O2 sensor output

[Tom Cloud]

Alex Nicu wrote:

>> Ok , I'll use the average between high and low , but , I have to set the 
>> perfect ( stoich ) point . The EFI unit will have to take a decision 
>> according with this point . 
>> It's one page on web where I read that this point for normal sensor it's 
>> arround 0.45V . OK , but yesterday on the list (#313) I read 
>> that the stoich point it's arround 0.6V . What's the real one ?
>> Alex

 and Seth replied:
 
>Alex- the curve is so steep at lambda=~1 that any value from 0.35to 0.75v 
>might be considered stoich, depending on your definition.
>
>-------------\       0.9v
>	      |
>	      |
>	      |
>	      \
>	       -------------   0.1v
>
>	      ^
>	      |
>	   the sacred lambda =1
>
>appologies for the ascii art
>
>Seth Allen


and then, I (oh sage one)  8^)  interjected:


below is graph for the EGO sensor (joltage on y-axis and A:F on x)
as I copied it, as best as I could (note, need to use a Text
editor or a non-proportional font to see this)


1000 ___|_______|_______|_______|_______|__
        |       |       |       |       |
        *   .   |       |       |       |
        |       *       |       |       |
 800 ___|_______|___.___|_______|_______|__
        |       |       |       |       |
        |       |     * |       |       |
        |       |       *       |       |
 600 ___|_______|_______|_._____|_______|__
        |       |       |       |       |
        |       |       |       |       |
        |       |       |       |       |
 400 ___|_______|_______|____.__|_______|__
        |       |       |       |       |
        |       |       |       |       |
        |       |       |       |       |
 200 ___|_______|_______|______.|_______|__
        |       |       |       *       |
        |       |       |       | .     |
        |       |       |       |    *  .
   0 ___|_______|_______|_______|_______|__
      14.4    14.5    14.7    14.8    15.0
  (richer)                             (leaner)

The X-axis nos aren't linear -- that's the way the graph was.

I interspersed "*" and "." to try to get the slopes on the
curve correct.  The curve flattens out below 14.4 and above
15.0 A:F .... the max output being ~ 1 volt and the least
theoretically being 0 (but not attainable in operation).
Also note that the "flat" parts beyond 14.4 and 15.0 vary
with temperature of the sensor -- and you're already working
with millivolts difference for each A:F point -- any kind
of accuracy is impossible in those regions !!

This really substantiates Seth's statement (about anything
between .35 and .7 jolts being "stoich" for all practical
porpoises)  ......  this is one reason the cute little LED
meters from MSD and Edelbrock, et al are just that -- "cute"

I use a $15 DPM from Hosfelt Electronics to read mine -- it
gives an actual joltage reading -- I've never expected it to
give me any qualitative data (i.e. no actual A:F values) but
rather an indication of what's going on -- lean or rich ....
REEEL lean or rich  ;-)

Also, Alex, I don't recall what you said your app was, but
if your EGO isn't reading anything else, a simple lo-pass filter
(capacitor across input to meter) should give you an average
of the excursions

Tom Cloud

    Just because you're paranoid doesn't mean they're NOT out to get you !!