Dipping idle after coming to stop

[phuz]

I'm trying to nail down a hunting idle issue that usually happens after coming to a stop. RPMs will drop to 300-400 and then eventually stabilize around 850.
I've datalogged and watched the airflow at steady-state idle and adjusted my Idle -> Airflow -> Base Running Airflow to match this as close as possible, but it still tends to do it.
What other parameters should I be looking to correct this?

[Alvin]

Quite often over fueling in decel is the cause of this. Check to make sure the VE table isn't rich in that area.

[RDF1]

Also datalog Throttle Follower and Throttle cracker to see if one of those falls off instantly when your rpms fall. Adjusting one of their timer delays can help substantially.

[phuz]

Also datalog Throttle Follower and Throttle cracker to see if one of those falls off instantly when your rpms fall. Adjusting one of their timer delays can help substantially.

Here's a simple log of just free-revving the engine a few times. It doesn't dip as low as it sometimes does, but the concept is still the same.
22-mar-2023-idle-rev.hpl

[Alvin]

The fuel trims show the engine is rich. This is the issue.

Throttle cracker and follower can mask this issue. The real fix is to fix fueling.

[phuz]

The fuel trims show the engine is rich. This is the issue.

Throttle cracker and follower can mask this issue. The real fix is to fix fueling.

Should I be starting with the Main VE table? I lowered the VE numbers by 3-5 in the idling portion of the map and interpolated out a few cells, but it's still sitting pretty rich. The STFT is still about the same, but the LTFT is 2% better.
Here's a fresh datalog from a cold start up to almost-operating temp.

• Also take a look at the various spots where the LTFT and STFT just reset to zero. What is causing this?
• What tables is it looking at to correct with both STFT and LTFT? I'm assuming the Open Loop EQ Ratio table is used for part of this.

25-mar-2023-warmup.hpl

Also, sharing this as maybe someone else will find it useful. I created a math channel which calculates actual VE from temperature, manifold pressure, RPM, and mass air flow rate. Then I created a table log with RPM vs. manifold pressure.
Actual VE.MathParameter.xml
ActualVEtable.png

UPDATE:
I applied the LTFT + STFT corrections to the VE table (multiply by half when pasting) and the LTFT still went to -31% and the STFT went to -10%. As if it had no effect.

UPDATE2: If I understand this correctly, the reason I'm not seeing any change is because the ECM is running off the MAF table, and not the VE table, and the VE table is only used in the event of a MAF failure.
That being said, should I be looking at adding adjustments to the MAF vs Frequency table? What else could I be looking at to eliminate these massive fuel trims at idle?

[Alvin]

Should I be starting with the Main VE table? I lowered the VE numbers by 3-5 in the idling portion of the map and interpolated out a few cells, but it's still sitting pretty rich. The STFT is still about the same, but the LTFT is 2% better.
Here's a fresh datalog from a cold start up to almost-operating temp.

• Also take a look at the various spots where the LTFT and STFT just reset to zero. What is causing this?
• What tables is it looking at to correct with both STFT and LTFT? I'm assuming the Open Loop EQ Ratio table is used for part of this.

25-mar-2023-warmup.hpl

Also, sharing this as maybe someone else will find it useful. I created a math channel which calculates actual VE from temperature, manifold pressure, RPM, and mass air flow rate. Then I created a table log with RPM vs. manifold pressure.
Actual VE.MathParameter.xml
ActualVEtable.png

UPDATE:
I applied the LTFT + STFT corrections to the VE table (multiply by half when pasting) and the LTFT still went to -31% and the STFT went to -10%. As if it had no effect.

UPDATE2: If I understand this correctly, the reason I'm not seeing any change is because the ECM is running off the MAF table, and not the VE table, and the VE table is only used in the event of a MAF failure.
That being said, should I be looking at adding adjustments to the MAF vs Frequency table? What else could I be looking at to eliminate these massive fuel trims at idle?

You cannot pull manifold temp which is a calculated value from a GEN 3 while logging. It is not using IAT for the fueling calculations. It's using manifold temp. On Gen4 and 5 cars where you can log this you still get poor results trying to tune the VE table from the MAF. Occasionally it works ok but mostly it's too noisy to get a good result from. I can look at the snapshot you posted and see without plotting that the results look like garbage in garbage out. No offense.

You need to fail the MAF to tune the VE table properly. Main and secondary if equipped will need to be tuned (matched)

The ECM will reference the VE table for quick transients among other things. It matters a lot to have a properly tuned VE table.

[phuz]

You cannot pull manifold temp which is a calculated value from a GEN 3 while logging. It is not using IAT for the fueling calculations. It's using manifold temp. On Gen4 and 5 cars where you can log this you still get poor results trying to tune the VE table from the MAF. Occasionally it works ok but mostly it's too noisy to get a good result from. I can look at the snapshot you posted and see without plotting that the results look like garbage in garbage out. No offense.

You need to fail the MAF to tune the VE table properly. Main and secondary if equipped will need to be tuned (matched)

The ECM will reference the VE table for quick transients among other things. It matters a lot to have a properly tuned VE table.

