O2 Sensor Open/Close Loop

[tom cloud]

>tom cloud wrote:
>> 
>> I don't know for sure, but some calculation has to be made (is
>> it basically in speed-density mode ??? at WOT ???)  Clearly, if
>> it's already speed-density, then it just goes with whatever calcs
>> or maps are set up for it.  
>
>What do the OEM guys use to setup the map values for pulse vs RPM vs
>temp vs load?
>
>
>Mark Mason

Mark,  I'll tell you what I know (very little):  To my way of thinking,
EFI is a sophisiticated extension of carburation.

Gots an engine.  Need air + fuel.  Assume the basic engine design is ok.

Carb gives A/F mixture based on venture principle for mid to WOT and
on atmospheric pressure / manifold pressure for idle (I welcome
corrections here).  The "calculations" are effective and have worked
for years, but they cannot be precise -- it's a mechanical kluge that
becomes far too complex trying to integrate all the engine variables
(temp, load, atmos pressure, altitude, humidity, etc) that affect
engine operation.

Carb has idle 'circuit', cold operation 'circuit' (choke), normal /
cruise 'circuit' (venturi), and acceleration / power 'circuit' (accel.
pump).

The Holley EFI I have duplicates these 'circuits' effectively -- better
than a carb -- using only engine rpm, temp, and TPS -- with the option
to add EGO feedback.

OEM EFI, IMHO, just fine tunes these functions.  I believe that we
hinder ourselves, many times, trying to duplicate their systems, which
are designed to meet EPA standards, and are created with massive research
budgets.

Duplicate a carb and your vehicle will run -- well.  Fine tune the
functions and it _might_ run better.  Fine tune it a little more, spend
an inordinate amount of time and money and achieve, maybe, some
questionable improvements (point of diminishing returns).  Add spark
control, data logging, etc.  All of these are probably fun.  What's
needed?  Whatta 'u want?  Power?  Nah!  Economy?  Legal?

Use EGO feedback, and you can reduce emissions.  Oscillate the EGO
output and you can use a three-way catalyst.  Run a little lean, increase
emissions and improve economy.  Run a little rich, increase emissions
and improve performance.  Unfortunately, the EGO is only accurate at
stoich (A:F = 14.7:1), so it is not an effective feedback for other fuel
mixtures.  For those, you're stuck with educated 'guesses'.  (I'd like
to build a system that stores parameters at various engine operating
conditions -- at stoich -- and extrapolates 'intelligently' to desired
leaner or richer mixtures.)

I 'spect the OEM guys do lots of testing to get the "mixture" (it's
a pun, get it?) just right.  That's why I like the little pots on
my EFI.  I can fine tune it rat now!  Don't need no $2000 laptop.
What do I think I want?  It'd be re-programmable via laptop,
but it would have 'analog' potentiometer-type controls to allow
fine tuning -- maybe even 'coarse' tuning.  Could then read selected
values from laptop and make program changes.  I also plan to use the
best of both analog and digital -- including external circuitry --
to simplify the design and programming process.  Avoid doing every-
thing with the processor.  For example, set the injector pulse width
buy loading a timer that's triggered from ignition pulse and then
leave it alone 'til it needs changing.  If doing ignition, do the
same thing with spark advance.  This frees up the processor (and,
more importantly, makes my programming chores easier -- don't have
to worry about tight interrupt loops, etc) to do things like figger
out what the injection p.w. needs to be.  Dunno, it's a bag-'a-
worms -- there are many factors.  That's why I have to kick my
cranial mass back to considering the simplicity of my Holley
Pro-Jection:  only uses rmp, TPS, and temp (for 'choke' function),
and it works great!  I would expect it to work with less adjustment for
atmospheric and altitude (same thing) changes if I added EGO
feedback.

Now, where are youse guys on this list that really know about this stuff?


Tom Cloud <cloud at peaches.ph.utexas.edu>