[alfa] More on bigger injectors!

["John Fielding" <johnf@xxxxxxxxxxxxxxx>]

Hi All,

With all these posts regarding the results to be expected from fitting bigger injectors, I
thought I would throw my 2c into the fray!

Eric Storhok gave a pretty good outline of the scenario but I feel he assumed most people
understand the EFI saga, which from some of the comments made does not seem to be the
case.  Most L-Jet & Motronic systems I have worked do run a rich mixture at WOT at
full load, the exact value varies between cars but if the A/F ratio is low then it is
usually a fuel pressure issue or a restricion somewhere in the fuel delivery circuit,
filters or injectors. Every job is slightly different, some it is just a weak pump, Nissan
seem to suffer more than most from this for some reason.

In the Bosch L-Jet and Motronic using the O2 sensor input the response time is
deliberately made slow.  From reading up on O2 sensors and the integrated circuits used to
proces the output signals the general consensus is that the response time to a change in
A/F ratio is of the order of 300ms for the O2 sensor output voltage to change.  This bears
out measurements I have made on various O2 sensors using both a propane torch method and
also from shooting a blast of carb cleaner into the throttle bodies of several vehicles
and observing the O2 voltage change on a 'scope.  I am unsure were the statement that an
O2 sensor can read individual exhaust pulses came from.  This does not seem the case from
measuremnts I have made, but perhaps I am missing something here?

In practice the ECU doesn't want to see rapid swings in O2 voltage as it will get into a
mess trying to make the correct decision based on O2 voltage.  In any case at idle the O2
sensor is probably out of its working temperature range even with a heated sensor type.
Running an engine with the O2 sensor active during idle usually gives a hunting idle
speed.  Switch off the O2 control and the idle settles down to being nice and steady,
perhaps a bit rich or lean but the engine seems happier that way.

On a cold start the O2 sensor is not in the loop.  The initial fueling is derived from the
engine temperature sensor, the water sensor, and it uses the base fuel map to determine
the injector pulse width.  So if it is a very cold condition the mixture needs to be very
rich to ensure adequate atomisation of the fuel, can be as high as 180%.

In most ECU's the software only reads the engine temperature every second or so.
It does not expect to see a rapid change in water temperaure.  (To prove this try
pulling off the water temp sensor connector on a fully warmed up Motronic system
and see how long it is before the exhaust turns black).  The same goes for the air
inlet temperaure, again it doesn't expect rapid changes in air temperature and only
reads the sensor about every 1/2s.

Only when the engine temperature crosses some temperature near to full operational
level will the O2 sensor be brought into the picture.  The ECU normally has a time
delay after start before it starts reading the O2 sensor, even when it is a hot start
condition.  This is because the initial mixture directly after a warm start needs to
be slightly richer for a short time to allow the engine to settle into a steady idle.
If the O2 sensor is not up to operating temperature the output voltage is below
the normal voltage and the ECU would interpret this as being too lean.  Only when
the ECU is absolutely certain the O2 sensor is giving a valid output will it be allowed
to have partial control of the fueling.  Some ECU's are cleverer than others.  Some
require the O2 voltage to run around stoich for a minimum time before allowing it to take
control.

So if you fitted, say, injectors 50% bigger than the ECU is mapped for the overall A/F
ratio will be 50% too rich and the O2 sensor correction is not expected be greater than
about 15%, you will end up with a rich mixture.  Why?  Because the ECU does not expect
such a magnitude of error and assumes the O2 sensor reading is out of range and simply
gives up trying to compensate.  If this happens sometimes the ECU will set an error code,
because the O2 reading is too high or too low for an extended period.  If this occurs the
ECU goes to "Plan B" and uses a limited fuel map to run the engine until the fault is
cleared, either by the O2 sensor voltage crossing the Stoich points so many times or the
manual restting of the error code.

So the only proper way to make bigger injector work is to re-map the ECU for the new
injector flow rates.  Not something you can easily do with the Bosch ECU's.  To the best
of my knowledge nobody, except Bosch, has the capability to re-map an L-Jet or Motronic
ECU, and plenty of people have tried!  You can buy "piggy-back" computers which take
some of the input signals from the engine sensors, modify them and then send these new
signals to the stock ECU.  This at best is a "gippo" and sometimes works and sometimes
doesn't.  A local company makes such a beast, however by their own admission they
haven't been able to make it work on the newer AR models.

An after-market ECU is a different story as the intention  is to allow the installer to
tune the ECU to suit the engine.  Even this is not a simple task and requires a fair bit
of time and commitment to get right.  At the least you will require an accurate A/F ratio
meter, preferably a wide-band type, and a means of measuring the torque and power.  I
normally use the Bosch 4 gas analyser and a chassis dyno to get the settings correct.
With experience this can be achieved in a half a day in most cases.  The Bosch gas
analyser is real slow!  It takes about 5s from making a change to see the results on the
display, this is because the tail pipe gasses have to travel along a long pipe to get to
the analyser input.  On the dyno I normally use a wide-band O2 sensor hooked into the
manifold down pipe which gives a faster response time.  There is a big interaction between
ignition timing and fueling which both effect the A/F ratio measured.  Run the ignition
too far retarded and the apparent A/F ratio goes up, because the mixture has insufficient
time to fully burn.  You have to juggle both the fuel and ignition settings to get the
best compomise of power/torque and emissions, not to mention keeping the engine out
of detonation under high loads.  A three handed job!

I have just completed such an exercise on a new engine type to be fitted in a series
production kit car to be sold in large numbers.  The emissions side takes a fair bit
longer to comply with the specifications.  I am sure Eric Storhok can eleborate on the
complexity of emissions, fortunately we don't have legislation in this country at present,
but this kit-car is destined for a foreign country where it is a major issue!  Running a
small cat-coverter, a Bosch O2 sensor and a fully adjustable ECU for fuel & ignition, I
managed to get figures of 100ppm for the HC, 0,3% for the CO, 15% for the CO2
and the O2 content was around 22%, all these are for idle.  This is acceptable
for the foreign market but can be tweaked if necessary.

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