KISS EFI

[Garfield Willis]

On Fri, 30 Jun 2000 22:29:45 -0700, garwillis at msn.com (Garfield Willis)
wrote:

>Sure, but there's "reliable cuz it's simple" and there's "reliable cuz
>it's failsoft/failsafe". In XA we're not much a fan of the "simpleton".
>It's just amazing how "un-simple" that "simple" system can be. :)  When
>your life's on the line, you want backup, not "simple" solutions.

Geez, I must be losin the paltry remnants of what's left of me mind!
Actually, we ARE all fans of what's simple, aren't we? FIRST and
foremost, you want "reliable cuz it's simple", THEN when it's really
critical, you ALSO want "reliable cuz it's failsoft/failsafe". Not good
to pit the two agin each other; it ain't either/or. And maybe another
point is, you don't wanna substitute redundancy for the additional
reliability that simplicity can give ya. But when all is saidNdone, and
the pucker meter pegs, there's nothing quite like a backup. That was my
point.

What I said about the sposedly "simple" system turning out not to be, is
true enough, but that dudnt negate the fact that there IS "simpler" vrs
"more complex". Didn't mean to sound like I was saying, "nothing's
simple, so all you can do to enhance reliability is make it redundant."
Twasn't me intent, tho it sure mighta sounded like it. 

>>BTW, I have this conceptual problem with redundancy and reliability.
>>Seems to me that the more redundancy is built in, the more inherently
>>unreliable the system will be. But that topic alone is a whole new
>>can-o-worms.
>
>Nah, no can-o-worms, just a matter of training and technical education.
>You heard o'the Space Shuttle haven't ya? Yeah, they got tripple
>redundant systems on-board. They did that cus they wanted the
>"inherently unreliable" systems you contemplate.
>
>Go fish, Peetie.

Ya know, I answered the way I did cuz I thot the intial retort was a
troll. Probly was. But in re-reading it, I realized that regardless of
where the comment is coming from, there's an altogether toooo common
misconception at work here.

Redundancy and simplicity have NOTHING to do with each other.
Reliability and simplicity DO. You can have a system that's just DIRT
simple, and then if you have TWO of them, you're still better off
(unless you do something really dumb). It's called "spare parts", eh?
There are two basic types of redundancy: passive and active. Passive is
where, when you believe you have a failure, you shut one down (or toss
one out, your choice of words) and swap another one in. "Spare parts",
sotaspeak. Active redundancy is where you have multiple "good ones"
running/working in parallel, and they all vote to throw the joker out if
someone runs amuck, and there's a disagreement about who's sane and
who's not. That's what Greg was referring to in the Shuttle.

What's all this gotta do with EFI? Wull, I find the complexity of modern
ECUs pretty overwhelming, but given all the smog requirements, I'm not
at all sure how to avoid it. OTOH, it's sure awful easy to miss the
forest for the trees. It IS really true what they say about "at some
level of complexity, technology is indistinguishable from magic". When
things get so friggin tangled that you have to be a member of the EFI
illuminati for years before it all comes together in your mind, geez
something's baroque (as in "broke").

I guess that's why I think KISS EFI also has a place even in the
automotive scheme of things, if for no other reason than to bring us
back to the roots of the thang. Probly would surprise alot of us to see
that you can actually implement a perfectly *driveable*  and
well-performing EFI with just a few chicklet parts. Well, it probably
wouldn't "theoretically" surprise us; we're all too smart and educated
for that, eh? But regardless of the General's Bogus Adventure contest,
it does seem like going back to basics does have it's attractions.

A lot get's lost in the translation of learning. For the sake of
understanding the basics and the subsequent frosting layers, anyone ever
disabled ALL transient (and static/temp-dependant) enrichments in their
ECU, and "felt" how the engine ran/vehicle performed? Criminee, ever
since the idle mix limits imposed on carbs when smog req's hit hard, a
person could grow up these days without knowing what rich stumble &
fouling or lean lope & misfire even FELT like! NOW, on the EFI cars,
there AIN'T even any idle mixture screw! Hee hee. Geez, there isn't even
any idle speed screw anymore! See my point? What's the world coming too?
:)

There just seems like there's gotta be something valuable, let alone
attractive for fun's sake, to starting out with some injectors, a
throttle body, and the worlds most primitive FI controller, and build up
the layers one by one. First you start with straight-line fueling MAP-n
say. NO transient enrichments, no TPS-based tip-ins. Do the same with
IGN/timing. Start with FIXED or manual advance (seriously; it used to be
that way in the old...really old...days). Measure AFRs out the tailpipe
to see how much variation you get over rpm & load (you knew there was
gonna be some reason you had to have an AFR meter, didn't ya? :) But
most of all, see how it FEELS. Do the same with Alpha-n fueling, using
just TPS and rpm alone; how many of us have ever driven an n-alpha
fuel-injected vehicle and know how they feel to drive? Let alone watched
what the mixture was doing (wow).

So much for the ramble. I guess I'm just feeling old and fossilish, and
wondering if we're not missing some valuable lessons by "growing up too
fast", EFI-wise.

Gar

----------------------------------------------------------------------------
To unsubscribe from diy_efi, send "unsubscribe diy_efi" (without the quotes)
in the body of a message (not the subject) to majordomo at lists.diy-efi.org