25% ltft

[KCG]

I'm pretty much in the dark when it comes to tuning. I'm starting up a new build and have very rough running and severe backfire....
SBE LQ4, L92 heads, LS3 stock intake and injectors. Custom cam and bench tune by EPS.
Upon startup, everything was good. Fired immediately, idled smooth, sounded good with open headers. This was only about a minutes worth of time. I didn't even have coolant in it yet. After a few short startups it started running rough and backfiring really bad. I thought maybe because open headers. Few days later, got loose ends tied up including the exhaust on. No change. It backfires and won't idle. I changed plugs. #2 & 4 were BLACK. Tried again. Holding rpm around 2000, and entering CL, the LTFT on bank 2 goes to 25%. Bank 1 at 12%.
Why would they do this?
Did I screw up by starting a few times with open headers?
If so, why would only 2 & 4 be so black?

[joecar]

It would do that with open headers (the low pressure in between each pulse causes fresh air to be pull in enough to hit the O2 sensors, causing trims to go positive).

Try clearing the trims (use the DVT tab), and try again.

If it keeps doing this, then check for air leaks between MAF and header collector.

[KCG]

Great. thank you for the info. I'll give it a shot.

[KCG]

I disabled LTFT at Fuel-Trims-Parameters thinking it would stop the 25% add till I can get this figured out. Also cleared Fuel Trims using DVT. Installed new plugs.
But I can't even get to Closed Loop now. The engine is running really rough and severe popping out the exhaust. I shut it off only about 20-25 seconds in. I did scan that if anyone wants to look at it. I don't see it telling me anything, but again, I'm new here.

Aside from it missing so bad, what causes an engine to pop so severely out the exhaust, AF ratio, bad gas, timing?

[ScarabEpic22]

If you disable LTFTs, I'm pretty sure you can't run in Closed Loop. The ECM needs the O2 feedback to make adjustments to the fueling based on output (LTFT/STFTs).

At LTFTs +25%, it means the tune is stupid lean. You'll need to figure out why, is there a MAF? Where is it located? Have you talked to EPS to see if they'll get you a tune update now that you've fired the engine up?

[KCG]

Above Joe says more than likely I caused this issue by starting the engine with open headers a few times. So I understand how it would think its lean and bumping LTFT to 25% when it gets to that point.
Standing back and looking at this....the engine fired right up, idled steady and sounded good for the first 4 or 5 times I started it. Only 15-30 seconds at a time. 5th & 6th time is when it started missing and popping severely thru the exhaust. That's the first time I saved the tune from the PCM. Did I save bad(?) data or does the PCM not change the actual tune that I would save?
On another suggestion, a fellow helped me get the LTFT disabled so that wouldn't be adding to the problem why I try to figure this out. Thats the only reason they were disabled.
I re-flashed the tune I saved from the PCM with LTFT disabled and clearing fuel trims like Joe suggested. I expected the car to start and run smooth just like it did on first srart. It didn't. So theres my questions....did I save a tune that was already altered in that first minute or so?

To answer your questions, 85mm MAF, its temp mounted to a 12" long 4" tube straight in front of the TB. A cone filter is clamped on the end of it. I have not talked to EPS. This tune and cam was ground years ago. Life got in the way and I couldn't complete this till now.

Again to back up a sec...it won't hardly run at all now. Severe popping and runs extremely rough. Shouldn't reflashing have at least got me back to square one?

Thanks for looking at my thread!

[5.7ute]

I would be double checking the valvetrain.

[KCG]

Really...what are you thinking is happening?

[5.7ute]

My thoughts are that if it started fine at first without a hiccup, and now has a miss, then there is the possibility that a mechanical failure has occurred.

[KCG]

My thoughts as well...not necessarily a mechanical failure, but maybe that, bad gas, loose grounds....just something other than the PCM causing this. That's why I'm directing my questions to anyone who can tell me that when I realized I needed to retrieve the tune from the PCM, did it already change itself due to the operating conditions of those first few minutes.

One of my replies in this thread is missing. I emailed Joecar about it. I guess this would make more sense if I typed it out again

[5.7ute]

All good.
The tune itself cannot change, but there are a few learned values that can mess with you. Disconnecting the battery should reset these, as will reflashing the tune.

