Index Home About Blog
From: John De Armond
Subject: Re: 98 Cruisemaster mystery
Date: Wed, 03 Nov 1999 12:44:13 EST
Newsgroups: rec.outdoors.rv-travel

sbourg wrote:
> In article <>, Neon John
> <> wrote:
> > Simple fuel pressure measurement won't work because the pressure is
> > regulated to a set pressure referenced to manifold pressure.  The
> > gauge pressure varies with load.
> I'm not sure I understand this - it certainly isn't that way on any
> vehicles I've worked on. Those maintain a set fuel pressure,
> mechanically regulated. 

Every modern-style EFI system, from the old Volkswagen Type III to
current production engines reference the fuel pressure to the intake
manifold pressure so that there is a constant differential pressure
across the injectors under all conditions.  This is accomplished by
connecting the reference side of the regulator diaphragm to the
intake manifold.  That's what the little rubber hose coming from the
top of the regulator is for.  The reason this is done is so that it
can be assumed for the purpose of calibrating the ECU that the
injector will always flow the same mass of fuel per unit open time
under all conditions.

>In order to achieve a calibrated injection of
> fuel into the manifold/port, the ECM measures the manifold pressure,
> and sometimes the air flow, to determine only the length of time the
> injector is on (or # of pulses, etc). Any pressure differential between
> the fuel line and manifold is accounted for in the ECM fixed program,
> and loop-corrected from the oxygen sensor if the mixture needs fine
> tuning.

No ECU I'm familiar with, and I'm familiar with all the major OEM
ECUs and several aftermarket ones, does this. The ECU MUST assume
the injector has a constant mass flow constant.  It is effectively
impossible to measure dynamic manifold pressure.  The reasons are
many.  First, for speed-density systems, the time constant of the
manifold pressure sensor, the connecting tubing and the ECU input
filtering is longer than the time between injections at anything
much above idle.  If a system was designed so that the ECU tried to
correct every injection cycle for manifold pressure using a signal
with a time constant longer than the injection period, at least one
and probably more injection periods would be uncontrolled because
the MAP signal lags manifold conditions.  Secondly, the dynamic
conditions (resonance, pulsations, reversions, etc) in the intake
manifold militate against accurate real time manifold pressure
measurement.  Mass flow systems have similar limitations and in
addition, for GM and companies that either use GM parts or copy
them, the mass flow signal is transmitted by a low variable
frequency AC signal.  Between the thermal time constant of the MAF
element, the sensor processing electronics and the time it takes for
the ECU to measure the MAF signal, the loop is simply too slow to do
real time measurements.  The simple solution is the one used:
Reference the fuel pressure to the manifold pressure so that real
time measurements are not necessary.

You can easily see this with a fuel pressure gauge.  Hook it up,
start the engine and blip the throttle.  The fuel pressure will rise
along with manifold pressure.  A differential pressure gauge with
one leg connected to the manifold and the other to the fuel rail
will indicate the actual pressure across the injectors.  This will
remain (almost) constant regardless of the throttle position.

> Steve
> * Sent from RemarQ The Internet's Discussion Network *
> The fastest and easiest way to search and participate in Usenet - Free!

Index Home About Blog