Originally Posted by
hufnagel
sadly i didn't record things like vehicle/engine speed and throttle position that time..
given it's a honda, and I was forced to use EOBD for data, and i'm not familiar with how honda O2 sensor traces should look, I have to rely on the general theory of operations of most O2 sensors is that they're narrow band sensors, so they switch between lean and rich instead of reporting a smoother output curve (wideband.) since the green one is almost always showing some form of switching on its graph, I'm taking that one to be, for the moment, expected performance. you can see from the blue curve that, at random times it stops switching, and if you look at the first graph right near the end it goes COMPLETELY dead. it should also be noted that, the blue curve is for Bank 1 sensor 1, which is the bank that has the catalyst performance code being reported (P0420).
a note on Catalyst System Efficiency Below Threshold codes, specifically P0420 and P0430, for those who don't know.
those 2 codes come about generally come about because the downstream sensor (sensor 2) starts switching, or showing high levels of unburned hydrocarbons. the predominant cause of that would be a failed or failing catalytic converter, hence the the name of the code. HOWEVER, that also requires that the vehicle be in proper running order, and that includes the upstream sensor reporting correctly to the ECU about HC emissions right after combustion. If the upstream sensor starts performing badly, the downstream will start to see problems. Basically, a garbage in garbage out problem.
Since we have a vehicle here that popped a MIL, i'm going to assume the dealership read codes and came up with the P0420 as an active code at that time, and reported back "bad cats." HOWEVER, the light turned itself OFF after some additional driving, which would indicate the OBD saw expected performance from the downstream sensor, and assumed the fault was corrected, and put the code into a STORED state. This would indicate a still functioning converter, and leads me to believe we are indeed dealing with a GIGO condition. Further research indicates that it's apparently NOT uncommon for Honda 3.5L v6's from this era to need O2 sensor replacement around the mileage she currently has on the 'van.
typing all this, I just had a thought...
again not being a Honda guy, I'm having to rely on experience and general O2 performance specifications.
I know that in the Chrysler world I deal with predominantly, the downstream O2 sensor DOES contribute to overall fuel tuning, which is NOT common in some other worlds (GM and Ford from what I remember do not take downstream data and use it to tweak performance.) I experienced that first hand one time when the downstream on the Neon had damaged wiring, would occasionally wig out, and the computer would go bonkers leaning the engine out thinking "holy crap I'm pissing fuel out the exhaust!" and make it run so poorly it scared the crap out of me that the motor was broke.
I don't think that's the case here, as there was no experienced performance issues in the 'van during the test drive, but it's something I should look into I guess.