Long post, hopefully it will help, copped from buick board....
The inj duty cycle is the time the ecm is trying to energize the injector coil divided by the time between intake valve events. This is going to get long, but let's go through everything that happens when an inj fires. When the ecm tells an inj to fire, first a couple of transistors have to turn on (very fast) so the voltage across the inj coil goes from zero to battery/alternator voltage, then the current begins to rise in the inj coil. The coil is an inductor and so time is required for the current to build up to it's max value. For a high impedance inj of say, 16 ohms, and a system voltage of 13.8 volts, the final current will be about (13.8-.2)/16=0.85 amps. The .2 V is a guess at the saturation voltage in the switching transistor. The coil inductance is about 2 mH (guessing; the Cherry Semi data sheet for their peaknhold drivers uses 4.0 mH for a 2.4 ohm low impedance injector) so the time constant t=L/R=2 mH/16 ohm=.125 msec, and the current will be at 95% of final at 3t=.375 msec. I don't know how much current flow is required to begin opening the injector but the pintle/disc/whatever has some inertia because of it's weight so there will be an extra time delay for this. In all, the stock GN injectors probably need on the order of 1-2 msec to reach full flow (and 72's more like 2-4 msec based on my observations making chips). So, the ecm says open, and 1-2 msec later the inj is at full flow. After the desired inj pw has passed the ecm says close and this all reverses. The current takes the same time to decay, at some fraction of full current the inj begins to close, and at some time later the flow stops. If the opening delays equal the closing delays then the length of time fuel actually flows will equal the pw commanded by the ecm and the only effect is a slight delay between ecm start of pw and delivery of fuel. However, the turn off delays are probably longer than the turn on delays and so fuel flows for longer than the ecm expected (my guess is that this is a msec or so, and this post would be really long if I tried to cover this too ). Anyway, that is all for a single injection event, with lots of time before and after so that the inj is completely closed from the previous shot. Now, onwards to more pulses. If the ecm fires an injector, and then fires it again before the injector can close, it will simply stay open and continuously hose fuel into the engine. Note that the ecm simply has to fire it again before flow stops; remember that turn off delay? If the ecm pulses the inj while the current is decaying away but before the pintle has started to move, the inj will stay open. If the pintle has started moving it will still require much less than the normal turn on time to get it back to full open and there will only be a slight, brief drop in fuel flow. Since the turn off delay is longer than the turn on delay (if you trust my earlier comments), then once the ecm has fired an injector it can actually stop firing the injector for a short while and then fire it again and the fuel flow will never stop. To put some numbers on this as an example, what if the ecm fires an inj for 5 msec every 30 msec? The ecm fires the inj at time zero, at time 2 msec (say) fuel flow is at maximum, at time 5 msec the ecm stops firing the inj, and at time 8 msec the fuel flow has stopped. The ecm fired the inj from time 0 to 5 msec, fuel flowed from time 2 to 8 msec so 6 msec of fuel actually flowed when the ecm only expected 5 msec worth, and then at time 30 msec the ecm fires the inj again. Now, remember the definition of duty cycle? Here, the dc as far as the ecm is concerned is 5 msec/30 msec - 5 msec of commanded flow divided by 30 msec between firings, or 16.7%. The actual fuel flow dc is 6 msec/30 msec or 20%. Okay, but what if the ecm fires the inj 29 msec every 30 msec? The ecm fires at time 0, at time 2 msec fuel flow is max, the ecm stops firing at time 29 msec, and then fires again at time 30 msec even though the fuel flow from the first firing won't have stopped until time 32 msec. The inj simply stays open. The commanded dc is 29 msec/30 msec=96.7% but the fuel flow never stops - then inj is static.
Now, on the engine in sequential mode the time between inj firings (for any one cylinder) is equal to two revolutions of the engine, which is equal to 120,000/rpm msec (24 msec at 5000 rpm). This is also the time between intake valve openings, the time between exhaust valve openings, etc. If the engine is at 5000 rpm and the ecm commands an inj pw of 10 msec then the duty cycle will be 10 msec/24 msec=41.7%, and if the ecm commands an inj pw of 24 msec than the dc will be 24 msec/24 msec=100%. Like I said above, if the ecm commands a pw of 30 msec, for a calculated dc of 30 msec/24 msec=125%, the inj will simply stay open and hose fuel. No additional fuel can go into the cylinder in any given intake event over and above what would have flowed at 100% dc because in both cases the inj was continuously flowing fuel the entire time between consecutive intake valve openings. Commanding 125% doesn't give any kind of a cushion over 100% dc. And, because of the turn off delay and depending on the injector and battery voltage, somewhere around 95-99% dc commanded by the ecm will actually result in continuous fuel delivery.
Now, batch fire. This is where the ecm fires the injectors once every engine revolution, or twice for every two revolutions, or twice for every period of time beteen intake valve openings on successive firings of the cylinder. The ecm uses half the inj pw it would in sequential mode so in theory if the turn on and turn off delays were the same and the dc well below 90% the amount of fuel will not change. At low pw's the entire fuel flow may be while the intake valve is open but near 100% dc the intake valve is closed for most of the injector event (for batch or sequential mode - two revolutions is 720 degrees of crank rotation while the intake valve duration is about 200-220 deg at .050 lift so about 2/3 of the time the injector is spraying fuel the intake valve is closed. Back to batch mode - at high dc's there will be two turn on and two turn off delays between intake valve openings and so a smaller commanded pw by the ecm will be required to get the inj to actually go static and just stay open, so if the chip is only commanding say 90% dc then in batch mode this may be just enough while in sequential mode it may take 95% or more and so a little more fuel will be delivered in batch mode. Note that this is only going to happen when the chip is commanding almost 100% dc - at lower dc's or dc's greater than something like 95-99% there should be no difference.
I know this was long and kind of wandered, but I hope I have convinced at least one person somewhere that commanding inj dc's over 100% does not result in any additional fuel flow. It may sound nice, like the amplifier knobs on Spinal Tap that go to 11 instead of 10, but all the way up is all the way up .
Oh, I completely agree with Jim that if you are using dc's of over 95% and (Jim didn't say this part) fuel pressures over maybe 47-50 psi you really need bigger injectors.
__________________
Regards,
Carl Ijames
ijames@netaxs.com
_________________
Scott
---------------------------------------------
As the light changed from red to green to yellow and back to red again, I sat there thinking about life. Was it nothing more than a bunch of honking and yelling? Sometimes it seemed that way.
<font size=-1>[ This Message was edited by: kwick6 on 2002-01-30 19:50 ]</font>