control algorithm for MAF setup

[Corey L. Cole]

Jeremy Edmondson wrote:
> 
> Dave Williams wrote:
> >
> > -> Remember with the MAF sensor it will measure airflow in both
> > -> directions and in the FORD strategy there is a lot of compensation
> > -> for backflow and pulsation at low engine speeds.
> >
> >  So will most MAFs.  So far I like the idea of sampling the MAF at a
> > particular crank position.  That would tend to even things out quite a
> > bit.
> 
> I suspect you are better to integrate it over an entire engine cycle, need high
> sample rates though.  Remember you are interested in the total quantity of air
> induced rather than some value at some point in the cycle.  Also because the intake
> system is tunned it will exhibit distinct resonant frequencies and hence
> considerable phase changes will occur throught the rpm range.  Obviously triggering
> at a specific crank angle will not always trigger with the same airflow event.
> 
> >  If backflow becomes too much of a hassle, I'm likely to just go down to
> > the motorcycle shop and get a bunch of two stroke reed blocks.
> 
> WHY????  This is the exact reason why efi is used, compensation for non-linear
> effects such as reversion.  It is easy, compensate for it in the fuel map, this is
> why the maps exist otherwise it would be really easy and a simple formula could be
> used.  The backflow/reversion ocurrs at exactly the same rpm/load point so just
> tune the map accordingly.  Carbs cannot easily do this.  Efi works great on race
> motors as far more wild cams can be used without the tractibility problems
> associated with poor fuel metering.
> 
> Alfa tried reed valves in the intake ports so that wild cams could be used but the
> parasitic losses associated with the obstruction in the air stream far out weighed
> the benefits.  Its a good idea, I had though of a similar scheme as it allows very
> radical cams to be used without losing low end torque.

Speaking of reed valves and MAF sensors...

I have an '87 Escort with EFI that I'd like to convert to MAF.  Is it possible
to 'trick' the computer into thinking that the MAF output is what it's
expecting?  I was thinking along the lines of some kind of circuit that would
correlate MAF output with what the computer was expecting.  But then again,
I'm an aero. engineer, not a spark chaser.

Any suggestions?
-- 
==============================================================================
Corey L. Cole                     | Cole's Law:  Thinly sliced cabbage
 M/S 19-HH                        | 
 E-mail: corey.l.cole at boeing.com  | Perpetual optimism is a force
 Phone: 206-662-3596              | multiplier  -Gen. Colin Powell
==============================================================================
Disclaimer:  If I'm not speaking FORTRAN, I'm not speaking for Boeing.