[KCG]

OK great. That was my first mistake, I didn't save the tune before I started the engine. But by disconnecting battery, which I do all the time, and clearing fuel trims, then reflash, and it still starts running rough and popping right away should be saying the tune isnt the problem. Am I correct saying that?

[5.7ute]

That is how I would see it, yes.
The actual tune file does not get modified while the engine is running. There is just a few learned values that get stored elsewhere in the PCM's memory. So the file you read out of the pcm will be the exact same file you flashed in. Sometimes there are a few minor variances due to the way efilive, (or any other software), converts the binary value to usable units but that is another story for another day.
I would be trying to isolate the cylinder causing the miss and swapping components around to isolate the cause.

[KCG]

Good deal. Thanks a lot for hanging with me here tonight.
Hopefully I'll update soon with what i've found.
Again, really appreciate your help.

[5.7ute]

No worries.
Daytime over here.

[KCG]

Ah right:thumb_yello:

[KCG]

So I had checked my fuel pressure to be sure and all seems well except when I pulled the guage off the rail I noticed how yellow looking the gas is. Some dripped on the floor, and even on concrete it still looks yellow. This was 93 octane I bought at a very busy location.
So i dripped a little bit of 87 octane for my mower on the floor. If you look closely, you can see the yellow at the top of the picture and the clear gas actually making that partial ring as it flows in to the yellow. WTH is the yellow? old gas still in the tank or just bad new gas. I'm 99% sure I completely emptied the tank years ago when I put the new 255lph pump in. I haven't put any new gas in till this start-up.

Is this my severe miss and backfiring?

21346

[KCG]

Guys, thanks for trying to help me out. I really appreciate it. Even tho my problem wasn't in the tune, it helped me to work towards finding the problem.
I'm new here but will be visiting often as I try to learn how to work on these tunes.
Long story short, my bank 2 coil harness was reversed. DOH!

[5.7ute]

Guys, thanks for trying to help me out. I really appreciate it. Even tho my problem wasn't in the tune, it helped me to work towards finding the problem.
I'm new here but will be visiting often as I try to learn how to work on these tunes.
Long story short, my bank 2 coil harness was reversed. DOH!

Thanks for the update.

[joecar]

I was away for a few days, what did I miss...?

Bank2 coil harness was reversed: so the cylinders were reversed...?

Fuel containing yellow substance: what is that...?

[KCG]

Sorry Joe, missed your reply. Yes my fault on the reverse coil wiring bank 2. No more popping. After 5 years I guess I was just too excited hearing it fire off right away to realize it really wasn't running smoothly.
Figured out the YELLOW gas was what the 93 octane here looks like.

Now part 2...

LTFT still at 25%. STFT ~10-12%. Working on the phone with my tuner and emailing logs and tunes, a lightbulb went off in his head. Trims adding 30-35% and my LS3 injectors being 30-35% larger than LS1, he sent a me a tune with LS1 injector data. Loaded and immediately change to trims. Much closer to zero. But why?
He feels the air the MAF is reporting makes sense and commanded afr is 14.63. I've went over this thing multiple times for unmetered air. Can't find any.

1 - MAF sensor reading incorrectly? bought used
2 - O2 sensors reading incorrectly? both new with first start

edit:
Been reading stickies...Could I follow this procedure to disable MAF and see what trims do?
Would the results tell me if my MAF is reading correctly or not?

A. C2901: MAF High Frequency Fail 1: Change to 1 Hz
B. C2903: MAF High Frequency Fail Limit: Change to 1
C. Copy B5913 High-Octane Spark to B5914 Low-Octane Spark
D. Insure MAF Sensor Circuit is displayed as a current DTC

[joecar]

Sorry Joe, missed your reply. Yes my fault on the reverse coil wiring bank 2. No more popping. After 5 years I guess I was just too excited hearing it fire off right away to realize it really wasn't running smoothly
No worries... I got so many projects in all different directions... yes, it's always exciting when you fire it up after putting it all back together :cheers:



Figured out the YELLOW gas was what the 93 octane here looks like.
where are you located...?



LTFT still at 25%. STFT ~10-12%. Working on the phone with my tuner and emailing logs and tunes, a lightbulb went off in his head. Trims adding 30-35% and my LS3 injectors being 30-35% larger than LS1, he sent a me a tune with LS1 injector data. Loaded and immediately change to trims. Much closer to zero. But why?Sounds like he edited the IFR table to match the larger injectors you have, so now the ECM opens the injectors for shorter period, so now your trims went down.



