Originally Posted by ajsalida
Originally Posted by Jake@PTF
The DME has logic in it that they work on a somewhat opposite pulsating trim. Basically - if they are post single turbo - they will read identical to each other since it will be flowing through a single exhaust (the turbine housing) before splitting back up. Since the DME will see identical readings it will "freak out" and enter wierd trimming modes. Without the opposite pulses that they get when separate the DME will not fuel properly. You would have to change the logic of the DME so that the sensors would be okay being identical.
This among other things are on the list of "to-do's" at Cobb. They do A LOT of custom logic and coding on top of just accessing and changing tables. Time is the issue and Jason and Josh are both working very hard to keep on top of as much as they can and delegate projects when needed.
I read up on this a while back trying to understand the logic. I believe this is part of their (BMW-Bosch's) way of ID'ing individual cylinder combustion health. That is they superimpose a low level coded pulse of fuel/timing onto each cylinder's ordinary fuel/timing every so often and then look for its signature variation out of the O2 sensors for the appropriate bank. I am not sure how they de-correlate it but they can then infer what is going on in each cylinder by time windowing or some other signal processing logic. So you'd capture the coded pulse and then everything near it time wise would contain info on that cylinder. Somehow they tease out the info.
It is there for a reason and a paper I read seemed to imply it was for some obscure CARB emissions requirement either in place now or was thought to be coming soon regarding individual cylinder combustion monitoring. I guess the point is before trying to spoof this it would be good to know why it is there.
On the other hand I may be completely full of shit.
So with this theory the cylinder bank individual runner lengths would have to be in the algorithm for that cylinder. If any aftermarket change in runner length for that cylinder would throw off the timing of the pulse capture. Sounds legit.