Decel / Idle Return / Tq ctrl Spark

[CCS86]

Hi guys,

I'm trying to sort out my last little glitch in using TQ based decel and return to idle.

From 1500 rpm or higher, it returns beautifully to idle. Swift and smooth.

The issue happens when I decelerate (in gear) to under 1500 rpm, then kick in the clutch. It doesn't stall, but can fall to 400 or 500 rpm before recovering to my 800 rpm idle.

The throttle angle control seems good. The moment I kick in the clutch, the throttle starts to open.

The spark control seems weird. It has held a steady 12.5* advance during decel, but the moment I kick in the clutch, spark source changes from 'closed throttle decel' to 'tq control' and advance drops to about -6*, then bounces between 0-6* as RPMs fall. Finally at 700 rpm (100 below set point), spark jumps up to 17*. This is also the moment that torque source changes from 'tipout limit' to 'target N'.

decel-dip.jpg

Any ideas? Is there a table controlling 'tipout limit'?

Log attached.

[MRRPMBRP]

Go to where it's doing that in torque table, reduce torque (&calculate inverse) by 50%, see if that helps.

[Slowleen]

Was this ever figured out I’m having the same issues and can’t seem to get the spark to stop going into torque control

[MRRPMBRP]

Have you reduced the values in the torque table? Mine was going into spark retard really bad, killing motor, lowered torque values, & fixed it.

[Thatwhite5.0]

Either throttle angle, spark, or fuel will be in torque control at all times.
You can adjust Target N in spark only and spark torque ratio as as the torque tables 0 and 1.
In the spark decel table, make sure to have spark at 17* (idle spark) a couple hundred RPM higher than desired idle RPM. If idle rpm is 700 with your idle spark at 17 but your spark is commanding 12 degrees at 700 RPM, it will use torque control to achieve Target N.

[CCS86]

Either throttle angle, spark, or fuel will be in torque control at all times.
You can adjust Target N in spark only and spark torque ratio as as the torque tables 0 and 1.
In the spark decel table, make sure to have spark at 17* (idle spark) a couple hundred RPM higher than desired idle RPM. If idle rpm is 700 with your idle spark at 17 but your spark is commanding 12 degrees at 700 RPM, it will use torque control to achieve Target N.

I'm not sure I understand completely. Here is my decel spark table:

decel-spark.jpg

I normalized this table to match spark values I was seeing at idle (~13*), so the transition was more smooth. It is quite a bit different than a stock GT table (all set to 60*). But I might try moving back towards the FPDX strategy values:

FPDX-decel-spark.jpg

These values seem like they will help the dips by adding a good amount of spark if RPMs drop. I'd just love to avoid RPMs dipping below commanded idle speed in the first place, instead of reacting.

I already have a big reduction in TQ values in the 0.10 load row of the MP0 table and a small reduction to 0.20. I can try going further.

[Slowleen]

OK, so I changed nothing in my tune. my program stopped working mid calibration write the other day and bricked my pcm for a few hours common issue. so I waited and then did a complete write this time. to my surprise the car now works with No issues now. idle no longer drops or stumbles when clutching in. but now the car is registering ipc wheel trq error.

[CCS86]

Here's what I am running into:

It seems logical to keep TQ vs load fairly linear. The inverse table seems to be to used more for functional aspects compared to the TQ table. My tables already point to 0 torque at 0.1 load, and they won't accept negative torque values (according to HPT warning). This means I can't raise the load value in the inverse table any more, or the calculated torque will be negative:


MP0-INV.jpg


The only thing I see that I could do here, is raise the inverse table row axis minimum from 0 ft*lbs to something larger. I'm not sure if this will cause issues though. It's possible that the PCM likes being bounded by 0 for calculations.

[CCS86]

I tried another revision with further reduced torque tables and some changes to idle spark. Still getting the dip.

Right when it switches to "torque control" spark, the advance jumps to -10* (retarded). Even with the RPMs 170 below setpoint, the spark is near zero. If I can find a way to reduce this spark retardation, I think the dip will be greatly reduced.

r15d-idle-dip.png


I still don't understand why my stable hot idle is averaging 10* advance, when my table is configured like this:

r15d-idle-spark.jpg

r15d-stable idle.jpg



I think it could also be running lean. In the condition during the dip, it is still running in OL, recovering from DFCO. Looking at the flow flow rate during the dip compared to steady idle, it is low. Any way I can force it a bit more rich in this after decel OL?

My [ECM] 11385 - DFCO Catalyst Reactivation Lambda: The lambda to reactive the catalyst, used after a fuel injector cut event is set to 0.85, which seems pretty rich already. But maybe the port walls are so dry after a long DFCO, there is a bunch of lost fuel following reactivation.


