Turbo MAP & MAT interaction

[Scot Sealander]

Webmaster wrote:
 
> I suppose it could be measuring the inlet manifold temp (i.e. the temp of
> metal manifold, not the air itself) and this could account for the slow
> rate of change, but I'm not sure why the ECU would want to know this value.
> There is already a separate Coolant Temperature Sensor (CTS), so I would
> have thought the ECU would use that to work out how hot the engine is.
> There is no ambiant air temp sensor (i.e. before the turbo).

In any EFI system the ECM needs to know the amount of air mass that
is in the cylinder.  With a MAF sensor, this is pretty easy.  With
a speed density system, it is a liitle harder to know what is in the
cylinder, but it can be calculated.  The ECM needs to know 4 things:

1) The swept volume of the cylinder               (Displacement/#cyl)
2) The temp of air in the cylinder                (IAT, ECT or combo)
3) The pressure in the cylinder                   (MAP)
3) How well the cylinder got filled last cycle.   (VolEff)

For most of these requirements we cannot measure them *in* the 
cylinder.  So we use sensors close to the cylinder to get a reasonable
number for needed values.

The first item is a known.  The swept volume can easily be calculated.

The temp of the air in the cylinder is tougher.  The IAT sensor in
your case (providing it is good) is the only temp sensor used to 
get the contribution to air density due to temp.  I have reservations
about mounting the IAT in the intake manifold.  In all the speed 
density turbo apps I have seen that use an intake mounted air temp
sensor, the air density table usually has very little variation from
cold to hot air temps.  My position is that this small variation 
confirms that the manifold mounted sensor may be a comprimise.

The IAT in this case is capable of measuring about -40 deg C to 
200 deg C.  Dry air density (at standard pressure) varies from 
about 1.5 g/L at -40 deg C to about 0.75 g/L at 200 deg C.
Don't shoot me if these numbers are wrong, they are off the top 
of my head...

There is a 2 to 1 variation in air density due to temp there.
In your case, the air temp density table has a value of 1.8 at
the -40 deg C, and 1.3 g/L at 200 degree C.  I did not plot it,
so I don't know if it is linear.  I my case, the Pontiac
Turbo GP, the -40 value (stock) is about 1.8 and the 200 deg
C value is 1.3....  Hmmm.  The TGP table is not at all linear
from cold to hot.

The SyTy is 1.5 g/L at -40 deg C, 1.2 g/L at 200 deg C.  I assume
this is done to be "safe".  

The cylinder pressure is taken from the MAP sensor, and finally,
the VE comes from a lookup table based on MAP and RPM.

Put all this into an equation, and you can determine how much 
air mass is in the cylinder.  Divide by AFR, and you have fuel 
mass needed.  Multiply fuel mass by the injector constant, and 
you have injector opening time needed to deliver the required 
fuel.

The air temp is a hard one to get right.  The air is heated
depending on how long it is in the intake & cylinder.  There
are some schemes to try and calculate this heating, but it
is not done in your case, or my case (TGP) or the SyTy.

I have corrected this in my car, and the results look good.
Very little correction ever needed by the closed loop
feedback system!

Scot Sealander
----------------------------------------------------------------------------
To unsubscribe from gmecm, send "unsubscribe gmecm" (without the quotes)
in the body of a message (not the subject) to majordomo at lists.diy-efi.org