Cutting Paper Scrolls With Frickin’ Lasers

This circuit illustration adds a scrolling paper feeder to the bed of a laser cutter. In the video after the break you can see that the actual assembly is put on the bed of the laser cutter. After the laser has cut out the specified pattern, the scroll is wound to move an un-cut portion into place. It uses a servo motor to drive one of the spools.

An Arduino Uno with a servo shield is being used for this application. It has one button which winds one spool for a pre-programmed period of time. There’s a few issues with this setup, namely that it’s not tied into the CNC program that runs the laser. There’s also a lack of precision when using a continuously rotating servo like this. If it were upgraded to use a stepper motor and patched into the CNC hardware this would make cutting new scrolls for your player piano a breeze.

Here’s a project that does the opposite, it takes old player piano rolls and digitizes them.


16 thoughts on “Cutting Paper Scrolls With Frickin’ Lasers

  1. It’s not just the lack of control or feedback from the motor driving the spool; the diameter of the spool with the paper gets larger as more paper is wound onto it, which would still make the setup inaccurate if a stepper motor was used.

    It’s probably better to use an incremental encoder with a separate wheel to measure the length of the paper being wound.

  2. SparkyGSX beat me to it. Needs some sort of registration marks on the paper itself. Pre-print black marks on the edge and use an optical encoder to determine how far it’s gone.

  3. Thats one hell of a ticker tape machine! As for measuring the paper, why not just gut a laser mouse and use those parts to measure the amount of paper dispensed.

  4. To add to Brad’s idea, rather than black marks, just use the laser to cut the next alignment mark on the edge of the paper use a typical optical switch. Of course this assumes this will be a framed cutting process like the video and not a continuous cut.

  5. The driver motor could be a series of roller, then the diameter of the roller would not change and you could hook it up to servos or steppers with not problem. Not sure how mach3 or any of those cnc programs could handle something like this natively though, may still need to be a seperate feeder system that recieves a pulse and sends a pulse at the end of the cutting program and the end of the new matterial feed.

  6. I always thought steppers were not as accurate as servos? Especially under high torque? Our (really high end) industrial (are there other kinds?) waterjet user servos, our “cheap” plasma cutter uses steppers and our high definition plasma cutter uses servos.

  7. Servos have closed-loop feedback controls (a potentiometer or rotary encoder that measures the position of the shaft) while steppers don’t. You’re right about torque — basically if there’s enough back-force that the stepper starts slipping and skipping coil pairs, all your accuracy is shot, while the servo will know that it’s off and attempt to realign.

    On the other hand the stepper may be more accurate for moving in small increments with little torque because the stepping gives it discrete, accurate, geometrically-fixed rotations rather than having to rely on an electronic part with its own set of errors.

  8. I remember reading once that there is a methodology for dealing with the increasing size of a spool of something.

    I found it when researching how to make film camera’s. I wish I could remember the details but it seemed like basic math that took into account length and thickness of the material and spool size.

  9. Does this account for the reduction in the size of the paper spool shown to the right?

    To get accurate cuts every time this needs to be accounted for, right?

    Or does the previously cut pattern stop using a sensor of some kind.

    Otherwise get halfway through the roll and your cuts will be well out!

  10. @Hackerspacer: your statement is correct, if we are talking about industrial servo amplifiers. These are basically a motor (used to be brushed DC, nowadays mostly BLDC or BLAC, but you could even use an asynchronous motor), a position feedback device (absolute or incremental encoder, resolver, etc.), and the amplifier itself. The amplifier is what makes it work; it continuously compares the current position to the desired position of the motor, and corrects for any deviations.

    This “servo”, however, is a hobby servo that has been modified for continuous rotation, which means you have to remove the position feedback potentiometer, so now it’s just a DC motor with a gearbox and a H-bridge that can be controlled by a 1-2ms pulse train. It’s no longer a servo because it lacks any type of feedback.

  11. Capstan drive not spool drive is the way to get it linear. However in the case of the player piano the take-up spool diameter is included in the design playing and cutting so it all balances out.
    Diamond shaped holes dumb. You want short on and off times.
    Does it leave hanging chads?

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.