MAF and SD: Who is telling the real deal?
02-18-2005, 09:25 AM
#1
Launching!
Thread Starter
Join Date: Nov 2002
Location: Miami
Posts: 275
Likes: 0
Received 0 Likes
on
0 Posts
MAF and SD: Who is telling the real deal? First I have a question. Would an aftermarket LID change the output of the MAF to a great degree? My MAF has my LTFT added fuel. With it disconnected I have fuel being taken away. If the aftermarket LID is causing the MAF to read uncorrectly do I need to recalibrate the MAF? Or just work with the VE?
Maybe I need a better understanding of how the readings for fuel is done. Now with the MAF unplugged does this just mean the computer simply isn't calculating wind density and temp? If so am is my LTFT really fixing my air fuel ratio or hurting it?
Maybe I need a better understanding of how the readings for fuel is done. Now with the MAF unplugged does this just mean the computer simply isn't calculating wind density and temp? If so am is my LTFT really fixing my air fuel ratio or hurting it?
02-18-2005, 10:31 AM
#2
Moderator
iTrader: (11)
Join Date: Mar 2002
Location: East Central Florida
Posts: 12,605
Likes: 0
Received 6 Likes
on
6 Posts
MAF works by hot wire sense elements and embedded
in the electronics calibration (and carried through the
rest of the mapping / computation chain) is an assumed
relation of total airflow to "sensible airflow" (the small
part of the air stream, that actually affects the thermal
operation of the sense element).
When you do things that bias the intake tract airflow,
like centrigugally crowd the air to the top (the 'vette
guys have issues with this, the Z06 MAF table reflects
the "air bridge" effects of this sort to some extent is
my take; trucks also sometimes respond badly to
descreening with their curved plumbing) you change this
total / sensed airflow ratio but the MAF has no clue about
that.
I think this is also why the nitrous guys see varying
enrichment based on how close they place the nozzle
before the MAF, etc. - you're creating various hard
biases in the system, for better or worse.
If you can wade through the various VE stickies you will
come away with a splitting headache and a couple of
nuggets o' knowledge (or, well-informed opinions from
people who dug through the code at least). What I got
from it is:
1) The airflow calculated below 4000RPM is a blend
of the two (SD, MAF) separate airflow calculations.
The blend favors SD down low where the MAF is less
accurate and tapers to all-MAF at 4000RPM+
2) -except- whenever the MAP is unsteady, it's all SD
there and for some period afterward (?)
Now, check your fuel trim cell boundaries. Many F-bodies
use Cell 0, 4, 8, 12 to cover 0-2500RPM and 1, 5, 9, 13
for 2500-6500.
Given that the true VE changes like mad between 650RPM
& 2500RPM (even stock; worse w/ many mods) and that the
"final airflow" computated also is changing the relative
contribution weight of SD and MAF terms over that range,
it stands to reason that doing a decent closed loop trim
over that space, with one cell value, is about as likely as
assembling an engine with a single open-end wrench.
If you can unplug the MAF and get your VE table righteous,
than you can plug the MAF back in, attribute any LTFT drift
to it and work the MAF table until you get back exactly to
where you were with it unplugged. But that's some work.
If you divvy up the closed loop driving space with more
sensible FTC boundaries then you will see a tighter trim
action, the 0/4/8/12 cells will respond almost 1:1 to VE
table tweakage (the 1/5/9/12 less so but somewhat, the
2/6/10/14 more to MAF and the 3/7/11/15 most likely
all-MAF). When you understand / expect this and have it
sliced more finely it should be easier to treat each cell
and each, being responsible for less of a changeable-basis
space, will be pushed around less by driving history.
in the electronics calibration (and carried through the
rest of the mapping / computation chain) is an assumed
relation of total airflow to "sensible airflow" (the small
part of the air stream, that actually affects the thermal
operation of the sense element).
When you do things that bias the intake tract airflow,
like centrigugally crowd the air to the top (the 'vette
guys have issues with this, the Z06 MAF table reflects
the "air bridge" effects of this sort to some extent is
my take; trucks also sometimes respond badly to
descreening with their curved plumbing) you change this
total / sensed airflow ratio but the MAF has no clue about
that.
I think this is also why the nitrous guys see varying
enrichment based on how close they place the nozzle
before the MAF, etc. - you're creating various hard
biases in the system, for better or worse.
If you can wade through the various VE stickies you will
come away with a splitting headache and a couple of
nuggets o' knowledge (or, well-informed opinions from
people who dug through the code at least). What I got
from it is:
1) The airflow calculated below 4000RPM is a blend
of the two (SD, MAF) separate airflow calculations.
The blend favors SD down low where the MAF is less
accurate and tapers to all-MAF at 4000RPM+
2) -except- whenever the MAP is unsteady, it's all SD
there and for some period afterward (?)
Now, check your fuel trim cell boundaries. Many F-bodies
use Cell 0, 4, 8, 12 to cover 0-2500RPM and 1, 5, 9, 13
for 2500-6500.
Given that the true VE changes like mad between 650RPM
& 2500RPM (even stock; worse w/ many mods) and that the
"final airflow" computated also is changing the relative
contribution weight of SD and MAF terms over that range,
it stands to reason that doing a decent closed loop trim
over that space, with one cell value, is about as likely as
assembling an engine with a single open-end wrench.
If you can unplug the MAF and get your VE table righteous,
than you can plug the MAF back in, attribute any LTFT drift
to it and work the MAF table until you get back exactly to
where you were with it unplugged. But that's some work.
If you divvy up the closed loop driving space with more
sensible FTC boundaries then you will see a tighter trim
action, the 0/4/8/12 cells will respond almost 1:1 to VE
table tweakage (the 1/5/9/12 less so but somewhat, the
2/6/10/14 more to MAF and the 3/7/11/15 most likely
all-MAF). When you understand / expect this and have it
sliced more finely it should be easier to treat each cell
and each, being responsible for less of a changeable-basis
space, will be pushed around less by driving history.
Last edited by jimmyblue; 02-18-2005 at 11:02 AM.
02-18-2005, 11:21 AM
#3
Jimmyblue:
what do you have your MAP/RPM boundries set to? in my 99 i had 1200/2000/6500rpm (or something like that, i don't have my laptop with me) and I was wondering if there's a better split. I was playing with 1200/2000/2800 to chop it up into idle/cruise/go teritories.
What's the rule of thumb for the boundries?
what do you have your MAP/RPM boundries set to? in my 99 i had 1200/2000/6500rpm (or something like that, i don't have my laptop with me) and I was wondering if there's a better split. I was playing with 1200/2000/2800 to chop it up into idle/cruise/go teritories.
What's the rule of thumb for the boundries?
02-18-2005, 11:37 AM
#4
Launching!
Thread Starter
Join Date: Nov 2002
Location: Miami
Posts: 275
Likes: 0
Received 0 Likes
on
0 Posts
There is a lot more to tuning that I thought. This is fun stuff I will admit. I am going ot go to they dyno anyway and see where I'm at. Make some small adjustments to the PE table if I am lean or too rich.
