Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Engineer a precision positioning setup for a paper on a roll

  1. Apr 24, 2014 #1
    There are several parts to this effort.

    Bond paper roll being stretched between 2 paper feeders on wheel bearings
    Precision positioning of the roll to a coordinate system devised for that roll That is if the paper roll is 35 meters long or 35'000 mm and 914.4mm wide I could choose a point on that roll say x=10567(where x=1 is 1mm offset) and have that point moved to a centre point between the two roll holders
    jNOGu.jpg

    Imagine the distance between the two rolls increased significantly and the first roll of paper tuck in or be held on the second spindle/shaft.... The idea is to be able to roll back and forth from the left roll to the right so as to achieve a precise positioning to a specific portion within the roll. The position requested by choosing a point on the entire roll i.e. 3'500 mm(3 metres and 50 cm) on x and have that be scrolled right in the centre between the 2 rolls. So if the distance between the 2 roll holders is 2 metres The point on the roll of 3'500mm should be positioned to the middle point which is a point spaced 1 metre from either scrolls.

    When I lay out the entire paper roll the surface gets the dimensions of about 35metres and 914mm.

    I have:

    Arduino compatible board
    Raspberry pi
    Laptop (Ubuntu Linux)
    I need ideas on the shaft , bearings and precision positioning of the paper through rotary or linear encoders etc...
     
  2. jcsd
  3. Apr 24, 2014 #2

    Baluncore

    User Avatar
    Science Advisor

    If you print a “Gray Code” on the back of the paper you will be able to read the position of the paper at any time and move the roll appropriately.
    http://en.wikipedia.org/wiki/Gray_code
     
  4. Apr 25, 2014 #3
    Grey code is indeed among the things I considered on the theoretical side.

    I know printers rely on optical rotary encoders and linear encoders which use grey code etc.

    How do I implement something like this ??
    What elements/parts do I need to make a reliable setup...
     
  5. Apr 25, 2014 #4

    Baluncore

    User Avatar
    Science Advisor

    Since you do not know how tightly the paper will be wound you must track the paper rather than the rotation of the rolls.

    With a 32 metre roll, and 1 metre pages you need a 5 bit code.
    With a 32.768 metre roll, and 1 mm resolution you need a 15 bit code.

    How you do it depends on the resolution you require. With a Gray code you can stop within 1mm of a transition. With 2 metres visible, do you actually need to position it to any 1 mm increment, or is fixed 1 metre pages OK? How many transitions do you need ?
     
  6. Apr 25, 2014 #5
    Have you thought of using a PLC, they are easily setup for encoder applications. Automation direct has an affordable product line compared to AB.

    On a side note how fast is the rollers turning, you will need to consider a pre-warning to stop the paper correctly to the desired position to prevent overshooting your desired position. Also to prevent tearing the paper with an immediate stop (if its a high speed application). One suggestion is to have a pulse encoder on your feed rollers. Take two rollers on each side of the paper, (rubber to prevent slippage) One of the roller will be a pulse encoder (count revolutions). Then prewarn prior to desired setpoint, slow down roller feed till desired location, then stop. With this method your just pulse counting which is easily done. Though you may need a timer control in case the desired set point is between pulses

    however as your using Arduino, here is an article covering encoder implememtation

    http://playground.arduino.cc/Main/RotaryEncoders

    keep in mind I don't use Arduino, however a quick google pulled that article up
     
    Last edited: Apr 25, 2014
  7. Apr 25, 2014 #6

    Baluncore

    User Avatar
    Science Advisor

    Rubber rollers will accumulate errors when moving repeatedly on paper. To track paper with an encoded roller you must use a grit wheel with rounded grit particles, opposed by a rubber pinch roller. When it first runs on the paper it puts dents in the paper that it will then repeatedly follow. The grit wheel was used originally by HP for their range of roller based plan plotters.
     
  8. Apr 25, 2014 #7
    Good point, I'll have to remember that, could come in handy one day
     
  9. Apr 25, 2014 #8

    AlephZero

    User Avatar
    Science Advisor
    Homework Helper

    You will not get high accuracy or consistency by trying to measure the paper movement with any type of pinch wheel or roller, because normal quality paper like bond is not dimensionally stable. Even a 0.1% change in length because of changes in humidity would add up to 35mm over your 35m scroll, and the length changes could be up to 1% depending on the conditions where this is being used.

    If you want reproducible positioning, you need to print some type of scale on the paper itself and use that to sense the position.

    With vertical cylinders you may get problems because of the weight of the paper. For example if the rolls are not kept under enough tension the outside coils will drop down under their own weight. Horizontal rolls and vertical scrolling would eliminate those types of problem (which is why high speed printing presses are designed that way!)
     
  10. Apr 25, 2014 #9

    Baluncore

    User Avatar
    Science Advisor

    HP invented the “gritwheel plotter” in the 1980s to make accurate pen plotting possible without a heavy flat bed. Maybe you don't yet understand the positive engagement and repeatability of HP gritwheels in the HP7470, HP7475 series of plotters.

    I would consider using an HP gritwheel or inkjet plotter to print a Gray code or barcode on the back of the paper.

    I agree the paper roll on its end will be a problem unless some form of differential paper guidance such as optical positioning is implemented. Maybe consider a 45° diameter guide roller to transform a wall of paper onto a horizontal axis roll. We do not yet know why this project is being attempted, projection might be a more economic alternative.
     
  11. Apr 25, 2014 #10

    AlephZero

    User Avatar
    Science Advisor
    Homework Helper

    I know very well (from using them) what the print quality was compared with professional flatbed plotters, and I'm not surprised they became obsolete. The current price on ebay (around $30) is about fair value IMO.

    They only plotted on single A4 size sheets, not continuous paper rolls. Dimensional stability of the paper isn't much of an issue if it only takes a couple of minutes to make a plot. You would need a lot more engineering to transport 35m of paper between two rolls in a controlled manner and use a grit wheel to monitor its position.
     
  12. Apr 25, 2014 #11

    Baluncore

    User Avatar
    Science Advisor

    The HP 7596 “Draftmaster” pen plotter series handled A0 sheet and continuous roll. Gritwheel plotters became obsolete because, HP ink jet technology combined with vector to raster converters, were faster as they no longer required multiple passes.

    The relative quality of flatbed and gritwheel plotter output is not relevant when positioning rather than drawing ink lines with a pen. I have never had a problem with alignment or repeatability with my gritwheel plotters.

    How do you print a code on the back of a 35 metre roll of paper if you do not use roll handling plotter technology?
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook




Similar Discussions: Engineer a precision positioning setup for a paper on a roll
Loading...