Affordable Wideband ECUs - How to?

[Garfield Willis]

On Sat, 1 Apr 2000 03:00:41 +0200, Jörgen Karlsson <jurg at pp.sbbs.se>
wrote:

<intriguing stuff on FelPro controller snipped...gonna have a look soon>

>Our sensor and fp box is matched to each other, the sensor is marked with
>the serial# of the box. The first box we got was defective (or as I see it
>they sent the wrong box) and the box and sensor had to be returned. This can
>of course be a mistake from the guys that sold the stuff... Scott's
>observations strongly suggest that the sensor has a calibration resistor on
>its connector.

Yes, it sounds like they've left the cal resistor there, but they AREN'T
bound to look at it electrically, and could easily have a mfg. step
where they burn that info into the controller board manually, so you
can't change the sensor without their help. This would be a sad
manuever, but totally possible. Other AFR meter mfgs. burn both the cal
and the transfer curve into EPROM, so you also are stuck with sending
the whole thing back to them if you need a replacement sensor. You can
argue with the business ethics of this, but the thing that burns me
about "property" techniques like this the most is technical: it prevents
people from swapping out the sensor vrs. controller to isolate a
suspected failure to either controller or sensor. NTK is the same
frigging way; you have a failure and want to send it back to them for
testing to tell you just WHAT has failed, sensor or interface?, and they
want to charge you, get this, "engineering bench time", which they of
course point out is going to be WAY too spendy for you, so they insist
you'd be better of to buy a whole new replacement box and sensor instead
(deduct $1000 from you bank account; do not pass GO). Cute, huh?

>I have noted that when the number in the VE table is to far from the correct
>number the o2 sometimes tend to oscillate. I think this happens when the
>value is more then 15-20% off and I think that it was only when the mixture
>was very rich (turbo rich). But this can of course be something about the
>car, syclones are pretty strange...

Curious, but somehow I doubt it's engine-specific. Sounds like
controller instability, perhaps from poorly set/tuned controller
parameters or phase lag from distance of O2 bung to exhaust ports. Just
wild guesses, really. But delays in a feedback situation can really snag
ya.

Thanks for the additional insights.

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