[GUMP]

Originally Posted by pbp1

.
I have a perfect example of a test I saw that proved this theory. I had a customer, running in Speed Density, Open Loop, who iced his manifold to pick up some ET. This car has a near perfect calibration and the actual A/F lays right on the Target A/F even though it is not using any O2 correction. When he iced the manifold, we did not think about the fact that the Air Temp Sensor was in front of the throttle body. The thing was so lean that it died as soon as he left. This was because the air temperature actually entering the engine was much colder than the sensor was reading. We took another car with the same combination, same exact tuning strategy, but we put the Air Temp Sensor in the manifold. We iced the manifold and this time, the car picked up .080 and the actual A/F laid right on top of the target A/F again (just like it is supposed to). I pointed out to him that when we looked at a datalog from a run with no ice compared to a run with ice, everything was the same except for on the run with ice, the air temp was lower and the pulse width was higher.

This is one of the reasons that I like to turn off any IAT corrections. This stuff is great for heads-up runs if your tune is pretty good, but it can make a car harder to dial.