Welcome to Our Community

Some features disabled for guests. Register Today.

The CNC cutting starts fine but starts loosing path after a few cuts

Discussion in 'CNC Mills/Routers' started by bunty, Jul 29, 2016.

  1. bunty

    bunty New
    Builder

    Joined:
    Aug 28, 2015
    Messages:
    23
    Likes Received:
    3
    Hi, I am facing issue with the cutting of the new CNC router that I have been building. The CNC cutting starts fine and cuts the material fine initially but a few parts are cuts it starts loosing path. Some pictures of the same are as below:

    The specs of my machine are:
    • I am using Arduino UNO with Arduino CNC Shield.
    • The drivers for the steppers that I am using are DRV8825 (running at 2 amps)
    • I have tried this ChiliPepper and GRBL Panel
    • Router - Makita RT0700C
    • I use Eased
    Some observations:
    • After the cuts start going wrong restarting the whole setup again, results in it going wrong almost right from the beginning.
    • Waiting for a while seems to make it better.
    • The drivers for the steppers that I am using are DRV8825 and they have heat sinks and I am using forced cooling. These do not seem to get hot during the operation
    • Additionally before I added a router to the setup, I did several dry runs by mounting a pen and never had any problems. Some of the drawings took a almost a couple of hours but this problem never happened
    • I rechecked the wiring, the smooth movement of the gantries, cut material clogged in between the bests but none of them seemed to be an issue.
    • The first cut I did was:
      • using a design in Easel and the GCode was generated using Easel only
      • Router used was a dremel
      • This cut went of fine
    • For the current cuts (where things are going wrong), I am using:
      • Makita RT0700C for the router
      • Drawing are in ACad and Estlcam 9 for CAM
      • CNC program generation is using GRBL
      • I have tried this Gcode generated by Estlcam using both ChiliPepper and GRBL Panel. however in both the cases, the issue is the same IMG_20160729_102138.jpg IMG_20160729_102150.jpg
    Appreciate help/ pointers from the forum to give ideas on how to resolve the issue.
     
  2. David the swarfer

    David the swarfer OpenBuilds Team
    Staff Member Moderator Builder Resident Builder

    Joined:
    Aug 6, 2013
    Messages:
    3,239
    Likes Received:
    1,815
    Looks like the cutting bit is overpowering the stepper motors.
    Also, that router is much heavier than the Dremel, indicating that it is losing steps when accelerating.
    So, reduce the acceleration settings in GRBL. Just halve them and try a cut. If that works you can try 3/4 of the current value. As you increase the acceleration you will get to where it goes wrong again, reduce to 75% of that value for reliable operation.

    Slightly more complicated is to get more power out of the steppers by reducing the microstepping setting.
    Lower microstepping gives more apparent power.
    After adjusting the jumpers for the 8825's you will need to change the steps/mm settings in GRBL, and possibly recalibrate.
     
  3. bunty

    bunty New
    Builder

    Joined:
    Aug 28, 2015
    Messages:
    23
    Likes Received:
    3
    Thanks David will try the following in the the order below:
    1. Reduce acceleration on X and Y axis
    2. I have 1/2 step on X & Y axis and up the current to the motors
    Does the quality and type of end mill have a big impact on the steppers missing a step?
     
  4. David the swarfer

    David the swarfer OpenBuilds Team
    Staff Member Moderator Builder Resident Builder

    Joined:
    Aug 6, 2013
    Messages:
    3,239
    Likes Received:
    1,815
    yes, the endmill will affect the amount of power required to push it through the work.
    If it is blunt it is harder to move it forward into the work.
    If it gets clogged with chips you will have problems.

    For wood, use 1 or 2 flute cutters, 3 and 4 flute will probably get stuck full of chips.
    CALCULATE your feedrate. guessing works, calculating is better. Plenty of online resources for this, even here on this site under 'resources'. There is such a thing as 'too slow' and this can cause problems, you want the bit to cut and make chips, not dust. Feeding too slowly makes dust and heat.

    straight or spiral. spiral up or down? You will have to read the manufacturers web sites about these choices.
    I have had good cuts with all of them, but down spiral can more easily get clogged and set your work on fire so rpm, depth of cut and feedrate become very important. (fire!? yep, 30000rpm on a 3kw spindle will do that, just like Bear Grylls making fire with a bow drill).

    sharp! sharp is more important than just about everything else. If you are cutting aluminum and wood, you need 2 sets of bits, 1 set for aluminum, and 1 set for wood. And never mix them up! Besides the wood making them blunt much quicker than aluminum does, the geometry is different and this makes a big difference in how well they cut and how long they last.

    microstepping: only 2x stepping? Most OX style belt drive machines are good at 4x or 8x. But then I don't know enough about your machine to really say. Just don't use 1x microstepping, that way lies resonance and other nasties.

    have fun
     
    GrayUK likes this.
  5. bunty

    bunty New
    Builder

    Joined:
    Aug 28, 2015
    Messages:
    23
    Likes Received:
    3
    I have been building my machine for over a year now... doing bits and pieces to progress it as and when I get time. Since it is a hobby, I have been taking the learning a step at a time. Till recently it was only the build and just recently, I have started experimenting with the cuts.

