LTFT 10+% lean at idle when cold, but goes to near 0% when warm

[1bad68ls]

Hey Guys,

VEHICLE: LSA engine with BTR Stage 3 cam, long tube headers, ID1050x injectors, 30% overdriven supercharger.
ISSUE: I am struggling with the LTFT at idle. When the engine is cold (InjTipTemp ~ 85) and the vehicle goes into closed loop, the idle LTFT is around 11%. When the engine warms up (InjTipTemp >100), the LTFT is at near 0%. I have looked through the calibrations, however I am requesting help on what parameter(s) need to be calibrated to correct for warmup. I have looked at the Open Loop [IVT Gain Gas] table, but this table states it is for when in open loop only and my issue is in closed loop. I believe this is an injector warmup issue, and I do see a table for an IVT offset, but I do not believe this is an injector offset issue.
Attached is my current tune with the log file (also shown in the pic).
Screenshot 2024-07-01 124816.png

I appreciate any help.
Jeff

[Lakegoat]

I would forget about LTFT at idle. At idle, your STFT's are fine---however, look at the rest of the chart. It's waaay lean in a lot of it.
This is my way of setting PE: Enable 85kpa, Delay 0, Enrich rate 2.0000. Just my preference. I don't want to wait until 5000 rpm to have PE kick in. You are in boost before that.
Are you logging Lambda with your wideband?

[1bad68ls]

Thanks for the feedback Lakegoat. I incorrectly started my thread stating "ISSUE: I am struggling with the LTFT at idle." however the issue is not only at idle, it is at all speeds/loads until the vehicle warms up. The data that I shared is a cold start only and I am getting this lean condition at cold temps, which I am trying to overcome.
When the vehicle is warm, my LTFT tables are within 2% throughout the speed/load range (even at idle), BUT when cold, the tables show lean.

I am looking for how to calibrate the fuel delivery when the ECT (or IVT) is at cold conditions.
Any thoughts?

[Lakegoat]

Turn off LTFT's and tune the MAF and VE with STFT's for your drive around rpm's. Use the wideband for boost areas---Log in Lambda since you doubled stoich.

[1bad68ls]

Lakegoat, thanks for the input. My LTFTs are within about 2% in all speeds and loads when the vehicle is warm for MAF and VEs. My initial calibration for VE and MAF was using a wide band, then I fine tuned them with the switching sensors.

The issue is when the vehicle is cold, the LTFTs start at about 10% lean at idle (less in other areas)... but when the vehicle warms up (ECT>170F), the LTFTs are dead on. I believe there should be a temperature based correction table for fuel or airflow to get the fueling correct at lower temps. Looking for input on this.
Jeff

[L1FTD]

Lakegoat, thanks for the input. My LTFTs are within about 2% in all speeds and loads when the vehicle is warm for MAF and VEs. My initial calibration for VE and MAF was using a wide band, then I fine tuned them with the switching sensors.

The issue is when the vehicle is cold, the LTFTs start at about 10% lean at idle (less in other areas)... but when the vehicle warms up (ECT>170F), the LTFTs are dead on. I believe there should be a temperature based correction table for fuel or airflow to get the fueling correct at lower temps. Looking for input on this.
Jeff

Stock injection timing is fairly advanced when cold, spraying early until everything warms up. If you look in the Injector Timing Normal ECT (ECM 13337), timing retards 135* from 68F to 155F.

If you lower the values in that table (before 155F cell), you'll be retarding the injection timing and should see it go richer.

There's also the Offset vs. Inj Temp (ECM 12303). You can add or subtract from that table. If you add to the zero values you'll be adding fuel and should see it go richer. Just a change of one step makes a difference. You can see when it's hot, factory adds a lot (rich after re-flash).

Both of those will let you go richer when cold, and not affect your commanded EQ (unlike IVT tables).

Or, you could start your log (or use data from your log) only when you're above ~155F. The cold leanness isn't much of an issue, could be made better, but not much to worry about.

And as Lakegoat mentioned, can do away with the LTFT at idle or permanently. Will probably be better for you on a cammed, supercharged engine.

[blindsquirrel]

Or, you could start your log (or use data from your log) only when you're above ~155F.

One thing to keep in mind, I have seen one where the entire first key/run cycle from a cold start shows FT weirdness even if the log isn't started until it's up to temp. Get it warm, turn it off and start right back up again and then log. Compare those logs against logs doing it 'normal' (without shutdown/restart) - if you don't compare logs done each way you might not notice you're getting wonky data.

