Shop OBEX P1 Docs P2 Docs Learn Events
4-axis digitizer — Parallax Forums

4-axis digitizer

toolmanMiketoolmanMike Posts: 5
edited 2005-04-01 00:02 in BASIC Stamp
I'm interested in building a 4-axis digitizer using rotary encoders. My experience is limited to the Boe-Bot which I'm still assembling. Would anyone care to point me in the right direction?

Thanks,
Mike

Comments

  • BeanBean Posts: 8,129
    edited 2005-03-27 02:43
    What is the 4th axis ? X, Y, Z and what ?
    Not trying to be smart, just trying to understand.
    Bean.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    "SX-Video Display Module" Available Now.

    www.sxvm.com

    "A problem well defined, is a problem·half solved."
    ·
  • Beau SchwabeBeau Schwabe Posts: 6,568
    edited 2005-03-27 03:53
    Bean, there could be a rotational axis on the X and Y plane. In another life
    when I worked on the R&D side of a prosthetic facility we had a CNC digitizer
    like this with the rotational base considered the 4th axis.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Beau Schwabe - Mask Designer III

    National Semiconductor Corporation
    Latest Company News
    (Communication Interface Division)
    500 Pinnacle Court, Suite 525
    Mail Stop GA1
    Norcross,GA 30071
  • toolmanMiketoolmanMike Posts: 5
    edited 2005-03-27 04:10
    There'll be waist, shoulder, elbow, and wrist rotation.

    Thanks,

    Mike
  • BeanBean Posts: 8,129
    edited 2005-03-27 16:05
    Just wondering, what's the difference between "4-axis" and "4 degrees of freedom" ?
    Bean.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    "SX-Video Display Module" Available Now.

    www.sxvm.com

    "A problem well defined, is a problem·half solved."
    ·
  • toolmanMiketoolmanMike Posts: 5
    edited 2005-03-27 17:14
    No difference, I think.

    Mike
  • toolmanMiketoolmanMike Posts: 5
    edited 2005-03-28 13:09
    Beau,

    You're right. The 4th axis is rotational. In case "digitizer" is an unfamiliar term to some, it's the reverse of a robot arm: Instead of telling the arm where to move, I want the arm to tell me where it is. The power to move the arm is supplied by the operator. He simply places the probe where he wants to register a point and pulls the trigger...and it displays on the monitor.

    Mike
  • kelvin jameskelvin james Posts: 531
    edited 2005-03-30 08:18
    Since this is a topic of interest of mine, i can give you some idea of where you can start. I will be attempting to do the same thing, but on a larger scale, and not using encoders. I assume you are trying to copy the likes of a Microscibe from Immersion ( and i guess they are pretty happy since they got that 90 mil$ patent judgement against Sony, that was the vibrator technology in their controllers), which is a contact probe for 3d modelling on an articulating arm. The Microscribe uses optical "sensors" , obviously encoders, so you are on the right track for the multi axis input. The big thing is though, is how to manipulate the positions from the different axis, and inverse calculate it into a 3d plane. And it isn't going to be easy, that is why these things cost mucho dollars. If you were just dealing with a simple x,y and z axis, an encoder for each movement of axis in a straight line ( such as a cnc), the math is pretty simple. But, when you add in the "rotational axis" of movement for each joint axis, it is a whole different scenerio. It is not like it can't be done, industrial robot controllers have been using this type of positioning software for years. Immersion actually has some free software called MUS, that will take the data from the encoders and input it into software like autocad, etc., but what kind of data comes out of the microscibe, i don't know. It may be just encoder positions, i don't have any idea. There is mathematical calculations you can do, like inputing the data into a matrix, this is some info on what it is about. http://www.gamedev.net/reference/articles/article695.asp
    You would have to set up a 3d plane, set all the encoders to zero at the "reference position", and calculate the movement of each axis together against that point. Clear as mud, right?
    Now, whether a stamp can handle this, it might be possible, maybe enough ram locations to use a matrix calculation. And as long as the encoders are in a fixed position ( not moving ), the stamp could take in the data no problem. Hopefully this will start you on your way, i am trying to find some of my old notes from prior schooling that has some more detailed info about this.

    kelvin

    Post Edited (kelvin james) : 3/30/2005 6:10:13 PM GMT
  • Bruce BatesBruce Bates Posts: 3,045
    edited 2005-03-30 08:38
    Folks -

    I'm not quite sure if indirect reference is being made here to what's known as a "tilt controller" or not, but if so, you may find some interesting and free information here in the SparkFun "KidBot-RC" Section:
    http://www.sparkfun.com/shop/index.php?shop=1&cart=211815&cat=81&

    Additionally, SparkFun also offers the Applied Digital ADXL202 accelerometer in a miniature breakout board format for easy integration:
    http://www.sparkfun.com/shop/index.php?shop=1&cart=211815&cat=71&

    There is a Basic Stamp example program for the ADXL202 there as well.

    Regards,

    Bruce Bates
  • toolmanMiketoolmanMike Posts: 5
    edited 2005-04-01 00:02
    Thanks for the good info.

    My plan is for 3-axis to move in the vertical plane and 1-axis (the waist) to move in the horizontal plane. I don't think the math will be overwhelming with that simple movement. If it's not clear, I intend for the math to be done on the desktop and have the PIC forwarding the encoder data through the serial connection.

    My little bit of research leads me to believe that the problem will be for the PIC to keep up with 4-encoders moving at once. I saw where something called a "decoder chip" was refered to in solving this problem. Also, there was a comment about using a USB chip.

    Mike
Sign In or Register to comment.