Alternative fuel (was Re: EV's vs HEV's)
Russ Cage
>e...@michael.apple.com (E. Michael Smith) writes:
>Closed loop didn't come into common use until the mid 80s and even
>then it was crap, usually implemented with some jerry-rigged
>electronic carburator.
Just FYI, in 1982 the Chevette/T-1000 had an electronic carburetor
which used a solenoid to vary the mixture by about 10%. Under most
conditions it could maintain stoich or close. When the charcoal
cannister was heavily loaded and the purge turned on, the solenoid
duty cycle would often be pegged at the lean limit...
The GM V8's had TBI that year, I think.
>You're proposing to set up a whole industry in order to use your
>elixir. It is VASTLY easier just to retrofit a closed loop adaptor.
That depends, John. Think for a minute about the rich-lean distribution
problems with what amounts to a big vacuum leak, plus the inherent
driveability problems associated with keeping airflow up during a
back-out prior to a shift (else you get a lot of fuel coming off the
manifold walls with nothing to burn it). Doing it right may not be
easy no matter which way you go. Instead of doing it right, you can
do it fast, which usually takes more time and turns out to be half-fast. ;-)
>I don't know what Russ has to say but I DO know what GM, at least, does
>because I have a system description and source code listing for their
>previous generation ECU (C3?) GM controls closed loop after warmup
>and only during part throttle and idle operation.
Yes, but you haven't said what the open-loop mixture is set by. Is
it hard-coded or does it use a learned-mixture multiplier? Injectors
vary from one unit to the next and change with age. Learning just
how much fuel a millisecond pulse gives is essential to smooth
operation of the system. You can't learn while the lambda sensor
is cold, but you can use historical data to set your mixture. Ford's
EEC's have done that for years. If the system control limits have not
been set too narrow to adjust to ether fuel, it should work just fine
if you can get it started the first time.
(This does presuppose some algorithm smarts. The algorithm has to vary
the mix rich before it will see the O2 sensor voltage switch; if it
runs perpetually lean, it will never switch "on". If it will vary it
far enough to go rich with ether fuel, it will work fine, otherwise it
will give up and declare the O2 sensor bad.)
---
Russ Cage | Forewarned is half an octopus
ru...@m-net.ann-arbor.mi.us
(313) 662-9259 | Software engineer for hire, real-time a specialty.