Randy, Just a quick follow up. After some delays due to weather, temperature, and logistics, I finally got out for a good tuning session again. Because I wanted to isolate the navigation tuning from the problems I have turning on grass, I wanted to find a nice large paved surface to test on, but also one without interference for my RTK setup. It can be hard to find such a space near me. I got thrown out by security for trying to test my rover in an abandoned and empty part of a mall parking lot!
I had been describing my rover as slowly weaving along the path between waypoints. And so in reading the navigation tuning page, I thought I should raise the Navl1_period. That did not seem to do much. So I tried zeroing out the Xtrack value, which had little effect, and raising the NavL1 damping similarly didn’t seem to help. But finally I realized I was possibly misunderstanding the relationship between how my rover was behaving and the NavL1 period, and lowered that sharply, and low and behold, that seemed to have a dramatic effect. I think now I can mess with changing Xtrack and damping values a little, but finally the rover seems to be trying to stay on the line better. I dont need it to be perfect, I really just need it to stick within about half a meter of each transect and finally I think its going to do that. Here is a very short time-lapse of some .5 m transects with extra overrun and lead in. Sometimes it is pivoting the wrong way at the end of a transect, but it seems to be getting sorted out by the lead in points.