    I have recently started going into the finer details of bit selection, feed rate ... Going by what you said, my initial cut with dremel was with a two flute bit. The one I am using now (the one causing the problem) is a four flute bit. May be clogging could be the reason for the deviation after some initial correct cuts ...

    Will check it out tomorrow...
     
    David the swarfer likes this.
  6. bunty

    bunty New
    Builder

    Joined:
    Aug 28, 2015
    Messages:
    23
    Likes Received:
    3
    Hi David, I tried the following over the weekend:
    1. Used a two flute up-cut end mill - However this too resulted in the machine loosing its path somewhere midway of the cut
    2. I then changed the acceleration of the X & Y axis from 20 mm/sec^2 to 10 mm/sec^2 - this lead to a successful cut
    I will try more cuts over the coming weekend to see the performance and fine tune the GRBL parameters.
     
  7. dean knipping

    Builder

    Joined:
    Mar 24, 2015
    Messages:
    84
    Likes Received:
    16
    something else to consider is how you were testing. For the first cut you used Easel only and a dremel and everything worked fine. Your current cuts are a Makita, ACad and Estlcam and something to generate the gCode and it doesn't work fine.

    Rather than trying change EVERYTHING at once, just change one thing at a time and see if it works (i.e. go back to what worked and try using the Makita instead of the dremel) and start changing the setup one item at a time. it takes longer but there could be more than one factor involved here
     
    GrayUK likes this.
  8. bunty

    bunty New
    Builder

    Joined:
    Aug 28, 2015
    Messages:
    23
    Likes Received:
    3
    Dean, I absolutely agree with you. Having too many variables all at once for some new stuff eventually may take a lot more time, re-work and be a cause of a lot of frustration :mad:. Guess it was just the excitement of getting things to work :) .. all the same lessons learnt.

    As of now I am back to reading/ learning a bit more about bit types, feed-rates etc before I attempt by next cuts. Guess I rather go slow than end in giving it up in frustration ...
     
  9. dean knipping

    Builder

    Joined:
    Mar 24, 2015
    Messages:
    84
    Likes Received:
    16
    bits, coatings, speeds and feeds are all good things to have down fairly well before going hog wild. Looking at the first cuts, they looked good so why not keep the same programing and just swap out the router and start there. find out what works best for what you are working on as that will give you good info on your feeds and speeds for a given material.

    I have been looking at these for the last year or so and have a 20+ year background in tech support / programing so I am used to taking it slow. Take your time, go step by step and you will get it working!
     
  10. GrayUK

    GrayUK Openbuilds Team Elder
    Staff Member Moderator Builder

    Joined:
    May 5, 2014
    Messages:
    1,668
    Likes Received:
    627
    Take a look at the CNC Cookbook. They pretty much specialise in feed rates etc.
    The program they have, they sell, but it has good reviews.

    CNCCookbook: Be a Better CNC'er

    Gray
     
    David the swarfer likes this.
  11. Christian Knüll

    Builder

    Joined:
    Mar 1, 2014
    Messages:
    27
    Likes Received:
    34
    Hi,

    • the DRV8825 has a thermal shutdown to prevent it from overheating. This shutdown often lasts only a fraction of a second and is almost not noticeable but causes step losses.
    • This fits well to your description: at startup the chip is cool so everything works fine - but after a while it gets too hot and the trouble starts and restarting does not help because the chip is still hot.
    • The chip is rated for 2.5A - but only with suitable cooling which is usually not the case on those little driver boards so it shuts down much earlier.
    • In case there are no heat sinks (or just very little ones) adding a reasonable heat sink will solve this kind of issue.

    Christian
     
    Mark Carew and GrayUK like this.
  12. bunty

    bunty New
    Builder

    Joined:
    Aug 28, 2015
    Messages:
    23
    Likes Received:
    3
    Hi Christian, thanks for the response and inputs.... I thought that may have been the problem. In fact one of Y axis steppers seem to shut down in the very initial cuts and thought that it may have been the problem. I replaced the DRV8825 with a spare one that I had. The motors that I have are NEMA23 rated for 2A, so I re-calibrated the current setting on the drivers, to ensure that the current draw was within range (current setting is 2A for the X& Y axis and about ~ 1.7 A for the Z axis Nema 17).

    My current setup is with small heat sinks and forced cooling with a fan. Though the heat sinks that I have there are the small stock heat sinks that camp with the drive. Additionally I stuck them with those think double sided tapes. Guess to be on the safer side, I need to put a bigger heat sink and a proper heat conductive glue to the drivers cool.

    This weekend I am more on a theoretical brush up on cutting speed, feed rate etc. But point taken... and another thing to add to my checklist of think before I do the next cut. One thing definitely to check, taking cue from your inputs above, before I do my next cut is to ensure that the current heat sinks do not have the cutting dust buildup on them.

    Thanks for all the inputs .. will keep posted on the next cut...
     
    Mark Carew likes this.

Share This Page

  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.
    Dismiss Notice