Quote:
Originally posted by deruvian
I also hate to interrupt, but I just thought of something interesting. And if I don't post it now, I'll forget.  
Just a scenario: One fills the SVX with 87, but forgets to change the code on the stage 1 over to the 87 version. Bad deal, or would the car simply run a lot richer than normal?
|
How to put this...hmm....not nearly as bad as running 87 on the factory code....probably not as bad as running 93 on the factory code...maybe not nearly as bad as running 93 on the factory code....very likely not nearly as bad as running 91 on the factory code.....but still should be avoided. The system by which the ecu determines that an ignition timing issue is serious enough to go into the hard to remove learned section is complex. It's hard to say how badly you have to screw up for how long before a correction is moved over from a volatile memory location, to a memory location retained when the car is turned off, to a memory location that is retained even without battery power. It's also hard to predict how hard it would be to remove a particular learned correction because many things can affect that.
In the end what it comes down to is things that are hard to remove were hard to learn to begin with.
Just to give you some idea how things can get intertwined and complex. Let's talk about where these adjustments that may get learned come from. The ecu uses the main maps to calculate base injector pulse width and ignition timing according to load and rpm. It then sequentially applies many adjustments to those base #'s according to many other #'s. Some of those #'s are current measurements, some of them are caculated from a # of current measurements, some of them are selected from tables according to current measurements, some of them are calculated from past learning, etc etc. And it's all done so that in some conditions the importance of some #'s is greater and in other conditions the importance of other #'s is greater.