Dead reckoning does not travel predictable distances

first95 / FRC2017

The codebase for our 2017 robots, Orville and Wilbur Wright.

GNU General Public License v3.0

3 stars 0 forks source link

Dead reckoning does not travel predictable distances #39

Open jwalthour opened 7 years ago

jwalthour commented 7 years ago

Seems to have ceased to work properly sometime between the MABOS event and the SNH event. Affects both driving straight and turning. Have checked the following:

Encoders appear properly mounted
Encoders appear connected
The new gear ratio doesn't affect it, because the encoders are on the gearbox output shaft
The automove is reporting that it's reached its destination

jwalthour commented 7 years ago

Here's an example of a different strategy, where PID positional control is delegated to the Talons: https://github.com/AquidneckIslandRobotics/2017/commit/3b4df11269cfa4c19e7b471545662c40c6b8efc9

I'm not sure if Follower mode works properly with this design - I believe the code shown above is run on a drivetrain without ganged motors.

RuggS commented 7 years ago

We have tested auto moves on Wilbur and all appears to be well. Enc ticks per foot has been changed and angle multiplier has been removed. We would like to test on Orville on carpet to confirm no differences between the 2 bots. Additionally we tested Boston code on Wilbur and it worked. Verifying this on Orville would also be good.