John

[John Appel]

I have  a question which is not directly Link related but someone may know the answer. In 1999 Harley Davidson adopted fuel injection for their V twin engines. For one year only they had a cam position sensor. After that there was no cam sensor. How did they run their injection system on a V twin without a cam sensor.

[Adamw]

The more recent Harley's that have a Delphi ecm synchronize using crankshaft acceleration to identify compression stroke. I'm not sure however if it is used only during cranking/low speed or full time.  The earlier Harley's that had a Marelli ecu I think didnt have any sync strategy and just ran batch fire/wasted spark.

[John Appel]

If the engine runs with batch fire each injector squirts once  every engine revolution which means twice per engine cycle (720 degrees). In the equivelant sequential system the injector squirts only once per engine cycle. This means the pulse width in batch fire would need to be half as long as in sequential to deliver the same amount of fuel per engine cycle. At idle the pulse width even in sequential would be very short if the injectors are sized to deliver sufficient fuel at full throttle with  a duty cycle of about 80%. In batch fire the pulse width would need to be half this time which would be getting pretty short. What would be the minimum pulse width in milliseconds which would still give accurate control of the fuel delivered.

[John Appel]

If the engine runs with batch fire each injector squirts once  every engine revolution which means twice per engine cycle (720 degrees). In the equivelant sequential system the injector squirts only once per engine cycle. This means the pulse width in batch fire would need to be half as long as in sequential to deliver the same amount of fuel per engine cycle. At idle the pulse width even in sequential would be very short if the injectors are sized to deliver sufficient fuel at full throttle with  a duty cycle of about 80%. In batch fire the pulse width would need to be half this time which would be getting pretty short. What would be the minimum pulse width in milliseconds which would still give accurate control of the fuel delivered.

[Adamw]

Typical minimum pulse width for most injectors is often considered around 2ms to keep them in the linear response region although you can run some as low as 1ms with the correct injector characterisation data and an ECU that can use that data.

I don't think minimum pulsewidth at idle will be too much of a problem in something like a Harley that only makes <50HP per injector.  I'm guessing they would have something like only 300cc/min injectors so idle PW will be >2ms in batch fire mode.