So how is it fueling now? MAF only? And what exactly is it targeting when it's in CL? Is it just Fuel -> General -> Stoich AFR?
I'm only familiar with my Haltech where I have a whole table of AFRs I can target, but the only thing I see here is that one parameter.

[Alvin]

Its blended MAF/VE

CL only goes by narrow bands which are Rich/lean. you can change the target all you want and it will not matter.

[phuz]

Its blended MAF/VE

CL only goes by narrow bands which are Rich/lean. you can change the target all you want and it will not matter.

Interesting. When I changed the VE table, it had zero effect on the LT/ST trims. I can suppose I can try larger increments, but the issue there is if there is a MAF failure, the VE table by itself won't provide enough fueling.

[Alvin]

I'm not sure what you want me to do to convince you.. Trust me. It matters. Its not a situation where half the fueling comes from the VE tables and the other from the MAF.

[phuz]

I'm not sure what you want me to do to convince you.. Trust me. It matters. Its not a situation where half the fueling comes from the VE tables and the other from the MAF.

Not calling you a liar, just trying to figure out why my idle is so rich (I know there's many possible reasons), and why modifying VE table appears to have no effect on it.

[phuz]

I pulled 25% out of the MAF from 0 to 4000Hz and it gave some pretty good results. I think I can pull another 5% and be OK with it.
But something I'm noticing from several logs, and I cannot find out why, is the trims are being ̶r̶e̶s̶e̶t̶ disabled by something. No change in TPS, RPM, MAP, or anything, so why is it randomly resetting? (Look at the 2:32 mark)

27-mar-2023-25%lessmaf.hpl

[bk2life]

Look at the 2:32 mark

maybe it is 'learned' at that point.
its been giving pulling 18 % of fuel for so long, it adjusts that cell and then resets and starts pulling 8.6% fuel.

im still learning as well, but need to pull some out of the VE table to get your idle not as rich.

Also, one vital thing i have learned, you cant really go by idle. Need to get out dirve it and then datalog.
I looked, but didnt see, anything about your combo? Is this stock?

If you search my name, youll see i fhave something simliar to what youre going through as i was real lean or rich and couldnt geti ti to add or remove fuel no matter what i tried. My issue, i wasnt driving far enough to get the computer to do its thing.

[blindsquirrel]

Also trim data is stored in multiple fuel trim cells, only one of which gets used at a time depending on operating conditions. Essentially multiple MAF/VE tables. However the tune file only has a single MAF table and single VE table.

[phuz]

maybe it is 'learned' at that point.
its been giving pulling 18 % of fuel for so long, it adjusts that cell and then resets and starts pulling 8.6% fuel.

im still learning as well, but need to pull some out of the VE table to get your idle not as rich.

Also, one vital thing i have learned, you cant really go by idle. Need to get out dirve it and then datalog.
I looked, but didnt see, anything about your combo? Is this stock?

If you search my name, youll see i fhave something simliar to what youre going through as i was real lean or rich and couldnt geti ti to add or remove fuel no matter what i tried. My issue, i wasnt driving far enough to get the computer to do its thing.

It's in my signature.
Full specs: Procharger with 8psi pulley, Torquer V2 cam, LS6 heads (and supporting mods), headers, no cats, Borla
Eventually this thing will go to the dyno and get professionally tuned, but I'm trying to do as much as I can ahead of time so we can focus on the important stuff. Drivability is pretty good right now with the exception of the dipping idle, which hunts and oscillates for several iterations before settling out.

I don't think it's learning and then storing it somewhere else and relearning, because when it does "reset to zero" my AFR goes rich again, and then slowly leans back out. If it was being learned and sticking, the AFRs would stay close to stoich.
I think I can figure out the rich idle part because I'm definitely making improvements on it, but this oscillating idle thing is killing me.
Here's another log from last night, where I blipped the throttle at the end a couple times. Look at the 11:58 mark and you can see the RPMs drop to less than 600 and it would've gone lower, but I hit the throttle slightly and it comes back up and steadies out. I don't think there's over-fueling happening here because the AFRs were around 13 and the fuel trims were not aggressive at this point. It just seems like the idle control isn't responding as fast as it should sometimes.

28-mar-2023.hpl

[reubone]

Post your tune file

[phuz]

Tune:
2001 Corvette_with_Procharger_MAF_changes.hpt

[Alvin]

You should consider doing a 2 bar OS with a 2 bar sensor and going away from the MAF. You will likely hit the limit of the MAF sensor on this setup.

[phuz]

You should consider doing a 2 bar OS with a 2 bar sensor and going away from the MAF. You will likely hit the limit of the MAF sensor on this setup.

Yep, and have the 2-bar sitting on the bench ready to go in, but trying to fix this hunting issue first. You can see from that datalog that I do exceed the 12,000Hz on the MAF on the one part of the datalog. I'm still breaking in my Monster Twin-Disc so for the most part I am driving conservatively, but I did want to "test" it a bit last night.

Side note on the switch to 2-bar: is the MAF used at all or is it completely disabled with the 2-bar OS? I really like the drivability of the MAF in the low end, but realize the fueling and spark accuracy won't be there in the high end.