Posted: Tue May 02, 2006 7:36 am
In vsn 2.6 there are algorithms to help you achieve your goal of stable ego readings. There is a 10 point barometric correction tabel you can use to dial in changing baro - assuming you have a dedicated baro sensor (2nd map sensor). You tune at a fixed altitude, then take the car to another altitude, note the ego change and the baro change, fill in a correction in the appropriate place in the table that brings your ego reading back to what it was at the initial altitude. Keep doing this for the whole range of altitudes you expect to encounter and that should take care of that. If you are really obsessive, you can tune your air temperature corrections in the same way.
Then we get to acceleration corrections. You will always see a lag in the ego when you accelerate. It is due to the fact that the mixture changes in the intake port due to fuel being dumped on the walls rather that going into the combustion chamber. You can't compensate for this until it is sensed and it won't be sensed until the fuel/ air mix reaches the ego sensor and the sensor has time to process and average the mix. So there is always going to be lag in the ego, and even jitter at steady state because even in the best prepared engine there is enormous variation from cylinder to cylinder and on consecutive exhausts from the same cylinder. This is averaged out in the ego, but if you average too much you will get a lag in response. There are WB prediction/ corrector algorithms in v2.6 to try to compensate at least to some extent for ego lag.
Then we get to acceleration corrections. You will always see a lag in the ego when you accelerate. It is due to the fact that the mixture changes in the intake port due to fuel being dumped on the walls rather that going into the combustion chamber. You can't compensate for this until it is sensed and it won't be sensed until the fuel/ air mix reaches the ego sensor and the sensor has time to process and average the mix. So there is always going to be lag in the ego, and even jitter at steady state because even in the best prepared engine there is enormous variation from cylinder to cylinder and on consecutive exhausts from the same cylinder. This is averaged out in the ego, but if you average too much you will get a lag in response. There are WB prediction/ corrector algorithms in v2.6 to try to compensate at least to some extent for ego lag.