Stomp or Impact

[mr.prick]

AFR is rich coming out of PE in the low MAP areas (<40kPa).
This is only when the throttle is shut quickly,
if the throttle is steadily reduced AFR is what it should be.
Will {B3426}_{B3428} or {B3406} correct this?

[swingtan]

None of those?

In the E38, we have

B2011: Wall Wetting Asymmetry Factor


Determines the amount of port wall wetting dependent on acceleration or deceleration states.
Lower number will provide no compensation when deceleration is occurring.


This has a marked effect on the close throttle rich spike and is what I have used to get rid of it.

But COS3 doesn't have that particular table so it's going to be a bit tricky. I'd probably try....


B3416: Wall Wetting Min TPS: You may be able to set this so that on throttle close it disables dynamics. It may upset the throttle opening though.
B3401: Evaporation Time: Given that MAP would be low, you should be able to adjust the evaporation times to help. I'm guessing that a longer evaporation time in the low MAP cells would make the PCM add less fuel on transients as it would think there is more fuel left over from previous injection events.


But if your dynamics are right for throttle opening, I don't think there is a lot you can do without an asymmetry table like the E38.

Simon.

[mr.prick]

None of those

After playing around with Impact factor this apparently is the answer because
I could not see any fueling changes after reducing the values in the lower kPa areas.

I'll try B3401 and see what happens.
I am filtering out non PE data and using BEN.

[swingtan]

Does that also alter the "throttle open" fueling? If not it's a good fix!

[mr.prick]

BENs are slightly lean in throttle tip and the cell count is low
My WBO2 is a bit far back in the exhaust,
however I have noticed BENs are near perfect even with MAP increase.
This leads me to believe that exhaust gas moves fast enough to almost keep up with
MAP data.

The few lean cells are not consistently lean they do see rich AFR as well
just not enough to balance out the lean.
On a side note: cells in the lower MAP kPa that are rich are also of low cell count.
but I know they are accurate because they are consistent for log to log.

[mr.prick]

No dice, {B3401} didn't change anything.

[mr.prick]

The lean tip in is not so bad when easing into the throttle,
when the throttle is stomped it is more pronounced.
AFR richens as TP% decreases slowly or abruptly.

6355

[hquick]

For the 'tip-in's'...I found using the 'impact factor modifier' {B9116} table works well.

[swingtan]

Remember that for tip in, air flow is initially low, therefore the WB AFR will lag a little until airflow pics up. You can see this if you set the stomp setting significantly higher and end up with a rich spike on tip in. If this is done you can see you have commanded too much fuel but it will still lag the commanded AFR in the log.

The rich AFR on throttle close is what the "asymmetry factor" table fixes in the E38. I'm not sure if any of the table available with the Ls1 controller will help here.

The final point is possibly being affected by two things...

B3248: Stomp reduction decay rate may be a bit slow to pull back the stomp commanded fuel.
As air flow speed has dropped again, the WB is possibly lagging the commanded and reading the left over rich mixture from PE.

[mr.prick]

The lean tip in does not worry me so much
it's the damn throttle slam rich AFR that I want to cure.

Tip in is less depending on what gear and/or RPM I'm in at the time.
WBO2 lag is going to be there but looking at my logs
I can see the lag really is minute.

IPW drops along with TP% :confused: so why would AFR spike so rich when the
throttle slams shut?
Maybe injector offsets need to be changed in the low MANVAC areas?:nixweiss:

Any suggestions on how to filter/apply data?
I made a frame count pid to narrow down the sections with the spikes.
6356

[swingtan]

AFR's go rich on throttle close because there is fuel left in the intake from the previous intake cycle. This is the evaporation tables. the PCM expects some amount of fuel to be left and available for the next intake cycle, but if you shut the throttle, that fuel load may be greater than expected and gives you a rich spike when the throttle snaps shut.


