[Diy_efi] Donegan ECU

[Bernd Felsche]

On Friday 05 January 2007 09:49, Steven P. Donegan wrote:
> On Fri, 2007-01-05 at 08:00 +0900, Bernd Felsche wrote:
> > On Friday 05 January 2007 00:57, Steven P. Donegan wrote:
> > > I had not thought of fuel temp at all - every vehicle I drive
> > > would consume the gas in the rails well before it got warm :-)
> > > However this does raise a point - perhaps my ECM/PCM/EFI
> > > computer needs a way to open a fuel return line and to sense
> > > fuel temp in the rails? Any ideas on how useful that would be
> > > in the 'real world' anyone?

> > Superflous if you're running in closed-loop; which you would be
> > if the fuel rail was warm enough to make a difference.

> > If the temperature in the fuel rail is a problem, then a
> > recirculating fuel pressure control system is IMNHSO a better
> > solution. In such a system, there's always "fresh" fuel that's
> > in excess to the amount required for injection from the tank
> > flushing the rail(s). The fuel tank is the cooling environment
> > for the fuel.

> > Pressure regulation also happens at the rail(s), whereas in
> > "dead-end" systems it's at the fuel pump, perhaps a several
> > metres from the rail and therefore the injectors. That increases
> > the difficulty in controlling the pressure (time delays that
> > depend on fuel pressure and temperature); especially if it's to
> > vary dynamically wrt manifold pressure.

> > The ability to vary the fuel rail pressure is at least desirable
> > to get consistent injected quantities due to a fairly constant
> > pressure difference across the injector; between the fuel rail
> > and the manifold where it's injecting.  Makes for simpler
> > calculations on injected quantity.

> > If you're stuck with a "dead-end" fuel delivery system, then you
> > need to add a return line and a valve that vents the rail(s)
> > back to the tank in the interval between the fuel pump running
> > and the engine actually being started. The time delay will
> > depend largely on the free-delivery rate of the fuel pump and
> > the volume of the fuel rail(s). That ensures that there's "cold"
> > fuel in the rail(s) before you start injecting it.

> I can't quite read this as a yes or a no to fuel pump pressure control -
> can you give me a binary response :-)

No. There isn't one. That should be obvious.

> I.E. does fuel pressure sensing/fuel pump control make sense for
> an EFI system or not.

It depends on the amount of time and money that you have to spend;
and the quality of control that you want out of the system.

It is very much more expensive (in time and money) to out-perform
the delivery control of a simple fuel pressure control referenced to
the inlet manifold at the fuel rail. The response of the mechanical
stuff at the rail is sharper than changing the fuel pump's speed;
and a bleed-back at the pump will always have to be against the
dominating flow of fuel.

The result is that the injectors will deliver a different quantity
of fuel to that intended and the closed-loop feedback at the exhaust
will tend to make your control system unstable as it attempts to
respond "statically" to a dynamic effect.

-- 
/"\ Bernd Felsche - Innovative Reckoning, Perth, Western Australia
\ /  ASCII ribbon campaign | "If we let things terrify us,
 X   against HTML mail     |  life will not be worth living."
/ \  and postings          | Lucius Annaeus Seneca, c. 4BC - 65AD.