replied to the thread DP Fix Bypass Switch
" [QUOTE=ajm8127;483476]As I understand it, the way the car monitors catalyst operation is by comparing the pre and post catalyst sensor data. During cruising, the pre catalyst senor will fluctuate around and lambda of 1. Some times above, sometime below. The three way catalyst has the ability to store oxygen, so it acts as a buffer. What you see at the rear sensor is fluctuation, but only about 25 or 50% as often. It's the slower fluctuation of the rear sensor that the DME uses to gauge the functionality of the catalyst.
With race downpipes, you remove the catalysts, thereby causing both the front and rear sensor to fluctuate at the same rate. What the DP Fix does is take the fluctuations of only one rear sensor and divides it, sending the resulting signal to both rear sensor inputs.
Innovate is fairly clear on their site that factors like age and altitude can make the Bosch wideband sensors inaccurate. Their LC-1 wideband system uses free air calibration to periodically reset the sensor's calibration and keep it accurate despite changing conditions. Obviously BMW does not expect the owner or the service department to do periodic free air calibrations of the two wideband sensors. The work around is using data from the rear sensor to keep the front wideband in check.
As noted above, the DP Fix takes the output of one sensor, divides it, and sends the signal to both sensor inputs on the DME. What happens is the DME never gets the proper rear sensor signal from one of the sensors. It gets the signal from the other banks's sensor. The fact that the signal has been divided is of no consequence because it is averaged over time. So what the DME does is apply corrections to Bank 2 wideband based on data from Bank 1 rear sensor. Obviously these corrections are erroneous.
As you can image, acting on erroneous data the Bank 2 sensor can no longer give an accurate reading. So you may see a divergence in lambda between banks because the closed loop control of the second bank is functioning with an inaccurate sensor, as it has had the wrong calibration applied to it. Or maybe you won't see the divergence, because the sensor is reading one thing, but the actual lambda is something different. "
Today, 03:02 PM