Pulsout and drive straight.
abpccpba
Posts: 3
Disregard I'm all wet.
Hello All
I have discovered a new way to make tests for long runs or for that mater any run that is straight.
I put a knife mark on each wheel rim face.
Now I turn the wheels manually so that both marks are at same reference point on the frame. I like to use the spaces between the two screw heads at the front holding the servos in place.
Run the program. Change the Pulsout function 2 units at a time. Until the two marks match across from each other regardless of where they end up compared to the original location at the screw head spaces. Or put another way the numbers are correct if the imaginary line created by the two wheel marks is perpendicular with the bot wheels plains.
Looks like I am with in one degree or less of each other stop point.
I have not run this program on the floor yet to see if it is near accurate.
Now I will try ten seconds backwards and see what happens.
With this system I have eliminated the plain it rides on and removed the wheel relationship with the surface ie. slip between the surface and wheel. Oh to the momentum of the bot in motion.
When I get to the real time test on the ground if bot does drift then I have eliminated the motor out of the system. Pretty much.
I have not looked into the right and left turn motion yet.
abpccpba
Post Edited (abpccpba) : 6/9/2010 12:58:35 AM GMT
Hello All
I have discovered a new way to make tests for long runs or for that mater any run that is straight.
I put a knife mark on each wheel rim face.
Now I turn the wheels manually so that both marks are at same reference point on the frame. I like to use the spaces between the two screw heads at the front holding the servos in place.
Run the program. Change the Pulsout function 2 units at a time. Until the two marks match across from each other regardless of where they end up compared to the original location at the screw head spaces. Or put another way the numbers are correct if the imaginary line created by the two wheel marks is perpendicular with the bot wheels plains.
Looks like I am with in one degree or less of each other stop point.
I have not run this program on the floor yet to see if it is near accurate.
Now I will try ten seconds backwards and see what happens.
With this system I have eliminated the plain it rides on and removed the wheel relationship with the surface ie. slip between the surface and wheel. Oh to the momentum of the bot in motion.
When I get to the real time test on the ground if bot does drift then I have eliminated the motor out of the system. Pretty much.
I have not looked into the right and left turn motion yet.
abpccpba
Post Edited (abpccpba) : 6/9/2010 12:58:35 AM GMT
Comments
Let us know how you make out! Lots of encoder projects here in the forums, when you're ready!
BTW are you using a BoeBot or Scribbler?
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·"If you build it, they will come."