Welcome to Our Community

Unlock hidden features. Sign Up for Free Today!

Backlash issue on X and Y - OX CNC

Discussion in 'CNC Mills/Routers' started by thecableguy, Dec 21, 2015.

  1. thecableguy

    thecableguy Well-Known
    Builder

    Joined:
    Oct 24, 2014
    Messages:
    11
    Likes Received:
    3
    Good day folks, I built a stock OX nearly a year ago, and since the beginning I've had 0.01" of backlash on the X and Y axes. It usually doesn't matter much since I use it to build guitars and I always add a small margin to the parts to allow for sanding. Where it does cause problems is when I'm cutting inlays and inlay pockets, as well as with small circles, say 1/4" diameter. I've addressed everything on the machine that could have been loose such as the sprockets on the stepper motors and I've set up a belt tensioning system to keep the belts tight. I've checked with a dial indicator and I get a repeatable 10 thousandths of backlash on both belt driven axes. I'd like to know if this is inherent in the belt drive system (wouldn't think so) and if not, has anyone else had this issue and how did you address it?

    Thanks very much folks, please let me know if you need more info from me.

    Brian
     
  2. Joe Santarsiero

    Joe Santarsiero OB addict
    Staff Member Moderator Builder

    Joined:
    Oct 30, 2014
    Messages:
    677
    Likes Received:
    142
    Hi Brian,

    What spindle, mount, step setting, and pulley diameters are you using? What are you indicating from? Are your motors tight?

    Joe
     
  3. Steve Fox

    Steve Fox Veteran
    Builder

    Joined:
    Feb 22, 2015
    Messages:
    122
    Likes Received:
    16
    Brian,

    Are you sure it's backlash?
    If you put a really small bit in it, does it zero to the same point every time?
    I only suggest these things because I had/have the same issue. I'm not sure mine is fixed yet.

    I seem to be losing steps every so often only on the Y-Axis.
    First thing, I marked the pulleys and shafts to make sure the pulleys weren't slipping.
    Then, I upped the current to my motors just a little, still within specs.
    I'm not sure that fixed it and haven't been in the same city with it since I made the changes.

    Someone suggested it's the motor controller, but I can't check that until I get back.
    Supposedly, the cheap controllers randomly skip steps.

    The weird thing to me, anyway is it only happens on the Y-Axis and is almost always 1/16".
    In my case, I believe that would be 64 steps, give or take a couple.
    There's never a problem on the X-Axis. I guess it could have something to do with two motors on the Y-Axis vs. one on the X-Axis.

    Any of this sound familiar?

    If it truly is backlash, there aren't many places it can occur. First, I would tighten the belts.
    I don't know how you have yours clamped, but I had to put brass shims under the screws holding mine to keep them from tearing up the belts. Once you make sure the pulleys aren't slipping, you can eliminate that. Are the carriage wheels so tight that there is too much friction?

    No idea what to do after that.
     
  4. thecableguy

    thecableguy Well-Known
    Builder

    Joined:
    Oct 24, 2014
    Messages:
    11
    Likes Received:
    3
    Hi Joe, I'm using a Bosch router mounted with a clamp I made on the machine. Running a TinyG controller at 1/8 microstepping, and the 20 tooth GT3 pulleys. I mounted a dial indicator to a vise and mounted that to the bed in order to make the measurement. I have the dial indicator measuring against the side of the collet nut. Everything is tight - motors, pulleys, belts etc. Doesn't make sense.

    Brian
     
  5. thecableguy

    thecableguy Well-Known
    Builder

    Joined:
    Oct 24, 2014
    Messages:
    11
    Likes Received:
    3
    Hi Steve, it does look off a bit when re-zeroing a small bit. I tried that before using the dial indicator.

    I did have problems initially with the pulleys loosening up on the motor shafts but I addressed that with larger set screws and loctite. I also changed the belt clamping method from stock to an adjustable tension mechanism. Didn't make any difference. When the belts are removed the X and Y move freely. I don't think I'm missing steps as I'm using a TinyG controller and I haven't heard that from any of the other users. Also I'm using 425 oz stepper motors with the current fairly high. I'm scratching my head here.

    Brian
     
  6. Steve Fox

    Steve Fox Veteran
    Builder

    Joined:
    Feb 22, 2015
    Messages:
    122
    Likes Received:
    16
    Is the problem in both X and Y?
     
  7. JustinTime

    JustinTime Master
    Builder

    Joined:
    Dec 18, 2013
    Messages:
    798
    Likes Received:
    158
    Brian, backlash is something we have to live with with our machines, especially when we use a belt. Using lead screws and anti-backlash nuts helps eliminate most of the problem. For control software I use Mach3. In there there is an option for backlash compensation. Once I've set it to the correct amount my parts cut perfect and the circles are round.
     
    snokid likes this.
  8. thecableguy

    thecableguy Well-Known
    Builder

    Joined:
    Oct 24, 2014
    Messages:
    11
    Likes Received:
    3
    Yeah it's both axes.
     
