As I mentioned before search does not work well on my system. So, if there is a definitive thread somewhere directions are welcome. If somebody knows Brian maybe they can get him in here to yell at me I'm going to post my understanding. This may, or likely may not be correct. I can not find the specific info I'm looking to explore and quite frankly it may not exist for good reason. I am a tinkerer and very driven to conquer mechanical things. In case you are wondering why I care. So, Modena, long tubes factory cats. CEL. Expected of course. Here is where we get sticky. Upstream O2s are sampling, egads O2! and providing feedback to trim fueling. Downstream narrowbands are reading O2 and used to determine CAT function. So, the addition of longtubes throws an efficiency code but why? Before we talk about how the ECU looks at downstream sensors, let me talk a bit on the output itself. So looking at the output after the longtube is installed, the downstream sensor is reading all over the place, from .06v to .900v. Clearly that low end is an issue. (Previously functioning fine). 02, precat car. The longtube without cats is surely providing less oxygen content than the stock, no? I mean we went from lots of scrubbing to no scrubbing. Before I get yelled at, I say scrubbing but realize there is a chemical reaction happening that is oxidizing, for lack of the proper terms, monoxide, and nitrogens into dioxides effectively adding oxygen and making the government and tree huggers happy. So why does this freak out the down stream sensor? My answer would have to be that we have overtaxed the stock cat to a point where it cannot get the exhaust to the previous level. I cant think of another way although I am hard pressed to believe this. My understanding of the oxidation process is weak and perhaps it is designed to reduce a certain percentage thus increasing one end has the same effect on the other. But in the end, it seems the cat output is no longer in range thus the efficiency codes. Wouldn't you think however, that whatever you threw at the cat, it would just make it right? So, the downstream sensor is supposed to see "x" does not see it and outputs a voltage representing that which angers Enzo who lives in the ECUs and shouts at me. But, is that all? Here is where I need more. Does the ECU compare the up and down stream sensors and how? Does it look for similar switching rates, or does it just look to see downstream is switching? If the latter, does it look for a specific freq? Is the ECU looking for specific changes between the two sensors, ie. if up stream is "x" downstream needs to be "x-something" and they need to correlate. My terrific hope is the ECU is simply looking for a non-constant voltage within a specific range, which I hope is .4-.9v But, with all the stories of people that cant get the CEL to go away I suspect there is more to it. Please don't say put the stock exhaust on. This is a challenge for my head that I really want to solve and understand. If it does turn out that Enzo the ECU is really looking for all the worse case scenario things with regard to downstream sensors, one would think a trimming circuit would almost work. Provide reference voltage and allow a regulator to take the input signal from the down stream sensor. If its below the lower limit it adds to bring it to the lower, where you set the lower some degree higher than the true low. That way a consistent low output would not show as a flat low limit output. Same thing for upper. In this manner the sensor would switch in sync with the upstream barring any delay in the circuit and you would always be producing a value in the intended range. But, what is the dang range really? If anyone at all has any insight into this I'd love it. I have a few more tricks I'm working on because I really want my CEL to come on if there is another problem. And, before anyone asks, yes the inefficiency is solely related to the addition of longtubes. Everything else is and was good to go.
Here is an image of a recent scan. With a mini cat the signal is much better, but I'm not sure its good enough. I have not driven far enough to get into ready state and throw a code. You can see the signal drops on a WOT punch and it stays there for a decent amount of time. It does recover with the mini cat, but I am guessing its not going to be fast enough, or the drop is too severe. The in-minicatted bank is low at idle. You can see its erratic but it does respond to a lean situation. I see a reference equation between the sensors listed in % but don't know if that is scanner driven or the ECU is looking at it. Having said that it seems that with the mini cat, I could trim the lower limit of the signal to bring it closer to average but I cant do much for recovery. I guess I will wait and see if that bank throws the efficiency code and go from there. Kinda neat to see what is going on. Image Unavailable, Please Login
So,not to confuse anyone I need to clarify my statement earlier. Low voltage output in the narrow band sensor is a lean condition not rich. So now nothing makes sense.