[L1FTD]

One thing to keep in mind, I have seen one where the entire first key/run cycle from a cold start shows FT weirdness even if the log isn't started until it's up to temp. Get it warm, turn it off and start right back up again and then log. Compare those logs against logs doing it 'normal' (without shutdown/restart) - if you don't compare logs done each way you might not notice you're getting wonky data.

Ah yes, I know what you mean, good point!

I think best time for a log would be from cold, using a tune that's already been driven on (i.e. flashed it one day and drove it, let it sit overnight, and didn't flash again the next morning), fuel trims cleared in scanner, and run all the way up to temp. without restart.

Second best, probably a morning flash when cold, trims cleared, and run up to temp. without restart.

Anytime you restart, all the 'virtual' temperature models try and do their thing. Particularly injector temp. going artificially high and working back down. Problem is, the models are usually still set up for factory parts, not the endless mods list.

The old '15-20min drive before logging' is always a good rule to follow. If you're making changes, try and use data from a consistent starting point each time, and adjust offsets etc. after your base data, to get it close in all conditions.

Not directed at you blindsquirrel, just the broader community if they happen to come upon this thread.

[1bad68ls]

Stock injection timing is fairly advanced when cold, spraying early until everything warms up. If you look in the Injector Timing Normal ECT (ECM 13337), timing retards 135* from 68F to 155F.

If you lower the values in that table (before 155F cell), you'll be retarding the injection timing and should see it go richer.

There's also the Offset vs. Inj Temp (ECM 12303). You can add or subtract from that table. If you add to the zero values you'll be adding fuel and should see it go richer. Just a change of one step makes a difference. You can see when it's hot, factory adds a lot (rich after re-flash).

Thanks guys, this is exactly what I needed. The Injector Timing Normal ECT (ECM 13337) had less of an impact on richening my cold start, but moving the EOI timing to 110 at the lower temps provided better idle at cold start. The Offset vs. Inj Temp (ECM 12303) completely cleaned up my cold start lean condition. THANKS!

[biggmac4]

Hey,

Curious how much did you have to add to fix the issue?

[1bad68ls]

Hey,

Curious how much did you have to add to fix the issue?

Hi Biggmac4,
For this [Injector Tip Temp(ITT) vs Inj Offset] table, I find that I can run with no additional offset for ITT in the 95F to 140F range. If my temps are higher than 140F, I gradually add fuel to counter the hot fuel handling effects (which is modified from what GM had in this table).
I also find that I need additional fuel for cold starts to keep my fuel trims stable. Below is a snapshot of a 75F cold start with my wide band sensor and I ran the vehicle in open loop and let it idle until the engine warmed up and the wideband sensor [EQ_ERR%] went to 0 error, which was about 4.5 minutes. When I ran this test, I started the car and I let it warm up with no throttle or load added. The chart #5 shows the Eq_Err starts at about ~25% lean, however slowly reaches 0% over the 4.5 minutes.

240911_12_28.png

To counter the cold start lean condition, I ramp in fuel in at the colder temps using the [Injector Tip Temp(ITT) vs Inj Offset] table, which resolves my lean cold start issue. Below is a snapshot of my current table. Note that our local temps have been warm therefore I still need to populate the table at temps below 68F.
ITTchart.png

[SiriusC1024]

Why don't you just set LTFT MinECT to something higher like 141F?

[1bad68ls]

Why don't you just set LTFT MinECT to something higher like 141F?

The issue is that I am lean at cold starts, which is causing a drive-ability issue. Setting the LTFT higher would not address the lean issue. When I run the wide band, the Eq_Err is lean at cold starts, which indicates the injector fueling is off. Some OEMs have a "lost fuel" (or "unburned fuel") model which accounts for the unburned fuel that only occurs during cold starts... If the unburned fuel calibration is tuned correctly and the WB sensor is warmed up, the wide band should always equal the Equivalence_Ratio_Commanded in open or closed loop. The GM [Injector Tip Temp vs. offset] table addresses both hot fuel handling (soak/lean restart) issues and cold (lean) start issues for me. It minimizes the Eq_Err in all conditions and keeps my drive-ability great.

[biggmac4]

Hey 1bad68ls,

This worked for me. Finally. My MAF and VVE tables being dialed in and not being able to add a little fuel in closed loop cold start was driving me crazy. I ended up running about .517ms up to 86* then tapered down to zero @104*. Nothing else I tried worked except this adjustment in [Injector Tip Temp(ITT) vs Inj Offset].

Thanks to all helped out .