C/L WB/WOT OEM ECUs...

[Garfield Willis]

On Sat, 8 Apr 2000 10:33:40 -0400, "nacelp" <nacelp at bright.net> wrote:

>All I see there is Accleration Enrichment.  I may be spacing it, but just
>done't see it

Right, that's what I was referring to. Are you saying those don't imply
O/L?

>You are clear (?) that you'll have to have the EGOR output voltages for O2v
>swing over, and under the O2 cross over reference settings in the ecm.

Yes, simulating the switch-type O2 sensor gives you the same phenom as
before; the controller criss-crosses that simulated stoich point. There
are lots of details to determine on the bench/road, tho. How tightly
this "fooled" controller will keep to within the set-point (the AFR you
have set as the new "stoich") is one question. When these controllers
hunt around real stoich, they overshoot some on either side. This slop
has to be taken into account when tuning this type of "fooled"
controller.

>Note 2:
>We know the ecm needs to shoot for a 14.7 for the EPA.  So we also know what
>14.7 is for EGOR. Then using those numbers we can set the kluge to generate
>a over/under O2 signal for the ecm to get that.  Meaning right output based
>on EGORs voltage.  This is what you've already mentioned, just a converter
>to run a oem ecm, with EGOR.  The big advantage I see here is that for none
>cat applications, can trash the 14.7 and go for min fuel consumption.  With
>that info., generate a DC based on the under/over voltages the ecm will
>like, and then it will do the AFR calc..  That is the trick for EGOR as I
>see things,

Yup, as you suggest above, you could also use the same scheme intended
to fix a nice well-controlled rich mix at WOT, to also home in on some
lean cruise AFR when lightly loaded. You can imagine a spectrum of AFR
setpoint repertoire, depending on how complex you wanna get:

(1) Go to "C/L precision rich" above a certain TPS, otherwise keep to
real stoich. [Unless of course you go to O/L during all the
aforementioned "special" conditions like cold-start/warmup, sensor
failure, etc.]

(2) Same as above, but replace "real stoich" with some other economy
setpoint, hopefully not too lean. :)

(3) Get crazy and "map" the TPS to a table of target AFRs. Now you have
as fine a control over the AFR as you have table space. You don't even
have to go digital here; you could make a simple piecewise linear curve
with a few op-amps, instead of table look-up, to create a rough curve of
TPS vrs AFR that the simulated O2 should follow. OTOH, this might be the
point of departure for including a proc into your O2 simulator circuit;
depends on how much you love/hate analog. :)

(4) Get crazier yet and tap into the MAP signal, 2-D map both TPS and
MAP to a table of commanded AFRs. Now you have C/L AFRs that depend both
on throt and load. Woohee.

Now you're starting to put nearly as much smarts in that lil goiter
that's fooling with the simulated stoich point, as you have in the OEM
controller. So you're approaching the point at which you might as well
just bite the bullet and create a custom controller, and dump the OEM
brain-stem? Well, maybe not quite, but you getmedrift. Depends on how
much you wanna play with the electrics.

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