O2 Sensors

[peter paul fenske]

>HI ALL
>Well here is my two bits about oxygen sensors.
>First off no matter what temp and what range of swing stoic
>14.7 to 1 occurs at the same cross point 0.5 volts. Do what
>the oems do, make sure that there is enough deviation from this
>about 300-700 mV before you accept 02 data. Then merely use the
>crossing to count your integrator up or down. Use the difference
>in up down counts to adjust your learn adaptive value for the
>cell you are operating in. You are quite correct. When oems
>set up their systems, they do it in open loop. Each load cell
>along with all the attendant corrections, CTS,MAT, Battery ect
>is set to the engine at stoic. Then the system is entered into
>closed loop. Adjustments of the learn cell should deviate only
>slightly from 128 or 1:1.
>
>And yes you can operate away from 14.7 to 1 a bit. Center
>your limit cycle oscillation at a different setpoint by using
>different integrator counts for up and down. GMs quad four operates
>at 14.6 to 1 by doing this. Mind you only a tiny fraction of lambda.
> 
>As for handling the interface use one of the National Interface
>chips. Then all you need to worry about is the output.
>Electrically heated sensors are nice but cost about twice as much.
>I only find them necessary when using long tube headers where 
>closed loop may drop out during extended idle.
>Yah you are probably right. The O2 is a bit much for DIY-EFI
>since it merely makes oems pass 50 thou emissions.
>GL all: peter