I hope your not as confused as I am. I see what Michael saw. The speed error isn't linear. I turned the errors in to percentages and the error increases by roughly .01% for each 10 mph. I can't see a mechanical speed difference causing that. interested in the ... Login to read more.