Post by @TechConnectify@mas.to

2025-10-02 20:30 UTC

@kwahoo Respectfully, I think you're confusing the algorithms necessary for narrowband sensors with those in use with a wideband sensor. Your original premise is that the engine should not need to oscillate between rich and lean based on the graph you have found. But that graph does not address the underlying chemistry of the reactions which take place in the catalyst. Narrowband feedback systems produced the necessary oscillations basically as a side-effect of how they worked.

Replies (2)

@TechConnectify@mas.to 2025-10-02 20:35

@kwahoo If I had a car with a narrowband upstream sensor, I would have written how even an analog control circuit could enrich fuel when O2 sensor signal shows lean, and once it was enriched enough for the O2 sensor to switch, it could pull fuel back. It would be attempting to stay at 1.0 but hysteresis and lag would cause it to oscillate. A wideband sensor, though it still has a Nernst cell at its heart, is precise enough for the computer to be proactive rather than reactive.

Open ##1407947
@kwahoo@101010.pl 2025-10-02 22:40

@TechConnectify Ok, after digging a bit I see your point about underlying chemistry. Definitely lambda ratio oscillations have an impact on a catalyst converter efficiency https://www.sciencedirect.com/science/article/pii/S0926337323013000 Both amplitude and frequency matters. This also explains why the frequency is similar in all engines. And the frequency is quite low, much lower than lambda sensor and ECU would theoretically allow - this indicates that ECU deliberately acts this way. Even with two state sensor we can call this a proactive action. The downside of the two state sensor is that, the lambda ratio can drift during low or high signal state period. As for the graph, it should have a disclaimer like "for average lambda in X period of time". I've learned something new here;)

Open ##1407950