[Diy_efi] Re: Q regarding placement of a pressure sensor

[Brian Dessent]

Bob Moon wrote:
 
> Since I am only interested in the amount of air
> flowing through the engine at that point in time, my
> question is this: for a given airflow count from a
> MAF, is there a given vac in the intake?  It seems to
> me that a given airflow will always have the same
> pressure signal, regardless of what's happening
> downstream of the turbo.

The MAP + RPM systems are called "speed density" for a reason.  At the
end of the day, you need to know how many air molecules (i.e. mass) have
entered the cylinder to calculate a pulsewidth.  Ignore VE (volumetric
efficiency) for a moment: the volume flow rate will be proportional to
RPM, and the density will depend on the pressure (at the manifold,
downstream of the throttle) and temperature of the air.  Hence, you need
to know all three of these things to calculate mass air flow.  Just
knowing pressure without RPM and temperature is not enough to calculate
mass.

Now in reality, you certainly can't ignore VE.  Fortunately, it's a
physical property of the mechanical system, i.e. a parameter, unlike
mass/volume/temperature, so once it's determined it's just a lookup
table factor (as a function of RPM.)  The technical definition of VE is
the ratio of actual air volume delivered to the cylinder divided by the
theoretical air volume that would be delivered by a constant-density,
inviscid, uncompressible gas.  In other words, it captures the
"breathing ability" of the engine at a given RPM.  A value of 1 is
ideal, in real life I'd say a well designed (naturally aspirated) engine
can achieve 0.6 - 0.8, with higher peaks (sometimes even above 1.0) for
tuned intakes.  On a forced induction engine it can be substantially
higher than 1.0.

So, in order to calculate mass flow rate you need to know volume flow,
temperature, and pressure *at the same place in the flow*.  You could
measure the pressure anywhere you want, but you'd also have to know
temperature and volume flow at that point in order to calculate the mass
flow rate.  Unless I'm quite mistaken, most every speed-density system
measures temperature and pressure in the manifold (hence the M in MAP
and MAT) and uses rpm+VE to estimate volume flow rate.

What I would suggest is install MAP and MAT sensors, and then datalog
these plus RPM, and the stock MAF.  With enough data and some analysis
and curve fitting, you should be able to formulate a transfer function
(or enough lookup tables) that you could implement on a microprocessor
to feed into the stock ECU.

Or am I missing something?

Brian

_______________________________________________
Diy_efi mailing list
Diy_efi at diy-efi.org
http://www.diy-efi.org/mailman/listinfo/diy_efi