I used the attached spread sheet that D-Stek put together. I should, and will re-visit this, as I did change injectors. Injector pulse width and offset are both input values used in the spread sheet.

Injector Timing Work Sheet.xlsx

After all this if your still missing due to a lean condition go to your stoichio table and set the first cell down around 14.1 to in the high 13's then slope or blend the corresponding cells from there. ...

Does that actually do anything? I tried that before, logging AFR with my wide band, and the O2?s would just bring AFR to 14.7 ? no matter what I put in this table. I don?t think it does anything, unless you have a flex fuel equipped system with dual fuel sensors?.. At least I did not see anything when I tried that before.

Side note, did adjusting the proportional table to a smaller value help or hurt? If I remember you seemed to think the larger numbers seemed to help. All of the above is to try and get your fueling constant and back to where it needs to be to fix the lean buck issue which the O2 settings can help with...

I was mistaken. After raising this value, I took the car for a drive. Seemed like it ran a bit better, but I think that was due to the ?rich after flash? thing?. After the car sat for a few hours, I drove it again and the miss was unbearable. I moved it back to stock, then lowered it. Maybe a slight improvement.

Yesterday afternoon, I re-verified VE. I failed the MAF, killed the fuel cut offs/adders and took a long drive. Left LTFT?s enabled. All cells still look good.
LTFT's still under 3%.

After the VE test, I did the same with SD disabled, and logged MAF LTFT error. Each gear, slowly and steadily populating low Hz table cells with LTFT data in the histogram. Error gets progressively worse from fifth gear to first gear. I really don?t get this. I zoom in on the Hz output in the scanner, and I?m not seeing any spikes or dips. Log is attached with config. As I know some will tell me to use either STFT's or WB / AFR Error, I already have. All three methods act the same.

[GHuggins]

It's because fuel is still going straight out the exhaust. You need more fuel in the combustion chamber which is why the rich after flash issue helped... Your fuel vapor is most likely burning in the tailpipe making you think your AF ratio is correct. This is somewhat rare, but have run into this a few times before. In fact I bet your AF ratio jumps up and down quite a bit possibly? Changing withen a .5 to .8 scale at idle? This explains your highly varying fuel trims in different gears too. I ran into this issue - the more I think about it - with the first cam tune I did. Mid range to High range aggressive cam that I tuned incorrectly using stoich only to get rid of the bucking. If I knew then what I know now about injection timing, I would have gone about it differently...

Forget adjusting injection timing via rpm or any of the other methods as most vary model to model and some parameters such as the "adder" table are only used under certain circumstances. Use the boundary table. ADD to this to retard the timing. Start off with 25ish min for your cam...

To get the stoich table to work you have to counter tune via maf and VE. Also helps to raise your O2 settings up around 550 instead of 450 then put your min max settings to 600 and 500 or even 525 and 575 for better control. You should be able to get your AF ratio around 14.5 even with O2's controlling it.

Once you get more fuel in the combustion chamber and keep it from going out the tailpipe, your problem should all but resolve itself. It would be really nice if you had the parameters for your O2's. Maybe HP will add this sometime to really get your AF under control.

SIDE NOTE - HP redid their forum website last night. Do NOT like having to be logged in to read the following pages of the forums or even general text in the forums themselves. If anyone is reading this - Bill, Mark, Keith - Is it possible to fix this? Thanks

Redo the injection timing. Don't bother with spreadsheets as they most often keep injection timing stock or spraying on a closed intake valve. Jason's is the only one on here I've ever seen take exhaust valve timing into account to spray the fuel into a cylinder with closed exhaust valve which is the whole point of changing the timing of the injector in the first place. Haven't looked at DStek's to see how he does it, so can't comment on his. I imagine with his knowledge, he would have set it up to spray with nearly closed exhaust valve, but like I said, I haven't looked at it.

[Michael_D]

Thanks again for your help..... Can't say how much I appreciate your time.

The spreadsheet I posted to above does in fact help you find the end of squirt to target AFTER the exhaust goes closed. But I didn't really move it all that much, cause I was a bit concerned about cylinder washout. I went ahead and moved the boundary +25. It did help. MAF is still different, same Hz, different gear though. So, I bumped it up another 15 for a total of 40 over OE setting. I don't think the extra 15 gained much, so that's as far as I went. Logged MAF in each gear, 1-5. Built a spread sheet, averaged the LTFT between each and flashed that cal to the low table. Seems to be a workable compromise. LTFT's are a +5-ish to a low of - 5-ish.... Screw it, good enough.

So then I reduced spark in the .28 g/cly rows, up to about 2600 rpm by another 4 deg. I'm now at 30 in these cells. Smoothed and matched the base / caost and low tables. Not much of a change, if any.

