LS2 FTC Boundaries?

[SMOKINV8]

I was just prodding through an LS2 vette bin file a little bit ago and I noticed the FTC RPM boundaries were all set to the same value which was in the 8400's. It seems as though the trend for GM tuning has been moving away from keeping the RPM boundaries functional, but I don't understand why. 98's used some reasonable values for them, 2000+ used the values where the mid and high ranges were 6200+, and now with this PCM they set them where they're not even reachable.

So what gives?

[PurplePiss]

I'd like to know the answer to this as well.

[EC_Tune]

Nothing wrong with changing them to something reasonable... GM did this to their calibrations back in 1995 as well. Never quite figured out why...
Does make tuning simpler as you don't bounce around in the fuel trim cell while "learning in" a cell.

[PurplePiss]

So what would you set a stock LS2 to? Can you explain these boundries a little?

[jimmyblue]

I haven't seen the LS2 tune files but expect the same
scheme as LS1/LS6. You're just working that old "cows
in the field" geometry problem (if any of you can both
remember elementary school, and went to one before
the New Math). You're trying to give each cell as little
work to do as possible, meaning within the MAP & RPM
boundaries of closed loop operation you want to divvy
up the trimming space finely (because the errors in air,
fuel, and system - such as thermal O2 issues - vary
with operating point). Driving-variable error gives a
"history tail" to the trims and it's not sane to expect
3000RPM, MAP=64.9 to match 1200RPM, MAP=40 in
any way.

You want the boundaries to not lie on top of your
main driving "pockets". You would like your everyday
normal modes (highway cruise, surface street roll,
etc.) to lie in the middle of a cell and get trimmed
all nice-nice, not bump across the line from (say)
a -5 trim to a -1 trim and back, never settling in.

If you simply accumulate a normal driving cycle and
then scatter-plot the RPM, MAP pairs in Excel you
will see where the meat is, and good places to set
your fences.

You probably want to crowd them lower in the
range, this is where the VE (real, and table) is the
most condition-variable and the system needs the
most resolution.

You probably want to put the upper RPM at the
same frequency as the dynamic airflow MAF cut-over
frequency, and the upper MAP boundary equal to your
PE enable MAP. There is no point in having fuel trim
cells allocated to open-loop MAP ranges.

You should find that fine-graining the trims makes
each cell more stable and segregates any gross trims
to a smaller number of them; this in itself is useful
as it shows you where, in the VE table, to focus any
fine-tuning at the low end.

[SMOKINV8]

Thanks for posting Jimmy, you're actually the guy that got me paying attention to it a while back with a similar post of yours on ls1tech.

Never quite figured out why...
Does make tuning simpler as you don't bounce around in the fuel trim cell while "learning in" a cell.

That's kind of the same boat I'm in, but I figured I'd ask just in case somebody figured it out.

Thanks guys.