  9. thecableguy

    thecableguy Well-Known
    Builder

    Joined:
    Oct 24, 2014
    Messages:
    11
    Likes Received:
    3
    Thanks for the info. I wouldn't have thought a belt driven machine would have backlash. Kind of surprising. What value did you have to enter in Mach 3?
     
  10. Steve Fox

    Steve Fox Veteran
    Builder

    Joined:
    Feb 22, 2015
    Messages:
    122
    Likes Received:
    16
    I don't think I have any backlash on my OX. The X-Axis always zeros out in the same place. I'm not sure about the Y-Axis.
    I have the problem I discussed earlier about missing steps, but no backlash I know of.
     
  11. JustinTime

    JustinTime Master
    Builder

    Joined:
    Dec 18, 2013
    Messages:
    798
    Likes Received:
    158
    @thecableguy, I do not recall how I did it in Mach3, been many years, but there is a place where you enter it. I measured it with a dial indicator. I ran the axis 5" and back to 0 and the dial indicator told me how much was missing. I think that's the number I have entered. If I recall correctly I had on X 0.003" or something like this. But don't quote me on this. :) It's a very small number anyway.
     
  12. Julius

    Julius Veteran
    Builder

    Joined:
    Nov 13, 2015
    Messages:
    156
    Likes Received:
    42
    Brian, are you aware that there is also a very high run out on the colt's collet, that may be causing inlay issues?
     
  13. thecableguy

    thecableguy Well-Known
    Builder

    Joined:
    Oct 24, 2014
    Messages:
    11
    Likes Received:
    3
    Hi Julius, I'm actually running a Bosch 1617 router with a precision collet from Precise Bits. I use 0.02" end mills to cut inlays so any appreciable runout would break them. I did measure the backlash on the machine with a dial indicator so I know it's there, just not sure if it can be fixed. I was thinking of retro fitting the machine to ball screws actually.
     
  14. thecableguy

    thecableguy Well-Known
    Builder

    Joined:
    Oct 24, 2014
    Messages:
    11
    Likes Received:
    3
    Interesting development; one of my belts recently broke, probably due to me over tightening them in an attempt to remove the backlash. I ordered 10 meters of GT3 belt from China through Ali Express. The belt arrived the other day and last night I replaced all three belts with new ones. The first thing I noticed was that the machine is quieter. It always had made a strange noise on the Y axis; that's now gone. I set a 1000 mm ruler on the bed in order to re-calibrate the controller. As I returned the machine to zero each time I saw that it is now returning very precisely to zero as far as my eye can tell. Previously the machine was visibly off and now it doesn't appear to be. I'll check later with a dial indicator to see what the actual backlash is, but I think these new belts made a difference.
     
  15. GrayUK

    GrayUK Master
    Staff Member Moderator Builder

    Joined:
    May 5, 2014
    Messages:
    951
    Likes Received:
    231
    The Cable Guy
    Sounds like there was stretch in your old belts. Perhaps over tightening your belts was the problem? Are you using the single or double method of belt drive?
     
  16. thecableguy

    thecableguy Well-Known
    Builder

    Joined:
    Oct 24, 2014
    Messages:
    11
    Likes Received:
    3
    The backlash was there the whole time, regardless of what the belt tension was. I use a single belt drive.
     
  17. GrayUK

    GrayUK Master
    Staff Member Moderator Builder

    Joined:
    May 5, 2014
    Messages:
    951
    Likes Received:
    231
    Just a thought. :rolleyes:
    You might consider the 2 belt system in the future. This is simply because, it reduces the length of belt, that is under immediate tension, to just the piece which leaves the junction of the other belt and then rejoins it! That only happens when it is forced to roll over the Pinion wheels. The rest of the belt is in, a mated, rack and pinion state and not being effected by the stress. :thumbsup:
     
  18. Joe Santarsiero

    Joe Santarsiero OB addict
    Staff Member Moderator Builder

    Joined:
    Oct 30, 2014
    Messages:
    677
    Likes Received:
    142
    Have you compared the teeth of the new belt against the old?
     

Share This Page

  • About Us

    The OpenBuilds Team is dedicated helping you to Dream it - Build it - Share it! Collaborate on our forums and be sure to visit the Part Store for all your Building needs!
  • Like us on Facebook

  • Support Open Source FairShare Program!

    OpenBuilds FairShare Give Back Program provide resources to Open Source projects, developers and schools around the world. Invest in your future by helping others develop theirs!

    Donate to FairShare!