Turbo control in a GM ECM

[James Boughton]

I just realized I have been screwing up my e-mail replies and sending
them only to specific people and not to the list.  OOPS!  Well you can read the 
comment I sent to dnult below as well as this response, since I think I am
sending this to the right place:)

Now, the only speed density system I have ever worked with that has a throttle term
to calculate base fuel (steady state) is a system that was put together strictly for racing.
The only reason we did this was because on an individual runner intake system there
isn't a convenient way to measure MAP.  All other systems I have worked with will ignore
throttle unless it is changing rapidly.  You talk about turning the MAP signal into load, well
most people would consider MAP and load to be interchangeable.  The engine couldn't care
less what position the throttle is in, it will only notice MAP.

You say you can have many different MAPs depending on between idle, decel, and WOT depending
on engine load.  This is exactly what I am saying.  MAP defines engine load.  Fueling is dependent
on MAP and RPM.  Throttle doesn't tell the ECU anything about load.

I will admit, for the sake of being thorough, that throttle is the indicator of idle and WOT.  This will 
activate different function in the ECU due to demands for good idle quality and rich WOT operation.
If you want to know why engines run rich at WOT it has been covered and is in the archives.

Believe me I know an engine must be calibrated for a MAP system it used to be my job (and sometimes
still is).  That is also why I KNOW that throttle is not used in any of the constant throttle fueling calculations.

How does this all apply to EFI?  If you don't know how an engine works you will have a hard time determining
the algorithms necessary for the software.  I can tell you that current systems by Chrysler and GM are not
the best way to run an engine.  They are a good starting point to learn from, but when you put 1000 engineers
together I guarantee you will not come up with what an engine really does.  I also am not claiming to know
exactly what an engine really does, but I try to learn a little every day.

As for MAF systems, try drilling a hole in the intake system of a MAF system or a MAP system.  Which will
handle the problem.  The automakers need a car to last 100,000 miles with vacuum leaks and all.  A MAP
system does a pretty damn good job.

Jim Boughton
jpb01 at ibm.net

----------
From:  dnult
Sent:  Friday, October 24, 1997 5:41 PM
To:  James Boughton
Subject:  Re: Turbo control in a GM ECM

Perhaps you are correct.  I neglected to observe that we were talking
steady state.  However you stated that the MAP could be directly read. 
Yes this is true, however, to turn this MAP signal into an engine load
or air consumption variable a calculation must be performed.  Wheather
that calculation is on-the-fly or from a lookup table (surface).  In
addition, throttle position is part of that calculation.  You can have
many different MAPs between idle, deacelleration, and WOT depending on
engine load.  Throttle position tells the ECU how to interpret the MAP
reading so it will properly adjust the fuel injection.

Here is the point I was trying to make.  A MAP system must be calibrated
to the engine being used based on the variables I listed earlier
(volumetric eff. etc.).  If you don't believe this, put headers, port
intake, or install a turbo charger on a MAP system and see what happens
to the fuel mixture.

A MAF system will easily adapt to these modifications since MAF is the
true variable needed to calculate fuel injection.  Ford for example
sells MAF conversion kits for this vary purpose - to allow the engine to
adapt to modifications that change the volumetric efficiency.  This
avoids the need for a new calibration (chip) when modifying the engine.

If I ventured off post.  Please forgive me.  However, I thought making
the distinction between MAF and MAP systems was important.

-dnult

James Boughton wrote:
> 
> On the MAP based engine control schemes I am used to working with you don't need
> any throttle information to calculate fueling for steady state loads.  Since MAP (manifold
> absolute pressure) is a direct function of the air entering the cylinders it is possible to calculate
> fueling based strictly on its value and RPM.  Typically, the only time throttle is used is to
> calculate the acceleration enrichments or the leanouts (deceleration enleanment to GM folk).
> In fact, if you could develop a control algorithm to determine true MAP changes and filter out the
> MAP oscillations due to manifold tuning you could completely fuel the engine based on only the
> MAP signal.
> 
> There is no calculation for MAP it is a directly measured variable.  Pulsewidths are typically calibrated
> into a surface of with x and y variables of MAP and RPM.  No formulas are necessary to determine the
> pulsewidth it is simply read straight from the surface.  There are modifications to the pulsewidth based
> on variables such as coolant temp., charge temp., and EGR, but again throttle is only used for transient
> control.
> 
> Jim Boughton
> jpb01 at ibm.net
> 
> ----------
> From:  dnult
> Sent:  Thursday, October 23, 1997 4:19 PM
> To:  diy_efi at coulomb.eng.ohio-state.edu
> Subject:  Re: Turbo control in a GM ECM
> 
> I could be wrong about this but, I thought that on MAP based engine
> control schemes, throttle position was an integral part of the formula
> for calculating engine load.  Keep in mind that MAP requires a
> Calculation based on engine size, volumetric efficiency of the engine as
> well as several other parameters I don't remember.  Basically a formula
> specific to a particular engine is required when using MAP.
> 
> On MAF systems the actual mass of air flowing into the engine is
> proportional to the MAF sensor output.  In that case throttle position
> is not important except for acceleration enrichment, idle and WOT
> detection.
> 
> Comments?
> 
> -dnult


-dnult