[Diy_efi] returnless fuel rails

[bcroe at juno.com]

I would like to suggest that while such a system could be 
developed, it might turn out not to be trivial at all.  To begin 
with, you have pulse width, pulses per second, and MAP 
readings all impacting the required pump drive.  You may 
have no idea of the transfer function of the pump, and the 
pump you chose may be different or much less predictable 
for this application than the one used by Fo*d.  A feedback 
system will need a fuel pressure sensor to work at all, and 
some kind of Proportional/Integral/Differential (PID) loop
will have to be developed to fit the particular components.  

A feed forward system would lessen the accuracy 
requirement of the above, but probably not eliminate the 
need.  Even if steady state values are reached within 
the required accuracy, transition situations could cause 
problems and "tuning" to cancel the fuel pressure errors.

Building a reliable pump driver without wiping out your 
radio (or ECU signals!) may not be simple either.  If
such problems are not in your line of experience, it 
might be best to just plumb in a fuel pressure regulator.

Bruce Roe

On Sun, 26 May 2002 19:00:49 -0700 Perry Harrington 
<pedward at apsoft.com> writes:
> FWIW...
> 
> My understanding of these systems (used on ford explorer 
> v8 and 99-up mustangs), is that they have an electronic fuel 
> pressure gauge that feeds back to the computer.  The 
> computer simply PWM controls the fuel pump to
> vary the voltage, and thus pressure/flow.
> 
> It seems to me that it would be very trivial to make a 
> circuit (or hack into the Megasquirt) that controls the 
> fuel pump via feedback.
> 
> The bonus on the Megasquirt is that since it's MAP 
> based, you already know the intake manifold pressure 
> and can use that to keep the pressure drop
> across the injectors linear.
> 
> --Perry

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