Closed loop Lambda (AFR)

[Rampant]

Hi guys

I'm toying with the idea of running a closed loop AFR fuelling system with the help of my existing wideband O2 system - an Innovate MTX-L.  I've had the wideband installed and running now for over 2 years, just as an engine health monitor.  

The thing that is making me hesitate is the thought of what might happen if the O2 sensor failed or began to lose it's calibration.  The sensor is mounted in the de-cat pipe on an Evo 9 (in a horizontal position in the side of the entrance end of the de-cat).

Are there safety strategies that can be programmed to prevent a failed O2 sensor leading ultimately to a engine kaboom death event?

Cheerz

Mark H

[Adturb]

Couple of plans of attack here- I scale mine inversely (0v = lean/ 22AFR, 5V = rich/ 8AFR) so if the MXTL dies then it'll overfuel rather than run lean. The ECU can be set to work on a narrow range, as most sensors do- 0.5 to 4.5V, either side of that (0 or 5V) will register as a sensor fault- I tried this but kept getting false ECU alarms, I don't think there is a time delay on the alarm so it activated each time the car started. Thirdly when you set up closed loop it is only active under certain conditions, typically cruising/ steady throttle. Furthermore there is trim limit, so it can't or subtract more fuel than that setting- I think I set mine to 8%. 

 

[Rampant]

Thanks Ads

So from what you're saying about selective active conditions - can it control a fuel target AFR during WOT acceleration?

Cheerz

[Simon]

Hi Rampant

It can do but it is not intended to be used this way. Its intended purpose is to keep the fueling as close as possible under cruise conditions. And in situations where emissions might be tested.

[Rampant]

Thanks for the response - in that case it is less of a concern on sensor failure. 

Cheerz

Mark H