584,802 active members*
4,988 visitors online*
Register for free
Login
IndustryArena Forum > OpenSource CNC Design Center > Arduino > Bad cut on angled lines and arc
Page 1 of 2 12
Results 1 to 20 of 37
  1. #1
    Join Date
    Jun 2022
    Posts
    30

    Bad cut on angled lines and arc

    Hello guys,
    I am new here, as the hobbist in CNC I use grbl few years now.
    I have some problems and think I am not able to solve it by myself.
    I have feeling that it must be somehow arduino or grbl related.

    I can cut very nice smooth surface finish on X and dual Y.
    Problem is when machine goes into something not straight, like 45° or Arc, nice surface
    finish is gone and edgy. If I run program with spindle off, just cutting air,
    I can feel by hand on non-straight lines these vibration shaking.
    I have to mention that I checked everything mechanical, machine
    is squared, I use preloaded C5 ballscrew, there is no backlash. As motors
    I use Clearpath from Teknic with proper tuning. There is no tension on
    machine, everything is super tight and free to move. If I push hard on gantries,
    there is no measurable bending. As controller I use Arduino Uno with Grbl shield (no stepper drivers),
    I do not need that shield, but it is easier wiring with it for clonning Y to A.
    I have to mention that also changing feed rates does not help and does
    not affect surface finish on these non straight lines, always edgy.
    I tried to generate g-code in different programs - same result, also different g-code senders - same result.
    I modified several times grbl setting, also tried different steps/mm, junction deviation... still same.

    Does anybody know what causes it, I suspect there is something with grbl that cannot synchronize somehow XY movement ?

    I add photo:
    First two are cutting on X and on Y, its nice smooth cut. Other two are edgy, 45° and Arc on circle.
    On 45° is visible repetitive pattern in waves.



    I would be glad for any help.

    thanks

  2. #2
    Join Date
    Mar 2015
    Posts
    409

    Re: Bad cut on angled lines and arc

    The X and YA finish isn't that bad.

    The 45° line is probably just 1 line of gcode. That rules out the gcode sender and the CAM package.
    The driver and controller are absolutely capable of keeping 2 steppers synchronized in motion.

    The stepper resolution could cause such a problem. At what resolution (Steps / mm) are you running ($100, $101, $102 grbl setting)?

  3. #3
    Join Date
    Jun 2022
    Posts
    30

    Re: Bad cut on angled lines and arc

    Hi, I use 640 steps/mm, also tried 320, 160, 80. Still same. Yes, X and YA is ok, It can be much smoother with slower speed. Could it cause when ballnut is not attached to plate precisely 90°, is it so sensitive for it ?
    I am not sure how I adjusted nut at X, so I have to check it.

  4. #4
    Join Date
    Mar 2015
    Posts
    409

    Re: Bad cut on angled lines and arc

    Quote Originally Posted by Arche View Post
    Hi, I use 640 steps/mm, also tried 320, 160, 80. Still same.
    Did you change the settings on the grbl controller and the stepper driver?

    Quote Originally Posted by Arche View Post
    Could it cause when ballnut is not attached to plate precisely 90°, is it so sensitive for it ?.
    I don't think this will influence the finish.

    You could try climb milling versus conventional milling. climb milling should give a better finish but is sensitive for play and flex. There is not only play in the ball nut, but also in the shaft coupler, guides, shaft bearings, etc.

    You could do a finish cut at a shallow depth of cut to get a better finish. This will also reduce the cutting forces and this helps if the mill / router isn't rigid enough.

    Positioning the part can make a difference. Routers are less rigid in the mid of the table.

    Can you post the gcode used to mill the 45° part

  5. #5
    Join Date
    Jun 2022
    Posts
    30

    Re: Bad cut on angled lines and arc

    Yes, I change setting on both, grbl and stepper driver. Also tried climb vs conventional, climb has nicer finish, but result is same wavy pattern. I milled cube 60x60mm, after measure its 60x60, so it is very precise.
    Shallow has same result. Machine is very solid, no bending, so it does not matter if I cut on left side center... I tried it.

    Could it be something that I use GRBL Shield attached to Arduino Uno and somehow pulses from XYA can influences themself ? (I have no steppers on them, I use only side pins for step/dir, cloning Y on Arduino Uno is not cable friendly). I do not understand electronics, so maybe...


    Here is codefor 45°:


    G0 Z2.
    G0 X100.621 Y199.616
    G1 Z-15. F400.0
    G1 X199.616 Y100.621 Z-15. F1800.0
    G2 X199.616 Y97.369 Z-15. I-1.626 J-1.626
    G1 X100.621 Y-1.626 Z-15.
    G2 X99.011 Y-2.3 Z-15. I-1.626 J1.625
    G2 X97.369 Y-1.626 Z-15. I-0.017 J2.299
    G1 X-1.626 Y97.369 Z-15.
    G2 X-1.626 Y100.621 Z-15. I1.626 J1.626
    G1 X97.369 Y199.616 Z-15.
    G2 X100.621 Y199.616 Z-15. I1.626 J-1.626
    G0 Z2.

