By Patrick Zambori, H-D tech at Wisconsin H-D
Webmaster comment: I beleive this to describe
the older Magneti-Morelli EFI system, not the new Delco system.
Having been through two EFI schools at harley, here's my take on the
EFI on the Harley, based on what I've learned from the motor company.
The oxygen content of the air is determined using the intake air temp [IAT], and
the barometric pressure [MAP]. Factor in the throttle position [TPS] and the computer can
roughly estimate total oxygen flow.
The crank position sensor [CKP] frequency establishes RPM and has one missing gap
in the wheel that times the spark and injection pulse. Which cylinder gets the
spark and which gets the shot of fuel is determined by the state of the cam
position sensor, either on or off. It's a hall effect sensor.
The cam sensor has no impact on the actual timing.
The fuel map that determines how long the injector is held open is a chart of total oxygen flow vs. RPM.
The ignition timing map is a chart of total oxygen flow vs. rpm
The engine temperature [ET] sensor only has one purpose. Managing the warm up sequence. Below operating temperature, the computer uses it's resistance to
determine how much fuel and how many steps to push the throttle open past the base idle. I expect there is a head temp vs throttle position map to
control injector pulsewidth.
Once the engine reaches operating temperature (285F), the head temp sensor reaches the limit of it's resistance band and only indicates that the
engine is warmed up. The idle speed control motor [ISCM] backs off and the idle
speed is controlled by the adjustment screw.
Idle speed control is odd, it doesn't use any RPM input, it only knows # of steps to position the throttle. During the shutdown sequence, after
turning the ignition off, the ECM turns the ISCM back for a length of time, theoretically long enough for it to reach the end of it's travel, and
let it's internal clutch slip for a few revolutions, then the ECM turns the motor back out a specific number of steps, it's the same every
time, regardless of temperature. When the ignition is re-energized, the ECM
reads cylinder head temperature and repositions the ISCM whatever number of steps in whatever direction is necessary for the throttle position that
is predetermined for that head temperature. Just # of steps whichever way.
If the ISCM clutch slips and it does not return to zero during the shutdown sequence, then the throttle will be open further than it actually is, and the
fast idle RPM will be higher, sometimes -much- higher than normal.
The bank angle sensor has only one job, if the bike falls over, it shuts
the engine off.
The fuel pressure is a fixed 48 psi give or take a few.
The ECM handles ignition timing differently when it's idling than when
it's not. It -remembers- the throttle position voltage at base idle, then
whenever the throttle is in that position, it recognizes that it's at idle,
and optimizes timing for a smooth idle.
I've noticed that after adjusting the throttle position sensor, if the
voltage won't drop low enough, the idle speed cannot be stabilized at
1000 rpm, it will jump back and forth between 900 and 1100 rpm. The only
solution I've found is to pull the ECM fuses or disconnect the battery
overnight. The ECM will forget the idle value, and next time you drive the
bike, it will re-learn the new one.