Originally Posted by
smokeshow
Lol. So, my initial idea for a part two was to address an issue that compromised the results of this procedure....MAF reversion. If you're not familiar, this is a phenomenon that occurs (usually) with aftermarket intake tubes. Reversion is exactly what it sounds like - reversal of flow. When you add a camshaft that has a late IVC angle and/or pair it up with a new intake tube (or if you are bone stock, in the case of the LS2 GTO), often times what you've done is create an environment for 'pneumatic communication' of the camshaft with the MAF sensor. The big cam closes the intake valve late, and at low RPM pushes some charge back out of the cylinder before the valve closes. This causes a pressure rise in the intake and the reversal of flow. The flow reversal can result in high OR low measured mass airflow relative to actual mass flow, depending on the hardware itself. The problem then became, how do I use sensor measurement as a reference for correction to an airflow model when the sensor may unpredictably yield inaccurate results? It is easy enough to pick the error out of a graph with your eyes, but not so easy within a set of collected data, and definitely not possible with HP Tuners due to the slow data collection rate and limitations of the scanner itself. I used to calibrate with a tool called Inca which was capable of data collection as fast as the controller execution loops updated the variables...but that doesn't happen on CAN that HPT uses. The data comes in scheduled and slow and doesn't provide a signal I can use that allows me to separate the error from it. Haven't figured out a way around that yet and not sure if I will. At least not without scripting in MATLAB, which few have access to.
That said, I've gotten requests a few times for a procedure on data filtering. Not really a 'part 2', but is still relevant and can help eliminate some error that you may find when calibrating with this procedure. Is there any interest for a 'how to' on data filtering?