The controller still has that prototype look:
I had a felling lately that it seems like there are very few that reads my thread. I saw a few days ago that there was 54.133 views, and still when I made the latest update it was the same number. Guess what, it is still 54.133 today. And then I have a like on that post without a single viewer
Anyway the biggest update on the bike was changing controller:
I had to make the bracket both a little longer and wider.
I was able to just move forward and down, so that feels pretty good anyway.
Making a new cover is pretty much work, but I was able to just modify it a bit.
The little ridge actually gives a better grip when lifting the bike
I made a cutout for the cooling fins, I didnt like the look at first.
I think it should be possible to leave the white part intact, maybe even a small ridge of the black if I cut off a lot from the inside.
But I dont think I care about it anymore.
An extra bracket for the sidecover
I dont think I can preview anymore, so I might just as well post before it gets to long
back to controller tuning. We have been struggeling to get it to run properly.
I got an explanation from Hackey this evening.
The problem seems to be that as I run hall sensors, vesc switches over to observer at a certain erpm.
If I had ran encoder it would have been accurate enough to stay on and not switch to observer.
That observer determines the position of the motor by reading the back emf from the motor.
But as it is an IPM motor with some salience, 30uH (to compare with 10uH on a qs 138 according to Hackey) it effect the back emf so that the observer dosent work very good.
But there should be a beta version of a newer vesc tool that should hopefully solve the problem.