He feels the air the MAF is reporting makes sense and commanded afr is 14.63. I've went over this thing multiple times for unmetered air. Can't find any.

1 - MAF sensor reading incorrectly? bought used
2 - O2 sensors reading incorrectly? both new with first start
Post pics of engine bay showing MAF plumbing.
Does your MAF have the honeycomb screen (which is an airflow straightener)...?
Post some logs, I can eyeball MAF wrt other pids to determine if it is correct.



edit:
Been reading stickies...Could I follow this procedure to disable MAF and see what trims do?
Would the results tell me if my MAF is reading correctly or not?

A. C2901: MAF High Frequency Fail 1: Change to 1 Hz
B. C2903: MAF High Frequency Fail Limit: Change to 1
C. Copy B5913 High-Octane Spark to B5914 Low-Octane Spark
D. Insure MAF Sensor Circuit is displayed as a current DTC
This would cause the PCM to run from the VVE/VE tables rather than the MAF, it would check how close VVE/VE is modelling air (i.e. if trims are not close to zero, then there is room for correction).

[KCG]

No worries... I got so many projects in all different directions... yes, it's always exciting when you fire it up after putting it all back together :cheers:

where are you located...?

Maryland


Sounds like he edited the IFR table to match the larger injectors you have, so now the ECM opens the injectors for shorter period, so now your trims went down.

Yes, numbers were ~5.x, now ~3.5


Post pics of engine bay showing MAF plumbing.

21597


Does your MAF have the honeycomb screen (which is an airflow straightener)...?

I think what you're referring to looks like this.. and mine is NOT like this

21594


Post some logs, I can eyeball MAF wrt other pids to determine if it is correct.

ok cool. Heres basically before & after the data change, EDIT: they're AFTER, then BEFORE.



This would cause the PCM to run from the VVE/VE tables rather than the MAF, it would check how close VVE/VE is modelling air (i.e. if trims are not close to zero, then there is room for correction).

I was looking at this as a way to test if the MAF is reporting correct data. I thought if its DISABLED, and my trims go somewhere other than with ENABLED MAF, then that would tell its reporting wrong. And the reason LS1 data has band-aided it even tho they're LS3 injectors. Did I explain clearly?

[joecar]

There is a 90 degree bend in plumbing immediately after the MAF... oin this bend, most of the airflow is on the outside, this skews/distorts the airflow that the sensor sees.

[joecar]

...
I was looking at this as a way to test if the MAF is reporting correct data. I thought if its DISABLED, and my trims go somewhere other than with ENABLED MAF, then that would tell its reporting wrong.
That would work, as long as the VE was close.


And the reason LS1 data has band-aided it even tho they're LS3 injectors. Did I explain clearly?Sorry, I'm not following you on this sentence (I just ran out of coffee)...?

[joecar]

From your logs, your MAF is reading low.

[KCG]

There is a 90 degree bend in plumbing immediately after the MAF... oin this bend, most of the airflow is on the outside, this skews/distorts the airflow that the sensor sees.

How much of a straight section is recommended on the output side?

[KCG]

That would work, as long as the VE was close.

Sorry, I'm not following you on this sentence (I just ran out of coffee)...?

I'm just trying to say since the injector data has been changed to create acceptable trims, this says something is wrong. Whether its the MAF, O2 sensors or possibly the injectors themselves.

From your logs, your MAF is reading low.

ok, reading low...from the 90 degree discussed above or faulty MAF?

[joecar]

I'm just trying to say since the injector data has been changed to create acceptable trims, this says something is wrong. Whether its the MAF, O2 sensors or possibly the injectors themselves.


ok, reading low...from the 90 degree discussed above or faulty MAF?From the bend being so close to the MAF sensor (the bulk of the air travels on the outside of the bend, starting from about 6-8 inches before the bend (because the airflow has to be continuous, i.e. it cannot have any discontinuities/steps), this causes the sensor to see less air as oriented in your pic.

