Welcome to Our Community

Some features disabled for guests. Register Today.

Control - Moves are .01mm off

Discussion in 'Control Software' started by Wagadougou, Aug 23, 2020.

Tags:
  1. Wagadougou

    Builder

    Joined:
    Oct 6, 2017
    Messages:
    4
    Likes Received:
    1
    Hello,
    I'm using Control 1.0.243 - Windows

    Command G0 X10 Y10 -> Randomly ends with X10 Y9.99
    A quick video is attached.
    Same for Z.
    Any clue ?

    Thanks !
     

    Attached Files:

    • 2.mp4
      File size:
      104.8 KB
      Views:
      1
  2. Rob Taylor

    Rob Taylor Master
    Builder

    Joined:
    Dec 15, 2013
    Messages:
    1,470
    Likes Received:
    746
    Happens all the time, just the way the mathematics of the motion calculations shake out when grbl reports them. Think of the command more like "Y10 +/-0.05" - in physical reality, it's the exact same position to within the tolerances of the machine.
     
  3. Wagadougou

    Builder

    Joined:
    Oct 6, 2017
    Messages:
    4
    Likes Received:
    1
    Understood.
    Thanks Rob.
     
  4. JustinTime

    JustinTime Veteran
    Builder

    Joined:
    Dec 18, 2013
    Messages:
    759
    Likes Received:
    253
    Just to add to Rob's correct explanation.

    Let's say your motion actuator, be it a screw or a pulley, advances , let's say, 3 per revolution (could be any other number). If you're command is 'go to 10' it just can't do it since 10/3=3.333..... If your DRO would display enough digits after the decimal point it will show 9.999.... which is darn close enough to 10. :) If your DRO displays less digits it may only show 9.99. The flip side is that if you tell the machine to go to 9.999 it will do it perfectly where as a machine that would go perfectly to 10 in the first example would still go to 10. :D
     
    Peter Van Der Walt likes this.
  5. Rob Taylor

    Rob Taylor Master
    Builder

    Joined:
    Dec 15, 2013
    Messages:
    1,470
    Likes Received:
    746
    Not only the limitations of base-10 math, but the limitations of binary math too; floating point arithmetic is far from precise, and some numbers literally cannot be represented in the floating point space. Add the conversion of decimal to binary and then back again, and you have several sets of rounding going on. And that's before you start considering the limitations of an 8-bit processor operating in real-time.
     
    Peter Van Der Walt likes this.

Share This Page

  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.
    Dismiss Notice