  6. #6
    Join Date
    Mar 2015
    Posts
    409

    Re: Bad cut on angled lines and arc

    According to the Gcode, you run at 640 steps/mm and 1800 mm feed rate. That are 19200 steps/sec about 19kHz. The max step rate of grbl running on an Uno is about 30 kHz. I prefer not running faster than 10 kHz. You could reduce the steps/mm because most routers don't benefit for more than 200 steps/mm . However, I don't think this is the cause.

    According to the Gcode, only one move is used for milling a 45° side so that is OK.

    Could it be something that I use GRBL Shield attached to Arduino Uno and somehow pulses from XYA can influences themself ? (I have no steppers on them, I use only side pins for step/dir, cloning Y on Arduino Uno is not cable friendly). I do not understand electronics, so maybe...
    If that is the case, your problems would be not so repeatable. I don't suspect the controller or wiring to be the cause of this problem

    I use Clearpath from Teknic with proper tuning.
    I use only side pins for step/dir, cloning Y
    I suspect that there is "resonance" in the feedback loop of one or both of the axis. As long as only one axis is moving, you won't notice this. You could run at 1/10 of the speed, that should change the pattern on the 45° side.

    Tuning is only optimal at one rate/load combination.
    If you drive one axis by 2 motors (cloning Y) you have to tune the motors for this situation. I have no experience in running 2 closed loop motors on 1 axis. I found this thread, maybe it helps.
    https://www.cnczone.com/forums/diy-c...wer-one-2.html

    The tuning software can probably show the position error during milling, in a graph. If so, you can locate the motor(s) that causes the problem.

    I am going to post your problem on the dutch CNC website. Maybe some one there recognizes the problem.

  7. #7
    Join Date
    Mar 2015
    Posts
    409

    Re: Bad cut on angled lines and arc

    It could also be something mechanical that has a relation to the spindle.
    What is the pitch of the spindles you are using?
    What is the "pitch" of the pattern in the 45° cut?
    How is de motor connected to the spindle?

  8. #8
    Join Date
    Jun 2022
    Posts
    30

    Re: Bad cut on angled lines and arc

    Hi, I will try look for this resonance errors, I have some diagnostic tool for it. I can try cut 1/10 speed if it helps.
    Also I ordered new screw for X because the source of it is uknown, after that I will have C5 preloaded all of them. I will recieve it in one month I think.
    On current screw I have zero backlash, I measured it. On dual Y are both C5 preloaded. As spindle I use regular 2,2 Kw water cooled spindle, chinese ones. I do not understand what spindle pitch means ?

  9. #9
    Join Date
    Mar 2015
    Posts
    409

    Re: Bad cut on angled lines and arc

    Quote Originally Posted by Arche View Post
    I can try cut 1/10 speed if it helps.
    It could help in finding the cause of the problems. If the pattern doesn't change, the cause is probably mechanical and related to the rotation of the spindle. If the distance between the pattern lines decreases, the problem is likely "time related" (tuning).
    Quote Originally Posted by Arche View Post
    I do not understand what spindle pitch means ?
    By spindle I mean the ball screw. The movement of 1 turn of this screw is the pitch.

    How are the motors attached to the ball screw?
    Is the Y-axis driven by 2 ball screws and 2 motors?

  10. #10
    Join Date
    Jun 2022
    Posts
    30

    Re: Bad cut on angled lines and arc

    10mm pitch, all of them, It is 1610. Motors are attached by couplers to screws and then through thick plates with FK and FF.

  11. #11
    Join Date
    Mar 2015
    Posts
    409

    Re: Bad cut on angled lines and arc

    10mm pitch, all of them, It is 1610. Motors are attached by couplers to screws and then through thick plates with FK and FF.
    At a feed of 1800 mm/min the screws will run 180 RPM. Long screws could resonate!!
    What is the size of your screws?
    What type of couplers are you using?

  12. #12
    Join Date
    Jun 2022
    Posts
    30

    Re: Bad cut on angled lines and arc

    All of them 1200mm, I think it is not resonate, still it is 3 turns per second, that is not many.
    Coupler type is similar to the picture.

    I used diagnostic for position errors and there is wave pattern. I contacted support, they will look at it.
    So I will know more next week.

  13. #13
    Join Date
    Mar 2015
    Posts
    409

    Re: Bad cut on angled lines and arc

    I got this document from the dutch CNC site for "auto tuning" a dual motor axis. It is from Teknic.
    Attachment 480958
    I used diagnostic for position errors and there is wave pattern. I contacted support, they will look at it.
    How large is the position deviation in that wave pattern?
    At witch axis did you found this pattern?

