Bob Allen wrote:
>As a computer geek, I hate to admit that I know nothing of the computer
>devices needed to run the deal, though. It would seem that with two sensors,
>180 out, you could fire both injectors at the same time. This will waste one
>pulse but, hell, the (old, mechanical) fuel injected Corvette wasted 7 of
>them.
I'm on the other end of the spectrum. I'm a computer geek who knows
very little about how TCP/IP works, but if you want some on the fly
(heh, with my last project, literally) digital signal processing done...
Bob, you will be suprised I bet to find out that most EFI systems
"waste" pulses. For instance, both the SD1 and Hotwire injection
systems I've scrounged for my Rover projects fire one entire
bank of injectors. Each bank fires twice for the intake stroke of
a given cylinder in that bank. Only in the recent GEMS system
did Rover go to sequential (one shot per intake pulse) injection.
Most EFI systems out there right now are bank instead of
sequential. The pulse isn't really "wasted", we're talking about
nicely atomized air-fuel in the intake runner a few milliseconds
before it is "needed".
The big problem isn't *when* to fire the injectors, but, for
how long. This is where you get into air-mass sensors, TPS,
manifold temp, 02, etc. However, you can get away with a lot
of super clean emissions is not your top priority: the prototype
Jags (IMSA GTPs) of the early '80s ran injection based on two
sensors: throttle position, and 02. (some had a third, a pot
in the cockpit the driver could tweak, remember seeing those
Jags at the start of the endurance races with rather black
exhaust? tweaking it rich so the engine would last for the
event...)
|