Fuel injection questions - -

[Charles S. Tosch]

>	Message 12:
>	From owner-diy_efi-outgoing at coulomb.eng.ohio-state.edu Thu Oct  5 11:14:10 1995
>	From: walter at roadster.sps.mot.com (thomas walter x5955)
>	To: DIY_EFI at coulomb.eng.ohio-state.edu
>	Subject: Re: Fuel injection questions - -
>	
>	> From: Andre Roy <ANDRE at einstein.unipissing.ca>
>	
>	> MPI ( Multi Port Injection) has one (or more) injector for each cylinder
>	>     The injectors fire in unison. The disadvantage is that there can be
>	>     a bit of fuel puddling (like on a carb) because air only flows while
>	>     the intake valv is open. The rest of the (admittadly short) time, the
>	>     air in the intake runner is not moving.
>	> 
>	> SPFI Sequential Point Fiel Injection (there are other acronyms, no doubt)
>	>    Same as above, but the injector for a particular cylinder only fires
>	>    when the intake valve is open (or just before) the rest of the time
>	>    there is just stationary air with no fuel in the intake runner.
>	>    Advantage, no fuel puddling. Disadvantage, slightly more complicated
>	>    control strategies.
>	> 
>	
>	Andre,
>	
>	Minor item. Ideally SPFI will turn off the fuel flow BEFORE the valve
>	closes. Keep in mind the intake valve is open, maybe 35%, of the time
>	where as the fuel injector will have a maximum duty cycle of 80-90%
>	at WOT. i.e. it will be INJECTING fuel longer than the valve can be
>	open.  [Ok, Ok, doesn't apply to a RX-7! ;-)].
>	
>	Still ideal to turn off fuel flow before the intake valve closes, and
>	then to turn it back on (assuming 90% duty cycle) for the fuel to hit
>	the back of the valve allowing better vaporization.
>	
>	So if you are after maximum part thottle economy, and low emissions,
>	SPFI is best. For maximum power, either approach will do!
>	
>	REVERSIONS: While we all realize the the closing of the intake valve
>	will cause a "back pulse" of air through the intake track, I was not
>	really aware of how critical it could be.  An engine builder could not
>	get enough fuel into an injected engine... at WOT the engine was running
>	LEAN, despite "having more than enough fuel to flood it". Hmm, problem
>	was a unique reversion pattern. Rather than the usual area around the
>	intake walls, there was ALSO an effected area in the middle. Dead center
>	of the intake area of the FUEL injector jet!!! The reversion pulse of this
>	radically cammed motor was enough to blow fuel back out the short intake 
>	track! [Haltech throttle bodies... look like Weber DCOE's with an injector 
>	port]. Result was interesting .  So when done with all the calculations, 
>	and theories, it is great fun to get it all blown back in your face! ;-)
>	
>	Andre has a good point on fuel puddling.
>	For those of you adapting "thottle bodies" to your home brew
>	system. While the phsical hardware may be easier to bolt up, the
>	software will have to correct for "pooling" the fuel in the manifold
>	under some conditions... a little more difficult task for home brewed
>	systems.  Hmm, seems like the port injection would be a simplier SOFTWARE
>	task to start with (even with the system allowing for SPFI control of
>	the injectors, just do rev 1.0 software for batch mode... then tweak it).
>	
>	Odd thought... with a SENSITIVE air pressure switch in the intake
>	manifold, I wonder if it would be possible to detect the closing of
>	the intake valve... i.e. for sequential timing WITHOUT using a camshaft
>	sensor. Any comments?
>	
>	Tom "How could the motor be too lean?" Walter
>	Austin, TX.
>	

The only problem is that it is difficult to distinguish the pressure pulses
from the valve you are trying to synchronize with. A standard MAP sensor
can detect the pressure pulses in the manifold. Typically they are filtered
by a calibrated orifice in the sensor or software. A length of vacuum
hose before the sensor also has the same effect.

FWIW

-- 
Chuck Tosch                     	c1icst1 at kocvrs01.delcoelect.com
Delco Electronics			(317) 451-0869
Kokomo, IN 46904	     		8-322-0869