[KCG]

ok, am I thinking this correctly...
The ECM thinks there is LESS air going into the engine. In other words, there's MORE air going in, the ECM doesn't know about.
So with O2 feedback trying to maintain commanded 14.63, it reads this additional air and therefore tells the ECM we need more fuel to maintain stoich?
That's why my trims were going so positive?

[joecar]

ok, am I thinking this correctly...
The ECM thinks there is LESS air going into the engine. In other words, there's MORE air going in, the ECM doesn't know about.
So with O2 feedback trying to maintain commanded 14.63, it reads this additional air and therefore tells the ECM we need more fuel to maintain stoich?
That's why my trims were going so positive?Yes.

[KCG]

ok cool. I really appreciate the lessons. Thank you.

So, is this procedure I posted earlier the correct way to force SD if I just want to get rid of the MAF at least temporarily? Not physically, I would still use the IAT right? If so, I don't understand what D is asking to do.

A. C2901: MAF High Frequency Fail 1: Change to 1 Hz
B. C2903: MAF High Frequency Fail Limit: Change to 1
C. Copy B5913 High-Octane Spark to B5914 Low-Octane Spark
D. Insure MAF Sensor Circuit is displayed as a current DTC

[joecar]

Yes, you always want to use IAT.

A and B are setting the MAF fail parameters to make the MAF "fail" immediately...
when the MAF "fails" it causes a DTC to show up, this is what D is asking you to verify;

the presence of a MAF DTC indicates that the PCM has abandoned the MAF and is running from VE only.



When the MAF DTC shows up, the PCM now does a few things differently:
- it now runs from the LO Spark table B5914,
- it now run from the Backup VE table B0103 if it exists,
- it may now ignore the knock sensors,
- it may now set the trans line pressure to max (see P0101-3 in tables E0102,3,4),
- a few other things possibly (most likely).

[KCG]

All great info Joe.

- B0103 does not exist in my tune (ECM?)
- Is there a "systems check" or way to verify KS are working?
- E0102 etc are all set to NO

[joecar]

B0103 is the Backup VE table that exists in all the LS1A and a few LS1B calibrations.


To verify that KS are working, try this (works only on some LS1B's):
ignition key on, engine off,
log the pids GM.ESC1DNL and GM.ESC2DNL,
knock sharply on the front of the cylinder head with a larger wrench (say 20mm combination) or the smallest sized ball-peen hammer...
sometimes you will see activity on one or both of those pids.


MAF DTC's in E010x set to No: ok this avoids the PCM setting trans max line pressure and/or trans fault state when the MAF is in failed state... but do monitor line pressure by logging the pid GM.TFMPRS.psi), make sure that goes up easily/quickly with throttle.

[KCG]

B0103 is the Backup VE table that exists in all the LS1A and a few LS1B calibrations.


To verify that KS are working, try this (works only on some LS1B's):
ignition key on, engine off,
log the pids GM.ESC1DNL and GM.ESC2DNL,
knock sharply on the front of the cylinder head with a larger wrench (say 20mm combination) or the smallest sized ball-peen hammer...
sometimes you will see activity on one or both of those pids.


MAF DTC's in E010x set to No: ok this avoids the PCM setting trans max line pressure and/or trans fault state when the MAF is in failed state... but do monitor line pressure by logging the pid GM.TFMPRS.psi), make sure that goes up easily/quickly with throttle.

Quick update..
Added the pids and rapped on the passenger side head a few times. Did not get any counts.
Added trans pressure pid and can see it jump quickly as throttle increases but drops to zero soon as input stops (steady state throttle). I assume this is correct response.
So, MAF is failed, Hi octane table copied to low. Started the engine and after warmup, STFT are in the -10% range. This is at idle. Wasn't able to drive for other reasons.
Replaced B3701, B4001, B4003-5, B4307 with data from the original tune (LS3 data instead of band-aid LS1 data). Now at idle, I'm about 13-14% positive
I'll be able to drive it some and log tomorrow.

[KCG]

Cold start. About 30 minute drive. Trims still plenty positive.

[joecar]

Look at the VE table where the LTFT's are positive, and remove about that much %.

[KCG]

Is it common for a mail order tune to be this far off? I'm at +25% a good part of that log..even with MAF failed

[joecar]

Usually they're closer than that (some MO tuners are very good, they have figured out the right questions to ask)... there are some bad ones... and then sometimes stuff just happens.