Then I got lost in the weeds screwing with VE again. Figured I better check it after moving injection timing around. Killed the MAF and took a drive. G-damn trims were so f'ing rich, they were pegged low. Drove about 15 minutes and it didn't get better, like it usually does in this time period. So I figured I'd just lower the whole VE table by 10%. Still way to rich. Lowered the table another 10%. Better, then low and behold, 15 minute later, I'm lean and KR is having a field day. Re-flashed the original VE tables and started over. This #@%@##%$@^%@ "rich after flash" BS is getting old!

I'll screw around with this some more over the next few days. Maybe one of these days I'll actually get to drive the car without constantly looking at a damn laptop.

[GHuggins]

Yes, I've beaten my head against the wall more than my fair share over the notorious rich after flash problem... Before tuning on the VE or MAF anymore, go ahead and change your O2 settings as described above, then change stoich down around 14.1 and then tune MAF. Leave VE rich for now as that may be close after all of this. Might also want to take 5 back out of your injection timing before all of this if you didn't notice any change with the last 15 degrees. If you have to you can continue lowering your stoich table and tuning MAF until bucking is completely gone or atleast much more less obvious. Might recommend if your indeed running headers to multiply your O2 airflow table by another .95 and see how that does. WE just need to get fueling under better control for that cam setup. I can't look at your tune, but if your min airflow table is too high it will cause additional problems. Just make sure your no higher than around 10g/s in your idle rpm area and there's no sudden increases from there.

OH.... Multiply your Idle - Proportional Airflow table by .8 and your idle integral air flow table by .8 and see if that doesn't help you out as well. The larger correction numbers can amplify the problem.

Get all of these changes put in and see how she does.

I would also recommend using this pid for tuning your fuel trims when using actual fuel trims and not a wideband. All it does is add long term and short term fuel trims together for both banks then devide by 2. Use this and run in as many different gears as possible for about 30 to 45 minutes at a time to get good averaged data. Just make sure your datalogging stft's and ltft's for both banks and it should work fine.

((([PID.6]+[PID.8])/2)+([PID.7]+[PID.9])/2)))

Sounds like we might be making some progress atleast...

[GHuggins]

Any updates? Better, No difference at all?

I know with your cam, you probably won't get rid of all of the bucking, but you should be able to help it out substantially...

[Michael_D]

I've been screwing around with injector timing and VE mostly. I wanted to get that out of the way first, so I installed the wide band and have been dialing it in as I get spare time. I'm sure I've moved injector timing as much as I should, and I have one final data logging trip to take yet, to verify my last move. But sadly, that has not yielded an improvement. The miss is still there.

I did also try the changes to the proportional and integral air tables you mentioned above. After I did that, fuel trims throughout the rpm range I was working in (idle - 4200), went -15 to -19. All gears. It was a global change, and it stuck. Even after an hour of driving and several starts. AFR held to 14.7, and the miss didn't mellow out any. Not really feeling an improvement, and not knowing exactly what these tables do, or why, I just moved them back to where they were before.

I did also multiply the O2 airflow table by .95. No noticeable improvement or change to speak of.

I have not tried moving the AFR table, yet, or changed O2 switch point voltage. Figured I'd try that later today.

I did however make a discovery, of sorts. And that's a complete elimination of this 'miss' while under load. There is a mountain at the end of my stretch of road that I use for tuning. It's a long, steep grade. Works great for WOT cal..... I drove up this grade in fifth, about 1800 rpm (where the miss is always present), and as a load was put on the engine, the miss completely disappeared. Spark did not change either. I'm still reviewing that particular log to see what changed, well other than the obvious - climbing a hill......

[GHuggins]

You'll most likely find the throttle opened up more which helped deminish the "egr" effect of your cam and injector pulsewidth increased helping to richen the fuel mix. This is why we've been trying to add more fuel. The only unfortunate way I know of to get rid of the egr effect is to reduce timing more, which has the possibility of making the engine less efficient in the areas. AND I mean reduce timing a lot, like 5 degrees or more...

The aftformentioned airflow tables are used for rpm corrections mostly at lower rpms, but never the less used for rpm corrections. Was trying to make them less sensitive...

Here's a thought since they drove your fueling really rich. Make the changes back to the airflow tables, then at the same time richen up your stoich table so the ecm doesn't turn around and try to take the fuel back out. You may also need to tune MAF at the same time and take airflow back out of it. SO change stoich to 14.1, change air tables back to their lesser numbers, change O2 switchpoints to 550, then tune MAF by taking air out of it.

See what that does. If your indeed fighting fueling, you should notice something here.

[Michael_D]

Weather turned to crap today.... No testing for me today. Just for reference, I am at 30 deg timing in these particular low load areas already. I have tried to move it further, down to 25, but that didn't seem to make any improvements. I've also gone the other direction, up to 50.

I'll play with the air tables and stoic tomorrow, should the weather cooperate.

It might be easier to just by you a plane ticket. LOL....