If it could get back into CL sooner, that would likely help. It takes about 3.4 seconds from fuel reactivation to closed loop engagement after DFCO.

[Witt]

I think it could also be running lean. In the condition during the dip, it is still running in OL, recovering from DFCO.

I've struggled with this in the past. The only real fix I found was to raise DFCO disable rpm.

[CCS86]

I've struggled with this in the past. The only real fix I found was to raise DFCO disable rpm.

Mine isn't responding as expected. Here are my DFCO rpm settings:

DFCO-rpm.jpg

But my fuel flow stays 0 until about 1140 rpm.

[CCS86]

Here is another view of the same dip.

Right at the point where RPMs cross my 800 rpm setpoint, the fuel flow (in green) is only 0.025 lb/min, compared to a steady idle flow rate of 0.064 lb/min (green dotted), despite the MAF rate essentially matching steady idle. That is definitely WAY low... ~60% too lean.

Injector PW is locked at 0.740, commanded EQ at 0.85. This seems weird, since MAF rate is varying slightly. PW finally kicks into a dynamic amount, right about the time that torque source switches from 'tipout limit' to 'target N'.

r15d-idle-dip2.jpg

[CCS86]

I tried a new revision where I fattened up the catalyst reactivation EQ, but the results were the same.

I have got to find out what tables are allowing spark to retard so far. In this one, at 130 rpm below my set point 'torque control' spark, with 'tipout limit' torque source is firing the plugs at 11* after TDC.

r15e-idle-dip.jpg

And this condition corresponds with an area of the torque table that has massive reductions compared to stock:

r15e-MP0-TQ-comp.jpg



Does anyone know which of these spark only torque ratio parameters matches with the "tipout limit" torque source?


spark-cut-tq-ratio.jpg

[MRRPMBRP]

Curious as to why your torque table doesn't have load labeled in the rows?

[CCS86]

Curious as to why your torque table doesn't have load labeled in the rows?

It does, but they blank out in compare display when every axis value doesn't match. I was comparing against a stock GT table, so my max load of 1.6 blanks out the load axis.

The row values start 0.1, 0.2, 0.3, 0.4...

[CCS86]

Still looking for answers:

• Why is DFCO staying enabled well below the deactivation RPM?

Here's a log of it happening.

DFCO Enable: 1550 rpm
DFCO Disable: 1400 rpm


DFCO appears to stay active (fuel flow rate 0) until ~1000rpm

[blackbolt22]

Still looking for answers:

• Why is DFCO staying enabled well below the deactivation RPM?

Here's a log of it happening.


DFCO Enable: 1550 rpm
DFCO Disable: 1400 rpm


DFCO appears to stay active (fuel flow rate 0) until ~1000rpm

How about this theory: Spark isn't dropping enough so it cuts fuel. Spark tip-out settings, torque ratio settings come to mind. Also is your DFCO MPH settings correct for what you want to occur?

[CCS86]

How about this theory: Spark isn't dropping enough so it cuts fuel. Spark tip-out settings, torque ratio settings come to mind. Also is your DFCO MPH settings correct for what you want to occur?

I like the thought, but here is my observation:

during that whole period of fuel cut, spark source is 'Closed thr. decel', so it pretty much tracks the 'tip-out / decel spark' table.

The instant that fuel flow kicks back on, spark source changes to 'TQ control', and advance immediately drops really low.

The MPH settings for DFCO are both zeroed, like the stock GT tune. I think you just have independant choice between using RPM, MPH, or both. Zeroed MPH values, seems like it would disable MPH control. That would be fine for me, because the real issue I'm having, is that if DFCO stays active at too low an RPM, then I kick in the clutch, it gets close to stalling. It could happen in any gear though, which changes the MPH it occurs at.

[MRRPMBRP]

CCS86, I just started another thread exactly like the problem you're having. I had it good, but as soon as I put a smaller pulley on, I can't get it tuned again, on takeoff, or going back to idle. I am having super low spark values, no idea why it wants them so low, before idle spark kicks in.
On takeoff, there's a small spot where I'm super lean, under 1000rpms, shows combustion stability limit, but briefly, briefly enough to cause some bucking, right where the clutch takes at, combined with it switching back to the target N super low timing, makes takeoff really nasty.

[blackbolt22]

What about DFCO enable VSS @ 128mph to rule that out..
Spark Decel Table: 1500+ rpms lower the spark to say 10-15*. Raise under 1500rpm to 20+ degrees progressively higher towards idle?

It really seems like the car thinks it should be decelerating faster so it cuts fuel because it doesn't have enough authority. These things are a pain sometimes hope you get it resolved...

I did away with torque based decel and went full dashpot. She drives like a dream now