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

Precision CNC router

  1. Aug 17, 2010 #1
    I am going to build a CNC router. Although that's misleading since I don't nessecarily plan on actually using a router as the main tool. I want to see how precise I can get and what kind of quality I can achieve without spending lots of money. So I have some basic questions. The plans I've seen rely on a chain and sprocket to drive router. Their precision relies on the gear ratio between the linear motion of the gantry and the stepper motor, and the tightness of the chain and sprocket system. Even though the CNC I saw was able to mill PCB boards without any apparent defects, I think precision can be improved.

    I want to incorporate a location tracking system. I'm thinking an optical detector and black and white lines across the axises. But i'm open to suggestions, in fact it's why i'm here.
  2. jcsd
  3. Aug 17, 2010 #2


    User Avatar

    Staff: Mentor

    The PCB milling machines that I've seen from LPKF use microstepper motors (or maybe servo motors?) and linear screw drives. That would seem to have much less of an accuracy/repeatability issue compared to a chain drive. It only uses end of travel limit switches to keep the absolute position information, I think.
  4. Aug 17, 2010 #3
    There is a great deal of information out there for building CNC machines of all types and sizes. One of the better forums I've found is http://www.cnczone.com/forums/index.php

    Some other links that may be helpful:
  5. Aug 18, 2010 #4
    I retrofitted my bench-top drill-mill for CNC. Replaced the (old,worn) acme screws with ball screws which have only a few mils of backlash. The screws are driven by servo motors and fairly tight timing belt/gears. I think chain and sprocket might not get tight enough to limit backlash, just a guess though.

    Limit switches are handy but I hardly ever run to the end of travel so I skipped them. I have puzzled over the idea of using digital readouts as feedback, but that's a lot of work and programming to get going. Generally, once you've compensated for the backlash issues things are pretty repeatable in the thousandth range.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook