Larry,
The closed loop only comes into play at higher rpms
and usually then only at cruise. The rest of the time most systems
rely on the mapping of the parameters in the ECU to control
what's going on.
Yes...a wideband is VASTLY different. That's what I'm running
in my car. Yes... a D-I-Y wbo2 will run you ~$400, but you basically
end up with unlimited dyno time (anywhere you're driving on the road)
Ha! I too...spend 30+ years designing embedded applications. Mostly
robotics for factory automation and manufacturing. Both in the military
sector (contractor) in the manufacturing of passive fiver optic components.
The Megasquirt is apparently a rather functional ECU considering its
about $150-$250 to assemble. (google search...there's a Yahoo group
on the Magasquirt and it's offshoots)
Paul Tegler
ptegler@cablespeed.com
www.teglerizer.com
----- Original Message -----
From: "Larry Colen" <lrcar@red4est.com>
To: <mgs@autox.team.net>
Sent: Friday, April 30, 2004 2:17 PM
Subject: Re: Fuel injection
> On Fri, Apr 30, 2004 at 09:50:20AM -0700, Paul M. wrote:
>
> > Fuel injection is great (and very modern cars, with
> > their incredible digital systems that can adjust
> > engine management in real time) are really good, but I
> > still love the throttle feel of a car with carbs.
> > Push the pedal in a little bit, and a little bit
> > happens. Push the pedal in more, and more happens.
> > Somewhere along the line, that was lost with fuel
> > injection. It's finally starting to come back now (I
> > just drove a buddy's BMW M3 and it feels wonderful),
> > but it's taken a long, long time.
> >
> > The only "vintage" exceptions to this I have found are
> > some of the mechanically-injected cars like the Alfa
> > Romeo SPICA system and some of the
> > Kugelfischer-injected Porsches.
> >
> > But for my money, I still love me some carburettors!
>
> I've not driven any "vintage" fuelie cars.
>
> Granted, the whole point of my post was an excuse for a cheap
> pun. However, when I was trying to dial in the carb on the blower, I
> installed an O2 sensor and would watch that while driving. It was
> amazingly educational, watching what happened, and noting things like
> how much richness affected fuel consumption.
>
> Unfortunately, it wasn't until after I had installed it I learned
> about the difference between a standard and a wide-band sensor. If you
> want to learn about what it going on in your motor, I highly reccomend
> installing one and watching it as you drive. I suspect that you can
> save yourself a lot of dyno time dialing in your car by watching one.
>
> The other bit of instrumentation I want to add is EGT probes down the
> air injection ports, so I can get per-cylinder data.
>
> I had always labored under the misconception that SUs do not lean out
> when you step on the gas. Unfortunately that is not at all true, and
> when running on the edge of detonation or preignition, it can be
> critical. I don't know if this leaning out on "tip in" is exacerbated
> by the blower, but I ended up following Joe Cuerto's advice and
> plugging off one of the ports on the carb piston.
>
> Since embedded systems and computer controls are what I do
> professionally, I'd love to have someone pay me to design and build an
> engine management system. But failing that, I can't help but notice
> the (at least theoretical) advantages of closed-loop control systems.
>
> A lot of the old fuel injection systems were open-loop. They'd measure
> various parameters and take a guess at how much gas was needed. Modern
> FI, reads the output of the O2 or AFR sensor and adjusts mixture
> accordingly.
|