Positive LTFT when upshifting/downshifting

[FLORIT]

I have a 2004 Pontiac GTO, 5.7L LS1 V8. I logged some data last night and I have been looking it over. The LTFT values go positive, as high as 4.7, in the following situations:
1) At launch (1st gear).
2) Between the 2-3 shift.
3) When downshifting (approaching a stop light).
4) When punching the throttle, any gear, around 2000 RPM. (Throttle position changes dramatically).

I have not done any tuning to the car, but I have installed a "High Flow Induction" kit (with a conical reusable filter) and NGK TR55 plugs with Taylor 10.4mm wires.

The MAF sensor and conical filter FELL OFF when a friend was driving my car on the Interstate. Since that incident, I feel like I have lost a lot of low end torque. The LTFT's going positive under the listed conditions seems to confirm my perception. Does anyone have any comments? Or would you expect the LTFT's to go positive under those conditions? Thanks.

P.S. The car is a manual 6-speed and yes, I DID reset the PCM after the mentioned incident. I have driven the car at least 1000 miles since then, so I would think it has had time to "re-learn." Thanks again.

[jimmyblue]

Have to check the FTC (Fuel Trim Cell) you
are in, when you get the + trims. If it is the
one for deceleration, throttle closed, then
I would not worry about it. Closed throttle
is a hard place for the MAF or the speed
density tune to work; air fuel ratio with no
air to speak of. Plus, the MAF does have a
thermal time constant to it which limits its
responsiveness - worse, at lower airflows.

Now, if you were in the habit of flipping
abruptly from decel/closed to WOT, you might
have some chance of locking in that + trim.
But this is a pretty abnormal driving style I
reckon.

Bottom line is, you should try to layout your
LTFTs in a matrix like this:

FTC00 FTC01 FTC02 FTC03
FTC04 FTC05 FTC06 FTC07
FTC08 FTC09 FTC10 FTC11
FTC12 FTC13 FTC14 FTC15

for the main open-throttle cells and see if
any are positive, and if so, is there a bias
that teaches you anything? (You may see,
for starters, that you never touch half of
them, in a late F-body). But round 'em all
up and map them out, this will show you
if you have a RPM or MAP (or airflow)
dependent mixture error.

[FLORIT]

Thanks for the response. Sorting the data by LTFT reveals that throttle position during the positive LTFT's was never higher than 15%, usually 0% (i.e. decelerating, downshifting, etc.)

Since then, I have collected more data and uploaded my first fuel trim mods. The LTFT's are much better, averaging around -2.7 and almost never positive. Sure wish I had a dyno to measure any changes.

Lastly, when the Fuel Trim Cell is 20 or 1 or whatever, which table does that refer to? Is the FTC number the "row" or "column" of a table? The "Injector Flow Rate vs. KPA VAC" table has one row and 17 columns, so I don't think that's the one the FTC is referencing. So which table is it?

[jimmyblue]

None of those I think. The FTC refers to the
Fuel Trim Cells. These are delineated by the
Lower, Middle, Upper boundaries you see in
the Engine>Fueling page little tables. These
cells are just two tables 4x4x2 for cells 0-15,
B1/B2, STFT and LTFTs per bank. They get
added into the fuel shot calculations (insert
handwaving & mumbling) but are learned only
during periods where the "desired AFR scaling"
is 1.000 (elsewhere the O2 feedback is whack).
You don't get to set these, the computer does.

The other cells (16-up) are "special", not in the
orderly MAPxRPM array.

[jimmyblue]

[quote author=FLORIT link=board=gmv8engine;num=1086198438;start=0#2 date=06/07/04 at 19:30:54]Sure wish I had a dyno to measure any changes.
quote]

You can log the Engine Delivered Torque PID and RPM,
scatter plot EDT vs RPM "strips" in Excel, and it sure
does look like a torque vs RPM dyno curve... it may or
may not be the most accurate, and it's flywheel, but as
a relative measure of improvement (or anti) it's pretty
useful.