I am concluding that if there is a difference it is well within the noise for my testing method and I am not going to bother differentiate between AC on and off for future tests when I want to see if a mod works or not.
The AC was on when it was warm enough that I needed it to be comfortable and off when it wasn't. I didn't do any ABA runs with/without the AC as I am not planning on altering my use of AC. I differentiated the data since I thought there would be a difference in consumption and I would need to factor in AC or no AC when testing a mod.
As far as ABA testing goes I generally tried to stick to the pattern of 90, 100, 110, 120, 96, 106, 116, 92, 102, 112, 98, 108, 118, 94, 104, 114, 90...
__________________
|