[GHuggins]

But we have so far seen some improvements with injection timing? AND when you said it was better, by how much possibly - 30 to 40%, more - less? Just wondering how much improvement we're possibly going to get. At what throttle percentages is the bucking present - only with very light throttle or all the way through 25% to 30% pedal travel? If it's light throttle, we can get rid of all of it...

Don't get me wrong, you should and will see improvements, but if it's 25 to 30% travel you'll possibly still have some present even after richening up fueling. Sorry, but that's just the nature of cams sometimes. Atleast we will be able to make it better and more pleasant to drive.

Really wish you had the O2 tables. That would have helped us that much more...

[Michael_D]

Your patience and help is very admirable.

I am pessimistic with any and all improvements, because I know that some may not be real, and only imaginary. In saying that, yes, the engine is behaving better today, than it was two weeks ago.

It is not unbearable either, just kinda irritating knowing it's there and not being able to get rid of it. I am somewhat of an anal retentive perfectionist.

I know it's reversion and all kinds of funkiness centered around the O2's and what they are telling the ECM. It just has to be. Seeing a consistent 6-9% higher fuel trims on the left bank, but no mechanical cause tells me that it's reversion. All within the same rpm/load areas that the misfire is occurring. I may just have to submit to the simple fact that tuning cannot fix a mechanically induced (reversion) problem.

Specifically, what I am trying to fix, occurs between 1600 - 2200, in gears fifth down. Sixth gear, 1200 - 1500 there is a very slight stumble, but nothing all that irritating, and I don't drive in sixth at the low of an rpm anyway. This isn't a bucking or surge problem, as that's just part of the beast of a cam with this much under curtain overlap and an open plenum. Right at 2200, in any gear, the engine just starts to smooth out. At 2500, it's beautiful. Trims are dead flat and it just comes alive. According to my scanner, TPS is 16% at idle. It is 25% at 1500, 27% at 2000 and 29% at 2300. Rarely does it get higher than 31%, light cruise.

Somebody else on this site described this in a way that is perfect. It "feels" as if I'm dragging a burlap bag filled with clothing behind me, and the bag hits and bounces along the road behind the car in a totally random manner, and every time it hits the road, it "tugs" on the car ever so slightly. That 'tug' is what I feel. And, I can also hear it in the exhaust note.

I've tried to pin it down to a particular cylinder, but didn't have any luck logging misfire events. I went so far as to build a spread sheet that I could populate with all the misfire events logged of every cylinder, then trend each to find problem cylinders (or one). Worked great, till the next log and a different cylinder would trend higher, then the next log and it was a different cylinder..... I gave up on that.

[GHuggins]

OK, that sounds like it might just be fixable... Not as bad as what I was taking you to say earlier. SO it's only with light throttle?

Do the O2, air tables, MAF and stoich as described above. See how that does.

I've even shifted the baro reversion table to the left "add 4" to the entire table before to help out with surging issues. You can try that, but wait until after the other tables are tried first.

You mind posting your latest tune so I or any others can take a look? Bad thing is that I just fixed this exact same problem on another car 3 weeks ago as it had to be fixed oddly too and I can't remember for the life of me exactly what I did to it. Guess I've got to many other things on my mind right now. I'll take a look at that tune and see what I did. Pretty sure it was just dialing in the fuel and MAF better that smoothed it right up... BUT it was as you stated - little "fish bites" up until around 2200 to 2500 rpms like your dragging a weight and then above that rpm it's like you break through a door. Sound about right?

[Michael_D]

5-23-2015_FINAL.hpt

My latest cal is attached. I put the ECM back into 'normal' with all the adders/fuel cuts enabled to test LTFT's and WOT AFR with the WB after working VE yesterday.

I was wondering what the VE reversion tables do. I asked a couple years ago, and no one seemed to know....

I have the beta HPT downloaded on my office PC too, and see they have some new VE tables. I have not used the beta to cal the ECM, however.... Beta always makes me nervous.....

Yes, what you describe is EXACTLY what's occurring! It's like an on/off switch as soon as the engine hits about 2500 - bang.....runs buttery smooth and snappy as a scalded dog.

Oh, and I love your PID that combines STFT's and LTFT's by the way..... I would have never thought to do that.

It's raining like cats and dogs.....grrrrrrrr

[Michael_D]

Here's the last config and log after the last calibration too...

Z06 CONFIG_AFR_ERROR_NTK_AFRM_WB_Custom.cfg5-23-15-FINAL.hpl

[realcanuk]

Just looked at the logs and tune out of curiosity. If the wideband is correct, it is really doing some strange stuff at times. I would try turning off DFCO just to see it anything changes. You got a lot of knock getting on the throttle there too. Maybe try lowering the PE throttle /pedal numbers so it can get into pe earlier. I am not sure but I don't thing those numbers really correspond with pedal percentage. It reads like it should, but I read somewhere it doesn't.