[alfa] Electronic Injector operation

["John Fielding" <johnf@xxxxxxxxxxxxxxx>]

Hi Peter,

I saw your explanation of the EFI given to another digester.

I think you misunderstand how the normal injector works?

The fuel pressure serves to hold the pintle valve on its seat.  The small compression
spring fitted inside the injector is simply to ensure the pintle is seated when the fuel
rail bleeds down and to assist with seating as the injector is turned off.  When the
injector is supplied with current it requires a lot of force to lift the pintle before
some fuel will start to flow around the pintle, across the seat and finally exit at the
injector tip.  Peak and Hold drivers are used for low Z injectors because they allow
faster opening times.  But the hold current being about 1/4 of the opening current should
have alerted you to one other very important factor.

The fuel pressure holds the pintle firmly seated until it is lifted a very small part of
its opening distance.  (Typically an injector pintle only moves about 0.2mm).   When the
pintle lifts, the fuel starts to flow around pintle and across the seat and the pressure
across the valve seat drops to a very low value.  It becomes essentially a
"balanced-valve" and the pressure across the seat being very low does not any longer force
the pintle valve onto the seat.  As the pintle lifts to its maximum the fuel pressure now
helps to hold the valve open with the smaller magnetic field generated in the solenoid.
This is why a peak and hold injector is a better proposition than a high Z injector, where
the solenoid current remains at a constant level.  The high peak current serves to bump
the pintle off the seat and then the fuel flowing serves to hold the injector open, hence
a much lower current is required to keep the injector open.  When the injector is turned
off, the fuel pressure then reverses its action and the small compression spring fitted to
the solenoid starts to close the valve, as this happens the pressure drop across the seat
rises, at some point the pressure drop has risen sufficiently to cause the valve to become
"unbalanced" and then the full fuel pressure causes the seat to bottom and the fuel is
turned off.

The way the injector operates also explains why it is imperative to maintain as constant a
fuel pressure as possible in the fuel rail.  If the pressure fluctuates over a wide range
as the injectors open and close the injection volume varies, hence the fuel/air ratio also
varies.

Standard turbo vehicles do not normally have a rising rate regulator.  The pressure
regulator most commonly used by Alfa is the same part number as used on the normally
aspirated Motronic EFI fitted vehicles.  The same part is also fitted to the Fiat and
Lancia normally aspirated and turbocharged EFI systems.  A rising rate regulator
complicates the fuel map as the ECU has to do a lot of extra calculations to quantify the
changing flow rates as the pressure varies.  In this is the assumption that the fuel
pressure follows the inlet manifold pressure pound for pound.  You need to input the
changing flow rate of the injector for each data point to get close to the real thing.

In a rising rate regulator it is specified by the pressure ratio.  For a 2:1 rising rate
regulator, if the boost rises by 10% the fuel pressure will rise by 20%.  So for a
boost of 1bar the fuel pressure will rise 2bar.  The type of regulator normally
used is a non-rising rate type. It varies the fuel pressure pound for pound with
the inlet manifold pressure  All this is factored in when the ECU designers
write the software for a particular engine map and the tuners confirm the
parameters on real engines on a dyno.

Flow rate is not a linear thing with pressure.  If the pressure increases by 50% the
flow only goes up about 22%, it is a quasi-sqare law equation.  If the fuel pressure
is raised too high the pintle valve takes longer to lift, hence the calculated injection
time is now too short and the engine leans out.  When the fuel pressure reaches
about 4bar, for a normal 3bar injector, the flow rate has dropped to about 75% of
that at 3bar, because the injector takes longer to open.  I have seen many adjustable
fuel regulators fitted by owners who think they are doing the right thing, where the
tuner has cranked the fuel pressure up to 5bar and the engine runs leaner than
when it is reset to 3bar, as the injectors were designed for!  On one turbo
vehicle, raising the fuel pressure made the engine run so dangerously lean it detonated
more and more as the fuel pressure was increased.  As the pressure rises the fuel no
longer acts like a liquid and becomes more like a viscous solid, akin to putty, and
hence harder to squeeze through the very small orifice. Jump into a swimming pool
from 10 feet and the water acts like a liquid, jump in from 1000 feet and the water
acts more like solid concrete, because the impact pressure is much higher.

Rising rate regulators impart a non-linear slope to the flow rate.  To double the flow
rate for a fixed orifice size means that you theoretically have to raise the pressure more
than 100%, this makes the pintle harder to lift and hence a leaner mixture.  Because the
ECU has no knowledge of the fuel pressure it cannot compensate the injector periods to get
the mixture right.  The other factor is the fuel pump flow rate.  Most fuel pumps are
specified at a fixed delivery pressure.  Taking a typical Bosch pump, it might be
specified at 195l/hr at 3bar but at 5bar the delivery volume has fallen to a little over
130l/hr.  At the specified "dead-head" pressure of about 8bar the flow rate has dropped to
zero! At zero pressure it can flow about 500l/hr.

Operating an injector at duty cycles above about 75% also brings about a strange effect.
If you plot the flow rate with a constant fuel pressure with different duty cycles on an
injector test fixture, you get a lot of conflicting data.. Simplistically injectors are
specified at xcc/min or xx lb/hr for a certain fuel pressure.  Most injector manufacturers
use a 3bar test pressure to characterise the injectors and a standard temperature of 20C.
If you operate an injector at WOT you would expect to get the manufacturers flow rate, let
us say it is 400cc/min.  At 50% duty cycle you should therefore get half this figure -
200cc/min.  In practice you do not.  This is because you have to factor in the opening and
closing time for the injector for a constant fuel pressure.  As mentioned earlier the fuel
pressure must be held absolutely constant, this requires a fuel pump with a flow rate
about 5 times more than a standard fuel pump and a very large reservoir chamber to
maintain the pressure.  Typically the volume of the normal fuel rails fitted to a vehicle
are far short of that required to do this test with any meaningful accuracy.

When the duty cycle exceeds about 80% the pintle valve shows a peculiar effect, it appears
to float on its seating and the flow rate drops like a stone.  High Z injectors are
generally worse than low Z.  On injectors I have tested on a flow rig the flow rate drops
to around 60% of the specified figure when run at 80%.  Some injectors are even worse,
especially the high Z types, these can dip into the 40% region at 85% duty cycle.

The way that most injector flow rigs do the test is to run the injector at an applied 50%
duty cycle and assume the opening and closing time is zero and the WOT flow rate is twice
the flow measured, this is not always the case.  I have measured new injectors from a
reputable manufacturer stamped with 700cc/min and some can barely deliver 300cc/min at WOT
for the specified test pressure.  Not knowing what the injector actually flows can make
you chase your tail when trying to set up an EFI ECU.  One of the most critical factors is
the flow rate of the injectors, if you do not know this with any certainty then the game
is off!

John
Durban
South Africa
--
to be removed from alfa, see /bin/digest-subs.cgi
or email "unsubscribe alfa" to majordomo@domain.elided