  14. #14
    Join Date
    Jun 2022
    Posts
    30

    Re: Bad cut on angled lines and arc

    I know this tuning, I did it several times but it seems does not work properly.
    Lets wait to next week, I will know more. The pattern is on all axis

  15. #15
    Join Date
    Jun 2022
    Posts
    30

    Re: Bad cut on angled lines and arc

    Hello,
    so I had servise assistance from engineer, they tuned
    both axis profiles and position error curve is now flattened. They
    told me that maybe I will not recognize difference in cut, because waveness of previous
    curve was not so extreme. They were right.
    I did test of 45° cut and it is little bit better, but still its not how
    I would image it should be. I would really imagine something close to straight cut
    of just X or Y, they are so smooth. I will make another test with new ballscrew when it comes
    for X. I add photo of slight improvement, but still it is not like solo X or Y.
    Attached Thumbnails Attached Thumbnails cut.jpg  

  16. #16
    Join Date
    Jun 2022
    Posts
    30

    Re: Bad cut on angled lines and arc

    Hello guys,
    so I installed new ballscrew on X and improvement of surface finish on 45° or arc cut is only tiny.
    I also tried to change 4x carriage HGH20 under X (not rails, do not have another ones) just for test, if surface finish change.
    It is the same, I added photo of it. Photo shows different speeds, slower is better but not so much.
    As I mentioned straight X cut is perfect and smooth, straight Y cut is smooth but sometimes it shows wave pattern, but still smooth.
    Problem are angled cuts and arcs.
    Is it possible that it could be that rails HGR20 with carriages are noname chinese ones, so quality is unknown ?
    Also my X with Z has around 85kg, is that weight ok for 4x HGH20 ?

  17. #17
    Join Date
    Mar 2015
    Posts
    409

    Re: Bad cut on angled lines and arc

    Originally Posted by Arche
    I can try cut 1/10 speed if it helps.
    It could help in finding the cause of the problems. If the pattern doesn't change, the cause is probably mechanical and related to the rotation of the spindle. If the distance between the pattern lines decreases, the problem is likely "time related" (tuning).
    It seems, can't see it clearly, that the pattern doesn't change when the feed changes?

    Also my X with Z has around 85kg, is that weight ok for 4x HGH20 ?
    That is only a low load!

    It took me 10 months, 200 thread cuts and 24 broken inserts to fix a problem threading M24x3 steel on my lathe. Just keep on finding the cause of your problem.

  18. #18
    Join Date
    Jun 2022
    Posts
    30

    Re: Bad cut on angled lines and arc

    Pattern seems to be same, just in higher speeds more visible.

  19. #19
    Join Date
    Jun 2022
    Posts
    30

    Re: Bad cut on angled lines and arc

    I have made another test. I removed one Y ballscrew and did cut on more rigid side. Same pattern like it is with two ballsrews.
    Is it possible if I had bad rails with carriages and since they are on machine horizontaly under X ( not on side), they can fight to sides with shaky pattern for its position ?

  20. #20
    Join Date
    Mar 2015
    Posts
    409

    Re: Bad cut on angled lines and arc

    Having 2 motors trying to control one axis is always a difficult (tricky) process and can result in a shaky pattern. If that was the case, I would expect the pattern to be depended on the feed rate and that is not the case.
    If you only have one motor, than a bad rail or carriage could result in a shaky pattern. If that was the case, I would expect the pattern would not be so repetitive and along the whole edge.

    You could mill two edges at an angle of 20° and 70°. I expect that on one side the pattern will be shallower and shorter and on the other side it will be deeper and longer (baddest edge). The axis opposite to the baddest edge is causing the trouble.
    If so, the movement on this axis is a bit irregular. This can be caused by bad bearings, crooked (not straight) spindle, crooked end machining, to lose fit of the spindle in the BK bearing, to lose fit of the bearing in the BK housing, etc.
    Swapping the BK/BF bearings (or removing the BF bearing) could help finding the cause.

Page 1 of 2 12

Similar Threads

  1. Machine can't cut straight angled lines.
    By shawnmelton in forum Avid CNC
    Replies: 3
    Last Post: 03-10-2019, 04:29 AM
  2. V-groove bearings on flat or 45 angled profile - bad idea?
    By Spikes in forum DIY CNC Router Table Machines
    Replies: 2
    Last Post: 03-28-2018, 03:34 PM
  3. MasterCam X arc or many lines...?
    By skmetal7 in forum Mastercam
    Replies: 5
    Last Post: 08-26-2014, 08:20 PM
  4. problem with angled cut
    By da_Rayman in forum CNC Plasma / Oxy Fuel Cutting Machines
    Replies: 19
    Last Post: 11-22-2012, 09:28 AM
  5. NOT break arc into lines
    By cncuser1 in forum Mastercam
    Replies: 8
    Last Post: 05-08-2008, 02:48 AM

Tags for this Thread

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •