torn - 8/1/2015 14:24 Ok. On its face, this seems to make at least some sense. However, I wouldn't have thought you'd see quite that variation in yield. The NH flow sensor is a linear (or single-point calibration) sensor, meaning you calibrate it at a single flow rate. Near that flow rate, you can expect very good accuracy if your calibration has been performed correctly. However, as you move away from the calibrated flow rate, you will see accuracy dropping off, as the software has to project what the incoming force values mean. Both Case IH and New Holland combines are promoted as being designed for "grain on grain" threshing. This means that grain kernels are primarily rubbed against other crop or MOG material in the rotor cage, rather than against the rotor itself, or the concaves. In theory, this means that a full machine would thresh more effectively than a machine that is partially empty. More effective threshing should equal a higher yield. Have you done a kill stall and performed grain loss exercises with this machine, checking for header loss, rotor loss, and sieve loss? If you find more rotor/sieve loss behind the 300mm swath, it would lend additional weight to the theory that the machine is threshing more effectively/efficiently when it actually has more non-crop material (MOG) for the rotors to work with. Another idea - can you cut two passes, one at 100mm and one at 300mm, but vary your speed to keep both passes at the same flow rate that was used for your weight calibration? This should result in consistent results from the flow sensor (i.e. eliminating error due to a flow rate that is different from the calibrated rate), and should help provide evidence to either support or contradict the idea that you get better yield when you keep the machine "fuller". This is very interesting. I'm curious to hear other theories.