Perhaps playing with the evaporation tables, indicating "more" fuel is left ( lower evaporation factor ) will make the PCm add less fuel on throttle close.

Simon

[LS1_Dragster]

For the 'tip-in's'...I found using the 'impact factor modifier' {B9116} table works well.

In what way? Can you give me more details on what to change? I don't have time to go out and make changes by guessing, so any info now would be very grateful.

Lee

[hquick]

In what way? Can you give me more details on what to change? I don't have time to go out and make changes by guessing, so any info now would be very grateful.

Lee

If you have a look at these two sets of numbers...you can see I increased the numbers in the lower grams/s area. That was what made most difference.
It was totally trial and error.

Stock numbers:
1.000002
1.000002
1.000002
1.000002
1.000002
1.000002
1.000002
1.000002
1.000002
1.000002
1.000002
0.958986
0.919924
0.879885
0.833986
0.793947
0.750002
0.700197
0.649904
0.609865
0.560060
0.560060
0.560060
0.560060
0.560060
0.560060
0.560060
0.560060
0.560060
0.560060
0.560060
0.560060
0.560060

New numbers:

1.000002
1.004885
1.129885
1.129885
1.138674
1.165042
1.189944
1.189944
1.189944
1.134768
1.087405
1.016604
1.000002
1.000002
1.000002
0.991701
0.923830
0.761232
0.659669
0.598634
0.518556
0.500001
0.500001
0.500001
0.500001
0.500001
0.500001
0.500001
0.500001
0.500001
0.500001
0.500001
0.500001

[LS1_Dragster]

Great! I really appreciate your trial and error!

Lee

[mr.prick]

I'm not having any luck with either of these tables.

[hquick]

Just as a side note to the whole 'Transient fuelling' issue.
I was talking to my GM calibrator buddy the other day and he mentioned that the guys who do all the work on the OEM Cal's 'transient fuelling' are NEVER happy.
So...maybe they have the same trouble we do.

[mr.prick]

Maybe thats the real reason for DFCO. :laugh:
LTFTs are good in the low MAP areas so I'm stumped.
I actually zeroed out the impact factor table and used RTACS with
evap time that had no effect on AFR.

[swingtan]

I have a theory..........

What is B4901 like? I wonder if this can be used to help with the initial mixture settings on throttle close? My thinking is....


Normal driving, the charge blend factor is correcting for a charge temp closer to IAT. So given that IAT will normally be lower than ECT ( when running at normal operating temp ) and the mixtures are normally set up for these conditions, fueling is correct ( BEN factor = 1 ).
On throttle close, the air flow drops to say 12gm/S from maybe 30gm/S when cruising.
If the blend factor is insufficient, the ECM will still think that the charge temp should be lower than it now it, as the slower moving air is warmed by the intake manifold. If the ECM thinks the charge temp is lower than it really is, then it would run rich.
B4902 is then used to "time" how fast the changes in the charge temp calculations work.


So, if B4902 is too low, then the charge temp correction will be too slow and you could run rich on initial throttle close, with a slow correction back to a BEN of "1". If B4902 is too high, then you would run lean on initial throttle close, with a slow correction back to a BEN of "1"

Does this make sense? I might try some work on this.....

Simon

[mr.prick]

Mine is stock, although I have thought about changing it.
Looking at an old log of {B4901} BENs are pretty good but
it has been hard to get a good filter for the throttle closed, PE to CL transition.
I've actually gone through frame ranges to look at all the averages.


#Units Low High Fmt Expression
#------------ ---------- ----- ------- ---------------------------------
*CLC-00-027
FN 0.0 99999 .0 "frame()"
================================================== =======
#Code PRN SLOT Units System Description
#--------------- ------- ---- ------------ ---------------- ---------------
CALC.FRAME F027 CLC-00-027 FN External "Frame Number"

[hquick]

That sounds interesting Simon.
I've played a little with the charge blending...but not seriously. I might need to re-visit it.