EGO Related

[garfield at pilgrimhouse.com]

On Wed, 10 Jun 1998 16:57:00 +1000, Wayne.MacDonald at zurich.com.au wrote:

> The approach I am taking is to put a 68HC11 based microcontroller between
>the ECU and the injectors measuring the injector pulse from the ECU and
>providing a new pulse for the injector using a compensation % from an
>RPM/Throttle position lookup table. I have also put an EGO sensor in the
>exhaust collector and planned to use the input from this to dynamically
>adjust the lookup table. Am I barking up the wrong tree trying to do this
>with a normal 4 wire EGO sensor ?.

Sounds like the right tree to me! It's not really a matter of whether or
not you can use a 2/4-wire sensor or not, but HOW you plan to use it.
You would surely be able to use that conventional sensor in the normal
bang-bang controller style used in most auto apps, where you would
program your micro to be constantly tweaking the mix so you're
continually crossing back and forth over stoich.

The difficulty comes if you're hoping to use those sensors as an "analog
measurement device" that your micro can just "read" the current mixture
from. The conventional sensors aren't really cut out for such a job.
But, hey, nobody uses them that way in an engine controller anyway, so
it's a "so what".

>I assume from a recent posting on EGOR
>that the Honda/NTK 5 wire sensor is quite different to a normal sensor ?.

Yup, their internal physics/chemistry is very similar, but they use a
compound device structure that required quite different circuitry to
read the mixture. BUT, that added complexity is accepted in return for
the device now being capable of a true measurement of mix.

>Will I get accurate enough information from a normal EGO sensor to do this
>type of calculation ?.

As per the above, depends on how you wanna use it. A straight
calculation, nope; but an adaptive controller program loop that "hunts"
around stoich, yup.

Gar